From 1c5363153dc7ae694404e7732b4ce36eecc94ca7 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Mon, 12 Sep 2005 09:15:14 -0500 Subject: [PATCH 001/204] [SCSI] blacklist REPORT LUNS usage on transtec arrays They report being SCSI-3 but seem to give back rubbish to a REPORT_LUNS command. Force them to be sequentially scanned. Signed-off-by: James Bottomley --- drivers/scsi/scsi_devinfo.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/scsi/scsi_devinfo.c b/drivers/scsi/scsi_devinfo.c index 07b554affcf..64fc9e21f35 100644 --- a/drivers/scsi/scsi_devinfo.c +++ b/drivers/scsi/scsi_devinfo.c @@ -110,6 +110,7 @@ static struct { {"RELISYS", "Scorpio", NULL, BLIST_NOLUN}, /* responds to all lun */ {"SANKYO", "CP525", "6.64", BLIST_NOLUN}, /* causes failed REQ SENSE, extra reset */ {"TEXEL", "CD-ROM", "1.06", BLIST_NOLUN}, + {"transtec", "T5008", "0001", BLIST_NOREPORTLUN }, {"YAMAHA", "CDR100", "1.00", BLIST_NOLUN}, /* locks up */ {"YAMAHA", "CDR102", "1.00", BLIST_NOLUN}, /* locks up */ {"YAMAHA", "CRW8424S", "1.0", BLIST_NOLUN}, /* locks up */ From 154fb614df83086ceb18a2c19908154e78d4dc98 Mon Sep 17 00:00:00 2001 From: Dave C Boutcher Date: Tue, 13 Sep 2005 10:09:02 -0500 Subject: [PATCH 002/204] [SCSI] ibmvscsi compatibility fix Linda Xie ever so gently pointed out that she had a patch to preserve compatibility with older SLES targets, and I told her we didn't need to push it to mainline. This patch explicitly checks the version of the IBMVSCSI target and ensures that large scatterlists are not sent to older targets. Signed-off-by: Linda Xie Signed-off-by: Dave Boutcher Signed-off-by: James Bottomley --- drivers/scsi/ibmvscsi/ibmvscsi.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/drivers/scsi/ibmvscsi/ibmvscsi.c b/drivers/scsi/ibmvscsi/ibmvscsi.c index 5b14934ba86..ff25210b00b 100644 --- a/drivers/scsi/ibmvscsi/ibmvscsi.c +++ b/drivers/scsi/ibmvscsi/ibmvscsi.c @@ -727,6 +727,16 @@ static void adapter_info_rsp(struct srp_event_struct *evt_struct) if (hostdata->madapter_info.port_max_txu[0]) hostdata->host->max_sectors = hostdata->madapter_info.port_max_txu[0] >> 9; + + if (hostdata->madapter_info.os_type == 3 && + strcmp(hostdata->madapter_info.srp_version, "1.6a") <= 0) { + printk("ibmvscsi: host (Ver. %s) doesn't support large" + "transfers\n", + hostdata->madapter_info.srp_version); + printk("ibmvscsi: limiting scatterlists to %d\n", + MAX_INDIRECT_BUFS); + hostdata->host->sg_tablesize = MAX_INDIRECT_BUFS; + } } } From a89f29f6ea8dba0b7c4ae5d1d0b43de6cb500ea6 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Tue, 30 Aug 2005 11:48:07 -0500 Subject: [PATCH 003/204] [SCSI] aic7xxx: move to dma_get_required_mask() and correct 39 bit assumptions This patch moves aic7xxx over to the dma_get_required_mask() API and dumps its open coded memory check. It also appears from this bug: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=167049 That 39 bit addressing doesn't work on older cards. I surmise that the AHC_LARGE_SCBS flag is the one that marks cards capable of using 39 bit addressing, so I also folded that check into the code. Signed-off-by: James Bottomley --- drivers/scsi/aic7xxx/aic7xxx_osm.c | 9 --------- drivers/scsi/aic7xxx/aic7xxx_osm.h | 2 -- drivers/scsi/aic7xxx/aic7xxx_osm_pci.c | 8 +++++--- 3 files changed, 5 insertions(+), 14 deletions(-) diff --git a/drivers/scsi/aic7xxx/aic7xxx_osm.c b/drivers/scsi/aic7xxx/aic7xxx_osm.c index c932b3b9449..876d1de8480 100644 --- a/drivers/scsi/aic7xxx/aic7xxx_osm.c +++ b/drivers/scsi/aic7xxx/aic7xxx_osm.c @@ -1109,15 +1109,6 @@ ahc_linux_register_host(struct ahc_softc *ahc, struct scsi_host_template *templa return (0); } -uint64_t -ahc_linux_get_memsize(void) -{ - struct sysinfo si; - - si_meminfo(&si); - return ((uint64_t)si.totalram << PAGE_SHIFT); -} - /* * Place the SCSI bus into a known state by either resetting it, * or forcing transfer negotiations on the next command to any diff --git a/drivers/scsi/aic7xxx/aic7xxx_osm.h b/drivers/scsi/aic7xxx/aic7xxx_osm.h index c5299626924..be9edbe26db 100644 --- a/drivers/scsi/aic7xxx/aic7xxx_osm.h +++ b/drivers/scsi/aic7xxx/aic7xxx_osm.h @@ -494,8 +494,6 @@ ahc_insb(struct ahc_softc * ahc, long port, uint8_t *array, int count) int ahc_linux_register_host(struct ahc_softc *, struct scsi_host_template *); -uint64_t ahc_linux_get_memsize(void); - /*************************** Pretty Printing **********************************/ struct info_str { char *buffer; diff --git a/drivers/scsi/aic7xxx/aic7xxx_osm_pci.c b/drivers/scsi/aic7xxx/aic7xxx_osm_pci.c index 0d44a6907dd..3ce77ddc889 100644 --- a/drivers/scsi/aic7xxx/aic7xxx_osm_pci.c +++ b/drivers/scsi/aic7xxx/aic7xxx_osm_pci.c @@ -180,6 +180,7 @@ ahc_linux_pci_dev_probe(struct pci_dev *pdev, const struct pci_device_id *ent) struct ahc_pci_identity *entry; char *name; int error; + struct device *dev = &pdev->dev; pci = pdev; entry = ahc_find_pci_device(pci); @@ -209,11 +210,12 @@ ahc_linux_pci_dev_probe(struct pci_dev *pdev, const struct pci_device_id *ent) pci_set_master(pdev); if (sizeof(dma_addr_t) > 4 - && ahc_linux_get_memsize() > 0x80000000 - && pci_set_dma_mask(pdev, mask_39bit) == 0) { + && ahc->features & AHC_LARGE_SCBS + && dma_set_mask(dev, mask_39bit) == 0 + && dma_get_required_mask(dev) > DMA_32BIT_MASK) { ahc->flags |= AHC_39BIT_ADDRESSING; } else { - if (pci_set_dma_mask(pdev, DMA_32BIT_MASK)) { + if (dma_set_mask(dev, DMA_32BIT_MASK)) { printk(KERN_WARNING "aic7xxx: No suitable DMA available.\n"); return (-ENODEV); } From 186d330e682210100c671355580a8592e4a21692 Mon Sep 17 00:00:00 2001 From: Timothy Thelin Date: Tue, 13 Sep 2005 19:56:28 -0700 Subject: [PATCH 004/204] [SCSI] scsi: sd, sr, st, and scsi_lib all fail to copy cmd_len to new cmd This fixes an issue in scsi command initialization from a request where sd, sr, st, and scsi_lib all fail to copy the request's cmd_len to the scsi command's cmd_len field. Signed-off-by: Timothy Thelin Signed-off-by: James Bottomley --- drivers/scsi/scsi_lib.c | 1 + drivers/scsi/sd.c | 1 + drivers/scsi/sr.c | 1 + drivers/scsi/st.c | 1 + 4 files changed, 4 insertions(+) diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 863bb6495da..f9026d08532 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -1268,6 +1268,7 @@ static int scsi_prep_fn(struct request_queue *q, struct request *req) } } else { memcpy(cmd->cmnd, req->cmd, sizeof(cmd->cmnd)); + cmd->cmd_len = req->cmd_len; if (rq_data_dir(req) == WRITE) cmd->sc_data_direction = DMA_TO_DEVICE; else if (req->data_len) diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index de564b38605..9a1dc0cea03 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -235,6 +235,7 @@ static int sd_init_command(struct scsi_cmnd * SCpnt) return 0; memcpy(SCpnt->cmnd, rq->cmd, sizeof(SCpnt->cmnd)); + SCpnt->cmd_len = rq->cmd_len; if (rq_data_dir(rq) == WRITE) SCpnt->sc_data_direction = DMA_TO_DEVICE; else if (rq->data_len) diff --git a/drivers/scsi/sr.c b/drivers/scsi/sr.c index ce63fc8312d..561901b1cf1 100644 --- a/drivers/scsi/sr.c +++ b/drivers/scsi/sr.c @@ -326,6 +326,7 @@ static int sr_init_command(struct scsi_cmnd * SCpnt) return 0; memcpy(SCpnt->cmnd, rq->cmd, sizeof(SCpnt->cmnd)); + SCpnt->cmd_len = rq->cmd_len; if (!rq->data_len) SCpnt->sc_data_direction = DMA_NONE; else if (rq_data_dir(rq) == WRITE) diff --git a/drivers/scsi/st.c b/drivers/scsi/st.c index a93308ae973..d001c046551 100644 --- a/drivers/scsi/st.c +++ b/drivers/scsi/st.c @@ -4206,6 +4206,7 @@ static int st_init_command(struct scsi_cmnd *SCpnt) return 0; memcpy(SCpnt->cmnd, rq->cmd, sizeof(SCpnt->cmnd)); + SCpnt->cmd_len = rq->cmd_len; if (rq_data_dir(rq) == WRITE) SCpnt->sc_data_direction = DMA_TO_DEVICE; From d39a942c3f4061d2218dcadf79ce10010dbcdb3c Mon Sep 17 00:00:00 2001 From: "Randy.Dunlap" Date: Tue, 13 Sep 2005 21:43:56 -0700 Subject: [PATCH 005/204] [SCSI] scsi: 2 drivers need MODULE_LICENSE() Modules need a license to prevent kernel tainting. Signed-off-by: Randy Dunlap Signed-off-by: James Bottomley --- drivers/scsi/fd_mcs.c | 2 ++ drivers/scsi/ibmmca.c | 2 ++ 2 files changed, 4 insertions(+) diff --git a/drivers/scsi/fd_mcs.c b/drivers/scsi/fd_mcs.c index fa652f8aa64..d59d449a9e4 100644 --- a/drivers/scsi/fd_mcs.c +++ b/drivers/scsi/fd_mcs.c @@ -1360,3 +1360,5 @@ static Scsi_Host_Template driver_template = { .use_clustering = DISABLE_CLUSTERING, }; #include "scsi_module.c" + +MODULE_LICENSE("GPL"); diff --git a/drivers/scsi/ibmmca.c b/drivers/scsi/ibmmca.c index 6e54c7d9b33..19392f65127 100644 --- a/drivers/scsi/ibmmca.c +++ b/drivers/scsi/ibmmca.c @@ -460,6 +460,8 @@ MODULE_PARM(adisplay, "1i"); MODULE_PARM(normal, "1i"); MODULE_PARM(ansi, "1i"); #endif + +MODULE_LICENSE("GPL"); #endif /*counter of concurrent disk read/writes, to turn on/off disk led */ static int disk_rw_in_progress = 0; From 59897dad98d63ac15e1e36fcc3a107c892b1826c Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Wed, 14 Sep 2005 12:57:42 -0400 Subject: [PATCH 006/204] [SCSI] fix sym scsi boot hang On Wed, 2005-09-14 at 18:06 +1000, Anton Blanchard wrote: > And in particular it looks like the scsi_unprep_request in > scsi_queue_insert is causing it. The following patch fixes the boot > problems on the vscsi machine: OK, my fault. Your fix is almost correct .. I was going to do this eventually, honest, because there's no need to unprep and reprep a command that comes in through scsi_queue_insert(). However, I decided to leave it in to exercise the scsi_unprep_request() path just to make sure it was working. What's happening, I think, is that we also use this path for retries. Since we kill and reget the command each time, the retries decrement is never seen, so we're retrying forever. Signed-off-by: James Bottomley --- drivers/scsi/scsi_lib.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index f9026d08532..0c6924eecf5 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -140,14 +140,12 @@ static void scsi_unprep_request(struct request *req) * commands. * Notes: This could be called either from an interrupt context or a * normal process context. - * Notes: Upon return, cmd is a stale pointer. */ int scsi_queue_insert(struct scsi_cmnd *cmd, int reason) { struct Scsi_Host *host = cmd->device->host; struct scsi_device *device = cmd->device; struct request_queue *q = device->request_queue; - struct request *req = cmd->request; unsigned long flags; SCSI_LOG_MLQUEUE(1, @@ -188,9 +186,8 @@ int scsi_queue_insert(struct scsi_cmnd *cmd, int reason) * function. The SCSI request function detects the blocked condition * and plugs the queue appropriately. */ - scsi_unprep_request(req); spin_lock_irqsave(q->queue_lock, flags); - blk_requeue_request(q, req); + blk_requeue_request(q, cmd->request); spin_unlock_irqrestore(q->queue_lock, flags); scsi_run_queue(q); From b5683557331b129658ab435391df527ef72d63dc Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Thu, 15 Sep 2005 08:59:36 -0500 Subject: [PATCH 007/204] [SCSI] atp870u: fix memory addressing bug From: Alan Cox The virt_to_bus() wasn't correctly taken out of this driver. It needs to be able to track both physical and virtual addresses for its prd table. Update the driver to do this with separate tracking entries. Signed-off-by: Alan Cox Signed-off-by: James Bottomley --- drivers/scsi/atp870u.c | 6 ++++-- drivers/scsi/atp870u.h | 5 +++-- 2 files changed, 7 insertions(+), 4 deletions(-) diff --git a/drivers/scsi/atp870u.c b/drivers/scsi/atp870u.c index e6153fe5842..a8cfbef304b 100644 --- a/drivers/scsi/atp870u.c +++ b/drivers/scsi/atp870u.c @@ -996,6 +996,7 @@ oktosend: #ifdef ED_DBGP printk("send_s870: prdaddr_2 0x%8x tmpcip %x target_id %d\n", dev->id[c][target_id].prdaddr,tmpcip,target_id); #endif + dev->id[c][target_id].prdaddr = dev->id[c][target_id].prd_bus; outl(dev->id[c][target_id].prdaddr, tmpcip); tmpcip = tmpcip - 2; outb(0x06, tmpcip); @@ -2572,7 +2573,7 @@ static void atp870u_free_tables(struct Scsi_Host *host) for (k = 0; k < 16; k++) { if (!atp_dev->id[j][k].prd_table) continue; - pci_free_consistent(atp_dev->pdev, 1024, atp_dev->id[j][k].prd_table, atp_dev->id[j][k].prdaddr); + pci_free_consistent(atp_dev->pdev, 1024, atp_dev->id[j][k].prd_table, atp_dev->id[j][k].prd_bus); atp_dev->id[j][k].prd_table = NULL; } } @@ -2584,12 +2585,13 @@ static int atp870u_init_tables(struct Scsi_Host *host) int c,k; for(c=0;c < 2;c++) { for(k=0;k<16;k++) { - atp_dev->id[c][k].prd_table = pci_alloc_consistent(atp_dev->pdev, 1024, &(atp_dev->id[c][k].prdaddr)); + atp_dev->id[c][k].prd_table = pci_alloc_consistent(atp_dev->pdev, 1024, &(atp_dev->id[c][k].prd_bus)); if (!atp_dev->id[c][k].prd_table) { printk("atp870u_init_tables fail\n"); atp870u_free_tables(host); return -ENOMEM; } + atp_dev->id[c][k].prdaddr = atp_dev->id[c][k].prd_bus; atp_dev->id[c][k].devsp=0x20; atp_dev->id[c][k].devtype = 0x7f; atp_dev->id[c][k].curr_req = NULL; diff --git a/drivers/scsi/atp870u.h b/drivers/scsi/atp870u.h index 89f43af39cf..62bae64a01c 100644 --- a/drivers/scsi/atp870u.h +++ b/drivers/scsi/atp870u.h @@ -54,8 +54,9 @@ struct atp_unit unsigned long tran_len; unsigned long last_len; unsigned char *prd_pos; - unsigned char *prd_table; - dma_addr_t prdaddr; + unsigned char *prd_table; /* Kernel address of PRD table */ + dma_addr_t prd_bus; /* Bus address of PRD */ + dma_addr_t prdaddr; /* Dynamically updated in driver */ struct scsi_cmnd *curr_req; } id[2][16]; struct Scsi_Host *host; From 541950027f8b7c96a639bc16e48930c590f1b98a Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Thu, 15 Sep 2005 21:52:51 -0400 Subject: [PATCH 008/204] [SCSI] fix use after potential free in scsi_remove_device Signed-off-by: Alan Stern Signed-off-by: James Bottomley --- drivers/scsi/scsi_sysfs.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c index b8052d5206c..83f87c41b18 100644 --- a/drivers/scsi/scsi_sysfs.c +++ b/drivers/scsi/scsi_sysfs.c @@ -707,9 +707,11 @@ void __scsi_remove_device(struct scsi_device *sdev) **/ void scsi_remove_device(struct scsi_device *sdev) { - down(&sdev->host->scan_mutex); + struct Scsi_Host *shost = sdev->host; + + down(&shost->scan_mutex); __scsi_remove_device(sdev); - up(&sdev->host->scan_mutex); + up(&shost->scan_mutex); } EXPORT_SYMBOL(scsi_remove_device); From 6cb1269b9607649b5edf1c4e7818e0cf34a9db71 Mon Sep 17 00:00:00 2001 From: Dave Kleikamp Date: Thu, 15 Sep 2005 23:25:41 -0500 Subject: [PATCH 009/204] JFS: Fix sparse warnings, including endian error The fix in inode.c is a real bug. It could result in undeleted, yet unconnected files on big-endian hardware. The others are trivial. Signed-off-by: Dave Kleikamp --- fs/jfs/inode.c | 3 +-- fs/jfs/jfs_dmap.c | 2 +- fs/jfs/jfs_txnmgr.c | 6 +++--- 3 files changed, 5 insertions(+), 6 deletions(-) diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c index 0ec62d5310d..9f942ca8e4e 100644 --- a/fs/jfs/inode.c +++ b/fs/jfs/inode.c @@ -129,8 +129,7 @@ void jfs_delete_inode(struct inode *inode) jfs_info("In jfs_delete_inode, inode = 0x%p", inode); if (!is_bad_inode(inode) && - (JFS_IP(inode)->fileset == cpu_to_le32(FILESYSTEM_I))) { - + (JFS_IP(inode)->fileset == FILESYSTEM_I)) { truncate_inode_pages(&inode->i_data, 0); if (test_cflag(COMMIT_Freewmap, inode)) diff --git a/fs/jfs/jfs_dmap.c b/fs/jfs/jfs_dmap.c index c739626f5bf..eadf319bee2 100644 --- a/fs/jfs/jfs_dmap.c +++ b/fs/jfs/jfs_dmap.c @@ -3055,7 +3055,7 @@ static int cntlz(u32 value) * RETURN VALUES: * log2 number of blocks */ -int blkstol2(s64 nb) +static int blkstol2(s64 nb) { int l2nb; s64 mask; /* meant to be signed */ diff --git a/fs/jfs/jfs_txnmgr.c b/fs/jfs/jfs_txnmgr.c index c7a92f9deb2..8132fce8099 100644 --- a/fs/jfs/jfs_txnmgr.c +++ b/fs/jfs/jfs_txnmgr.c @@ -2070,8 +2070,8 @@ static void xtLog(struct jfs_log * log, struct tblock * tblk, struct lrd * lrd, * * function: log from maplock of freed data extents; */ -void mapLog(struct jfs_log * log, struct tblock * tblk, struct lrd * lrd, - struct tlock * tlck) +static void mapLog(struct jfs_log * log, struct tblock * tblk, struct lrd * lrd, + struct tlock * tlck) { struct pxd_lock *pxdlock; int i, nlock; @@ -2209,7 +2209,7 @@ void txEA(tid_t tid, struct inode *ip, dxd_t * oldea, dxd_t * newea) * function: synchronously write pages locked by transaction * after txLog() but before txUpdateMap(); */ -void txForce(struct tblock * tblk) +static void txForce(struct tblock * tblk) { struct tlock *tlck; lid_t lid, next; From 7a83e90b32a2b2500e0be6a5317ab411b39222c9 Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Thu, 15 Sep 2005 15:24:59 +0100 Subject: [PATCH 010/204] [PATCH] PATCH: silly in piix driver Signed-off-by: Alan Cox Signed-off-by: Jeff Garzik --- drivers/scsi/ata_piix.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/scsi/ata_piix.c b/drivers/scsi/ata_piix.c index 87e0c36f155..d71cef767ce 100644 --- a/drivers/scsi/ata_piix.c +++ b/drivers/scsi/ata_piix.c @@ -442,7 +442,6 @@ static void piix_sata_phy_reset(struct ata_port *ap) * piix_set_piomode - Initialize host controller PATA PIO timings * @ap: Port whose timings we are configuring * @adev: um - * @pio: PIO mode, 0 - 4 * * Set PIO mode for device, in host controller PCI config space. * From 17b14451fd2b187ddd6303726755a3af0a926b6c Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Thu, 15 Sep 2005 15:44:00 +0100 Subject: [PATCH 011/204] [PATCH] PATCH: remove function for non-PCI as requested Signed-off-by: Alan Cox Signed-off-by: Jeff Garzik --- drivers/scsi/libata-core.c | 81 +++++++++++++++++++++++--------------- include/linux/libata.h | 1 + 2 files changed, 50 insertions(+), 32 deletions(-) diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index 5cc53cd9323..72bdc91e148 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -4122,6 +4122,53 @@ err_out: return 0; } +/** + * ata_host_set_remove - PCI layer callback for device removal + * @host_set: ATA host set that was removed + * + * Unregister all objects associated with this host set. Free those + * objects. + * + * LOCKING: + * Inherited from calling layer (may sleep). + */ + + +void ata_host_set_remove(struct ata_host_set *host_set) +{ + struct ata_port *ap; + unsigned int i; + + for (i = 0; i < host_set->n_ports; i++) { + ap = host_set->ports[i]; + scsi_remove_host(ap->host); + } + + free_irq(host_set->irq, host_set); + + for (i = 0; i < host_set->n_ports; i++) { + ap = host_set->ports[i]; + + ata_scsi_release(ap->host); + + if ((ap->flags & ATA_FLAG_NO_LEGACY) == 0) { + struct ata_ioports *ioaddr = &ap->ioaddr; + + if (ioaddr->cmd_addr == 0x1f0) + release_region(0x1f0, 8); + else if (ioaddr->cmd_addr == 0x170) + release_region(0x170, 8); + } + + scsi_host_put(ap->host); + } + + if (host_set->ops->host_stop) + host_set->ops->host_stop(host_set); + + kfree(host_set); +} + /** * ata_scsi_release - SCSI layer callback hook for host unload * @host: libata host to be unloaded @@ -4462,39 +4509,8 @@ void ata_pci_remove_one (struct pci_dev *pdev) { struct device *dev = pci_dev_to_dev(pdev); struct ata_host_set *host_set = dev_get_drvdata(dev); - struct ata_port *ap; - unsigned int i; - - for (i = 0; i < host_set->n_ports; i++) { - ap = host_set->ports[i]; - - scsi_remove_host(ap->host); - } - - free_irq(host_set->irq, host_set); - - for (i = 0; i < host_set->n_ports; i++) { - ap = host_set->ports[i]; - - ata_scsi_release(ap->host); - - if ((ap->flags & ATA_FLAG_NO_LEGACY) == 0) { - struct ata_ioports *ioaddr = &ap->ioaddr; - - if (ioaddr->cmd_addr == 0x1f0) - release_region(0x1f0, 8); - else if (ioaddr->cmd_addr == 0x170) - release_region(0x170, 8); - } - - scsi_host_put(ap->host); - } - - if (host_set->ops->host_stop) - host_set->ops->host_stop(host_set); - - kfree(host_set); + ata_host_set_remove(host_set); pci_release_regions(pdev); pci_disable_device(pdev); dev_set_drvdata(dev, NULL); @@ -4564,6 +4580,7 @@ module_exit(ata_exit); EXPORT_SYMBOL_GPL(ata_std_bios_param); EXPORT_SYMBOL_GPL(ata_std_ports); EXPORT_SYMBOL_GPL(ata_device_add); +EXPORT_SYMBOL_GPL(ata_host_set_remove); EXPORT_SYMBOL_GPL(ata_sg_init); EXPORT_SYMBOL_GPL(ata_sg_init_one); EXPORT_SYMBOL_GPL(ata_qc_complete); diff --git a/include/linux/libata.h b/include/linux/libata.h index 022105c745f..ceee1fc42c6 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -393,6 +393,7 @@ extern int ata_pci_init_one (struct pci_dev *pdev, struct ata_port_info **port_i extern void ata_pci_remove_one (struct pci_dev *pdev); #endif /* CONFIG_PCI */ extern int ata_device_add(struct ata_probe_ent *ent); +extern void ata_host_set_remove(struct ata_host_set *host_set); extern int ata_scsi_detect(Scsi_Host_Template *sht); extern int ata_scsi_ioctl(struct scsi_device *dev, int cmd, void __user *arg); extern int ata_scsi_queuecmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *)); From 1832a5862f2e1b4e5835611ee14bc30a8ed3cad5 Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Fri, 16 Sep 2005 11:01:14 +0200 Subject: [PATCH 012/204] [SCSI] change port speed definitions for scsi_transport_fc obviously FC Port Speeds in scsi_transport_fc.h are defined according to FC-HBA: #define FC_PORTSPEED_1GBIT 1 #define FC_PORTSPEED_2GBIT 2 #define FC_PORTSPEED_10GBIT 4 #define FC_PORTSPEED_4GBIT 8 Problem is, whoever invented FC-HBA did not care about FC-FS or FC-GS-x. Following FC-FS/FC-GS-x defintions of port speeds would look like: 1 GBit: 0x0001 2 GBit: 0x0002 4 GBit: 0x0004 10GBit: 0x0008 (and new in FC-LS: 8 Gbit: 0x0010 16GBit: 0x0020) I really appreciate if scsi_transport_fc.h would define port speeds according to FC-GS-x/FC-FS. Thus mapping of port speed capabilities to values defined in scsi_transport_fc.h can be avoided in the LLDD. Attached is a patch to change the definitions. Signed-off-by: James Bottomley --- include/scsi/scsi_transport_fc.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/include/scsi/scsi_transport_fc.h b/include/scsi/scsi_transport_fc.h index 115db056dc6..b0d44543737 100644 --- a/include/scsi/scsi_transport_fc.h +++ b/include/scsi/scsi_transport_fc.h @@ -103,8 +103,8 @@ enum fc_port_state { incapable of reporting */ #define FC_PORTSPEED_1GBIT 1 #define FC_PORTSPEED_2GBIT 2 -#define FC_PORTSPEED_10GBIT 4 -#define FC_PORTSPEED_4GBIT 8 +#define FC_PORTSPEED_4GBIT 4 +#define FC_PORTSPEED_10GBIT 8 #define FC_PORTSPEED_NOT_NEGOTIATED (1 << 15) /* Speed not established */ /* From b95be99d52ce4f9db9ff0bd5f10e9e2066da6d2e Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Thu, 15 Sep 2005 13:52:51 -0400 Subject: [PATCH 013/204] [SCSI] fix oops in scsi_release_buffers() I found one other thing that needs to be fixed. The call to scsi_release_buffers in scsi_unprep_request causes an oops, because the sgtable has already been freed in scsi_io_completion. The following patch is needed. Signed-off-by: Alan Stern Signed-off-by: James Bottomley --- drivers/scsi/scsi_lib.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 0c6924eecf5..f065cbc1a6e 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -118,7 +118,6 @@ static void scsi_unprep_request(struct request *req) req->flags &= ~REQ_DONTPREP; req->special = (req->flags & REQ_SPECIAL) ? cmd->sc_request : NULL; - scsi_release_buffers(cmd); scsi_put_command(cmd); } @@ -1512,7 +1511,6 @@ static void scsi_request_fn(struct request_queue *q) * cases (host limits or settings) should run the queue at some * later time. */ - scsi_unprep_request(req); spin_lock_irq(q->queue_lock); blk_requeue_request(q, req); sdev->device_busy--; From 7f8c0fd78dccaf30e60cb4303bd7a21c7d9e6da3 Mon Sep 17 00:00:00 2001 From: Russell King Date: Sun, 18 Sep 2005 21:11:08 +0100 Subject: [PATCH 014/204] [ARM] Fix warning in asm/futex.h The recently added futex.h contains an unused variable, which gcc naturally warns about. Remove this unused variable. Signed-off-by: Russell King --- include/asm-arm/futex.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/asm-arm/futex.h b/include/asm-arm/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-arm/futex.h +++ b/include/asm-arm/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; From e0a200894efa33ea1ac9957717ae151afa02bb6f Mon Sep 17 00:00:00 2001 From: Deepak Saxena Date: Sun, 18 Sep 2005 21:11:56 +0100 Subject: [PATCH 015/204] [ARM] 2917/1: Make IXP4xx mach_desc's depend on config options Patch from Deepak Saxena Building a kernel for IXDP425 currently includes the machine descriptors for IXDP465 and PRPMC1100 even if those machines are not configured. This means we can build a kernel that boots on those machines even though the machine_is_xxx() macro will always return 0 and other bits such as PCI won't be compiled in. This can lead to many wasted hours wondering what you have done to your kernel to make it randomly crash thus requireing large quantities of beer to be consumed. While I am all for consumption of large quantities of beer, there are better reasons to do so then stupid kernel bugs. Signed-off-by: Deepak Saxena Signed-off-by: Russell King --- arch/arm/mach-ixp4xx/ixdp425-setup.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/arch/arm/mach-ixp4xx/ixdp425-setup.c b/arch/arm/mach-ixp4xx/ixdp425-setup.c index ae1fa099d5f..39b06ed8064 100644 --- a/arch/arm/mach-ixp4xx/ixdp425-setup.c +++ b/arch/arm/mach-ixp4xx/ixdp425-setup.c @@ -123,6 +123,7 @@ static void __init ixdp425_init(void) platform_add_devices(ixdp425_devices, ARRAY_SIZE(ixdp425_devices)); } +#ifdef CONFIG_ARCH_IXDP465 MACHINE_START(IXDP425, "Intel IXDP425 Development Platform") /* Maintainer: MontaVista Software, Inc. */ .phys_ram = PHYS_OFFSET, @@ -134,7 +135,9 @@ MACHINE_START(IXDP425, "Intel IXDP425 Development Platform") .boot_params = 0x0100, .init_machine = ixdp425_init, MACHINE_END +#endif +#ifdef CONFIG_MACH_IXDP465 MACHINE_START(IXDP465, "Intel IXDP465 Development Platform") /* Maintainer: MontaVista Software, Inc. */ .phys_ram = PHYS_OFFSET, @@ -146,7 +149,9 @@ MACHINE_START(IXDP465, "Intel IXDP465 Development Platform") .boot_params = 0x0100, .init_machine = ixdp425_init, MACHINE_END +#endif +#ifdef CONFIG_ARCH_PRPMC1100 MACHINE_START(IXCDP1100, "Intel IXCDP1100 Development Platform") /* Maintainer: MontaVista Software, Inc. */ .phys_ram = PHYS_OFFSET, @@ -158,6 +163,7 @@ MACHINE_START(IXCDP1100, "Intel IXCDP1100 Development Platform") .boot_params = 0x0100, .init_machine = ixdp425_init, MACHINE_END +#endif /* * Avila is functionally equivalent to IXDP425 except that it adds From a64358db1253b35d508a411e80a3ad23b859ec88 Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Tue, 26 Jul 2005 10:27:10 -0400 Subject: [PATCH 016/204] [SCSI] SCSI scanning and removal fixes This patch (as545) fixes the list traversals in __scsi_remove_target and scsi_forget_host. In each case the existing code list_for_each_entry_safe in an _unsafe_ manner, because the list was not protected from outside modification while the iteration was running. The new scsi_forget_host routine takes the moderately controversial step of iterating over devices for removal rather than iterating over targets. This makes more sense to me because the current scheme treats targets as second-class citizens, created and removed on demand, rather than as objects corresponding to actual hardware. (Also I couldn't figure out any safe way to iterate over the target list, since it's not so easy to tell when a target has already been removed.) Signed-off-by: Alan Stern Signed-off-by: James Bottomley --- drivers/scsi/scsi_scan.c | 20 +++++++------------- drivers/scsi/scsi_sysfs.c | 9 ++++++--- 2 files changed, 13 insertions(+), 16 deletions(-) diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c index b86f170fa8e..fcf9f6cbb14 100644 --- a/drivers/scsi/scsi_scan.c +++ b/drivers/scsi/scsi_scan.c @@ -1466,23 +1466,17 @@ EXPORT_SYMBOL(scsi_scan_single_target); void scsi_forget_host(struct Scsi_Host *shost) { - struct scsi_target *starget, *tmp; + struct scsi_device *sdev; unsigned long flags; - /* - * Ok, this look a bit strange. We always look for the first device - * on the list as scsi_remove_device removes them from it - thus we - * also have to release the lock. - * We don't need to get another reference to the device before - * releasing the lock as we already own the reference from - * scsi_register_device that's release in scsi_remove_device. And - * after that we don't look at sdev anymore. - */ + restart: spin_lock_irqsave(shost->host_lock, flags); - list_for_each_entry_safe(starget, tmp, &shost->__targets, siblings) { + list_for_each_entry(sdev, &shost->__devices, siblings) { + if (sdev->sdev_state == SDEV_DEL) + continue; spin_unlock_irqrestore(shost->host_lock, flags); - scsi_remove_target(&starget->dev); - spin_lock_irqsave(shost->host_lock, flags); + __scsi_remove_device(sdev); + goto restart; } spin_unlock_irqrestore(shost->host_lock, flags); } diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c index 83f87c41b18..1e47b7eddef 100644 --- a/drivers/scsi/scsi_sysfs.c +++ b/drivers/scsi/scsi_sysfs.c @@ -719,17 +719,20 @@ void __scsi_remove_target(struct scsi_target *starget) { struct Scsi_Host *shost = dev_to_shost(starget->dev.parent); unsigned long flags; - struct scsi_device *sdev, *tmp; + struct scsi_device *sdev; spin_lock_irqsave(shost->host_lock, flags); starget->reap_ref++; - list_for_each_entry_safe(sdev, tmp, &shost->__devices, siblings) { + restart: + list_for_each_entry(sdev, &shost->__devices, siblings) { if (sdev->channel != starget->channel || - sdev->id != starget->id) + sdev->id != starget->id || + sdev->sdev_state == SDEV_DEL) continue; spin_unlock_irqrestore(shost->host_lock, flags); scsi_remove_device(sdev); spin_lock_irqsave(shost->host_lock, flags); + goto restart; } spin_unlock_irqrestore(shost->host_lock, flags); scsi_target_reap(starget); From 63150fcf73dbfb752a850a5e2fe3aed570154433 Mon Sep 17 00:00:00 2001 From: Russell King Date: Sun, 18 Sep 2005 21:49:29 +0100 Subject: [PATCH 017/204] [ARM] Fix warning in arch/arm/kernel/semaphore.c Newer binutils complains: /tmp/cc07pbI9.s:146: Warning: ignoring changed section type for .sched.text Fix this warning by adding %progbits to the .section. Signed-off-by: Russell King --- arch/arm/kernel/semaphore.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/kernel/semaphore.c b/arch/arm/kernel/semaphore.c index ac423e3e224..4c31f292305 100644 --- a/arch/arm/kernel/semaphore.c +++ b/arch/arm/kernel/semaphore.c @@ -178,7 +178,7 @@ int __down_trylock(struct semaphore * sem) * registers (r0 to r3 and lr), but not ip, as we use it as a return * value in some cases.. */ -asm(" .section .sched.text,\"ax\" \n\ +asm(" .section .sched.text,\"ax\",%progbits \n\ .align 5 \n\ .globl __down_failed \n\ __down_failed: \n\ From 118ec0b3e02e7d22cf27b9f5a2836fa6f395b61a Mon Sep 17 00:00:00 2001 From: Russell King Date: Sun, 18 Sep 2005 22:16:41 +0100 Subject: [PATCH 018/204] [ARM] Fix warning in arch/arm/mach-sa1100/generic.c Fix: arch/arm/mach-sa1100/generic.c:224: warning: 'struct mcp_plat_data' declared inside parameter list caused by mussing structure and function declaration. Signed-off-by: Russell King --- arch/arm/mach-sa1100/generic.h | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/arm/mach-sa1100/generic.h b/arch/arm/mach-sa1100/generic.h index 279e3afa3c3..f085d68e568 100644 --- a/arch/arm/mach-sa1100/generic.h +++ b/arch/arm/mach-sa1100/generic.h @@ -39,3 +39,6 @@ extern void sa11x0_set_ssp_data(struct sa11x0_ssp_plat_ops *ops); struct irda_platform_data; void sa11x0_set_irda_data(struct irda_platform_data *irda); + +struct mcp_plat_data; +void sa11x0_set_mcp_data(struct mcp_plat_data *data); From 51574e0398a2d93cbf7f26e36b673cd919062268 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Mon, 12 Sep 2005 09:52:28 -0700 Subject: [PATCH 019/204] [PATCH] IPoIB: fix module removal race Since ipoib uses queue_delayed_work to run flush task on port state events, it must flush scheduled work after unregistering the event handler. Signed-off-by: Michael S. Tsirkin Signed-off-by: Roland Dreier --- drivers/infiniband/ulp/ipoib/ipoib_main.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index 49d120d2b92..704f48e0b6a 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -1005,6 +1005,7 @@ debug_failed: register_failed: ib_unregister_event_handler(&priv->event_handler); + flush_scheduled_work(); event_failed: ipoib_dev_cleanup(priv->dev); @@ -1057,6 +1058,7 @@ static void ipoib_remove_one(struct ib_device *device) list_for_each_entry_safe(priv, tmp, dev_list, list) { ib_unregister_event_handler(&priv->event_handler); + flush_scheduled_work(); unregister_netdev(priv->dev); ipoib_dev_cleanup(priv->dev); From bb4a7f0da75ce9f3e933880428d39b2aa1f16961 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Mon, 12 Sep 2005 14:08:51 -0700 Subject: [PATCH 020/204] [PATCH] IB/mthca: assign ACK timeout field correctly The hardware reads the ACK timeout field from the most significant 5 bits of struct mthca_qp_path's ackto field, not the least significant bits. This fix has the driver put the timeout in the right place. Without this, we get a timeout that is 2^8 times too small. Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_qp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c index bcef06bf15e..2c883c1fd53 100644 --- a/drivers/infiniband/hw/mthca/mthca_qp.c +++ b/drivers/infiniband/hw/mthca/mthca_qp.c @@ -687,7 +687,7 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask) } if (attr_mask & IB_QP_TIMEOUT) { - qp_context->pri_path.ackto = attr->timeout; + qp_context->pri_path.ackto = attr->timeout << 3; qp_param->opt_param_mask |= cpu_to_be32(MTHCA_QP_OPTPAR_ACK_TIMEOUT); } From d6cff021e24515255b296d399ec517a68bf2ed08 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Tue, 13 Sep 2005 10:41:03 -0700 Subject: [PATCH 021/204] [PATCH] IB/mthca: fix posting of first work request Fix posting first WQE for mem-free HCAs: we need to link to previous WQE even in that case. While we're at it, simplify code for Tavor-mode HCAs. We don't really need the conditional test there either; we can similarly always link to the previous WQE. Based on Michael S. Tsirkin's analogous fix for userspace libmthca. Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_qp.c | 48 ++++++++++++------------- drivers/infiniband/hw/mthca/mthca_srq.c | 14 ++++---- 2 files changed, 28 insertions(+), 34 deletions(-) diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c index 2c883c1fd53..eaa52185763 100644 --- a/drivers/infiniband/hw/mthca/mthca_qp.c +++ b/drivers/infiniband/hw/mthca/mthca_qp.c @@ -227,7 +227,6 @@ static void mthca_wq_init(struct mthca_wq *wq) wq->last_comp = wq->max - 1; wq->head = 0; wq->tail = 0; - wq->last = NULL; } void mthca_qp_event(struct mthca_dev *dev, u32 qpn, @@ -1103,6 +1102,9 @@ static int mthca_alloc_qp_common(struct mthca_dev *dev, } } + qp->sq.last = get_send_wqe(qp, qp->sq.max - 1); + qp->rq.last = get_recv_wqe(qp, qp->rq.max - 1); + return 0; } @@ -1583,15 +1585,13 @@ int mthca_tavor_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto out; } - if (prev_wqe) { - ((struct mthca_next_seg *) prev_wqe)->nda_op = - cpu_to_be32(((ind << qp->sq.wqe_shift) + - qp->send_wqe_offset) | - mthca_opcode[wr->opcode]); - wmb(); - ((struct mthca_next_seg *) prev_wqe)->ee_nds = - cpu_to_be32((size0 ? 0 : MTHCA_NEXT_DBD) | size); - } + ((struct mthca_next_seg *) prev_wqe)->nda_op = + cpu_to_be32(((ind << qp->sq.wqe_shift) + + qp->send_wqe_offset) | + mthca_opcode[wr->opcode]); + wmb(); + ((struct mthca_next_seg *) prev_wqe)->ee_nds = + cpu_to_be32((size0 ? 0 : MTHCA_NEXT_DBD) | size); if (!size0) { size0 = size; @@ -1688,13 +1688,11 @@ int mthca_tavor_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr, qp->wrid[ind] = wr->wr_id; - if (likely(prev_wqe)) { - ((struct mthca_next_seg *) prev_wqe)->nda_op = - cpu_to_be32((ind << qp->rq.wqe_shift) | 1); - wmb(); - ((struct mthca_next_seg *) prev_wqe)->ee_nds = - cpu_to_be32(MTHCA_NEXT_DBD | size); - } + ((struct mthca_next_seg *) prev_wqe)->nda_op = + cpu_to_be32((ind << qp->rq.wqe_shift) | 1); + wmb(); + ((struct mthca_next_seg *) prev_wqe)->ee_nds = + cpu_to_be32(MTHCA_NEXT_DBD | size); if (!size0) size0 = size; @@ -1905,15 +1903,13 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto out; } - if (likely(prev_wqe)) { - ((struct mthca_next_seg *) prev_wqe)->nda_op = - cpu_to_be32(((ind << qp->sq.wqe_shift) + - qp->send_wqe_offset) | - mthca_opcode[wr->opcode]); - wmb(); - ((struct mthca_next_seg *) prev_wqe)->ee_nds = - cpu_to_be32(MTHCA_NEXT_DBD | size); - } + ((struct mthca_next_seg *) prev_wqe)->nda_op = + cpu_to_be32(((ind << qp->sq.wqe_shift) + + qp->send_wqe_offset) | + mthca_opcode[wr->opcode]); + wmb(); + ((struct mthca_next_seg *) prev_wqe)->ee_nds = + cpu_to_be32(MTHCA_NEXT_DBD | size); if (!size0) { size0 = size; diff --git a/drivers/infiniband/hw/mthca/mthca_srq.c b/drivers/infiniband/hw/mthca/mthca_srq.c index 75cd2d84ef1..fe06cc0df93 100644 --- a/drivers/infiniband/hw/mthca/mthca_srq.c +++ b/drivers/infiniband/hw/mthca/mthca_srq.c @@ -189,7 +189,6 @@ int mthca_alloc_srq(struct mthca_dev *dev, struct mthca_pd *pd, srq->max = attr->max_wr; srq->max_gs = attr->max_sge; - srq->last = NULL; srq->counter = 0; if (mthca_is_memfree(dev)) @@ -264,6 +263,7 @@ int mthca_alloc_srq(struct mthca_dev *dev, struct mthca_pd *pd, srq->first_free = 0; srq->last_free = srq->max - 1; + srq->last = get_wqe(srq, srq->max - 1); return 0; @@ -446,13 +446,11 @@ int mthca_tavor_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, ((struct mthca_data_seg *) wqe)->addr = 0; } - if (likely(prev_wqe)) { - ((struct mthca_next_seg *) prev_wqe)->nda_op = - cpu_to_be32((ind << srq->wqe_shift) | 1); - wmb(); - ((struct mthca_next_seg *) prev_wqe)->ee_nds = - cpu_to_be32(MTHCA_NEXT_DBD); - } + ((struct mthca_next_seg *) prev_wqe)->nda_op = + cpu_to_be32((ind << srq->wqe_shift) | 1); + wmb(); + ((struct mthca_next_seg *) prev_wqe)->ee_nds = + cpu_to_be32(MTHCA_NEXT_DBD); srq->wrid[ind] = wr->wr_id; srq->first_free = next_ind; From ce5b65cc9626feac0d4ffb96f798407e50c45575 Mon Sep 17 00:00:00 2001 From: Hal Rosenstock Date: Sun, 18 Sep 2005 13:47:53 -0700 Subject: [PATCH 022/204] [PATCH] IPoIB: Fix SA client retransmission strategy We got a little mixed up with what the backoff member holds in the IPoIB multicast group structure: sometimes it was used as a number of seconds, and sometimes it was used as a number of jiffies. Fix the code so that backoff is always in seconds. Signed-off-by: Hal Rosenstock Signed-off-by: Roland Dreier --- drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index aca7aea18a6..3aab4d85abf 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -145,7 +145,7 @@ static struct ipoib_mcast *ipoib_mcast_alloc(struct net_device *dev, mcast->dev = dev; mcast->created = jiffies; - mcast->backoff = HZ; + mcast->backoff = 1; mcast->logcount = 0; INIT_LIST_HEAD(&mcast->list); @@ -396,7 +396,7 @@ static void ipoib_mcast_join_complete(int status, IPOIB_GID_ARG(mcast->mcmember.mgid), status); if (!status && !ipoib_mcast_join_finish(mcast, mcmember)) { - mcast->backoff = HZ; + mcast->backoff = 1; down(&mcast_mutex); if (test_bit(IPOIB_MCAST_RUN, &priv->flags)) queue_work(ipoib_workqueue, &priv->mcast_task); @@ -496,7 +496,7 @@ static void ipoib_mcast_join(struct net_device *dev, struct ipoib_mcast *mcast, if (test_bit(IPOIB_MCAST_RUN, &priv->flags)) queue_delayed_work(ipoib_workqueue, &priv->mcast_task, - mcast->backoff); + mcast->backoff * HZ); up(&mcast_mutex); } else mcast->query_id = ret; From c915033fc62d7186d243d89f88782d6be33fd8f6 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Sun, 18 Sep 2005 13:52:06 -0700 Subject: [PATCH 023/204] [PATCH] IB/mthca: Initialize eq->nent before we use it In mthca_create_eq(), we call get_eqe() before setting eq->nent. This is wrong, because get_eqe() uses eq->nent. Fix this, and clean up the code a little while we're at it. (We got lucky with the current code, because eq->nent was cleared to 0, which get_eqe() made happen to do the right thing) Pointed out by Michael S. Tsirkin Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_eq.c | 16 +++++----------- 1 file changed, 5 insertions(+), 11 deletions(-) diff --git a/drivers/infiniband/hw/mthca/mthca_eq.c b/drivers/infiniband/hw/mthca/mthca_eq.c index 18f0981eb0c..78152a8ad17 100644 --- a/drivers/infiniband/hw/mthca/mthca_eq.c +++ b/drivers/infiniband/hw/mthca/mthca_eq.c @@ -476,12 +476,8 @@ static int __devinit mthca_create_eq(struct mthca_dev *dev, int i; u8 status; - /* Make sure EQ size is aligned to a power of 2 size. */ - for (i = 1; i < nent; i <<= 1) - ; /* nothing */ - nent = i; - - eq->dev = dev; + eq->dev = dev; + eq->nent = roundup_pow_of_two(max(nent, 2)); eq->page_list = kmalloc(npages * sizeof *eq->page_list, GFP_KERNEL); @@ -512,7 +508,7 @@ static int __devinit mthca_create_eq(struct mthca_dev *dev, memset(eq->page_list[i].buf, 0, PAGE_SIZE); } - for (i = 0; i < nent; ++i) + for (i = 0; i < eq->nent; ++i) set_eqe_hw(get_eqe(eq, i)); eq->eqn = mthca_alloc(&dev->eq_table.alloc); @@ -528,8 +524,6 @@ static int __devinit mthca_create_eq(struct mthca_dev *dev, if (err) goto err_out_free_eq; - eq->nent = nent; - memset(eq_context, 0, sizeof *eq_context); eq_context->flags = cpu_to_be32(MTHCA_EQ_STATUS_OK | MTHCA_EQ_OWNER_HW | @@ -538,7 +532,7 @@ static int __devinit mthca_create_eq(struct mthca_dev *dev, if (mthca_is_memfree(dev)) eq_context->flags |= cpu_to_be32(MTHCA_EQ_STATE_ARBEL); - eq_context->logsize_usrpage = cpu_to_be32((ffs(nent) - 1) << 24); + eq_context->logsize_usrpage = cpu_to_be32((ffs(eq->nent) - 1) << 24); if (mthca_is_memfree(dev)) { eq_context->arbel_pd = cpu_to_be32(dev->driver_pd.pd_num); } else { @@ -569,7 +563,7 @@ static int __devinit mthca_create_eq(struct mthca_dev *dev, dev->eq_table.arm_mask |= eq->eqn_mask; mthca_dbg(dev, "Allocated EQ %d with %d entries\n", - eq->eqn, nent); + eq->eqn, eq->nent); return err; From 3853194c2e174cee4da093c67bd54cbf9a38559a Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Sun, 18 Sep 2005 14:00:17 -0700 Subject: [PATCH 024/204] [PATCH] IB/mthca: Fix posting work requests to shared receive queues The error handling paths in mthca_tavor_post_srq_recv() and mthca_arbel_post_srq_recv() are quite bogus, the result of a screwed up merge. Fix them so they work as intended. Pointed out by Michael S. Tsirkin Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_srq.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/drivers/infiniband/hw/mthca/mthca_srq.c b/drivers/infiniband/hw/mthca/mthca_srq.c index fe06cc0df93..45dedd2787a 100644 --- a/drivers/infiniband/hw/mthca/mthca_srq.c +++ b/drivers/infiniband/hw/mthca/mthca_srq.c @@ -409,7 +409,7 @@ int mthca_tavor_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, mthca_err(dev, "SRQ %06x full\n", srq->srqn); err = -ENOMEM; *bad_wr = wr; - return nreq; + break; } wqe = get_wqe(srq, ind); @@ -427,7 +427,7 @@ int mthca_tavor_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, err = -EINVAL; *bad_wr = wr; srq->last = prev_wqe; - return nreq; + break; } for (i = 0; i < wr->num_sge; ++i) { @@ -456,8 +456,6 @@ int mthca_tavor_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, srq->first_free = next_ind; } - return nreq; - if (likely(nreq)) { __be32 doorbell[2]; @@ -501,7 +499,7 @@ int mthca_arbel_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, mthca_err(dev, "SRQ %06x full\n", srq->srqn); err = -ENOMEM; *bad_wr = wr; - return nreq; + break; } wqe = get_wqe(srq, ind); @@ -517,7 +515,7 @@ int mthca_arbel_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, if (unlikely(wr->num_sge > srq->max_gs)) { err = -EINVAL; *bad_wr = wr; - return nreq; + break; } for (i = 0; i < wr->num_sge; ++i) { From 5c9f6de3b80ca46000bd1b63d892820f9ee32138 Mon Sep 17 00:00:00 2001 From: Anton Altaparmakov Date: Mon, 19 Sep 2005 09:33:40 +0100 Subject: [PATCH 025/204] NTFS: Fix various bugs in the runlist merging code. (Based on libntfs changes by Richard Russon.) Signed-off-by: Anton Altaparmakov --- fs/ntfs/ChangeLog | 2 + fs/ntfs/runlist.c | 132 ++++++++++++++++++++++++---------------------- 2 files changed, 70 insertions(+), 64 deletions(-) diff --git a/fs/ntfs/ChangeLog b/fs/ntfs/ChangeLog index 49eafbdb15c..c7e9237379c 100644 --- a/fs/ntfs/ChangeLog +++ b/fs/ntfs/ChangeLog @@ -92,6 +92,8 @@ ToDo/Notes: an octal number to conform to how chmod(1) works, too. Thanks to Giuseppe Bilotta and Horst von Brand for pointing out the errors of my ways. + - Fix various bugs in the runlist merging code. (Based on libntfs + changes by Richard Russon.) 2.1.23 - Implement extension of resident files and make writing safe as well as many bug fixes, cleanups, and enhancements... diff --git a/fs/ntfs/runlist.c b/fs/ntfs/runlist.c index f5b2ac92908..e2665d011d7 100644 --- a/fs/ntfs/runlist.c +++ b/fs/ntfs/runlist.c @@ -2,7 +2,7 @@ * runlist.c - NTFS runlist handling code. Part of the Linux-NTFS project. * * Copyright (c) 2001-2005 Anton Altaparmakov - * Copyright (c) 2002 Richard Russon + * Copyright (c) 2002-2005 Richard Russon * * This program/include file is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License as published @@ -214,8 +214,8 @@ static inline void __ntfs_rl_merge(runlist_element *dst, runlist_element *src) static inline runlist_element *ntfs_rl_append(runlist_element *dst, int dsize, runlist_element *src, int ssize, int loc) { - BOOL right; - int magic; + BOOL right; /* Right end of @src needs merging. */ + int marker; /* End of the inserted runs. */ BUG_ON(!dst); BUG_ON(!src); @@ -236,18 +236,19 @@ static inline runlist_element *ntfs_rl_append(runlist_element *dst, if (right) __ntfs_rl_merge(src + ssize - 1, dst + loc + 1); - magic = loc + ssize; + /* First run after the @src runs that have been inserted. */ + marker = loc + ssize + 1; /* Move the tail of @dst out of the way, then copy in @src. */ - ntfs_rl_mm(dst, magic + 1, loc + 1 + right, dsize - loc - 1 - right); + ntfs_rl_mm(dst, marker, loc + 1 + right, dsize - (loc + 1 + right)); ntfs_rl_mc(dst, loc + 1, src, 0, ssize); /* Adjust the size of the preceding hole. */ dst[loc].length = dst[loc + 1].vcn - dst[loc].vcn; /* We may have changed the length of the file, so fix the end marker */ - if (dst[magic + 1].lcn == LCN_ENOENT) - dst[magic + 1].vcn = dst[magic].vcn + dst[magic].length; + if (dst[marker].lcn == LCN_ENOENT) + dst[marker].vcn = dst[marker - 1].vcn + dst[marker - 1].length; return dst; } @@ -279,18 +280,17 @@ static inline runlist_element *ntfs_rl_append(runlist_element *dst, static inline runlist_element *ntfs_rl_insert(runlist_element *dst, int dsize, runlist_element *src, int ssize, int loc) { - BOOL left = FALSE; - BOOL disc = FALSE; /* Discontinuity */ - BOOL hole = FALSE; /* Following a hole */ - int magic; + BOOL left = FALSE; /* Left end of @src needs merging. */ + BOOL disc = FALSE; /* Discontinuity between @dst and @src. */ + int marker; /* End of the inserted runs. */ BUG_ON(!dst); BUG_ON(!src); - /* disc => Discontinuity between the end of @dst and the start of @src. - * This means we might need to insert a hole. - * hole => @dst ends with a hole or an unmapped region which we can - * extend to match the discontinuity. */ + /* + * disc => Discontinuity between the end of @dst and the start of @src. + * This means we might need to insert a "not mapped" run. + */ if (loc == 0) disc = (src[0].vcn > 0); else { @@ -303,58 +303,49 @@ static inline runlist_element *ntfs_rl_insert(runlist_element *dst, merged_length += src->length; disc = (src[0].vcn > dst[loc - 1].vcn + merged_length); - if (disc) - hole = (dst[loc - 1].lcn == LCN_HOLE); } - - /* Space required: @dst size + @src size, less one if we merged, plus - * one if there was a discontinuity, less one for a trailing hole. */ - dst = ntfs_rl_realloc(dst, dsize, dsize + ssize - left + disc - hole); + /* + * Space required: @dst size + @src size, less one if we merged, plus + * one if there was a discontinuity. + */ + dst = ntfs_rl_realloc(dst, dsize, dsize + ssize - left + disc); if (IS_ERR(dst)) return dst; /* * We are guaranteed to succeed from here so can start modifying the * original runlist. */ - if (left) __ntfs_rl_merge(dst + loc - 1, src); - - magic = loc + ssize - left + disc - hole; + /* + * First run after the @src runs that have been inserted. + * Nominally, @marker equals @loc + @ssize, i.e. location + number of + * runs in @src. However, if @left, then the first run in @src has + * been merged with one in @dst. And if @disc, then @dst and @src do + * not meet and we need an extra run to fill the gap. + */ + marker = loc + ssize - left + disc; /* Move the tail of @dst out of the way, then copy in @src. */ - ntfs_rl_mm(dst, magic, loc, dsize - loc); - ntfs_rl_mc(dst, loc + disc - hole, src, left, ssize - left); + ntfs_rl_mm(dst, marker, loc, dsize - loc); + ntfs_rl_mc(dst, loc + disc, src, left, ssize - left); - /* Adjust the VCN of the last run ... */ - if (dst[magic].lcn <= LCN_HOLE) - dst[magic].vcn = dst[magic - 1].vcn + dst[magic - 1].length; + /* Adjust the VCN of the first run after the insertion... */ + dst[marker].vcn = dst[marker - 1].vcn + dst[marker - 1].length; /* ... and the length. */ - if (dst[magic].lcn == LCN_HOLE || dst[magic].lcn == LCN_RL_NOT_MAPPED) - dst[magic].length = dst[magic + 1].vcn - dst[magic].vcn; + if (dst[marker].lcn == LCN_HOLE || dst[marker].lcn == LCN_RL_NOT_MAPPED) + dst[marker].length = dst[marker + 1].vcn - dst[marker].vcn; - /* Writing beyond the end of the file and there's a discontinuity. */ + /* Writing beyond the end of the file and there is a discontinuity. */ if (disc) { - if (hole) - dst[loc - 1].length = dst[loc].vcn - dst[loc - 1].vcn; - else { - if (loc > 0) { - dst[loc].vcn = dst[loc - 1].vcn + - dst[loc - 1].length; - dst[loc].length = dst[loc + 1].vcn - - dst[loc].vcn; - } else { - dst[loc].vcn = 0; - dst[loc].length = dst[loc + 1].vcn; - } - dst[loc].lcn = LCN_RL_NOT_MAPPED; + if (loc > 0) { + dst[loc].vcn = dst[loc - 1].vcn + dst[loc - 1].length; + dst[loc].length = dst[loc + 1].vcn - dst[loc].vcn; + } else { + dst[loc].vcn = 0; + dst[loc].length = dst[loc + 1].vcn; } - - magic += hole; - - if (dst[magic].lcn == LCN_ENOENT) - dst[magic].vcn = dst[magic - 1].vcn + - dst[magic - 1].length; + dst[loc].lcn = LCN_RL_NOT_MAPPED; } return dst; } @@ -385,9 +376,10 @@ static inline runlist_element *ntfs_rl_insert(runlist_element *dst, static inline runlist_element *ntfs_rl_replace(runlist_element *dst, int dsize, runlist_element *src, int ssize, int loc) { - BOOL left = FALSE; - BOOL right; - int magic; + BOOL left = FALSE; /* Left end of @src needs merging. */ + BOOL right; /* Right end of @src needs merging. */ + int tail; /* Start of tail of @dst. */ + int marker; /* End of the inserted runs. */ BUG_ON(!dst); BUG_ON(!src); @@ -396,9 +388,10 @@ static inline runlist_element *ntfs_rl_replace(runlist_element *dst, right = ntfs_are_rl_mergeable(src + ssize - 1, dst + loc + 1); if (loc > 0) left = ntfs_are_rl_mergeable(dst + loc - 1, src); - - /* Allocate some space. We'll need less if the left, right, or both - * ends were merged. */ + /* + * Allocate some space. We will need less if the left, right, or both + * ends were merged. + */ dst = ntfs_rl_realloc(dst, dsize, dsize + ssize - left - right); if (IS_ERR(dst)) return dst; @@ -410,17 +403,28 @@ static inline runlist_element *ntfs_rl_replace(runlist_element *dst, __ntfs_rl_merge(src + ssize - 1, dst + loc + 1); if (left) __ntfs_rl_merge(dst + loc - 1, src); - - /* FIXME: What does this mean? (AIA) */ - magic = loc + ssize - left; + /* + * First run of @dst that needs to be moved out of the way to make + * space for the runs to be copied from @src, i.e. the first run of the + * tail of @dst. + */ + tail = loc + right + 1; + /* + * First run after the @src runs that have been inserted, i.e. where + * the tail of @dst needs to be moved to. + * Nominally, marker equals @loc + @ssize, i.e. location + number of + * runs in @src). However, if @left, then the first run in @src has + * been merged with one in @dst. + */ + marker = loc + ssize - left; /* Move the tail of @dst out of the way, then copy in @src. */ - ntfs_rl_mm(dst, magic, loc + right + 1, dsize - loc - right - 1); + ntfs_rl_mm(dst, marker, tail, dsize - tail); ntfs_rl_mc(dst, loc, src, left, ssize - left); - /* We may have changed the length of the file, so fix the end marker */ - if (dst[magic].lcn == LCN_ENOENT) - dst[magic].vcn = dst[magic - 1].vcn + dst[magic - 1].length; + /* We may have changed the length of the file, so fix the end marker. */ + if (dsize - tail > 0 && dst[marker].lcn == LCN_ENOENT) + dst[marker].vcn = dst[marker - 1].vcn + dst[marker - 1].length; return dst; } From 4e64c88693fde1b1cbaa4cfecad43a0c3fad354e Mon Sep 17 00:00:00 2001 From: Anton Altaparmakov Date: Mon, 19 Sep 2005 09:38:41 +0100 Subject: [PATCH 026/204] NTFS: Fix handling of compressed directories that I broke in earlier changeset. Signed-off-by: Anton Altaparmakov --- fs/ntfs/aops.c | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/fs/ntfs/aops.c b/fs/ntfs/aops.c index b6cc8cf2462..5fd516f42ee 100644 --- a/fs/ntfs/aops.c +++ b/fs/ntfs/aops.c @@ -389,9 +389,11 @@ retry_readpage: * Only $DATA attributes can be encrypted and only unnamed $DATA * attributes can be compressed. Index root can have the flags set but * this means to create compressed/encrypted files, not that the - * attribute is compressed/encrypted. + * attribute is compressed/encrypted. Note we need to check for + * AT_INDEX_ALLOCATION since this is the type of both directory and + * index inodes. */ - if (ni->type != AT_INDEX_ROOT) { + if (ni->type != AT_INDEX_ALLOCATION) { /* If attribute is encrypted, deny access, just like NT4. */ if (NInoEncrypted(ni)) { BUG_ON(ni->type != AT_DATA); @@ -1341,9 +1343,11 @@ retry_writepage: * Only $DATA attributes can be encrypted and only unnamed $DATA * attributes can be compressed. Index root can have the flags set but * this means to create compressed/encrypted files, not that the - * attribute is compressed/encrypted. + * attribute is compressed/encrypted. Note we need to check for + * AT_INDEX_ALLOCATION since this is the type of both directory and + * index inodes. */ - if (ni->type != AT_INDEX_ROOT) { + if (ni->type != AT_INDEX_ALLOCATION) { /* If file is encrypted, deny access, just like NT4. */ if (NInoEncrypted(ni)) { unlock_page(page); From f6098cf449b81c14a51e48dd22ae47d03126a1de Mon Sep 17 00:00:00 2001 From: Anton Altaparmakov Date: Mon, 19 Sep 2005 09:41:39 +0100 Subject: [PATCH 027/204] NTFS: Fix ntfs_{read,write}page() to cope with concurrent truncates better. Signed-off-by: Anton Altaparmakov --- fs/ntfs/aops.c | 108 ++++++++++++++++++++++++++++++++--------------- fs/ntfs/inode.c | 9 ++-- fs/ntfs/malloc.h | 2 +- 3 files changed, 79 insertions(+), 40 deletions(-) diff --git a/fs/ntfs/aops.c b/fs/ntfs/aops.c index 5fd516f42ee..5e80c07c6a4 100644 --- a/fs/ntfs/aops.c +++ b/fs/ntfs/aops.c @@ -59,39 +59,49 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) unsigned long flags; struct buffer_head *first, *tmp; struct page *page; + struct inode *vi; ntfs_inode *ni; int page_uptodate = 1; page = bh->b_page; - ni = NTFS_I(page->mapping->host); + vi = page->mapping->host; + ni = NTFS_I(vi); if (likely(uptodate)) { - s64 file_ofs, initialized_size; + loff_t i_size; + s64 file_ofs, init_size; set_buffer_uptodate(bh); file_ofs = ((s64)page->index << PAGE_CACHE_SHIFT) + bh_offset(bh); read_lock_irqsave(&ni->size_lock, flags); - initialized_size = ni->initialized_size; + init_size = ni->initialized_size; + i_size = i_size_read(vi); read_unlock_irqrestore(&ni->size_lock, flags); + if (unlikely(init_size > i_size)) { + /* Race with shrinking truncate. */ + init_size = i_size; + } /* Check for the current buffer head overflowing. */ - if (file_ofs + bh->b_size > initialized_size) { - char *addr; - int ofs = 0; + if (unlikely(file_ofs + bh->b_size > init_size)) { + u8 *kaddr; + int ofs; - if (file_ofs < initialized_size) - ofs = initialized_size - file_ofs; - addr = kmap_atomic(page, KM_BIO_SRC_IRQ); - memset(addr + bh_offset(bh) + ofs, 0, bh->b_size - ofs); + ofs = 0; + if (file_ofs < init_size) + ofs = init_size - file_ofs; + kaddr = kmap_atomic(page, KM_BIO_SRC_IRQ); + memset(kaddr + bh_offset(bh) + ofs, 0, + bh->b_size - ofs); + kunmap_atomic(kaddr, KM_BIO_SRC_IRQ); flush_dcache_page(page); - kunmap_atomic(addr, KM_BIO_SRC_IRQ); } } else { clear_buffer_uptodate(bh); SetPageError(page); - ntfs_error(ni->vol->sb, "Buffer I/O error, logical block %llu.", - (unsigned long long)bh->b_blocknr); + ntfs_error(ni->vol->sb, "Buffer I/O error, logical block " + "0x%llx.", (unsigned long long)bh->b_blocknr); } first = page_buffers(page); local_irq_save(flags); @@ -124,7 +134,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) if (likely(page_uptodate && !PageError(page))) SetPageUptodate(page); } else { - char *addr; + u8 *kaddr; unsigned int i, recs; u32 rec_size; @@ -132,12 +142,12 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate) recs = PAGE_CACHE_SIZE / rec_size; /* Should have been verified before we got here... */ BUG_ON(!recs); - addr = kmap_atomic(page, KM_BIO_SRC_IRQ); + kaddr = kmap_atomic(page, KM_BIO_SRC_IRQ); for (i = 0; i < recs; i++) - post_read_mst_fixup((NTFS_RECORD*)(addr + + post_read_mst_fixup((NTFS_RECORD*)(kaddr + i * rec_size), rec_size); + kunmap_atomic(kaddr, KM_BIO_SRC_IRQ); flush_dcache_page(page); - kunmap_atomic(addr, KM_BIO_SRC_IRQ); if (likely(page_uptodate && !PageError(page))) SetPageUptodate(page); } @@ -168,8 +178,11 @@ still_busy: */ static int ntfs_read_block(struct page *page) { + loff_t i_size; VCN vcn; LCN lcn; + s64 init_size; + struct inode *vi; ntfs_inode *ni; ntfs_volume *vol; runlist_element *rl; @@ -180,7 +193,8 @@ static int ntfs_read_block(struct page *page) int i, nr; unsigned char blocksize_bits; - ni = NTFS_I(page->mapping->host); + vi = page->mapping->host; + ni = NTFS_I(vi); vol = ni->vol; /* $MFT/$DATA must have its complete runlist in memory at all times. */ @@ -199,11 +213,28 @@ static int ntfs_read_block(struct page *page) bh = head = page_buffers(page); BUG_ON(!bh); + /* + * We may be racing with truncate. To avoid some of the problems we + * now take a snapshot of the various sizes and use those for the whole + * of the function. In case of an extending truncate it just means we + * may leave some buffers unmapped which are now allocated. This is + * not a problem since these buffers will just get mapped when a write + * occurs. In case of a shrinking truncate, we will detect this later + * on due to the runlist being incomplete and if the page is being + * fully truncated, truncate will throw it away as soon as we unlock + * it so no need to worry what we do with it. + */ iblock = (s64)page->index << (PAGE_CACHE_SHIFT - blocksize_bits); read_lock_irqsave(&ni->size_lock, flags); lblock = (ni->allocated_size + blocksize - 1) >> blocksize_bits; - zblock = (ni->initialized_size + blocksize - 1) >> blocksize_bits; + init_size = ni->initialized_size; + i_size = i_size_read(vi); read_unlock_irqrestore(&ni->size_lock, flags); + if (unlikely(init_size > i_size)) { + /* Race with shrinking truncate. */ + init_size = i_size; + } + zblock = (init_size + blocksize - 1) >> blocksize_bits; /* Loop through all the buffers in the page. */ rl = NULL; @@ -366,6 +397,8 @@ handle_zblock: */ static int ntfs_readpage(struct file *file, struct page *page) { + loff_t i_size; + struct inode *vi; ntfs_inode *ni, *base_ni; u8 *kaddr; ntfs_attr_search_ctx *ctx; @@ -384,7 +417,8 @@ retry_readpage: unlock_page(page); return 0; } - ni = NTFS_I(page->mapping->host); + vi = page->mapping->host; + ni = NTFS_I(vi); /* * Only $DATA attributes can be encrypted and only unnamed $DATA * attributes can be compressed. Index root can have the flags set but @@ -458,7 +492,12 @@ retry_readpage: read_lock_irqsave(&ni->size_lock, flags); if (unlikely(attr_len > ni->initialized_size)) attr_len = ni->initialized_size; + i_size = i_size_read(vi); read_unlock_irqrestore(&ni->size_lock, flags); + if (unlikely(attr_len > i_size)) { + /* Race with shrinking truncate. */ + attr_len = i_size; + } kaddr = kmap_atomic(page, KM_USER0); /* Copy the data to the page. */ memcpy(kaddr, (u8*)ctx->attr + @@ -1383,8 +1422,8 @@ retry_writepage: unsigned int ofs = i_size & ~PAGE_CACHE_MASK; kaddr = kmap_atomic(page, KM_USER0); memset(kaddr + ofs, 0, PAGE_CACHE_SIZE - ofs); - flush_dcache_page(page); kunmap_atomic(kaddr, KM_USER0); + flush_dcache_page(page); } /* Handle mst protected attributes. */ if (NInoMstProtected(ni)) @@ -1447,34 +1486,33 @@ retry_writepage: BUG_ON(PageWriteback(page)); set_page_writeback(page); unlock_page(page); - /* - * Here, we do not need to zero the out of bounds area everytime - * because the below memcpy() already takes care of the - * mmap-at-end-of-file requirements. If the file is converted to a - * non-resident one, then the code path use is switched to the - * non-resident one where the zeroing happens on each ntfs_writepage() - * invocation. - */ attr_len = le32_to_cpu(ctx->attr->data.resident.value_length); i_size = i_size_read(vi); if (unlikely(attr_len > i_size)) { + /* Race with shrinking truncate or a failed truncate. */ attr_len = i_size; - ctx->attr->data.resident.value_length = cpu_to_le32(attr_len); + /* + * If the truncate failed, fix it up now. If a concurrent + * truncate, we do its job, so it does not have to do anything. + */ + err = ntfs_resident_attr_value_resize(ctx->mrec, ctx->attr, + attr_len); + /* Shrinking cannot fail. */ + BUG_ON(err); } kaddr = kmap_atomic(page, KM_USER0); /* Copy the data from the page to the mft record. */ memcpy((u8*)ctx->attr + le16_to_cpu(ctx->attr->data.resident.value_offset), kaddr, attr_len); - flush_dcache_mft_record_page(ctx->ntfs_ino); /* Zero out of bounds area in the page cache page. */ memset(kaddr + attr_len, 0, PAGE_CACHE_SIZE - attr_len); - flush_dcache_page(page); kunmap_atomic(kaddr, KM_USER0); - + flush_dcache_mft_record_page(ctx->ntfs_ino); + flush_dcache_page(page); + /* We are done with the page. */ end_page_writeback(page); - - /* Mark the mft record dirty, so it gets written back. */ + /* Finally, mark the mft record dirty, so it gets written back. */ mark_mft_record_dirty(ctx->ntfs_ino); ntfs_attr_put_search_ctx(ctx); unmap_mft_record(base_ni); diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c index dc4bbe3acf5..7ec04513180 100644 --- a/fs/ntfs/inode.c +++ b/fs/ntfs/inode.c @@ -1166,6 +1166,8 @@ err_out: * * Return 0 on success and -errno on error. In the error case, the inode will * have had make_bad_inode() executed on it. + * + * Note this cannot be called for AT_INDEX_ALLOCATION. */ static int ntfs_read_locked_attr_inode(struct inode *base_vi, struct inode *vi) { @@ -1242,8 +1244,8 @@ static int ntfs_read_locked_attr_inode(struct inode *base_vi, struct inode *vi) } } /* - * The encryption flag set in an index root just means to - * compress all files. + * The compressed/sparse flag set in an index root just means + * to compress all files. */ if (NInoMstProtected(ni) && ni->type != AT_INDEX_ROOT) { ntfs_error(vi->i_sb, "Found mst protected attribute " @@ -1319,8 +1321,7 @@ static int ntfs_read_locked_attr_inode(struct inode *base_vi, struct inode *vi) "the mapping pairs array."); goto unm_err_out; } - if ((NInoCompressed(ni) || NInoSparse(ni)) && - ni->type != AT_INDEX_ROOT) { + if (NInoCompressed(ni) || NInoSparse(ni)) { if (a->data.non_resident.compression_unit != 4) { ntfs_error(vi->i_sb, "Found nonstandard " "compression unit (%u instead " diff --git a/fs/ntfs/malloc.h b/fs/ntfs/malloc.h index 3288bcc2c4a..006946efca8 100644 --- a/fs/ntfs/malloc.h +++ b/fs/ntfs/malloc.h @@ -1,7 +1,7 @@ /* * malloc.h - NTFS kernel memory handling. Part of the Linux-NTFS project. * - * Copyright (c) 2001-2004 Anton Altaparmakov + * Copyright (c) 2001-2005 Anton Altaparmakov * * This program/include file is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License as published From 939647ee308e0ad924e776657704c7bedd498664 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Sun, 18 Sep 2005 15:05:20 -0500 Subject: [PATCH 028/204] [SCSI] fix oops on usb storage device disconnect We fix the oops by enforcing the host state model. There have also been two extra states added: SHOST_CANCEL_RECOVERY and SHOST_DEL_RECOVERY so we can take the model through host removal while the recovery thread is active. Signed-off-by: James Bottomley --- drivers/scsi/hosts.c | 35 +++++++++++++++++++++++++++++++++-- drivers/scsi/scsi.c | 5 ++--- drivers/scsi/scsi_error.c | 23 ++++++++++++++++------- drivers/scsi/scsi_ioctl.c | 2 +- drivers/scsi/scsi_lib.c | 4 ++-- drivers/scsi/scsi_sysfs.c | 2 ++ drivers/scsi/sg.c | 2 +- include/scsi/scsi_host.h | 9 +++++++++ 8 files changed, 66 insertions(+), 16 deletions(-) diff --git a/drivers/scsi/hosts.c b/drivers/scsi/hosts.c index 85503fad789..f2a72d33132 100644 --- a/drivers/scsi/hosts.c +++ b/drivers/scsi/hosts.c @@ -98,6 +98,7 @@ int scsi_host_set_state(struct Scsi_Host *shost, enum scsi_host_state state) switch (oldstate) { case SHOST_CREATED: case SHOST_RUNNING: + case SHOST_CANCEL_RECOVERY: break; default: goto illegal; @@ -107,12 +108,31 @@ int scsi_host_set_state(struct Scsi_Host *shost, enum scsi_host_state state) case SHOST_DEL: switch (oldstate) { case SHOST_CANCEL: + case SHOST_DEL_RECOVERY: break; default: goto illegal; } break; + case SHOST_CANCEL_RECOVERY: + switch (oldstate) { + case SHOST_CANCEL: + case SHOST_RECOVERY: + break; + default: + goto illegal; + } + break; + + case SHOST_DEL_RECOVERY: + switch (oldstate) { + case SHOST_CANCEL_RECOVERY: + break; + default: + goto illegal; + } + break; } shost->shost_state = state; return 0; @@ -134,13 +154,24 @@ EXPORT_SYMBOL(scsi_host_set_state); **/ void scsi_remove_host(struct Scsi_Host *shost) { + unsigned long flags; down(&shost->scan_mutex); - scsi_host_set_state(shost, SHOST_CANCEL); + spin_lock_irqsave(shost->host_lock, flags); + if (scsi_host_set_state(shost, SHOST_CANCEL)) + if (scsi_host_set_state(shost, SHOST_CANCEL_RECOVERY)) { + spin_unlock_irqrestore(shost->host_lock, flags); + up(&shost->scan_mutex); + return; + } + spin_unlock_irqrestore(shost->host_lock, flags); up(&shost->scan_mutex); scsi_forget_host(shost); scsi_proc_host_rm(shost); - scsi_host_set_state(shost, SHOST_DEL); + spin_lock_irqsave(shost->host_lock, flags); + if (scsi_host_set_state(shost, SHOST_DEL)) + BUG_ON(scsi_host_set_state(shost, SHOST_DEL_RECOVERY)); + spin_unlock_irqrestore(shost->host_lock, flags); transport_unregister_device(&shost->shost_gendev); class_device_unregister(&shost->shost_classdev); diff --git a/drivers/scsi/scsi.c b/drivers/scsi/scsi.c index a780546eda9..1f0ebabf6d4 100644 --- a/drivers/scsi/scsi.c +++ b/drivers/scsi/scsi.c @@ -1265,9 +1265,8 @@ int scsi_device_cancel(struct scsi_device *sdev, int recovery) list_for_each_safe(lh, lh_sf, &active_list) { scmd = list_entry(lh, struct scsi_cmnd, eh_entry); list_del_init(lh); - if (recovery) { - scsi_eh_scmd_add(scmd, SCSI_EH_CANCEL_CMD); - } else { + if (recovery && + !scsi_eh_scmd_add(scmd, SCSI_EH_CANCEL_CMD)) { scmd->result = (DID_ABORT << 16); scsi_finish_command(scmd); } diff --git a/drivers/scsi/scsi_error.c b/drivers/scsi/scsi_error.c index 895c9452be4..af589fac814 100644 --- a/drivers/scsi/scsi_error.c +++ b/drivers/scsi/scsi_error.c @@ -68,19 +68,24 @@ int scsi_eh_scmd_add(struct scsi_cmnd *scmd, int eh_flag) { struct Scsi_Host *shost = scmd->device->host; unsigned long flags; + int ret = 0; if (shost->eh_wait == NULL) return 0; spin_lock_irqsave(shost->host_lock, flags); + if (scsi_host_set_state(shost, SHOST_RECOVERY)) + if (scsi_host_set_state(shost, SHOST_CANCEL_RECOVERY)) + goto out_unlock; + ret = 1; scmd->eh_eflags |= eh_flag; list_add_tail(&scmd->eh_entry, &shost->eh_cmd_q); - scsi_host_set_state(shost, SHOST_RECOVERY); shost->host_failed++; scsi_eh_wakeup(shost); + out_unlock: spin_unlock_irqrestore(shost->host_lock, flags); - return 1; + return ret; } /** @@ -176,8 +181,8 @@ void scsi_times_out(struct scsi_cmnd *scmd) } if (unlikely(!scsi_eh_scmd_add(scmd, SCSI_EH_CANCEL_CMD))) { - panic("Error handler thread not present at %p %p %s %d", - scmd, scmd->device->host, __FILE__, __LINE__); + scmd->result |= DID_TIME_OUT << 16; + __scsi_done(scmd); } } @@ -196,8 +201,7 @@ int scsi_block_when_processing_errors(struct scsi_device *sdev) { int online; - wait_event(sdev->host->host_wait, (sdev->host->shost_state != - SHOST_RECOVERY)); + wait_event(sdev->host->host_wait, !scsi_host_in_recovery(sdev->host)); online = scsi_device_online(sdev); @@ -1441,6 +1445,7 @@ static void scsi_eh_lock_door(struct scsi_device *sdev) static void scsi_restart_operations(struct Scsi_Host *shost) { struct scsi_device *sdev; + unsigned long flags; /* * If the door was locked, we need to insert a door lock request @@ -1460,7 +1465,11 @@ static void scsi_restart_operations(struct Scsi_Host *shost) SCSI_LOG_ERROR_RECOVERY(3, printk("%s: waking up host to restart\n", __FUNCTION__)); - scsi_host_set_state(shost, SHOST_RUNNING); + spin_lock_irqsave(shost->host_lock, flags); + if (scsi_host_set_state(shost, SHOST_RUNNING)) + if (scsi_host_set_state(shost, SHOST_CANCEL)) + BUG_ON(scsi_host_set_state(shost, SHOST_DEL)); + spin_unlock_irqrestore(shost->host_lock, flags); wake_up(&shost->host_wait); diff --git a/drivers/scsi/scsi_ioctl.c b/drivers/scsi/scsi_ioctl.c index b7fddac8134..de7f98cc38f 100644 --- a/drivers/scsi/scsi_ioctl.c +++ b/drivers/scsi/scsi_ioctl.c @@ -458,7 +458,7 @@ int scsi_nonblockable_ioctl(struct scsi_device *sdev, int cmd, * error processing, as long as the device was opened * non-blocking */ if (filp && filp->f_flags & O_NONBLOCK) { - if (sdev->host->shost_state == SHOST_RECOVERY) + if (scsi_host_in_recovery(sdev->host)) return -ENODEV; } else if (!scsi_block_when_processing_errors(sdev)) return -ENODEV; diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index f065cbc1a6e..dc9c772bc87 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -447,7 +447,7 @@ void scsi_device_unbusy(struct scsi_device *sdev) spin_lock_irqsave(shost->host_lock, flags); shost->host_busy--; - if (unlikely((shost->shost_state == SHOST_RECOVERY) && + if (unlikely(scsi_host_in_recovery(shost) && shost->host_failed)) scsi_eh_wakeup(shost); spin_unlock(shost->host_lock); @@ -1339,7 +1339,7 @@ static inline int scsi_host_queue_ready(struct request_queue *q, struct Scsi_Host *shost, struct scsi_device *sdev) { - if (shost->shost_state == SHOST_RECOVERY) + if (scsi_host_in_recovery(shost)) return 0; if (shost->host_busy == 0 && shost->host_blocked) { /* diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c index 1e47b7eddef..72a6550a056 100644 --- a/drivers/scsi/scsi_sysfs.c +++ b/drivers/scsi/scsi_sysfs.c @@ -57,6 +57,8 @@ static struct { { SHOST_CANCEL, "cancel" }, { SHOST_DEL, "deleted" }, { SHOST_RECOVERY, "recovery" }, + { SHOST_CANCEL_RECOVERY, "cancel/recovery" }, + { SHOST_DEL_RECOVERY, "deleted/recovery", }, }; const char *scsi_host_state_name(enum scsi_host_state state) { diff --git a/drivers/scsi/sg.c b/drivers/scsi/sg.c index 9ea4765d1d1..4d09a6e4dd2 100644 --- a/drivers/scsi/sg.c +++ b/drivers/scsi/sg.c @@ -1027,7 +1027,7 @@ sg_ioctl(struct inode *inode, struct file *filp, if (sdp->detached) return -ENODEV; if (filp->f_flags & O_NONBLOCK) { - if (sdp->device->host->shost_state == SHOST_RECOVERY) + if (scsi_host_in_recovery(sdp->device->host)) return -EBUSY; } else if (!scsi_block_when_processing_errors(sdp->device)) return -EBUSY; diff --git a/include/scsi/scsi_host.h b/include/scsi/scsi_host.h index 916144be208..540369ff815 100644 --- a/include/scsi/scsi_host.h +++ b/include/scsi/scsi_host.h @@ -439,6 +439,8 @@ enum scsi_host_state { SHOST_CANCEL, SHOST_DEL, SHOST_RECOVERY, + SHOST_CANCEL_RECOVERY, + SHOST_DEL_RECOVERY, }; struct Scsi_Host { @@ -621,6 +623,13 @@ static inline struct Scsi_Host *dev_to_shost(struct device *dev) return container_of(dev, struct Scsi_Host, shost_gendev); } +static inline int scsi_host_in_recovery(struct Scsi_Host *shost) +{ + return shost->shost_state == SHOST_RECOVERY || + shost->shost_state == SHOST_CANCEL_RECOVERY || + shost->shost_state == SHOST_DEL_RECOVERY; +} + extern int scsi_queue_work(struct Scsi_Host *, struct work_struct *); extern void scsi_flush_work(struct Scsi_Host *); From 3ed7a4704beb66a155acd67b78b7e9a5674d55fb Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Mon, 19 Sep 2005 09:50:04 -0500 Subject: [PATCH 029/204] [SCSI] Fix thread termination for the SCSI error handle From: Alan Stern This patch (as561) fixes the error handler's thread-exit code. The kthread_stop call won't wake the thread from a down_interruptible, so the patch gets rid of the semaphore and simply does set_current_state(TASK_INTERRUPTIBLE); Signed-off-by: Alan Stern Modified to simplify the termination loop and correct the sleep condition. Signed-off-by: James Bottomley --- drivers/scsi/scsi_error.c | 55 ++++++++++++++++----------------------- include/scsi/scsi_host.h | 2 -- 2 files changed, 23 insertions(+), 34 deletions(-) diff --git a/drivers/scsi/scsi_error.c b/drivers/scsi/scsi_error.c index af589fac814..ad534216507 100644 --- a/drivers/scsi/scsi_error.c +++ b/drivers/scsi/scsi_error.c @@ -50,7 +50,7 @@ void scsi_eh_wakeup(struct Scsi_Host *shost) { if (shost->host_busy == shost->host_failed) { - up(shost->eh_wait); + wake_up_process(shost->ehandler); SCSI_LOG_ERROR_RECOVERY(5, printk("Waking error handler thread\n")); } @@ -70,7 +70,7 @@ int scsi_eh_scmd_add(struct scsi_cmnd *scmd, int eh_flag) unsigned long flags; int ret = 0; - if (shost->eh_wait == NULL) + if (!shost->ehandler) return 0; spin_lock_irqsave(shost->host_lock, flags); @@ -1591,40 +1591,31 @@ int scsi_error_handler(void *data) { struct Scsi_Host *shost = (struct Scsi_Host *) data; int rtn; - DECLARE_MUTEX_LOCKED(sem); current->flags |= PF_NOFREEZE; - shost->eh_wait = &sem; + /* - * Wake up the thread that created us. + * Note - we always use TASK_INTERRUPTIBLE even if the module + * was loaded as part of the kernel. The reason is that + * UNINTERRUPTIBLE would cause this thread to be counted in + * the load average as a running process, and an interruptible + * wait doesn't. */ - SCSI_LOG_ERROR_RECOVERY(3, printk("Wake up parent of" - " scsi_eh_%d\n",shost->host_no)); - - while (1) { - /* - * If we get a signal, it means we are supposed to go - * away and die. This typically happens if the user is - * trying to unload a module. - */ - SCSI_LOG_ERROR_RECOVERY(1, printk("Error handler" - " scsi_eh_%d" - " sleeping\n",shost->host_no)); - - /* - * Note - we always use down_interruptible with the semaphore - * even if the module was loaded as part of the kernel. The - * reason is that down() will cause this thread to be counted - * in the load average as a running process, and down - * interruptible doesn't. Given that we need to allow this - * thread to die if the driver was loaded as a module, using - * semaphores isn't unreasonable. - */ - down_interruptible(&sem); - if (kthread_should_stop()) - break; + set_current_state(TASK_INTERRUPTIBLE); + while (!kthread_should_stop()) { + if (shost->host_failed == 0 || + shost->host_failed != shost->host_busy) { + SCSI_LOG_ERROR_RECOVERY(1, printk("Error handler" + " scsi_eh_%d" + " sleeping\n", + shost->host_no)); + schedule(); + set_current_state(TASK_INTERRUPTIBLE); + continue; + } + __set_current_state(TASK_RUNNING); SCSI_LOG_ERROR_RECOVERY(1, printk("Error handler" " scsi_eh_%d waking" " up\n",shost->host_no)); @@ -1651,7 +1642,7 @@ int scsi_error_handler(void *data) * which are still online. */ scsi_restart_operations(shost); - + set_current_state(TASK_INTERRUPTIBLE); } SCSI_LOG_ERROR_RECOVERY(1, printk("Error handler scsi_eh_%d" @@ -1660,7 +1651,7 @@ int scsi_error_handler(void *data) /* * Make sure that nobody tries to wake us up again. */ - shost->eh_wait = NULL; + shost->ehandler = NULL; return 0; } diff --git a/include/scsi/scsi_host.h b/include/scsi/scsi_host.h index 540369ff815..69313ba7505 100644 --- a/include/scsi/scsi_host.h +++ b/include/scsi/scsi_host.h @@ -467,8 +467,6 @@ struct Scsi_Host { struct list_head eh_cmd_q; struct task_struct * ehandler; /* Error recovery thread. */ - struct semaphore * eh_wait; /* The error recovery thread waits - on this. */ struct semaphore * eh_action; /* Wait for specific actions on the host. */ unsigned int eh_active:1; /* Indicates the eh thread is awake and active if From 6577ae51cf52f5fb0e4a85e673dd7bf2d0074e3e Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Mon, 19 Sep 2005 09:17:56 -0700 Subject: [PATCH 030/204] [PATCH] IB/mthca: Don't try to set srq->last for userspace SRQs Userspace SRQs don't have a buffer allocated for them in the kernel, so it doesn't make sense to set srq->last during initialization. In fact, this can crash trying to follow a nonexistent buffer pointer. Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_srq.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/mthca/mthca_srq.c b/drivers/infiniband/hw/mthca/mthca_srq.c index 45dedd2787a..18998d48c53 100644 --- a/drivers/infiniband/hw/mthca/mthca_srq.c +++ b/drivers/infiniband/hw/mthca/mthca_srq.c @@ -172,6 +172,8 @@ static int mthca_alloc_srq_buf(struct mthca_dev *dev, struct mthca_pd *pd, scatter->lkey = cpu_to_be32(MTHCA_INVAL_LKEY); } + srq->last = get_wqe(srq, srq->max - 1); + return 0; } @@ -263,7 +265,6 @@ int mthca_alloc_srq(struct mthca_dev *dev, struct mthca_pd *pd, srq->first_free = 0; srq->last_free = srq->max - 1; - srq->last = get_wqe(srq, srq->max - 1); return 0; From 997a51ae373df6484cdd4a5fc61a9c9bec82cd68 Mon Sep 17 00:00:00 2001 From: Ivan Kokshaysky Date: Mon, 19 Sep 2005 18:55:51 +0400 Subject: [PATCH 031/204] [PATCH] Alpha: ISA IRQs fixup for dp264 Basically, this extends original dp264 fixup to all dp264 variations. Here is one minor change: mask out bits 4-7 of a value assigned by SRM, because - newer consoles report ISA IRQs with offset 0xe0; - even if console IRQ value is bogus, we'll have a value < 16 so it should be harmless as it won't clash with native IRQs. Particularly this fixes USB interrupt problem on xp1000 and es40. Signed-off-by: Ivan Kokshaysky Acked-by: Richard Henderson Signed-off-by: Linus Torvalds --- arch/alpha/kernel/sys_dp264.c | 41 ++++++++++++++++++++--------------- 1 file changed, 23 insertions(+), 18 deletions(-) diff --git a/arch/alpha/kernel/sys_dp264.c b/arch/alpha/kernel/sys_dp264.c index 9e36b07fa94..d5da6b1b28e 100644 --- a/arch/alpha/kernel/sys_dp264.c +++ b/arch/alpha/kernel/sys_dp264.c @@ -394,6 +394,22 @@ clipper_init_irq(void) * 10 64 bit PCI option slot 3 (not bus 0) */ +static int __init +isa_irq_fixup(struct pci_dev *dev, int irq) +{ + u8 irq8; + + if (irq > 0) + return irq; + + /* This interrupt is routed via ISA bridge, so we'll + just have to trust whatever value the console might + have assigned. */ + pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &irq8); + + return irq8 & 0xf; +} + static int __init dp264_map_irq(struct pci_dev *dev, u8 slot, u8 pin) { @@ -407,25 +423,13 @@ dp264_map_irq(struct pci_dev *dev, u8 slot, u8 pin) { 16+ 3, 16+ 3, 16+ 2, 16+ 1, 16+ 0} /* IdSel 10 slot 3 */ }; const long min_idsel = 5, max_idsel = 10, irqs_per_slot = 5; - struct pci_controller *hose = dev->sysdata; int irq = COMMON_TABLE_LOOKUP; - if (irq > 0) { + if (irq > 0) irq += 16 * hose->index; - } else { - /* ??? The Contaq IDE controller on the ISA bridge uses - "legacy" interrupts 14 and 15. I don't know if anything - can wind up at the same slot+pin on hose1, so we'll - just have to trust whatever value the console might - have assigned. */ - u8 irq8; - pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &irq8); - irq = irq8; - } - - return irq; + return isa_irq_fixup(dev, irq); } static int __init @@ -453,7 +457,8 @@ monet_map_irq(struct pci_dev *dev, u8 slot, u8 pin) { 24, 24, 25, 26, 27} /* IdSel 15 slot 5 PCI2*/ }; const long min_idsel = 3, max_idsel = 15, irqs_per_slot = 5; - return COMMON_TABLE_LOOKUP; + + return isa_irq_fixup(dev, COMMON_TABLE_LOOKUP); } static u8 __init @@ -507,7 +512,8 @@ webbrick_map_irq(struct pci_dev *dev, u8 slot, u8 pin) { 47, 47, 46, 45, 44}, /* IdSel 17 slot 3 */ }; const long min_idsel = 7, max_idsel = 17, irqs_per_slot = 5; - return COMMON_TABLE_LOOKUP; + + return isa_irq_fixup(dev, COMMON_TABLE_LOOKUP); } static int __init @@ -524,14 +530,13 @@ clipper_map_irq(struct pci_dev *dev, u8 slot, u8 pin) { -1, -1, -1, -1, -1} /* IdSel 7 ISA Bridge */ }; const long min_idsel = 1, max_idsel = 7, irqs_per_slot = 5; - struct pci_controller *hose = dev->sysdata; int irq = COMMON_TABLE_LOOKUP; if (irq > 0) irq += 16 * hose->index; - return irq; + return isa_irq_fixup(dev, irq); } static void __init From 82ffb67164064752a56669511545316075b41e1d Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 9 Sep 2005 16:25:54 +0200 Subject: [PATCH 032/204] [SCSI] fusion core changes for SAS support - various bits for SAS support from the LSI driver. - use the device private data for the fusion target private data. this should be using the midlayer target data framework, but we can't move over to that until fusion has been switched to the generic DV code - use target ID and channel from the fusion target private data, because those in scsi_device will be different for mptsas Signed-off-by: Christoph Hellwig Signed-off-by: James Bottomley --- drivers/message/fusion/mptbase.c | 514 +++++++++++++++++++++++++----- drivers/message/fusion/mptbase.h | 28 +- drivers/message/fusion/mptscsih.c | 15 +- 3 files changed, 471 insertions(+), 86 deletions(-) diff --git a/drivers/message/fusion/mptbase.c b/drivers/message/fusion/mptbase.c index f517d0692d5..14d62d96ca4 100644 --- a/drivers/message/fusion/mptbase.c +++ b/drivers/message/fusion/mptbase.c @@ -141,7 +141,7 @@ static int GetPortFacts(MPT_ADAPTER *ioc, int portnum, int sleepFlag); static int SendIocInit(MPT_ADAPTER *ioc, int sleepFlag); static int SendPortEnable(MPT_ADAPTER *ioc, int portnum, int sleepFlag); static int mpt_do_upload(MPT_ADAPTER *ioc, int sleepFlag); -static int mpt_downloadboot(MPT_ADAPTER *ioc, int sleepFlag); +static int mpt_downloadboot(MPT_ADAPTER *ioc, MpiFwHeader_t *pFwHeader, int sleepFlag); static int mpt_diag_reset(MPT_ADAPTER *ioc, int ignore, int sleepFlag); static int KickStart(MPT_ADAPTER *ioc, int ignore, int sleepFlag); static int SendIocReset(MPT_ADAPTER *ioc, u8 reset_type, int sleepFlag); @@ -152,6 +152,7 @@ static int WaitForDoorbellReply(MPT_ADAPTER *ioc, int howlong, int sleepFlag); static int GetLanConfigPages(MPT_ADAPTER *ioc); static int GetFcPortPage0(MPT_ADAPTER *ioc, int portnum); static int GetIoUnitPage2(MPT_ADAPTER *ioc); +int mptbase_sas_persist_operation(MPT_ADAPTER *ioc, u8 persist_opcode); static int mpt_GetScsiPortSettings(MPT_ADAPTER *ioc, int portnum); static int mpt_readScsiDevicePageHeaders(MPT_ADAPTER *ioc, int portnum); static void mpt_read_ioc_pg_1(MPT_ADAPTER *ioc); @@ -159,6 +160,8 @@ static void mpt_read_ioc_pg_4(MPT_ADAPTER *ioc); static void mpt_timer_expired(unsigned long data); static int SendEventNotification(MPT_ADAPTER *ioc, u8 EvSwitch); static int SendEventAck(MPT_ADAPTER *ioc, EventNotificationReply_t *evnp); +static int mpt_host_page_access_control(MPT_ADAPTER *ioc, u8 access_control_value, int sleepFlag); +static int mpt_host_page_alloc(MPT_ADAPTER *ioc, pIOCInit_t ioc_init); #ifdef CONFIG_PROC_FS static int procmpt_summary_read(char *buf, char **start, off_t offset, @@ -509,6 +512,14 @@ mpt_base_reply(MPT_ADAPTER *ioc, MPT_FRAME_HDR *mf, MPT_FRAME_HDR *reply) pCfg->wait_done = 1; wake_up(&mpt_waitq); } + } else if (func == MPI_FUNCTION_SAS_IO_UNIT_CONTROL) { + /* we should be always getting a reply frame */ + memcpy(ioc->persist_reply_frame, reply, + min(MPT_DEFAULT_FRAME_SIZE, + 4*reply->u.reply.MsgLength)); + del_timer(&ioc->persist_timer); + ioc->persist_wait_done = 1; + wake_up(&mpt_waitq); } else { printk(MYIOC_s_ERR_FMT "Unexpected msg function (=%02Xh) reply received!\n", ioc->name, func); @@ -750,6 +761,7 @@ mpt_get_msg_frame(int handle, MPT_ADAPTER *ioc) mf = list_entry(ioc->FreeQ.next, MPT_FRAME_HDR, u.frame.linkage.list); list_del(&mf->u.frame.linkage.list); + mf->u.frame.linkage.arg1 = 0; mf->u.frame.hwhdr.msgctxu.fld.cb_idx = handle; /* byte */ req_offset = (u8 *)mf - (u8 *)ioc->req_frames; /* u16! */ @@ -845,6 +857,7 @@ mpt_free_msg_frame(MPT_ADAPTER *ioc, MPT_FRAME_HDR *mf) /* Put Request back on FreeQ! */ spin_lock_irqsave(&ioc->FreeQlock, flags); + mf->u.frame.linkage.arg1 = 0xdeadbeaf; /* signature to know if this mf is freed */ list_add_tail(&mf->u.frame.linkage.list, &ioc->FreeQ); #ifdef MFCNT ioc->mfcnt--; @@ -971,10 +984,121 @@ mpt_send_handshake_request(int handle, MPT_ADAPTER *ioc, int reqBytes, u32 *req, /* Make sure there are no doorbells */ CHIPREG_WRITE32(&ioc->chip->IntStatus, 0); - + return r; } +/*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ +/** + * mpt_host_page_access_control - provides mechanism for the host + * driver to control the IOC's Host Page Buffer access. + * @ioc: Pointer to MPT adapter structure + * @access_control_value: define bits below + * + * Access Control Value - bits[15:12] + * 0h Reserved + * 1h Enable Access { MPI_DB_HPBAC_ENABLE_ACCESS } + * 2h Disable Access { MPI_DB_HPBAC_DISABLE_ACCESS } + * 3h Free Buffer { MPI_DB_HPBAC_FREE_BUFFER } + * + * Returns 0 for success, non-zero for failure. + */ + +static int +mpt_host_page_access_control(MPT_ADAPTER *ioc, u8 access_control_value, int sleepFlag) +{ + int r = 0; + + /* return if in use */ + if (CHIPREG_READ32(&ioc->chip->Doorbell) + & MPI_DOORBELL_ACTIVE) + return -1; + + CHIPREG_WRITE32(&ioc->chip->IntStatus, 0); + + CHIPREG_WRITE32(&ioc->chip->Doorbell, + ((MPI_FUNCTION_HOST_PAGEBUF_ACCESS_CONTROL + <HostPageBuffer) { + + host_page_buffer_sz = + le32_to_cpu(ioc->facts.HostPageBufferSGE.FlagsLength) & 0xFFFFFF; + + if(!host_page_buffer_sz) + return 0; /* fw doesn't need any host buffers */ + + /* spin till we get enough memory */ + while(host_page_buffer_sz > 0) { + + if((ioc->HostPageBuffer = pci_alloc_consistent( + ioc->pcidev, + host_page_buffer_sz, + &ioc->HostPageBuffer_dma)) != NULL) { + + dinitprintk((MYIOC_s_INFO_FMT + "host_page_buffer @ %p, dma @ %x, sz=%d bytes\n", + ioc->name, + ioc->HostPageBuffer, + ioc->HostPageBuffer_dma, + hst_page_buffer_sz)); + ioc->alloc_total += host_page_buffer_sz; + ioc->HostPageBuffer_sz = host_page_buffer_sz; + break; + } + + host_page_buffer_sz -= (4*1024); + } + } + + if(!ioc->HostPageBuffer) { + printk(MYIOC_s_ERR_FMT + "Failed to alloc memory for host_page_buffer!\n", + ioc->name); + return -999; + } + + psge = (char *)&ioc_init->HostPageBufferSGE; + flags_length = MPI_SGE_FLAGS_SIMPLE_ELEMENT | + MPI_SGE_FLAGS_SYSTEM_ADDRESS | + MPI_SGE_FLAGS_32_BIT_ADDRESSING | + MPI_SGE_FLAGS_HOST_TO_IOC | + MPI_SGE_FLAGS_END_OF_BUFFER; + if (sizeof(dma_addr_t) == sizeof(u64)) { + flags_length |= MPI_SGE_FLAGS_64_BIT_ADDRESSING; + } + flags_length = flags_length << MPI_SGE_FLAGS_SHIFT; + flags_length |= ioc->HostPageBuffer_sz; + mpt_add_sge(psge, flags_length, ioc->HostPageBuffer_dma); + ioc->facts.HostPageBufferSGE = ioc_init->HostPageBufferSGE; + +return 0; +} + /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ /** * mpt_verify_adapter - Given a unique IOC identifier, set pointer to @@ -1213,6 +1337,33 @@ mpt_attach(struct pci_dev *pdev, const struct pci_device_id *id) ioc->prod_name = "LSI53C1035"; ioc->bus_type = SCSI; } + else if (pdev->device == MPI_MANUFACTPAGE_DEVID_SAS1064) { + ioc->prod_name = "LSISAS1064"; + ioc->bus_type = SAS; + ioc->errata_flag_1064 = 1; + } + else if (pdev->device == MPI_MANUFACTPAGE_DEVID_SAS1066) { + ioc->prod_name = "LSISAS1066"; + ioc->bus_type = SAS; + ioc->errata_flag_1064 = 1; + } + else if (pdev->device == MPI_MANUFACTPAGE_DEVID_SAS1068) { + ioc->prod_name = "LSISAS1068"; + ioc->bus_type = SAS; + ioc->errata_flag_1064 = 1; + } + else if (pdev->device == MPI_MANUFACTPAGE_DEVID_SAS1064E) { + ioc->prod_name = "LSISAS1064E"; + ioc->bus_type = SAS; + } + else if (pdev->device == MPI_MANUFACTPAGE_DEVID_SAS1066E) { + ioc->prod_name = "LSISAS1066E"; + ioc->bus_type = SAS; + } + else if (pdev->device == MPI_MANUFACTPAGE_DEVID_SAS1068E) { + ioc->prod_name = "LSISAS1068E"; + ioc->bus_type = SAS; + } if (ioc->errata_flag_1064) pci_disable_io_access(pdev); @@ -1640,7 +1791,22 @@ mpt_do_ioc_recovery(MPT_ADAPTER *ioc, u32 reason, int sleepFlag) * and we try GetLanConfigPages again... */ if ((ret == 0) && (reason == MPT_HOSTEVENT_IOC_BRINGUP)) { - if (ioc->bus_type == FC) { + if (ioc->bus_type == SAS) { + + /* clear persistency table */ + if(ioc->facts.IOCExceptions & + MPI_IOCFACTS_EXCEPT_PERSISTENT_TABLE_FULL) { + ret = mptbase_sas_persist_operation(ioc, + MPI_SAS_OP_CLEAR_NOT_PRESENT); + if(ret != 0) + return -1; + } + + /* Find IM volumes + */ + mpt_findImVolumes(ioc); + + } else if (ioc->bus_type == FC) { /* * Pre-fetch FC port WWN and stuff... * (FCPortPage0_t stuff) @@ -1783,7 +1949,7 @@ mpt_adapter_disable(MPT_ADAPTER *ioc) if (ioc->cached_fw != NULL) { ddlprintk((KERN_INFO MYNAM ": mpt_adapter_disable: Pushing FW onto adapter\n")); - if ((ret = mpt_downloadboot(ioc, NO_SLEEP)) < 0) { + if ((ret = mpt_downloadboot(ioc, (MpiFwHeader_t *)ioc->cached_fw, NO_SLEEP)) < 0) { printk(KERN_WARNING MYNAM ": firmware downloadboot failure (%d)!\n", ret); } @@ -1852,6 +2018,23 @@ mpt_adapter_disable(MPT_ADAPTER *ioc) kfree(ioc->ChainToChain); ioc->ChainToChain = NULL; + + if (ioc->HostPageBuffer != NULL) { + if((ret = mpt_host_page_access_control(ioc, + MPI_DB_HPBAC_FREE_BUFFER, NO_SLEEP)) != 0) { + printk(KERN_ERR MYNAM + ": %s: host page buffers free failed (%d)!\n", + __FUNCTION__, ret); + } + dexitprintk((KERN_INFO MYNAM ": %s HostPageBuffer free @ %p, sz=%d bytes\n", + ioc->name, ioc->HostPageBuffer, ioc->HostPageBuffer_sz)); + pci_free_consistent(ioc->pcidev, ioc->HostPageBuffer_sz, + ioc->HostPageBuffer, + ioc->HostPageBuffer_dma); + ioc->HostPageBuffer = NULL; + ioc->HostPageBuffer_sz = 0; + ioc->alloc_total -= ioc->HostPageBuffer_sz; + } } /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ @@ -2034,7 +2217,7 @@ MakeIocReady(MPT_ADAPTER *ioc, int force, int sleepFlag) * Loop here waiting for IOC to come READY. */ ii = 0; - cntdn = ((sleepFlag == CAN_SLEEP) ? HZ : 1000) * 15; /* 15 seconds */ + cntdn = ((sleepFlag == CAN_SLEEP) ? HZ : 1000) * 5; /* 5 seconds */ while ((ioc_state = mpt_GetIocState(ioc, 1)) != MPI_IOC_STATE_READY) { if (ioc_state == MPI_IOC_STATE_OPERATIONAL) { @@ -2212,6 +2395,7 @@ GetIocFacts(MPT_ADAPTER *ioc, int sleepFlag, int reason) le32_to_cpu(facts->CurrentSenseBufferHighAddr); facts->CurReplyFrameSize = le16_to_cpu(facts->CurReplyFrameSize); + facts->IOCCapabilities = le32_to_cpu(facts->IOCCapabilities); /* * Handle NEW (!) IOCFactsReply fields in MPI-1.01.xx @@ -2383,13 +2567,25 @@ SendIocInit(MPT_ADAPTER *ioc, int sleepFlag) ddlprintk((MYIOC_s_INFO_FMT "upload_fw %d facts.Flags=%x\n", ioc->name, ioc->upload_fw, ioc->facts.Flags)); - if (ioc->bus_type == FC) + if(ioc->bus_type == SAS) + ioc_init.MaxDevices = ioc->facts.MaxDevices; + else if(ioc->bus_type == FC) ioc_init.MaxDevices = MPT_MAX_FC_DEVICES; else ioc_init.MaxDevices = MPT_MAX_SCSI_DEVICES; - ioc_init.MaxBuses = MPT_MAX_BUS; + dinitprintk((MYIOC_s_INFO_FMT "facts.MsgVersion=%x\n", + ioc->name, ioc->facts.MsgVersion)); + if (ioc->facts.MsgVersion >= MPI_VERSION_01_05) { + // set MsgVersion and HeaderVersion host driver was built with + ioc_init.MsgVersion = cpu_to_le16(MPI_VERSION); + ioc_init.HeaderVersion = cpu_to_le16(MPI_HEADER_VERSION); + if (ioc->facts.Flags & MPI_IOCFACTS_FLAGS_HOST_PAGE_BUFFER_PERSISTENT) { + ioc_init.HostPageBufferSGE = ioc->facts.HostPageBufferSGE; + } else if(mpt_host_page_alloc(ioc, &ioc_init)) + return -99; + } ioc_init.ReplyFrameSize = cpu_to_le16(ioc->reply_sz); /* in BYTES */ if (sizeof(dma_addr_t) == sizeof(u64)) { @@ -2403,17 +2599,21 @@ SendIocInit(MPT_ADAPTER *ioc, int sleepFlag) ioc_init.HostMfaHighAddr = cpu_to_le32(0); ioc_init.SenseBufferHighAddr = cpu_to_le32(0); } - + ioc->facts.CurrentHostMfaHighAddr = ioc_init.HostMfaHighAddr; ioc->facts.CurrentSenseBufferHighAddr = ioc_init.SenseBufferHighAddr; + ioc->facts.MaxDevices = ioc_init.MaxDevices; + ioc->facts.MaxBuses = ioc_init.MaxBuses; dhsprintk((MYIOC_s_INFO_FMT "Sending IOCInit (req @ %p)\n", ioc->name, &ioc_init)); r = mpt_handshake_req_reply_wait(ioc, sizeof(IOCInit_t), (u32*)&ioc_init, sizeof(MPIDefaultReply_t), (u16*)&init_reply, 10 /*seconds*/, sleepFlag); - if (r != 0) + if (r != 0) { + printk(MYIOC_s_ERR_FMT "Sending IOCInit failed(%d)!\n",ioc->name, r); return r; + } /* No need to byte swap the multibyte fields in the reply * since we don't even look at it's contents. @@ -2472,7 +2672,7 @@ SendPortEnable(MPT_ADAPTER *ioc, int portnum, int sleepFlag) { PortEnable_t port_enable; MPIDefaultReply_t reply_buf; - int ii; + int rc; int req_sz; int reply_sz; @@ -2494,22 +2694,15 @@ SendPortEnable(MPT_ADAPTER *ioc, int portnum, int sleepFlag) /* RAID FW may take a long time to enable */ - if (ioc->bus_type == FC) { - ii = mpt_handshake_req_reply_wait(ioc, req_sz, (u32*)&port_enable, - reply_sz, (u16*)&reply_buf, 65 /*seconds*/, sleepFlag); - } else { - ii = mpt_handshake_req_reply_wait(ioc, req_sz, (u32*)&port_enable, + if ( (ioc->facts.ProductID & MPI_FW_HEADER_PID_PROD_MASK) + > MPI_FW_HEADER_PID_PROD_TARGET_SCSI ) { + rc = mpt_handshake_req_reply_wait(ioc, req_sz, (u32*)&port_enable, reply_sz, (u16*)&reply_buf, 300 /*seconds*/, sleepFlag); + } else { + rc = mpt_handshake_req_reply_wait(ioc, req_sz, (u32*)&port_enable, + reply_sz, (u16*)&reply_buf, 30 /*seconds*/, sleepFlag); } - - if (ii != 0) - return ii; - - /* We do not even look at the reply, so we need not - * swap the multi-byte fields. - */ - - return 0; + return rc; } /* @@ -2666,9 +2859,8 @@ mpt_do_upload(MPT_ADAPTER *ioc, int sleepFlag) * <0 for fw upload failure. */ static int -mpt_downloadboot(MPT_ADAPTER *ioc, int sleepFlag) +mpt_downloadboot(MPT_ADAPTER *ioc, MpiFwHeader_t *pFwHeader, int sleepFlag) { - MpiFwHeader_t *pFwHeader; MpiExtImageHeader_t *pExtImage; u32 fwSize; u32 diag0val; @@ -2679,18 +2871,8 @@ mpt_downloadboot(MPT_ADAPTER *ioc, int sleepFlag) u32 load_addr; u32 ioc_state=0; - ddlprintk((MYIOC_s_INFO_FMT "downloadboot: fw size 0x%x, ioc FW Ptr %p\n", - ioc->name, ioc->facts.FWImageSize, ioc->cached_fw)); - - if ( ioc->facts.FWImageSize == 0 ) - return -1; - - if (ioc->cached_fw == NULL) - return -2; - - /* prevent a second downloadboot and memory free with alt_ioc */ - if (ioc->alt_ioc && ioc->alt_ioc->cached_fw) - ioc->alt_ioc->cached_fw = NULL; + ddlprintk((MYIOC_s_INFO_FMT "downloadboot: fw size 0x%x (%d), FW Ptr %p\n", + ioc->name, pFwHeader->ImageSize, pFwHeader->ImageSize, pFwHeader)); CHIPREG_WRITE32(&ioc->chip->WriteSequence, 0xFF); CHIPREG_WRITE32(&ioc->chip->WriteSequence, MPI_WRSEQ_1ST_KEY_VALUE); @@ -2718,16 +2900,17 @@ mpt_downloadboot(MPT_ADAPTER *ioc, int sleepFlag) ioc->name, count)); break; } - /* wait 1 sec */ + /* wait .1 sec */ if (sleepFlag == CAN_SLEEP) { - msleep_interruptible (1000); + msleep_interruptible (100); } else { - mdelay (1000); + mdelay (100); } } if ( count == 30 ) { - ddlprintk((MYIOC_s_INFO_FMT "downloadboot failed! Unable to RESET_ADAPTER diag0val=%x\n", + ddlprintk((MYIOC_s_INFO_FMT "downloadboot failed! " + "Unable to get MPI_DIAG_DRWE mode, diag0val=%x\n", ioc->name, diag0val)); return -3; } @@ -2742,7 +2925,6 @@ mpt_downloadboot(MPT_ADAPTER *ioc, int sleepFlag) /* Set the DiagRwEn and Disable ARM bits */ CHIPREG_WRITE32(&ioc->chip->Diagnostic, (MPI_DIAG_RW_ENABLE | MPI_DIAG_DISABLE_ARM)); - pFwHeader = (MpiFwHeader_t *) ioc->cached_fw; fwSize = (pFwHeader->ImageSize + 3)/4; ptrFw = (u32 *) pFwHeader; @@ -2792,19 +2974,38 @@ mpt_downloadboot(MPT_ADAPTER *ioc, int sleepFlag) /* Clear the internal flash bad bit - autoincrementing register, * so must do two writes. */ - CHIPREG_PIO_WRITE32(&ioc->pio_chip->DiagRwAddress, 0x3F000000); - diagRwData = CHIPREG_PIO_READ32(&ioc->pio_chip->DiagRwData); - diagRwData |= 0x4000000; - CHIPREG_PIO_WRITE32(&ioc->pio_chip->DiagRwAddress, 0x3F000000); - CHIPREG_PIO_WRITE32(&ioc->pio_chip->DiagRwData, diagRwData); + if (ioc->bus_type == SCSI) { + /* + * 1030 and 1035 H/W errata, workaround to access + * the ClearFlashBadSignatureBit + */ + CHIPREG_PIO_WRITE32(&ioc->pio_chip->DiagRwAddress, 0x3F000000); + diagRwData = CHIPREG_PIO_READ32(&ioc->pio_chip->DiagRwData); + diagRwData |= 0x40000000; + CHIPREG_PIO_WRITE32(&ioc->pio_chip->DiagRwAddress, 0x3F000000); + CHIPREG_PIO_WRITE32(&ioc->pio_chip->DiagRwData, diagRwData); + + } else /* if((ioc->bus_type == SAS) || (ioc->bus_type == FC)) */ { + diag0val = CHIPREG_READ32(&ioc->chip->Diagnostic); + CHIPREG_WRITE32(&ioc->chip->Diagnostic, diag0val | + MPI_DIAG_CLEAR_FLASH_BAD_SIG); + + /* wait 1 msec */ + if (sleepFlag == CAN_SLEEP) { + msleep_interruptible (1); + } else { + mdelay (1); + } + } if (ioc->errata_flag_1064) pci_disable_io_access(ioc->pcidev); diag0val = CHIPREG_READ32(&ioc->chip->Diagnostic); - ddlprintk((MYIOC_s_INFO_FMT "downloadboot diag0val=%x, turning off PREVENT_IOC_BOOT, DISABLE_ARM\n", + ddlprintk((MYIOC_s_INFO_FMT "downloadboot diag0val=%x, " + "turning off PREVENT_IOC_BOOT, DISABLE_ARM, RW_ENABLE\n", ioc->name, diag0val)); - diag0val &= ~(MPI_DIAG_PREVENT_IOC_BOOT | MPI_DIAG_DISABLE_ARM); + diag0val &= ~(MPI_DIAG_PREVENT_IOC_BOOT | MPI_DIAG_DISABLE_ARM | MPI_DIAG_RW_ENABLE); ddlprintk((MYIOC_s_INFO_FMT "downloadboot now diag0val=%x\n", ioc->name, diag0val)); CHIPREG_WRITE32(&ioc->chip->Diagnostic, diag0val); @@ -2812,10 +3013,23 @@ mpt_downloadboot(MPT_ADAPTER *ioc, int sleepFlag) /* Write 0xFF to reset the sequencer */ CHIPREG_WRITE32(&ioc->chip->WriteSequence, 0xFF); + if (ioc->bus_type == SAS) { + ioc_state = mpt_GetIocState(ioc, 0); + if ( (GetIocFacts(ioc, sleepFlag, + MPT_HOSTEVENT_IOC_BRINGUP)) != 0 ) { + ddlprintk((MYIOC_s_INFO_FMT "GetIocFacts failed: IocState=%x\n", + ioc->name, ioc_state)); + return -EFAULT; + } + } + for (count=0; countname, count, ioc_state)); + if (ioc->bus_type == SAS) { + return 0; + } if ((SendIocInit(ioc, sleepFlag)) != 0) { ddlprintk((MYIOC_s_INFO_FMT "downloadboot: SendIocInit failed\n", ioc->name)); @@ -3049,12 +3263,13 @@ mpt_diag_reset(MPT_ADAPTER *ioc, int ignore, int sleepFlag) /* wait 1 sec */ if (sleepFlag == CAN_SLEEP) { - ssleep(1); + msleep_interruptible (1000); } else { mdelay (1000); } } - if ((count = mpt_downloadboot(ioc, sleepFlag)) < 0) { + if ((count = mpt_downloadboot(ioc, + (MpiFwHeader_t *)ioc->cached_fw, sleepFlag)) < 0) { printk(KERN_WARNING MYNAM ": firmware downloadboot failure (%d)!\n", count); } @@ -3999,6 +4214,85 @@ GetFcPortPage0(MPT_ADAPTER *ioc, int portnum) return rc; } +/*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ +/* + * mptbase_sas_persist_operation - Perform operation on SAS Persitent Table + * @ioc: Pointer to MPT_ADAPTER structure + * @sas_address: 64bit SAS Address for operation. + * @target_id: specified target for operation + * @bus: specified bus for operation + * @persist_opcode: see below + * + * MPI_SAS_OP_CLEAR_NOT_PRESENT - Free all persist TargetID mappings for + * devices not currently present. + * MPI_SAS_OP_CLEAR_ALL_PERSISTENT - Clear al persist TargetID mappings + * + * NOTE: Don't use not this function during interrupt time. + * + * Returns: 0 for success, non-zero error + */ + +/*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ +int +mptbase_sas_persist_operation(MPT_ADAPTER *ioc, u8 persist_opcode) +{ + SasIoUnitControlRequest_t *sasIoUnitCntrReq; + SasIoUnitControlReply_t *sasIoUnitCntrReply; + MPT_FRAME_HDR *mf = NULL; + MPIHeader_t *mpi_hdr; + + + /* insure garbage is not sent to fw */ + switch(persist_opcode) { + + case MPI_SAS_OP_CLEAR_NOT_PRESENT: + case MPI_SAS_OP_CLEAR_ALL_PERSISTENT: + break; + + default: + return -1; + break; + } + + printk("%s: persist_opcode=%x\n",__FUNCTION__, persist_opcode); + + /* Get a MF for this command. + */ + if ((mf = mpt_get_msg_frame(mpt_base_index, ioc)) == NULL) { + printk("%s: no msg frames!\n",__FUNCTION__); + return -1; + } + + mpi_hdr = (MPIHeader_t *) mf; + sasIoUnitCntrReq = (SasIoUnitControlRequest_t *)mf; + memset(sasIoUnitCntrReq,0,sizeof(SasIoUnitControlRequest_t)); + sasIoUnitCntrReq->Function = MPI_FUNCTION_SAS_IO_UNIT_CONTROL; + sasIoUnitCntrReq->MsgContext = mpi_hdr->MsgContext; + sasIoUnitCntrReq->Operation = persist_opcode; + + init_timer(&ioc->persist_timer); + ioc->persist_timer.data = (unsigned long) ioc; + ioc->persist_timer.function = mpt_timer_expired; + ioc->persist_timer.expires = jiffies + HZ*10 /* 10 sec */; + ioc->persist_wait_done=0; + add_timer(&ioc->persist_timer); + mpt_put_msg_frame(mpt_base_index, ioc, mf); + wait_event(mpt_waitq, ioc->persist_wait_done); + + sasIoUnitCntrReply = + (SasIoUnitControlReply_t *)ioc->persist_reply_frame; + if (le16_to_cpu(sasIoUnitCntrReply->IOCStatus) != MPI_IOCSTATUS_SUCCESS) { + printk("%s: IOCStatus=0x%X IOCLogInfo=0x%X\n", + __FUNCTION__, + sasIoUnitCntrReply->IOCStatus, + sasIoUnitCntrReply->IOCLogInfo); + return -1; + } + + printk("%s: success\n",__FUNCTION__); + return 0; +} + /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ /* * GetIoUnitPage2 - Retrieve BIOS version and boot order information. @@ -5366,8 +5660,8 @@ mpt_HardResetHandler(MPT_ADAPTER *ioc, int sleepFlag) } /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ -static char * -EventDescriptionStr(u8 event, u32 evData0) +static void +EventDescriptionStr(u8 event, u32 evData0, char *evStr) { char *ds; @@ -5420,8 +5714,95 @@ EventDescriptionStr(u8 event, u32 evData0) ds = "Events(OFF) Change"; break; case MPI_EVENT_INTEGRATED_RAID: - ds = "Integrated Raid"; + { + u8 ReasonCode = (u8)(evData0 >> 16); + switch (ReasonCode) { + case MPI_EVENT_RAID_RC_VOLUME_CREATED : + ds = "Integrated Raid: Volume Created"; + break; + case MPI_EVENT_RAID_RC_VOLUME_DELETED : + ds = "Integrated Raid: Volume Deleted"; + break; + case MPI_EVENT_RAID_RC_VOLUME_SETTINGS_CHANGED : + ds = "Integrated Raid: Volume Settings Changed"; + break; + case MPI_EVENT_RAID_RC_VOLUME_STATUS_CHANGED : + ds = "Integrated Raid: Volume Status Changed"; + break; + case MPI_EVENT_RAID_RC_VOLUME_PHYSDISK_CHANGED : + ds = "Integrated Raid: Volume Physdisk Changed"; + break; + case MPI_EVENT_RAID_RC_PHYSDISK_CREATED : + ds = "Integrated Raid: Physdisk Created"; + break; + case MPI_EVENT_RAID_RC_PHYSDISK_DELETED : + ds = "Integrated Raid: Physdisk Deleted"; + break; + case MPI_EVENT_RAID_RC_PHYSDISK_SETTINGS_CHANGED : + ds = "Integrated Raid: Physdisk Settings Changed"; + break; + case MPI_EVENT_RAID_RC_PHYSDISK_STATUS_CHANGED : + ds = "Integrated Raid: Physdisk Status Changed"; + break; + case MPI_EVENT_RAID_RC_DOMAIN_VAL_NEEDED : + ds = "Integrated Raid: Domain Validation Needed"; + break; + case MPI_EVENT_RAID_RC_SMART_DATA : + ds = "Integrated Raid; Smart Data"; + break; + case MPI_EVENT_RAID_RC_REPLACE_ACTION_STARTED : + ds = "Integrated Raid: Replace Action Started"; + break; + default: + ds = "Integrated Raid"; break; + } + break; + } + case MPI_EVENT_SCSI_DEVICE_STATUS_CHANGE: + ds = "SCSI Device Status Change"; + break; + case MPI_EVENT_SAS_DEVICE_STATUS_CHANGE: + { + u8 ReasonCode = (u8)(evData0 >> 16); + switch (ReasonCode) { + case MPI_EVENT_SAS_DEV_STAT_RC_ADDED: + ds = "SAS Device Status Change: Added"; + break; + case MPI_EVENT_SAS_DEV_STAT_RC_NOT_RESPONDING: + ds = "SAS Device Status Change: Deleted"; + break; + case MPI_EVENT_SAS_DEV_STAT_RC_SMART_DATA: + ds = "SAS Device Status Change: SMART Data"; + break; + case MPI_EVENT_SAS_DEV_STAT_RC_NO_PERSIST_ADDED: + ds = "SAS Device Status Change: No Persistancy Added"; + break; + default: + ds = "SAS Device Status Change: Unknown"; + break; + } + break; + } + case MPI_EVENT_ON_BUS_TIMER_EXPIRED: + ds = "Bus Timer Expired"; + break; + case MPI_EVENT_QUEUE_FULL: + ds = "Queue Full"; + break; + case MPI_EVENT_SAS_SES: + ds = "SAS SES Event"; + break; + case MPI_EVENT_PERSISTENT_TABLE_FULL: + ds = "Persistent Table Full"; + break; + case MPI_EVENT_SAS_PHY_LINK_STATUS: + ds = "SAS PHY Link Status"; + break; + case MPI_EVENT_SAS_DISCOVERY_ERROR: + ds = "SAS Discovery Error"; + break; + /* * MPT base "custom" events may be added here... */ @@ -5429,7 +5810,7 @@ EventDescriptionStr(u8 event, u32 evData0) ds = "Unknown"; break; } - return ds; + strcpy(evStr,ds); } /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ @@ -5451,7 +5832,7 @@ ProcessEventNotification(MPT_ADAPTER *ioc, EventNotificationReply_t *pEventReply int ii; int r = 0; int handlers = 0; - char *evStr; + char evStr[100]; u8 event; /* @@ -5464,7 +5845,7 @@ ProcessEventNotification(MPT_ADAPTER *ioc, EventNotificationReply_t *pEventReply evData0 = le32_to_cpu(pEventReply->Data[0]); } - evStr = EventDescriptionStr(event, evData0); + EventDescriptionStr(event, evData0, evStr); devtprintk((MYIOC_s_INFO_FMT "MPT event (%s=%02Xh) detected!\n", ioc->name, evStr, @@ -5481,20 +5862,6 @@ ProcessEventNotification(MPT_ADAPTER *ioc, EventNotificationReply_t *pEventReply * Do general / base driver event processing */ switch(event) { - case MPI_EVENT_NONE: /* 00 */ - case MPI_EVENT_LOG_DATA: /* 01 */ - case MPI_EVENT_STATE_CHANGE: /* 02 */ - case MPI_EVENT_UNIT_ATTENTION: /* 03 */ - case MPI_EVENT_IOC_BUS_RESET: /* 04 */ - case MPI_EVENT_EXT_BUS_RESET: /* 05 */ - case MPI_EVENT_RESCAN: /* 06 */ - case MPI_EVENT_LINK_STATUS_CHANGE: /* 07 */ - case MPI_EVENT_LOOP_STATE_CHANGE: /* 08 */ - case MPI_EVENT_LOGOUT: /* 09 */ - case MPI_EVENT_INTEGRATED_RAID: /* 0B */ - case MPI_EVENT_SCSI_DEVICE_STATUS_CHANGE: /* 0C */ - default: - break; case MPI_EVENT_EVENT_CHANGE: /* 0A */ if (evDataLen) { u8 evState = evData0 & 0xFF; @@ -5507,6 +5874,8 @@ ProcessEventNotification(MPT_ADAPTER *ioc, EventNotificationReply_t *pEventReply } } break; + default: + break; } /* @@ -5814,6 +6183,7 @@ EXPORT_SYMBOL(mpt_findImVolumes); EXPORT_SYMBOL(mpt_read_ioc_pg_3); EXPORT_SYMBOL(mpt_alloc_fw_memory); EXPORT_SYMBOL(mpt_free_fw_memory); +EXPORT_SYMBOL(mptbase_sas_persist_operation); /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ diff --git a/drivers/message/fusion/mptbase.h b/drivers/message/fusion/mptbase.h index f4827d92373..bbd21d74ce5 100644 --- a/drivers/message/fusion/mptbase.h +++ b/drivers/message/fusion/mptbase.h @@ -65,6 +65,7 @@ #include "lsi/mpi_fc.h" /* Fibre Channel (lowlevel) support */ #include "lsi/mpi_targ.h" /* SCSI/FCP Target protcol support */ #include "lsi/mpi_tool.h" /* Tools support */ +#include "lsi/mpi_sas.h" /* SAS support */ /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ @@ -477,6 +478,14 @@ typedef struct _ScsiCfgData { u8 rsvd[1]; } ScsiCfgData; +typedef struct _SasCfgData { + u8 ptClear; /* 1 to automatically clear the + * persistent table. + * 0 to disable + * automatic clearing. + */ +}SasCfgData; + /* * Adapter Structure - pci_dev specific. Maximum: MPT_MAX_ADAPTERS */ @@ -530,11 +539,15 @@ typedef struct _MPT_ADAPTER u8 *sense_buf_pool; dma_addr_t sense_buf_pool_dma; u32 sense_buf_low_dma; + u8 *HostPageBuffer; /* SAS - host page buffer support */ + u32 HostPageBuffer_sz; + dma_addr_t HostPageBuffer_dma; int mtrr_reg; struct pci_dev *pcidev; /* struct pci_dev pointer */ u8 __iomem *memmap; /* mmap address */ struct Scsi_Host *sh; /* Scsi Host pointer */ ScsiCfgData spi_data; /* Scsi config. data */ + SasCfgData sas_data; /* Sas config. data */ MPT_IOCTL *ioctl; /* ioctl data pointer */ struct proc_dir_entry *ioc_dentry; struct _MPT_ADAPTER *alt_ioc; /* ptr to 929 bound adapter port */ @@ -554,31 +567,35 @@ typedef struct _MPT_ADAPTER #else u32 mfcnt; #endif - u32 NB_for_64_byte_frame; + u32 NB_for_64_byte_frame; u32 hs_req[MPT_MAX_FRAME_SIZE/sizeof(u32)]; u16 hs_reply[MPT_MAX_FRAME_SIZE/sizeof(u16)]; IOCFactsReply_t facts; PortFactsReply_t pfacts[2]; FCPortPage0_t fc_port_page0[2]; + struct timer_list persist_timer; /* persist table timer */ + int persist_wait_done; /* persist completion flag */ + u8 persist_reply_frame[MPT_DEFAULT_FRAME_SIZE]; /* persist reply */ LANPage0_t lan_cnfg_page0; LANPage1_t lan_cnfg_page1; - /* + /* * Description: errata_flag_1064 * If a PCIX read occurs within 1 or 2 cycles after the chip receives * a split completion for a read data, an internal address pointer incorrectly * increments by 32 bytes */ - int errata_flag_1064; + int errata_flag_1064; u8 FirstWhoInit; u8 upload_fw; /* If set, do a fw upload */ u8 reload_fw; /* Force a FW Reload on next reset */ - u8 NBShiftFactor; /* NB Shift Factor based on Block Size (Facts) */ + u8 NBShiftFactor; /* NB Shift Factor based on Block Size (Facts) */ u8 pad1[4]; int DoneCtx; int TaskCtx; int InternalCtx; - struct list_head list; + struct list_head list; struct net_device *netdev; + struct list_head sas_topology; } MPT_ADAPTER; /* @@ -964,6 +981,7 @@ extern void mpt_alloc_fw_memory(MPT_ADAPTER *ioc, int size); extern void mpt_free_fw_memory(MPT_ADAPTER *ioc); extern int mpt_findImVolumes(MPT_ADAPTER *ioc); extern int mpt_read_ioc_pg_3(MPT_ADAPTER *ioc); +extern int mptbase_sas_persist_operation(MPT_ADAPTER *ioc, u8 persist_opcode); /* * Public data decl's... diff --git a/drivers/message/fusion/mptscsih.c b/drivers/message/fusion/mptscsih.c index 4a003dc5fde..58b5fdee009 100644 --- a/drivers/message/fusion/mptscsih.c +++ b/drivers/message/fusion/mptscsih.c @@ -1256,8 +1256,7 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *)) MPT_SCSI_HOST *hd; MPT_FRAME_HDR *mf; SCSIIORequest_t *pScsiReq; - VirtDevice *pTarget; - int target; + VirtDevice *pTarget = SCpnt->device->hostdata; int lun; u32 datalen; u32 scsictl; @@ -1267,12 +1266,9 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *)) int ii; hd = (MPT_SCSI_HOST *) SCpnt->device->host->hostdata; - target = SCpnt->device->id; lun = SCpnt->device->lun; SCpnt->scsi_done = done; - pTarget = hd->Targets[target]; - dmfprintk((MYIOC_s_INFO_FMT "qcmd: SCpnt=%p, done()=%p\n", (hd && hd->ioc) ? hd->ioc->name : "ioc?", SCpnt, done)); @@ -1315,7 +1311,7 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *)) /* Default to untagged. Once a target structure has been allocated, * use the Inquiry data to determine if device supports tagged. */ - if ( pTarget + if (pTarget && (pTarget->tflags & MPT_TARGET_FLAGS_Q_YES) && (SCpnt->device->tagged_supported)) { scsictl = scsidir | MPI_SCSIIO_CONTROL_SIMPLEQ; @@ -1325,8 +1321,8 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *)) /* Use the above information to set up the message frame */ - pScsiReq->TargetID = (u8) target; - pScsiReq->Bus = (u8) SCpnt->device->channel; + pScsiReq->TargetID = (u8) pTarget->target_id; + pScsiReq->Bus = pTarget->bus_id; pScsiReq->ChainOffset = 0; pScsiReq->Function = MPI_FUNCTION_SCSI_IO_REQUEST; pScsiReq->CDBLength = SCpnt->cmd_len; @@ -1378,7 +1374,7 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *)) #ifdef MPTSCSIH_ENABLE_DOMAIN_VALIDATION if (hd->ioc->bus_type == SCSI) { - int dvStatus = hd->ioc->spi_data.dvStatus[target]; + int dvStatus = hd->ioc->spi_data.dvStatus[pTarget->target_id]; int issueCmd = 1; if (dvStatus || hd->ioc->spi_data.forceDv) { @@ -2180,6 +2176,7 @@ mptscsih_slave_alloc(struct scsi_device *device) out: vdev->num_luns++; + device->hostdata = vdev; return 0; } From 0c33b27deb93178f10778b3d2669af1674793cef Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 9 Sep 2005 16:27:19 +0200 Subject: [PATCH 033/204] [SCSI] fusion SAS support (mptsas driver) Adds the actual mptsas driver, based upon the LSI driver with new work for SAS transport class integration from Eric Moore and me. This obviously depends on the SAS transport class. Signed-off-by: Christoph Hellwig Signed-off-by: James Bottomley --- drivers/message/fusion/Kconfig | 17 + drivers/message/fusion/Makefile | 1 + drivers/message/fusion/mptsas.c | 1235 +++++++++++++++++++++++++++++++ 3 files changed, 1253 insertions(+) create mode 100644 drivers/message/fusion/mptsas.c diff --git a/drivers/message/fusion/Kconfig b/drivers/message/fusion/Kconfig index 33f209a39cb..1883d22cffe 100644 --- a/drivers/message/fusion/Kconfig +++ b/drivers/message/fusion/Kconfig @@ -35,6 +35,23 @@ config FUSION_FC LSIFC929X LSIFC929XL +config FUSION_SAS + tristate "Fusion MPT ScsiHost drivers for SAS" + depends on PCI && SCSI + select FUSION + select SCSI_SAS_ATTRS + ---help--- + SCSI HOST support for a SAS host adapters. + + List of supported controllers: + + LSISAS1064 + LSISAS1066 + LSISAS1068 + LSISAS1064E + LSISAS1066E + LSISAS1068E + config FUSION_MAX_SGE int "Maximum number of scatter gather entries (16 - 128)" depends on FUSION diff --git a/drivers/message/fusion/Makefile b/drivers/message/fusion/Makefile index 1d2f9db813c..8a2e2657f4c 100644 --- a/drivers/message/fusion/Makefile +++ b/drivers/message/fusion/Makefile @@ -34,5 +34,6 @@ obj-$(CONFIG_FUSION_SPI) += mptbase.o mptscsih.o mptspi.o obj-$(CONFIG_FUSION_FC) += mptbase.o mptscsih.o mptfc.o +obj-$(CONFIG_FUSION_SAS) += mptbase.o mptscsih.o mptsas.o obj-$(CONFIG_FUSION_CTL) += mptctl.o obj-$(CONFIG_FUSION_LAN) += mptlan.o diff --git a/drivers/message/fusion/mptsas.c b/drivers/message/fusion/mptsas.c new file mode 100644 index 00000000000..0d9a192e1bd --- /dev/null +++ b/drivers/message/fusion/mptsas.c @@ -0,0 +1,1235 @@ +/* + * linux/drivers/message/fusion/mptsas.c + * For use with LSI Logic PCI chip/adapter(s) + * running LSI Logic Fusion MPT (Message Passing Technology) firmware. + * + * Copyright (c) 1999-2005 LSI Logic Corporation + * (mailto:mpt_linux_developer@lsil.com) + * Copyright (c) 2005 Dell + */ +/*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ +/* + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; version 2 of the License. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + NO WARRANTY + THE PROGRAM IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OR + CONDITIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED INCLUDING, WITHOUT + LIMITATION, ANY WARRANTIES OR CONDITIONS OF TITLE, NON-INFRINGEMENT, + MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Each Recipient is + solely responsible for determining the appropriateness of using and + distributing the Program and assumes all risks associated with its + exercise of rights under this Agreement, including but not limited to + the risks and costs of program errors, damage to or loss of data, + programs or equipment, and unavailability or interruption of operations. + + DISCLAIMER OF LIABILITY + NEITHER RECIPIENT NOR ANY CONTRIBUTORS SHALL HAVE ANY LIABILITY FOR ANY + DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + DAMAGES (INCLUDING WITHOUT LIMITATION LOST PROFITS), HOWEVER CAUSED AND + ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR + TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE + USE OR DISTRIBUTION OF THE PROGRAM OR THE EXERCISE OF ANY RIGHTS GRANTED + HEREUNDER, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA +*/ +/*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ + +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include + +#include "mptbase.h" +#include "mptscsih.h" + + +#define my_NAME "Fusion MPT SAS Host driver" +#define my_VERSION MPT_LINUX_VERSION_COMMON +#define MYNAM "mptsas" + +MODULE_AUTHOR(MODULEAUTHOR); +MODULE_DESCRIPTION(my_NAME); +MODULE_LICENSE("GPL"); + +static int mpt_pq_filter; +module_param(mpt_pq_filter, int, 0); +MODULE_PARM_DESC(mpt_pq_filter, + "Enable peripheral qualifier filter: enable=1 " + "(default=0)"); + +static int mpt_pt_clear; +module_param(mpt_pt_clear, int, 0); +MODULE_PARM_DESC(mpt_pt_clear, + "Clear persistency table: enable=1 " + "(default=MPTSCSIH_PT_CLEAR=0)"); + +static int mptsasDoneCtx = -1; +static int mptsasTaskCtx = -1; +static int mptsasInternalCtx = -1; /* Used only for internal commands */ + + +/* + * SAS topology structures + * + * The MPT Fusion firmware interface spreads information about the + * SAS topology over many manufacture pages, thus we need some data + * structure to collect it and process it for the SAS transport class. + */ + +struct mptsas_devinfo { + u16 handle; /* unique id to address this device */ + u8 phy_id; /* phy number of parent device */ + u8 port_id; /* sas physical port this device + is assoc'd with */ + u8 target; /* logical target id of this device */ + u8 bus; /* logical bus number of this device */ + u64 sas_address; /* WWN of this device, + SATA is assigned by HBA,expander */ + u32 device_info; /* bitfield detailed info about this device */ +}; + +struct mptsas_phyinfo { + u8 phy_id; /* phy index */ + u8 port_id; /* port number this phy is part of */ + u8 negotiated_link_rate; /* nego'd link rate for this phy */ + u8 hw_link_rate; /* hardware max/min phys link rate */ + u8 programmed_link_rate; /* programmed max/min phy link rate */ + struct mptsas_devinfo identify; /* point to phy device info */ + struct mptsas_devinfo attached; /* point to attached device info */ + struct sas_rphy *rphy; +}; + +struct mptsas_portinfo { + struct list_head list; + u16 handle; /* unique id to address this */ + u8 num_phys; /* number of phys */ + struct mptsas_phyinfo *phy_info; +}; + +/* + * This is pretty ugly. We will be able to seriously clean it up + * once the DV code in mptscsih goes away and we can properly + * implement ->target_alloc. + */ +static int +mptsas_slave_alloc(struct scsi_device *device) +{ + struct Scsi_Host *host = device->host; + MPT_SCSI_HOST *hd = (MPT_SCSI_HOST *)host->hostdata; + struct sas_rphy *rphy; + struct mptsas_portinfo *p; + VirtDevice *vdev; + uint target = device->id; + int i; + + if ((vdev = hd->Targets[target]) != NULL) + goto out; + + vdev = kmalloc(sizeof(VirtDevice), GFP_KERNEL); + if (!vdev) { + printk(MYIOC_s_ERR_FMT "slave_alloc kmalloc(%zd) FAILED!\n", + hd->ioc->name, sizeof(VirtDevice)); + return -ENOMEM; + } + + memset(vdev, 0, sizeof(VirtDevice)); + vdev->tflags = MPT_TARGET_FLAGS_Q_YES|MPT_TARGET_FLAGS_VALID_INQUIRY; + vdev->ioc_id = hd->ioc->id; + + rphy = dev_to_rphy(device->sdev_target->dev.parent); + list_for_each_entry(p, &hd->ioc->sas_topology, list) { + for (i = 0; i < p->num_phys; i++) { + if (p->phy_info[i].attached.sas_address == + rphy->identify.sas_address) { + vdev->target_id = + p->phy_info[i].attached.target; + vdev->bus_id = p->phy_info[i].attached.bus; + hd->Targets[device->id] = vdev; + goto out; + } + } + } + + printk("No matching SAS device found!!\n"); + kfree(vdev); + return -ENODEV; + + out: + vdev->num_luns++; + device->hostdata = vdev; + return 0; +} + +static struct scsi_host_template mptsas_driver_template = { + .proc_name = "mptsas", + .proc_info = mptscsih_proc_info, + .name = "MPT SPI Host", + .info = mptscsih_info, + .queuecommand = mptscsih_qcmd, + .slave_alloc = mptsas_slave_alloc, + .slave_configure = mptscsih_slave_configure, + .slave_destroy = mptscsih_slave_destroy, + .change_queue_depth = mptscsih_change_queue_depth, + .eh_abort_handler = mptscsih_abort, + .eh_device_reset_handler = mptscsih_dev_reset, + .eh_bus_reset_handler = mptscsih_bus_reset, + .eh_host_reset_handler = mptscsih_host_reset, + .bios_param = mptscsih_bios_param, + .can_queue = MPT_FC_CAN_QUEUE, + .this_id = -1, + .sg_tablesize = MPT_SCSI_SG_DEPTH, + .max_sectors = 8192, + .cmd_per_lun = 7, + .use_clustering = ENABLE_CLUSTERING, +}; + +static struct sas_function_template mptsas_transport_functions = { +}; + +static struct scsi_transport_template *mptsas_transport_template; + +#ifdef SASDEBUG +static void mptsas_print_phy_data(MPI_SAS_IO_UNIT0_PHY_DATA *phy_data) +{ + printk("---- IO UNIT PAGE 0 ------------\n"); + printk("Handle=0x%X\n", + le16_to_cpu(phy_data->AttachedDeviceHandle)); + printk("Controller Handle=0x%X\n", + le16_to_cpu(phy_data->ControllerDevHandle)); + printk("Port=0x%X\n", phy_data->Port); + printk("Port Flags=0x%X\n", phy_data->PortFlags); + printk("PHY Flags=0x%X\n", phy_data->PhyFlags); + printk("Negotiated Link Rate=0x%X\n", phy_data->NegotiatedLinkRate); + printk("Controller PHY Device Info=0x%X\n", + le32_to_cpu(phy_data->ControllerPhyDeviceInfo)); + printk("DiscoveryStatus=0x%X\n", + le32_to_cpu(phy_data->DiscoveryStatus)); + printk("\n"); +} + +static void mptsas_print_phy_pg0(SasPhyPage0_t *pg0) +{ + __le64 sas_address; + + memcpy(&sas_address, &pg0->SASAddress, sizeof(__le64)); + + printk("---- SAS PHY PAGE 0 ------------\n"); + printk("Attached Device Handle=0x%X\n", + le16_to_cpu(pg0->AttachedDevHandle)); + printk("SAS Address=0x%llX\n", + (unsigned long long)le64_to_cpu(sas_address)); + printk("Attached PHY Identifier=0x%X\n", pg0->AttachedPhyIdentifier); + printk("Attached Device Info=0x%X\n", + le32_to_cpu(pg0->AttachedDeviceInfo)); + printk("Programmed Link Rate=0x%X\n", pg0->ProgrammedLinkRate); + printk("Change Count=0x%X\n", pg0->ChangeCount); + printk("PHY Info=0x%X\n", le32_to_cpu(pg0->PhyInfo)); + printk("\n"); +} + +static void mptsas_print_device_pg0(SasDevicePage0_t *pg0) +{ + __le64 sas_address; + + memcpy(&sas_address, &pg0->SASAddress, sizeof(__le64)); + + printk("---- SAS DEVICE PAGE 0 ---------\n"); + printk("Handle=0x%X\n" ,le16_to_cpu(pg0->DevHandle)); + printk("Enclosure Handle=0x%X\n", le16_to_cpu(pg0->EnclosureHandle)); + printk("Slot=0x%X\n", le16_to_cpu(pg0->Slot)); + printk("SAS Address=0x%llX\n", le64_to_cpu(sas_address)); + printk("Target ID=0x%X\n", pg0->TargetID); + printk("Bus=0x%X\n", pg0->Bus); + printk("PhyNum=0x%X\n", pg0->PhyNum); + printk("AccessStatus=0x%X\n", le16_to_cpu(pg0->AccessStatus)); + printk("Device Info=0x%X\n", le32_to_cpu(pg0->DeviceInfo)); + printk("Flags=0x%X\n", le16_to_cpu(pg0->Flags)); + printk("Physical Port=0x%X\n", pg0->PhysicalPort); + printk("\n"); +} + +static void mptsas_print_expander_pg1(SasExpanderPage1_t *pg1) +{ + printk("---- SAS EXPANDER PAGE 1 ------------\n"); + + printk("Physical Port=0x%X\n", pg1->PhysicalPort); + printk("PHY Identifier=0x%X\n", pg1->Phy); + printk("Negotiated Link Rate=0x%X\n", pg1->NegotiatedLinkRate); + printk("Programmed Link Rate=0x%X\n", pg1->ProgrammedLinkRate); + printk("Hardware Link Rate=0x%X\n", pg1->HwLinkRate); + printk("Owner Device Handle=0x%X\n", + le16_to_cpu(pg1->OwnerDevHandle)); + printk("Attached Device Handle=0x%X\n", + le16_to_cpu(pg1->AttachedDevHandle)); +} +#else +#define mptsas_print_phy_data(phy_data) do { } while (0) +#define mptsas_print_phy_pg0(pg0) do { } while (0) +#define mptsas_print_device_pg0(pg0) do { } while (0) +#define mptsas_print_expander_pg1(pg1) do { } while (0) +#endif + +static int +mptsas_sas_io_unit_pg0(MPT_ADAPTER *ioc, struct mptsas_portinfo *port_info) +{ + ConfigExtendedPageHeader_t hdr; + CONFIGPARMS cfg; + SasIOUnitPage0_t *buffer; + dma_addr_t dma_handle; + int error, i; + + hdr.PageVersion = MPI_SASIOUNITPAGE0_PAGEVERSION; + hdr.ExtPageLength = 0; + hdr.PageNumber = 0; + hdr.Reserved1 = 0; + hdr.Reserved2 = 0; + hdr.PageType = MPI_CONFIG_PAGETYPE_EXTENDED; + hdr.ExtPageType = MPI_CONFIG_EXTPAGETYPE_SAS_IO_UNIT; + + cfg.cfghdr.ehdr = &hdr; + cfg.physAddr = -1; + cfg.pageAddr = 0; + cfg.action = MPI_CONFIG_ACTION_PAGE_HEADER; + cfg.dir = 0; /* read */ + cfg.timeout = 10; + + error = mpt_config(ioc, &cfg); + if (error) + goto out; + if (!hdr.ExtPageLength) { + error = -ENXIO; + goto out; + } + + buffer = pci_alloc_consistent(ioc->pcidev, hdr.ExtPageLength * 4, + &dma_handle); + if (!buffer) { + error = -ENOMEM; + goto out; + } + + cfg.physAddr = dma_handle; + cfg.action = MPI_CONFIG_ACTION_PAGE_READ_CURRENT; + + error = mpt_config(ioc, &cfg); + if (error) + goto out_free_consistent; + + port_info->num_phys = buffer->NumPhys; + port_info->phy_info = kcalloc(port_info->num_phys, + sizeof(struct mptsas_phyinfo),GFP_KERNEL); + if (!port_info->phy_info) { + error = -ENOMEM; + goto out_free_consistent; + } + + for (i = 0; i < port_info->num_phys; i++) { + mptsas_print_phy_data(&buffer->PhyData[i]); + port_info->phy_info[i].phy_id = i; + port_info->phy_info[i].port_id = + buffer->PhyData[i].Port; + port_info->phy_info[i].negotiated_link_rate = + buffer->PhyData[i].NegotiatedLinkRate; + } + + out_free_consistent: + pci_free_consistent(ioc->pcidev, hdr.ExtPageLength * 4, + buffer, dma_handle); + out: + return error; +} + +static int +mptsas_sas_phy_pg0(MPT_ADAPTER *ioc, struct mptsas_phyinfo *phy_info, + u32 form, u32 form_specific) +{ + ConfigExtendedPageHeader_t hdr; + CONFIGPARMS cfg; + SasPhyPage0_t *buffer; + dma_addr_t dma_handle; + int error; + + hdr.PageVersion = MPI_SASPHY0_PAGEVERSION; + hdr.ExtPageLength = 0; + hdr.PageNumber = 0; + hdr.Reserved1 = 0; + hdr.Reserved2 = 0; + hdr.PageType = MPI_CONFIG_PAGETYPE_EXTENDED; + hdr.ExtPageType = MPI_CONFIG_EXTPAGETYPE_SAS_PHY; + + cfg.cfghdr.ehdr = &hdr; + cfg.dir = 0; /* read */ + cfg.timeout = 10; + + /* Get Phy Pg 0 for each Phy. */ + cfg.physAddr = -1; + cfg.pageAddr = form + form_specific; + cfg.action = MPI_CONFIG_ACTION_PAGE_HEADER; + + error = mpt_config(ioc, &cfg); + if (error) + goto out; + + if (!hdr.ExtPageLength) { + error = -ENXIO; + goto out; + } + + buffer = pci_alloc_consistent(ioc->pcidev, hdr.ExtPageLength * 4, + &dma_handle); + if (!buffer) { + error = -ENOMEM; + goto out; + } + + cfg.physAddr = dma_handle; + cfg.action = MPI_CONFIG_ACTION_PAGE_READ_CURRENT; + + error = mpt_config(ioc, &cfg); + if (error) + goto out_free_consistent; + + mptsas_print_phy_pg0(buffer); + + phy_info->hw_link_rate = buffer->HwLinkRate; + phy_info->programmed_link_rate = buffer->ProgrammedLinkRate; + phy_info->identify.handle = le16_to_cpu(buffer->OwnerDevHandle); + phy_info->attached.handle = le16_to_cpu(buffer->AttachedDevHandle); + + out_free_consistent: + pci_free_consistent(ioc->pcidev, hdr.ExtPageLength * 4, + buffer, dma_handle); + out: + return error; +} + +static int +mptsas_sas_device_pg0(MPT_ADAPTER *ioc, struct mptsas_devinfo *device_info, + u32 form, u32 form_specific) +{ + ConfigExtendedPageHeader_t hdr; + CONFIGPARMS cfg; + SasDevicePage0_t *buffer; + dma_addr_t dma_handle; + __le64 sas_address; + int error; + + hdr.PageVersion = MPI_SASDEVICE0_PAGEVERSION; + hdr.ExtPageLength = 0; + hdr.PageNumber = 0; + hdr.Reserved1 = 0; + hdr.Reserved2 = 0; + hdr.PageType = MPI_CONFIG_PAGETYPE_EXTENDED; + hdr.ExtPageType = MPI_CONFIG_EXTPAGETYPE_SAS_DEVICE; + + cfg.cfghdr.ehdr = &hdr; + cfg.pageAddr = form + form_specific; + cfg.physAddr = -1; + cfg.action = MPI_CONFIG_ACTION_PAGE_HEADER; + cfg.dir = 0; /* read */ + cfg.timeout = 10; + + error = mpt_config(ioc, &cfg); + if (error) + goto out; + if (!hdr.ExtPageLength) { + error = -ENXIO; + goto out; + } + + buffer = pci_alloc_consistent(ioc->pcidev, hdr.ExtPageLength * 4, + &dma_handle); + if (!buffer) { + error = -ENOMEM; + goto out; + } + + cfg.physAddr = dma_handle; + cfg.action = MPI_CONFIG_ACTION_PAGE_READ_CURRENT; + + error = mpt_config(ioc, &cfg); + if (error) + goto out_free_consistent; + + mptsas_print_device_pg0(buffer); + + device_info->handle = le16_to_cpu(buffer->DevHandle); + device_info->phy_id = buffer->PhyNum; + device_info->port_id = buffer->PhysicalPort; + device_info->target = buffer->TargetID; + device_info->bus = buffer->Bus; + memcpy(&sas_address, &buffer->SASAddress, sizeof(__le64)); + device_info->sas_address = le64_to_cpu(sas_address); + device_info->device_info = + le32_to_cpu(buffer->DeviceInfo); + + out_free_consistent: + pci_free_consistent(ioc->pcidev, hdr.ExtPageLength * 4, + buffer, dma_handle); + out: + return error; +} + +static int +mptsas_sas_expander_pg0(MPT_ADAPTER *ioc, struct mptsas_portinfo *port_info, + u32 form, u32 form_specific) +{ + ConfigExtendedPageHeader_t hdr; + CONFIGPARMS cfg; + SasExpanderPage0_t *buffer; + dma_addr_t dma_handle; + int error; + + hdr.PageVersion = MPI_SASEXPANDER0_PAGEVERSION; + hdr.ExtPageLength = 0; + hdr.PageNumber = 0; + hdr.Reserved1 = 0; + hdr.Reserved2 = 0; + hdr.PageType = MPI_CONFIG_PAGETYPE_EXTENDED; + hdr.ExtPageType = MPI_CONFIG_EXTPAGETYPE_SAS_EXPANDER; + + cfg.cfghdr.ehdr = &hdr; + cfg.physAddr = -1; + cfg.pageAddr = form + form_specific; + cfg.action = MPI_CONFIG_ACTION_PAGE_HEADER; + cfg.dir = 0; /* read */ + cfg.timeout = 10; + + error = mpt_config(ioc, &cfg); + if (error) + goto out; + + if (!hdr.ExtPageLength) { + error = -ENXIO; + goto out; + } + + buffer = pci_alloc_consistent(ioc->pcidev, hdr.ExtPageLength * 4, + &dma_handle); + if (!buffer) { + error = -ENOMEM; + goto out; + } + + cfg.physAddr = dma_handle; + cfg.action = MPI_CONFIG_ACTION_PAGE_READ_CURRENT; + + error = mpt_config(ioc, &cfg); + if (error) + goto out_free_consistent; + + /* save config data */ + port_info->num_phys = buffer->NumPhys; + port_info->handle = le16_to_cpu(buffer->DevHandle); + port_info->phy_info = kcalloc(port_info->num_phys, + sizeof(struct mptsas_phyinfo),GFP_KERNEL); + if (!port_info->phy_info) { + error = -ENOMEM; + goto out_free_consistent; + } + + out_free_consistent: + pci_free_consistent(ioc->pcidev, hdr.ExtPageLength * 4, + buffer, dma_handle); + out: + return error; +} + +static int +mptsas_sas_expander_pg1(MPT_ADAPTER *ioc, struct mptsas_phyinfo *phy_info, + u32 form, u32 form_specific) +{ + ConfigExtendedPageHeader_t hdr; + CONFIGPARMS cfg; + SasExpanderPage1_t *buffer; + dma_addr_t dma_handle; + int error; + + hdr.PageVersion = MPI_SASEXPANDER0_PAGEVERSION; + hdr.ExtPageLength = 0; + hdr.PageNumber = 1; + hdr.Reserved1 = 0; + hdr.Reserved2 = 0; + hdr.PageType = MPI_CONFIG_PAGETYPE_EXTENDED; + hdr.ExtPageType = MPI_CONFIG_EXTPAGETYPE_SAS_EXPANDER; + + cfg.cfghdr.ehdr = &hdr; + cfg.physAddr = -1; + cfg.pageAddr = form + form_specific; + cfg.action = MPI_CONFIG_ACTION_PAGE_HEADER; + cfg.dir = 0; /* read */ + cfg.timeout = 10; + + error = mpt_config(ioc, &cfg); + if (error) + goto out; + + if (!hdr.ExtPageLength) { + error = -ENXIO; + goto out; + } + + buffer = pci_alloc_consistent(ioc->pcidev, hdr.ExtPageLength * 4, + &dma_handle); + if (!buffer) { + error = -ENOMEM; + goto out; + } + + cfg.physAddr = dma_handle; + cfg.action = MPI_CONFIG_ACTION_PAGE_READ_CURRENT; + + error = mpt_config(ioc, &cfg); + if (error) + goto out_free_consistent; + + + mptsas_print_expander_pg1(buffer); + + /* save config data */ + phy_info->phy_id = buffer->Phy; + phy_info->port_id = buffer->PhysicalPort; + phy_info->negotiated_link_rate = buffer->NegotiatedLinkRate; + phy_info->programmed_link_rate = buffer->ProgrammedLinkRate; + phy_info->hw_link_rate = buffer->HwLinkRate; + phy_info->identify.handle = le16_to_cpu(buffer->OwnerDevHandle); + phy_info->attached.handle = le16_to_cpu(buffer->AttachedDevHandle); + + + out_free_consistent: + pci_free_consistent(ioc->pcidev, hdr.ExtPageLength * 4, + buffer, dma_handle); + out: + return error; +} + +static void +mptsas_parse_device_info(struct sas_identify *identify, + struct mptsas_devinfo *device_info) +{ + u16 protocols; + + identify->sas_address = device_info->sas_address; + identify->phy_identifier = device_info->phy_id; + + /* + * Fill in Phy Initiator Port Protocol. + * Bits 6:3, more than one bit can be set, fall through cases. + */ + protocols = device_info->device_info & 0x78; + identify->initiator_port_protocols = 0; + if (protocols & MPI_SAS_DEVICE_INFO_SSP_INITIATOR) + identify->initiator_port_protocols |= SAS_PROTOCOL_SSP; + if (protocols & MPI_SAS_DEVICE_INFO_STP_INITIATOR) + identify->initiator_port_protocols |= SAS_PROTOCOL_STP; + if (protocols & MPI_SAS_DEVICE_INFO_SMP_INITIATOR) + identify->initiator_port_protocols |= SAS_PROTOCOL_SMP; + if (protocols & MPI_SAS_DEVICE_INFO_SATA_HOST) + identify->initiator_port_protocols |= SAS_PROTOCOL_SATA; + + /* + * Fill in Phy Target Port Protocol. + * Bits 10:7, more than one bit can be set, fall through cases. + */ + protocols = device_info->device_info & 0x780; + identify->target_port_protocols = 0; + if (protocols & MPI_SAS_DEVICE_INFO_SSP_TARGET) + identify->target_port_protocols |= SAS_PROTOCOL_SSP; + if (protocols & MPI_SAS_DEVICE_INFO_STP_TARGET) + identify->target_port_protocols |= SAS_PROTOCOL_STP; + if (protocols & MPI_SAS_DEVICE_INFO_SMP_TARGET) + identify->target_port_protocols |= SAS_PROTOCOL_SMP; + if (protocols & MPI_SAS_DEVICE_INFO_SATA_DEVICE) + identify->target_port_protocols |= SAS_PROTOCOL_SATA; + + /* + * Fill in Attached device type. + */ + switch (device_info->device_info & + MPI_SAS_DEVICE_INFO_MASK_DEVICE_TYPE) { + case MPI_SAS_DEVICE_INFO_NO_DEVICE: + identify->device_type = SAS_PHY_UNUSED; + break; + case MPI_SAS_DEVICE_INFO_END_DEVICE: + identify->device_type = SAS_END_DEVICE; + break; + case MPI_SAS_DEVICE_INFO_EDGE_EXPANDER: + identify->device_type = SAS_EDGE_EXPANDER_DEVICE; + break; + case MPI_SAS_DEVICE_INFO_FANOUT_EXPANDER: + identify->device_type = SAS_FANOUT_EXPANDER_DEVICE; + break; + } +} + +static int mptsas_probe_one_phy(struct device *dev, + struct mptsas_phyinfo *phy_info, int index) +{ + struct sas_phy *port; + int error; + + port = sas_phy_alloc(dev, index); + if (!port) + return -ENOMEM; + + port->port_identifier = phy_info->port_id; + mptsas_parse_device_info(&port->identify, &phy_info->identify); + + /* + * Set Negotiated link rate. + */ + switch (phy_info->negotiated_link_rate) { + case MPI_SAS_IOUNIT0_RATE_PHY_DISABLED: + port->negotiated_linkrate = SAS_PHY_DISABLED; + break; + case MPI_SAS_IOUNIT0_RATE_FAILED_SPEED_NEGOTIATION: + port->negotiated_linkrate = SAS_LINK_RATE_FAILED; + break; + case MPI_SAS_IOUNIT0_RATE_1_5: + port->negotiated_linkrate = SAS_LINK_RATE_1_5_GBPS; + break; + case MPI_SAS_IOUNIT0_RATE_3_0: + port->negotiated_linkrate = SAS_LINK_RATE_3_0_GBPS; + break; + case MPI_SAS_IOUNIT0_RATE_SATA_OOB_COMPLETE: + case MPI_SAS_IOUNIT0_RATE_UNKNOWN: + default: + port->negotiated_linkrate = SAS_LINK_RATE_UNKNOWN; + break; + } + + /* + * Set Max hardware link rate. + */ + switch (phy_info->hw_link_rate & MPI_SAS_PHY0_PRATE_MAX_RATE_MASK) { + case MPI_SAS_PHY0_HWRATE_MAX_RATE_1_5: + port->maximum_linkrate_hw = SAS_LINK_RATE_1_5_GBPS; + break; + case MPI_SAS_PHY0_PRATE_MAX_RATE_3_0: + port->maximum_linkrate_hw = SAS_LINK_RATE_3_0_GBPS; + break; + default: + break; + } + + /* + * Set Max programmed link rate. + */ + switch (phy_info->programmed_link_rate & + MPI_SAS_PHY0_PRATE_MAX_RATE_MASK) { + case MPI_SAS_PHY0_PRATE_MAX_RATE_1_5: + port->maximum_linkrate = SAS_LINK_RATE_1_5_GBPS; + break; + case MPI_SAS_PHY0_PRATE_MAX_RATE_3_0: + port->maximum_linkrate = SAS_LINK_RATE_3_0_GBPS; + break; + default: + break; + } + + /* + * Set Min hardware link rate. + */ + switch (phy_info->hw_link_rate & MPI_SAS_PHY0_HWRATE_MIN_RATE_MASK) { + case MPI_SAS_PHY0_HWRATE_MIN_RATE_1_5: + port->minimum_linkrate_hw = SAS_LINK_RATE_1_5_GBPS; + break; + case MPI_SAS_PHY0_PRATE_MIN_RATE_3_0: + port->minimum_linkrate_hw = SAS_LINK_RATE_3_0_GBPS; + break; + default: + break; + } + + /* + * Set Min programmed link rate. + */ + switch (phy_info->programmed_link_rate & + MPI_SAS_PHY0_PRATE_MIN_RATE_MASK) { + case MPI_SAS_PHY0_PRATE_MIN_RATE_1_5: + port->minimum_linkrate = SAS_LINK_RATE_1_5_GBPS; + break; + case MPI_SAS_PHY0_PRATE_MIN_RATE_3_0: + port->minimum_linkrate = SAS_LINK_RATE_3_0_GBPS; + break; + default: + break; + } + + error = sas_phy_add(port); + if (error) { + sas_phy_free(port); + return error; + } + + if (phy_info->attached.handle) { + struct sas_rphy *rphy; + + rphy = sas_rphy_alloc(port); + if (!rphy) + return 0; /* non-fatal: an rphy can be added later */ + + mptsas_parse_device_info(&rphy->identify, &phy_info->attached); + error = sas_rphy_add(rphy); + if (error) { + sas_rphy_free(rphy); + return error; + } + + phy_info->rphy = rphy; + } + + return 0; +} + +static int +mptsas_probe_hba_phys(MPT_ADAPTER *ioc, int *index) +{ + struct mptsas_portinfo *port_info; + u32 handle = 0xFFFF; + int error = -ENOMEM, i; + + port_info = kmalloc(sizeof(*port_info), GFP_KERNEL); + if (!port_info) + goto out; + memset(port_info, 0, sizeof(*port_info)); + + error = mptsas_sas_io_unit_pg0(ioc, port_info); + if (error) + goto out_free_port_info; + + list_add_tail(&port_info->list, &ioc->sas_topology); + + for (i = 0; i < port_info->num_phys; i++) { + mptsas_sas_phy_pg0(ioc, &port_info->phy_info[i], + (MPI_SAS_PHY_PGAD_FORM_PHY_NUMBER << + MPI_SAS_PHY_PGAD_FORM_SHIFT), i); + + mptsas_sas_device_pg0(ioc, &port_info->phy_info[i].identify, + (MPI_SAS_DEVICE_PGAD_FORM_GET_NEXT_HANDLE << + MPI_SAS_DEVICE_PGAD_FORM_SHIFT), handle); + handle = port_info->phy_info[i].identify.handle; + + if (port_info->phy_info[i].attached.handle) { + mptsas_sas_device_pg0(ioc, + &port_info->phy_info[i].attached, + (MPI_SAS_DEVICE_PGAD_FORM_HANDLE << + MPI_SAS_DEVICE_PGAD_FORM_SHIFT), + port_info->phy_info[i].attached.handle); + } + + mptsas_probe_one_phy(&ioc->sh->shost_gendev, + &port_info->phy_info[i], *index); + (*index)++; + } + + return 0; + + out_free_port_info: + kfree(port_info); + out: + return error; +} + +static int +mptsas_probe_expander_phys(MPT_ADAPTER *ioc, u32 *handle, int *index) +{ + struct mptsas_portinfo *port_info, *p; + int error = -ENOMEM, i, j; + + port_info = kmalloc(sizeof(*port_info), GFP_KERNEL); + if (!port_info) + goto out; + memset(port_info, 0, sizeof(*port_info)); + + error = mptsas_sas_expander_pg0(ioc, port_info, + (MPI_SAS_EXPAND_PGAD_FORM_GET_NEXT_HANDLE << + MPI_SAS_EXPAND_PGAD_FORM_SHIFT), *handle); + if (error) + goto out_free_port_info; + + *handle = port_info->handle; + + list_add_tail(&port_info->list, &ioc->sas_topology); + for (i = 0; i < port_info->num_phys; i++) { + struct device *parent; + + mptsas_sas_expander_pg1(ioc, &port_info->phy_info[i], + (MPI_SAS_EXPAND_PGAD_FORM_HANDLE_PHY_NUM << + MPI_SAS_EXPAND_PGAD_FORM_SHIFT), (i << 16) + *handle); + + if (port_info->phy_info[i].identify.handle) { + mptsas_sas_device_pg0(ioc, + &port_info->phy_info[i].identify, + (MPI_SAS_DEVICE_PGAD_FORM_HANDLE << + MPI_SAS_DEVICE_PGAD_FORM_SHIFT), + port_info->phy_info[i].identify.handle); + } + + if (port_info->phy_info[i].attached.handle) { + mptsas_sas_device_pg0(ioc, + &port_info->phy_info[i].attached, + (MPI_SAS_DEVICE_PGAD_FORM_HANDLE << + MPI_SAS_DEVICE_PGAD_FORM_SHIFT), + port_info->phy_info[i].attached.handle); + } + + /* + * If we find a parent port handle this expander is + * attached to another expander, else it hangs of the + * HBA phys. + */ + parent = &ioc->sh->shost_gendev; + list_for_each_entry(p, &ioc->sas_topology, list) { + for (j = 0; j < p->num_phys; j++) { + if (port_info->phy_info[i].identify.handle == + p->phy_info[j].attached.handle) + parent = &p->phy_info[j].rphy->dev; + } + } + + mptsas_probe_one_phy(parent, &port_info->phy_info[i], *index); + (*index)++; + } + + return 0; + + out_free_port_info: + kfree(port_info); + out: + return error; +} + +static void +mptsas_scan_sas_topology(MPT_ADAPTER *ioc) +{ + u32 handle = 0xFFFF; + int index = 0; + + mptsas_probe_hba_phys(ioc, &index); + while (!mptsas_probe_expander_phys(ioc, &handle, &index)) + ; +} + +static int +mptsas_probe(struct pci_dev *pdev, const struct pci_device_id *id) +{ + struct Scsi_Host *sh; + MPT_SCSI_HOST *hd; + MPT_ADAPTER *ioc; + unsigned long flags; + int sz, ii; + int numSGE = 0; + int scale; + int ioc_cap; + u8 *mem; + int error=0; + int r; + + r = mpt_attach(pdev,id); + if (r) + return r; + + ioc = pci_get_drvdata(pdev); + ioc->DoneCtx = mptsasDoneCtx; + ioc->TaskCtx = mptsasTaskCtx; + ioc->InternalCtx = mptsasInternalCtx; + + /* Added sanity check on readiness of the MPT adapter. + */ + if (ioc->last_state != MPI_IOC_STATE_OPERATIONAL) { + printk(MYIOC_s_WARN_FMT + "Skipping because it's not operational!\n", + ioc->name); + return -ENODEV; + } + + if (!ioc->active) { + printk(MYIOC_s_WARN_FMT "Skipping because it's disabled!\n", + ioc->name); + return -ENODEV; + } + + /* Sanity check - ensure at least 1 port is INITIATOR capable + */ + ioc_cap = 0; + for (ii = 0; ii < ioc->facts.NumberOfPorts; ii++) { + if (ioc->pfacts[ii].ProtocolFlags & + MPI_PORTFACTS_PROTOCOL_INITIATOR) + ioc_cap++; + } + + if (!ioc_cap) { + printk(MYIOC_s_WARN_FMT + "Skipping ioc=%p because SCSI Initiator mode " + "is NOT enabled!\n", ioc->name, ioc); + return -ENODEV; + } + + sh = scsi_host_alloc(&mptsas_driver_template, sizeof(MPT_SCSI_HOST)); + if (!sh) { + printk(MYIOC_s_WARN_FMT + "Unable to register controller with SCSI subsystem\n", + ioc->name); + return -1; + } + + spin_lock_irqsave(&ioc->FreeQlock, flags); + + /* Attach the SCSI Host to the IOC structure + */ + ioc->sh = sh; + + sh->io_port = 0; + sh->n_io_port = 0; + sh->irq = 0; + + /* set 16 byte cdb's */ + sh->max_cmd_len = 16; + + sh->max_id = ioc->pfacts->MaxDevices + 1; + + sh->transportt = mptsas_transport_template; + + sh->max_lun = MPT_LAST_LUN + 1; + sh->max_channel = 0; + sh->this_id = ioc->pfacts[0].PortSCSIID; + + /* Required entry. + */ + sh->unique_id = ioc->id; + + INIT_LIST_HEAD(&ioc->sas_topology); + + /* Verify that we won't exceed the maximum + * number of chain buffers + * We can optimize: ZZ = req_sz/sizeof(SGE) + * For 32bit SGE's: + * numSGE = 1 + (ZZ-1)*(maxChain -1) + ZZ + * + (req_sz - 64)/sizeof(SGE) + * A slightly different algorithm is required for + * 64bit SGEs. + */ + scale = ioc->req_sz/(sizeof(dma_addr_t) + sizeof(u32)); + if (sizeof(dma_addr_t) == sizeof(u64)) { + numSGE = (scale - 1) * + (ioc->facts.MaxChainDepth-1) + scale + + (ioc->req_sz - 60) / (sizeof(dma_addr_t) + + sizeof(u32)); + } else { + numSGE = 1 + (scale - 1) * + (ioc->facts.MaxChainDepth-1) + scale + + (ioc->req_sz - 64) / (sizeof(dma_addr_t) + + sizeof(u32)); + } + + if (numSGE < sh->sg_tablesize) { + /* Reset this value */ + dprintk((MYIOC_s_INFO_FMT + "Resetting sg_tablesize to %d from %d\n", + ioc->name, numSGE, sh->sg_tablesize)); + sh->sg_tablesize = numSGE; + } + + spin_unlock_irqrestore(&ioc->FreeQlock, flags); + + hd = (MPT_SCSI_HOST *) sh->hostdata; + hd->ioc = ioc; + + /* SCSI needs scsi_cmnd lookup table! + * (with size equal to req_depth*PtrSz!) + */ + sz = ioc->req_depth * sizeof(void *); + mem = kmalloc(sz, GFP_ATOMIC); + if (mem == NULL) { + error = -ENOMEM; + goto mptsas_probe_failed; + } + + memset(mem, 0, sz); + hd->ScsiLookup = (struct scsi_cmnd **) mem; + + dprintk((MYIOC_s_INFO_FMT "ScsiLookup @ %p, sz=%d\n", + ioc->name, hd->ScsiLookup, sz)); + + /* Allocate memory for the device structures. + * A non-Null pointer at an offset + * indicates a device exists. + * max_id = 1 + maximum id (hosts.h) + */ + sz = sh->max_id * sizeof(void *); + mem = kmalloc(sz, GFP_ATOMIC); + if (mem == NULL) { + error = -ENOMEM; + goto mptsas_probe_failed; + } + + memset(mem, 0, sz); + hd->Targets = (VirtDevice **) mem; + + dprintk((KERN_INFO + " Targets @ %p, sz=%d\n", hd->Targets, sz)); + + /* Clear the TM flags + */ + hd->tmPending = 0; + hd->tmState = TM_STATE_NONE; + hd->resetPending = 0; + hd->abortSCpnt = NULL; + + /* Clear the pointer used to store + * single-threaded commands, i.e., those + * issued during a bus scan, dv and + * configuration pages. + */ + hd->cmdPtr = NULL; + + /* Initialize this SCSI Hosts' timers + * To use, set the timer expires field + * and add_timer + */ + init_timer(&hd->timer); + hd->timer.data = (unsigned long) hd; + hd->timer.function = mptscsih_timer_expired; + + hd->mpt_pq_filter = mpt_pq_filter; + ioc->sas_data.ptClear = mpt_pt_clear; + + if (ioc->sas_data.ptClear==1) { + mptbase_sas_persist_operation( + ioc, MPI_SAS_OP_CLEAR_ALL_PERSISTENT); + } + + ddvprintk((MYIOC_s_INFO_FMT + "mpt_pq_filter %x mpt_pq_filter %x\n", + ioc->name, + mpt_pq_filter, + mpt_pq_filter)); + + init_waitqueue_head(&hd->scandv_waitq); + hd->scandv_wait_done = 0; + hd->last_queue_full = 0; + + error = scsi_add_host(sh, &ioc->pcidev->dev); + if (error) { + dprintk((KERN_ERR MYNAM + "scsi_add_host failed\n")); + goto mptsas_probe_failed; + } + + mptsas_scan_sas_topology(ioc); + + return 0; + +mptsas_probe_failed: + + mptscsih_remove(pdev); + return error; +} + +static void __devexit mptsas_remove(struct pci_dev *pdev) +{ + MPT_ADAPTER *ioc = pci_get_drvdata(pdev); + struct mptsas_portinfo *p, *n; + + sas_remove_host(ioc->sh); + + list_for_each_entry_safe(p, n, &ioc->sas_topology, list) { + list_del(&p->list); + kfree(p); + } + + mptscsih_remove(pdev); +} + +static struct pci_device_id mptsas_pci_table[] = { + { PCI_VENDOR_ID_LSI_LOGIC, PCI_DEVICE_ID_LSI_SAS1064, + PCI_ANY_ID, PCI_ANY_ID }, + { PCI_VENDOR_ID_LSI_LOGIC, PCI_DEVICE_ID_LSI_SAS1066, + PCI_ANY_ID, PCI_ANY_ID }, + { PCI_VENDOR_ID_LSI_LOGIC, PCI_DEVICE_ID_LSI_SAS1068, + PCI_ANY_ID, PCI_ANY_ID }, + { PCI_VENDOR_ID_LSI_LOGIC, PCI_DEVICE_ID_LSI_SAS1064E, + PCI_ANY_ID, PCI_ANY_ID }, + { PCI_VENDOR_ID_LSI_LOGIC, PCI_DEVICE_ID_LSI_SAS1066E, + PCI_ANY_ID, PCI_ANY_ID }, + { PCI_VENDOR_ID_LSI_LOGIC, PCI_DEVICE_ID_LSI_SAS1068E, + PCI_ANY_ID, PCI_ANY_ID }, + {0} /* Terminating entry */ +}; +MODULE_DEVICE_TABLE(pci, mptsas_pci_table); + + +static struct pci_driver mptsas_driver = { + .name = "mptsas", + .id_table = mptsas_pci_table, + .probe = mptsas_probe, + .remove = __devexit_p(mptsas_remove), + .shutdown = mptscsih_shutdown, +#ifdef CONFIG_PM + .suspend = mptscsih_suspend, + .resume = mptscsih_resume, +#endif +}; + +static int __init +mptsas_init(void) +{ + show_mptmod_ver(my_NAME, my_VERSION); + + mptsas_transport_template = + sas_attach_transport(&mptsas_transport_functions); + if (!mptsas_transport_template) + return -ENODEV; + + mptsasDoneCtx = mpt_register(mptscsih_io_done, MPTSAS_DRIVER); + mptsasTaskCtx = mpt_register(mptscsih_taskmgmt_complete, MPTSAS_DRIVER); + mptsasInternalCtx = + mpt_register(mptscsih_scandv_complete, MPTSAS_DRIVER); + + if (mpt_event_register(mptsasDoneCtx, mptscsih_event_process) == 0) { + devtprintk((KERN_INFO MYNAM + ": Registered for IOC event notifications\n")); + } + + if (mpt_reset_register(mptsasDoneCtx, mptscsih_ioc_reset) == 0) { + dprintk((KERN_INFO MYNAM + ": Registered for IOC reset notifications\n")); + } + + return pci_register_driver(&mptsas_driver); +} + +static void __exit +mptsas_exit(void) +{ + pci_unregister_driver(&mptsas_driver); + sas_release_transport(mptsas_transport_template); + + mpt_reset_deregister(mptsasDoneCtx); + mpt_event_deregister(mptsasDoneCtx); + + mpt_deregister(mptsasInternalCtx); + mpt_deregister(mptsasTaskCtx); + mpt_deregister(mptsasDoneCtx); +} + +module_init(mptsas_init); +module_exit(mptsas_exit); From 466544d8898fc87ed6e2e62ac14af7c50ab7a1a4 Mon Sep 17 00:00:00 2001 From: "Moore, Eric Dean" Date: Wed, 14 Sep 2005 18:09:10 -0600 Subject: [PATCH 034/204] [SCSI] fusion SAS support (mptsas driver) updates Summary of Changes: * splitting mpt_interrupt per Christophs suggestion about a month ago * rename ScsiCfgData to SpiCfgData structure, then move all the raid related info into new structure called RaidCfgData. This is done because SAS supports RAID, as well as SPI, so the raid stuff should be seperate. * incorrect timeout calculation for cntdn inside WaitForDoorbellAck and WaitForDoortbellInt * add support for interpreting SAS Log Info * Increase Event Log Size from 0xA to 0x32 * Fix bug in mptsas/mptfc/mptspi - when controller has Initiator Mode Disabled, and only running in TargetMode, the mptctl would panic when loading. The fix is to return 0, instead of -ENODEV, in SCSI LLD respective probe routines * Fix bug in mptlan.c - driver will panic if there is host reset, due to dev being set to zero in mpt_lan_ioc_reset * Fix's for SPI - Echo Buffer * Several fix's in mptscsih_io_done - FCP Response info, RESIDUAL_MISMATCH, Data Underrun, etc. * Cleanup Error Handling - EH handlers, mptscsih_flush_cmds, and zeroing out ScsiLookup from mptscsih_qcmd * Cleanup asyn event handling from mptscsih -> mptscsih_event_process. Also added support for SAS Persistent Table Full, an asyn event Signed-off-by: Eric Moore Signed-off-by: James Bottomley --- drivers/message/fusion/mptbase.c | 449 +++++++++++++++++++----------- drivers/message/fusion/mptbase.h | 28 +- drivers/message/fusion/mptctl.c | 4 +- drivers/message/fusion/mptfc.c | 2 +- drivers/message/fusion/mptlan.c | 7 +- drivers/message/fusion/mptsas.c | 2 +- drivers/message/fusion/mptscsih.c | 447 +++++++++++++++-------------- drivers/message/fusion/mptscsih.h | 8 +- drivers/message/fusion/mptspi.c | 2 +- 9 files changed, 549 insertions(+), 400 deletions(-) diff --git a/drivers/message/fusion/mptbase.c b/drivers/message/fusion/mptbase.c index 14d62d96ca4..790a2932ded 100644 --- a/drivers/message/fusion/mptbase.c +++ b/drivers/message/fusion/mptbase.c @@ -135,7 +135,6 @@ static void mpt_adapter_dispose(MPT_ADAPTER *ioc); static void MptDisplayIocCapabilities(MPT_ADAPTER *ioc); static int MakeIocReady(MPT_ADAPTER *ioc, int force, int sleepFlag); -//static u32 mpt_GetIocState(MPT_ADAPTER *ioc, int cooked); static int GetIocFacts(MPT_ADAPTER *ioc, int sleepFlag, int reason); static int GetPortFacts(MPT_ADAPTER *ioc, int portnum, int sleepFlag); static int SendIocInit(MPT_ADAPTER *ioc, int sleepFlag); @@ -178,6 +177,7 @@ static int ProcessEventNotification(MPT_ADAPTER *ioc, EventNotificationReply_t * static void mpt_sp_ioc_info(MPT_ADAPTER *ioc, u32 ioc_status, MPT_FRAME_HDR *mf); static void mpt_fc_log_info(MPT_ADAPTER *ioc, u32 log_info); static void mpt_sp_log_info(MPT_ADAPTER *ioc, u32 log_info); +static void mpt_sas_log_info(MPT_ADAPTER *ioc, u32 log_info); /* module entry point */ static int __init fusion_init (void); @@ -209,6 +209,144 @@ pci_enable_io_access(struct pci_dev *pdev) pci_write_config_word(pdev, PCI_COMMAND, command_reg); } +/* + * Process turbo (context) reply... + */ +static void +mpt_turbo_reply(MPT_ADAPTER *ioc, u32 pa) +{ + MPT_FRAME_HDR *mf = NULL; + MPT_FRAME_HDR *mr = NULL; + int req_idx = 0; + int cb_idx; + + dmfprintk((MYIOC_s_INFO_FMT "Got TURBO reply req_idx=%08x\n", + ioc->name, pa)); + + switch (pa >> MPI_CONTEXT_REPLY_TYPE_SHIFT) { + case MPI_CONTEXT_REPLY_TYPE_SCSI_INIT: + req_idx = pa & 0x0000FFFF; + cb_idx = (pa & 0x00FF0000) >> 16; + mf = MPT_INDEX_2_MFPTR(ioc, req_idx); + break; + case MPI_CONTEXT_REPLY_TYPE_LAN: + cb_idx = mpt_lan_index; + /* + * Blind set of mf to NULL here was fatal + * after lan_reply says "freeme" + * Fix sort of combined with an optimization here; + * added explicit check for case where lan_reply + * was just returning 1 and doing nothing else. + * For this case skip the callback, but set up + * proper mf value first here:-) + */ + if ((pa & 0x58000000) == 0x58000000) { + req_idx = pa & 0x0000FFFF; + mf = MPT_INDEX_2_MFPTR(ioc, req_idx); + mpt_free_msg_frame(ioc, mf); + mb(); + return; + break; + } + mr = (MPT_FRAME_HDR *) CAST_U32_TO_PTR(pa); + break; + case MPI_CONTEXT_REPLY_TYPE_SCSI_TARGET: + cb_idx = mpt_stm_index; + mr = (MPT_FRAME_HDR *) CAST_U32_TO_PTR(pa); + break; + default: + cb_idx = 0; + BUG(); + } + + /* Check for (valid) IO callback! */ + if (cb_idx < 1 || cb_idx >= MPT_MAX_PROTOCOL_DRIVERS || + MptCallbacks[cb_idx] == NULL) { + printk(MYIOC_s_WARN_FMT "%s: Invalid cb_idx (%d)!\n", + __FUNCTION__, ioc->name, cb_idx); + goto out; + } + + if (MptCallbacks[cb_idx](ioc, mf, mr)) + mpt_free_msg_frame(ioc, mf); + out: + mb(); +} + +static void +mpt_reply(MPT_ADAPTER *ioc, u32 pa) +{ + MPT_FRAME_HDR *mf; + MPT_FRAME_HDR *mr; + int req_idx; + int cb_idx; + int freeme; + + u32 reply_dma_low; + u16 ioc_stat; + + /* non-TURBO reply! Hmmm, something may be up... + * Newest turbo reply mechanism; get address + * via left shift 1 (get rid of MPI_ADDRESS_REPLY_A_BIT)! + */ + + /* Map DMA address of reply header to cpu address. + * pa is 32 bits - but the dma address may be 32 or 64 bits + * get offset based only only the low addresses + */ + + reply_dma_low = (pa <<= 1); + mr = (MPT_FRAME_HDR *)((u8 *)ioc->reply_frames + + (reply_dma_low - ioc->reply_frames_low_dma)); + + req_idx = le16_to_cpu(mr->u.frame.hwhdr.msgctxu.fld.req_idx); + cb_idx = mr->u.frame.hwhdr.msgctxu.fld.cb_idx; + mf = MPT_INDEX_2_MFPTR(ioc, req_idx); + + dmfprintk((MYIOC_s_INFO_FMT "Got non-TURBO reply=%p req_idx=%x cb_idx=%x Function=%x\n", + ioc->name, mr, req_idx, cb_idx, mr->u.hdr.Function)); + DBG_DUMP_REPLY_FRAME(mr) + + /* Check/log IOC log info + */ + ioc_stat = le16_to_cpu(mr->u.reply.IOCStatus); + if (ioc_stat & MPI_IOCSTATUS_FLAG_LOG_INFO_AVAILABLE) { + u32 log_info = le32_to_cpu(mr->u.reply.IOCLogInfo); + if (ioc->bus_type == FC) + mpt_fc_log_info(ioc, log_info); + else if (ioc->bus_type == SCSI) + mpt_sp_log_info(ioc, log_info); + else if (ioc->bus_type == SAS) + mpt_sas_log_info(ioc, log_info); + } + if (ioc_stat & MPI_IOCSTATUS_MASK) { + if (ioc->bus_type == SCSI && + cb_idx != mpt_stm_index && + cb_idx != mpt_lan_index) + mpt_sp_ioc_info(ioc, (u32)ioc_stat, mf); + } + + + /* Check for (valid) IO callback! */ + if (cb_idx < 1 || cb_idx >= MPT_MAX_PROTOCOL_DRIVERS || + MptCallbacks[cb_idx] == NULL) { + printk(MYIOC_s_WARN_FMT "%s: Invalid cb_idx (%d)!\n", + __FUNCTION__, ioc->name, cb_idx); + freeme = 0; + goto out; + } + + freeme = MptCallbacks[cb_idx](ioc, mf, mr); + + out: + /* Flush (non-TURBO) reply with a WRITE! */ + CHIPREG_WRITE32(&ioc->chip->ReplyFifo, pa); + + if (freeme) + mpt_free_msg_frame(ioc, mf); + mb(); +} + /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ /* * mpt_interrupt - MPT adapter (IOC) specific interrupt handler. @@ -230,164 +368,21 @@ pci_enable_io_access(struct pci_dev *pdev) static irqreturn_t mpt_interrupt(int irq, void *bus_id, struct pt_regs *r) { - MPT_ADAPTER *ioc; - MPT_FRAME_HDR *mf; - MPT_FRAME_HDR *mr; - u32 pa; - int req_idx; - int cb_idx; - int type; - int freeme; - - ioc = (MPT_ADAPTER *)bus_id; + MPT_ADAPTER *ioc = bus_id; + u32 pa; /* * Drain the reply FIFO! - * - * NOTES: I've seen up to 10 replies processed in this loop, so far... - * Update: I've seen up to 9182 replies processed in this loop! ?? - * Update: Limit ourselves to processing max of N replies - * (bottom of loop). */ while (1) { - - if ((pa = CHIPREG_READ32_dmasync(&ioc->chip->ReplyFifo)) == 0xFFFFFFFF) + pa = CHIPREG_READ32_dmasync(&ioc->chip->ReplyFifo); + if (pa == 0xFFFFFFFF) return IRQ_HANDLED; - - cb_idx = 0; - freeme = 0; - - /* - * Check for non-TURBO reply! - */ - if (pa & MPI_ADDRESS_REPLY_A_BIT) { - u32 reply_dma_low; - u16 ioc_stat; - - /* non-TURBO reply! Hmmm, something may be up... - * Newest turbo reply mechanism; get address - * via left shift 1 (get rid of MPI_ADDRESS_REPLY_A_BIT)! - */ - - /* Map DMA address of reply header to cpu address. - * pa is 32 bits - but the dma address may be 32 or 64 bits - * get offset based only only the low addresses - */ - reply_dma_low = (pa = (pa << 1)); - mr = (MPT_FRAME_HDR *)((u8 *)ioc->reply_frames + - (reply_dma_low - ioc->reply_frames_low_dma)); - - req_idx = le16_to_cpu(mr->u.frame.hwhdr.msgctxu.fld.req_idx); - cb_idx = mr->u.frame.hwhdr.msgctxu.fld.cb_idx; - mf = MPT_INDEX_2_MFPTR(ioc, req_idx); - - dmfprintk((MYIOC_s_INFO_FMT "Got non-TURBO reply=%p req_idx=%x cb_idx=%x Function=%x\n", - ioc->name, mr, req_idx, cb_idx, mr->u.hdr.Function)); - DBG_DUMP_REPLY_FRAME(mr) - - /* Check/log IOC log info - */ - ioc_stat = le16_to_cpu(mr->u.reply.IOCStatus); - if (ioc_stat & MPI_IOCSTATUS_FLAG_LOG_INFO_AVAILABLE) { - u32 log_info = le32_to_cpu(mr->u.reply.IOCLogInfo); - if (ioc->bus_type == FC) - mpt_fc_log_info(ioc, log_info); - else if (ioc->bus_type == SCSI) - mpt_sp_log_info(ioc, log_info); - } - if (ioc_stat & MPI_IOCSTATUS_MASK) { - if (ioc->bus_type == SCSI) - mpt_sp_ioc_info(ioc, (u32)ioc_stat, mf); - } - } else { - /* - * Process turbo (context) reply... - */ - dmfprintk((MYIOC_s_INFO_FMT "Got TURBO reply req_idx=%08x\n", ioc->name, pa)); - type = (pa >> MPI_CONTEXT_REPLY_TYPE_SHIFT); - if (type == MPI_CONTEXT_REPLY_TYPE_SCSI_TARGET) { - cb_idx = mpt_stm_index; - mf = NULL; - mr = (MPT_FRAME_HDR *) CAST_U32_TO_PTR(pa); - } else if (type == MPI_CONTEXT_REPLY_TYPE_LAN) { - cb_idx = mpt_lan_index; - /* Blind set of mf to NULL here was fatal - * after lan_reply says "freeme" - * Fix sort of combined with an optimization here; - * added explicit check for case where lan_reply - * was just returning 1 and doing nothing else. - * For this case skip the callback, but set up - * proper mf value first here:-) - */ - if ((pa & 0x58000000) == 0x58000000) { - req_idx = pa & 0x0000FFFF; - mf = MPT_INDEX_2_MFPTR(ioc, req_idx); - freeme = 1; - /* - * IMPORTANT! Invalidate the callback! - */ - cb_idx = 0; - } else { - mf = NULL; - } - mr = (MPT_FRAME_HDR *) CAST_U32_TO_PTR(pa); - } else { - req_idx = pa & 0x0000FFFF; - cb_idx = (pa & 0x00FF0000) >> 16; - mf = MPT_INDEX_2_MFPTR(ioc, req_idx); - mr = NULL; - } - pa = 0; /* No reply flush! */ - } - -#ifdef MPT_DEBUG_IRQ - if (ioc->bus_type == SCSI) { - /* Verify mf, mr are reasonable. - */ - if ((mf) && ((mf >= MPT_INDEX_2_MFPTR(ioc, ioc->req_depth)) - || (mf < ioc->req_frames)) ) { - printk(MYIOC_s_WARN_FMT - "mpt_interrupt: Invalid mf (%p)!\n", ioc->name, (void *)mf); - cb_idx = 0; - pa = 0; - freeme = 0; - } - if ((pa) && (mr) && ((mr >= MPT_INDEX_2_RFPTR(ioc, ioc->req_depth)) - || (mr < ioc->reply_frames)) ) { - printk(MYIOC_s_WARN_FMT - "mpt_interrupt: Invalid rf (%p)!\n", ioc->name, (void *)mr); - cb_idx = 0; - pa = 0; - freeme = 0; - } - if (cb_idx > (MPT_MAX_PROTOCOL_DRIVERS-1)) { - printk(MYIOC_s_WARN_FMT - "mpt_interrupt: Invalid cb_idx (%d)!\n", ioc->name, cb_idx); - cb_idx = 0; - pa = 0; - freeme = 0; - } - } -#endif - - /* Check for (valid) IO callback! */ - if (cb_idx) { - /* Do the callback! */ - freeme = (*(MptCallbacks[cb_idx]))(ioc, mf, mr); - } - - if (pa) { - /* Flush (non-TURBO) reply with a WRITE! */ - CHIPREG_WRITE32(&ioc->chip->ReplyFifo, pa); - } - - if (freeme) { - /* Put Request back on FreeQ! */ - mpt_free_msg_frame(ioc, mf); - } - - mb(); - } /* drain reply FIFO */ + else if (pa & MPI_ADDRESS_REPLY_A_BIT) + mpt_reply(ioc, pa); + else + mpt_turbo_reply(ioc, pa); + } return IRQ_HANDLED; } @@ -1065,7 +1060,7 @@ mpt_host_page_alloc(MPT_ADAPTER *ioc, pIOCInit_t ioc_init) ioc->name, ioc->HostPageBuffer, ioc->HostPageBuffer_dma, - hst_page_buffer_sz)); + host_page_buffer_sz)); ioc->alloc_total += host_page_buffer_sz; ioc->HostPageBuffer_sz = host_page_buffer_sz; break; @@ -1208,7 +1203,7 @@ mpt_attach(struct pci_dev *pdev, const struct pci_device_id *id) /* Initilize SCSI Config Data structure */ - memset(&ioc->spi_data, 0, sizeof(ScsiCfgData)); + memset(&ioc->spi_data, 0, sizeof(SpiCfgData)); /* Initialize the running configQ head. */ @@ -1755,8 +1750,23 @@ mpt_do_ioc_recovery(MPT_ADAPTER *ioc, u32 reason, int sleepFlag) */ if (ret == 0) { rc = mpt_do_upload(ioc, sleepFlag); - if (rc != 0) + if (rc == 0) { + if (ioc->alt_ioc && ioc->alt_ioc->cached_fw) { + /* + * Maintain only one pointer to FW memory + * so there will not be two attempt to + * downloadboot onboard dual function + * chips (mpt_adapter_disable, + * mpt_diag_reset) + */ + ioc->cached_fw = NULL; + ddlprintk((MYIOC_s_INFO_FMT ": mpt_upload: alt_%s has cached_fw=%p \n", + ioc->name, ioc->alt_ioc->name, ioc->alt_ioc->cached_fw)); + } + } else { printk(KERN_WARNING MYNAM ": firmware upload failure!\n"); + ret = -5; + } } } } @@ -1997,9 +2007,9 @@ mpt_adapter_disable(MPT_ADAPTER *ioc) } kfree(ioc->spi_data.nvram); - kfree(ioc->spi_data.pIocPg3); + kfree(ioc->raid_data.pIocPg3); ioc->spi_data.nvram = NULL; - ioc->spi_data.pIocPg3 = NULL; + ioc->raid_data.pIocPg3 = NULL; if (ioc->spi_data.pIocPg4 != NULL) { sz = ioc->spi_data.IocPg4Sz; @@ -3852,7 +3862,7 @@ WaitForDoorbellAck(MPT_ADAPTER *ioc, int howlong, int sleepFlag) int count = 0; u32 intstat=0; - cntdn = ((sleepFlag == CAN_SLEEP) ? HZ : 1000) * howlong; + cntdn = 1000 * howlong; if (sleepFlag == CAN_SLEEP) { while (--cntdn) { @@ -3902,7 +3912,7 @@ WaitForDoorbellInt(MPT_ADAPTER *ioc, int howlong, int sleepFlag) int count = 0; u32 intstat=0; - cntdn = ((sleepFlag == CAN_SLEEP) ? HZ : 1000) * howlong; + cntdn = 1000 * howlong; if (sleepFlag == CAN_SLEEP) { while (--cntdn) { intstat = CHIPREG_READ32(&ioc->chip->IntStatus); @@ -4634,10 +4644,10 @@ mpt_findImVolumes(MPT_ADAPTER *ioc) if (mpt_config(ioc, &cfg) != 0) goto done_and_free; - if ( (mem = (u8 *)ioc->spi_data.pIocPg2) == NULL ) { + if ( (mem = (u8 *)ioc->raid_data.pIocPg2) == NULL ) { mem = kmalloc(iocpage2sz, GFP_ATOMIC); if (mem) { - ioc->spi_data.pIocPg2 = (IOCPage2_t *) mem; + ioc->raid_data.pIocPg2 = (IOCPage2_t *) mem; } else { goto done_and_free; } @@ -4654,7 +4664,7 @@ mpt_findImVolumes(MPT_ADAPTER *ioc) /* At least 1 RAID Volume */ pIocRv = pIoc2->RaidVolume; - ioc->spi_data.isRaid = 0; + ioc->raid_data.isRaid = 0; for (jj = 0; jj < nVols; jj++, pIocRv++) { vid = pIocRv->VolumeID; vbus = pIocRv->VolumeBus; @@ -4663,7 +4673,7 @@ mpt_findImVolumes(MPT_ADAPTER *ioc) /* find the match */ if (vbus == 0) { - ioc->spi_data.isRaid |= (1 << vid); + ioc->raid_data.isRaid |= (1 << vid); } else { /* Error! Always bus 0 */ @@ -4698,8 +4708,8 @@ mpt_read_ioc_pg_3(MPT_ADAPTER *ioc) /* Free the old page */ - kfree(ioc->spi_data.pIocPg3); - ioc->spi_data.pIocPg3 = NULL; + kfree(ioc->raid_data.pIocPg3); + ioc->raid_data.pIocPg3 = NULL; /* There is at least one physical disk. * Read and save IOC Page 3 @@ -4736,7 +4746,7 @@ mpt_read_ioc_pg_3(MPT_ADAPTER *ioc) mem = kmalloc(iocpage3sz, GFP_ATOMIC); if (mem) { memcpy(mem, (u8 *)pIoc3, iocpage3sz); - ioc->spi_data.pIocPg3 = (IOCPage3_t *) mem; + ioc->raid_data.pIocPg3 = (IOCPage3_t *) mem; } } @@ -6022,6 +6032,111 @@ mpt_sp_log_info(MPT_ADAPTER *ioc, u32 log_info) printk(MYIOC_s_INFO_FMT "LogInfo(0x%08x): F/W: %s\n", ioc->name, log_info, desc); } +/* strings for sas loginfo */ + static char *originator_str[] = { + "IOP", /* 00h */ + "PL", /* 01h */ + "IR" /* 02h */ + }; + static char *iop_code_str[] = { + NULL, /* 00h */ + "Invalid SAS Address", /* 01h */ + NULL, /* 02h */ + "Invalid Page", /* 03h */ + NULL, /* 04h */ + "Task Terminated" /* 05h */ + }; + static char *pl_code_str[] = { + NULL, /* 00h */ + "Open Failure", /* 01h */ + "Invalid Scatter Gather List", /* 02h */ + "Wrong Relative Offset or Frame Length", /* 03h */ + "Frame Transfer Error", /* 04h */ + "Transmit Frame Connected Low", /* 05h */ + "SATA Non-NCQ RW Error Bit Set", /* 06h */ + "SATA Read Log Receive Data Error", /* 07h */ + "SATA NCQ Fail All Commands After Error", /* 08h */ + "SATA Error in Receive Set Device Bit FIS", /* 09h */ + "Receive Frame Invalid Message", /* 0Ah */ + "Receive Context Message Valid Error", /* 0Bh */ + "Receive Frame Current Frame Error", /* 0Ch */ + "SATA Link Down", /* 0Dh */ + "Discovery SATA Init W IOS", /* 0Eh */ + "Config Invalid Page", /* 0Fh */ + "Discovery SATA Init Timeout", /* 10h */ + "Reset", /* 11h */ + "Abort", /* 12h */ + "IO Not Yet Executed", /* 13h */ + "IO Executed", /* 14h */ + NULL, /* 15h */ + NULL, /* 16h */ + NULL, /* 17h */ + NULL, /* 18h */ + NULL, /* 19h */ + NULL, /* 1Ah */ + NULL, /* 1Bh */ + NULL, /* 1Ch */ + NULL, /* 1Dh */ + NULL, /* 1Eh */ + NULL, /* 1Fh */ + "Enclosure Management" /* 20h */ + }; + +/*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ +/* + * mpt_sas_log_info - Log information returned from SAS IOC. + * @ioc: Pointer to MPT_ADAPTER structure + * @log_info: U32 LogInfo reply word from the IOC + * + * Refer to lsi/mpi_log_sas.h. + */ +static void +mpt_sas_log_info(MPT_ADAPTER *ioc, u32 log_info) +{ +union loginfo_type { + u32 loginfo; + struct { + u32 subcode:16; + u32 code:8; + u32 originator:4; + u32 bus_type:4; + }dw; +}; + union loginfo_type sas_loginfo; + char *code_desc = NULL; + + sas_loginfo.loginfo = log_info; + if ((sas_loginfo.dw.bus_type != 3 /*SAS*/) && + (sas_loginfo.dw.originator < sizeof(originator_str)/sizeof(char*))) + return; + if ((sas_loginfo.dw.originator == 0 /*IOP*/) && + (sas_loginfo.dw.code < sizeof(iop_code_str)/sizeof(char*))) { + code_desc = iop_code_str[sas_loginfo.dw.code]; + }else if ((sas_loginfo.dw.originator == 1 /*PL*/) && + (sas_loginfo.dw.code < sizeof(pl_code_str)/sizeof(char*) )) { + code_desc = pl_code_str[sas_loginfo.dw.code]; + } + + if (code_desc != NULL) + printk(MYIOC_s_INFO_FMT + "LogInfo(0x%08x): Originator={%s}, Code={%s}," + " SubCode(0x%04x)\n", + ioc->name, + log_info, + originator_str[sas_loginfo.dw.originator], + code_desc, + sas_loginfo.dw.subcode); + else + printk(MYIOC_s_INFO_FMT + "LogInfo(0x%08x): Originator={%s}, Code=(0x%02x)," + " SubCode(0x%04x)\n", + ioc->name, + log_info, + originator_str[sas_loginfo.dw.originator], + sas_loginfo.dw.code, + sas_loginfo.dw.subcode); +} + /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ /* * mpt_sp_ioc_info - IOC information returned from SCSI Parallel IOC. diff --git a/drivers/message/fusion/mptbase.h b/drivers/message/fusion/mptbase.h index bbd21d74ce5..75105277e22 100644 --- a/drivers/message/fusion/mptbase.h +++ b/drivers/message/fusion/mptbase.h @@ -77,8 +77,8 @@ #define COPYRIGHT "Copyright (c) 1999-2005 " MODULEAUTHOR #endif -#define MPT_LINUX_VERSION_COMMON "3.03.02" -#define MPT_LINUX_PACKAGE_NAME "@(#)mptlinux-3.03.02" +#define MPT_LINUX_VERSION_COMMON "3.03.03" +#define MPT_LINUX_PACKAGE_NAME "@(#)mptlinux-3.03.03" #define WHAT_MAGIC_STRING "@" "(" "#" ")" #define show_mptmod_ver(s,ver) \ @@ -424,7 +424,7 @@ typedef struct _MPT_IOCTL { /* * Event Structure and define */ -#define MPTCTL_EVENT_LOG_SIZE (0x0000000A) +#define MPTCTL_EVENT_LOG_SIZE (0x000000032) typedef struct _mpt_ioctl_events { u32 event; /* Specified by define above */ u32 eventContext; /* Index or counter */ @@ -452,16 +452,13 @@ typedef struct _mpt_ioctl_events { #define MPT_SCSICFG_ALL_IDS 0x02 /* WriteSDP1 to all IDS */ /* #define MPT_SCSICFG_BLK_NEGO 0x10 WriteSDP1 with WDTR and SDTR disabled */ -typedef struct _ScsiCfgData { +typedef struct _SpiCfgData { u32 PortFlags; int *nvram; /* table of device NVRAM values */ - IOCPage2_t *pIocPg2; /* table of Raid Volumes */ - IOCPage3_t *pIocPg3; /* table of physical disks */ IOCPage4_t *pIocPg4; /* SEP devices addressing */ dma_addr_t IocPg4_dma; /* Phys Addr of IOCPage4 data */ int IocPg4Sz; /* IOCPage4 size */ u8 dvStatus[MPT_MAX_SCSI_DEVICES]; - int isRaid; /* bit field, 1 if RAID */ u8 minSyncFactor; /* 0xFF if async */ u8 maxSyncOffset; /* 0 if async */ u8 maxBusWidth; /* 0 if narrow, 1 if wide */ @@ -473,10 +470,14 @@ typedef struct _ScsiCfgData { u8 dvScheduled; /* 1 if scheduled */ u8 forceDv; /* 1 to force DV scheduling */ u8 noQas; /* Disable QAS for this adapter */ - u8 Saf_Te; /* 1 to force all Processors as SAF-TE if Inquiry data length is too short to check for SAF-TE */ + u8 Saf_Te; /* 1 to force all Processors as + * SAF-TE if Inquiry data length + * is too short to check for SAF-TE + */ u8 mpt_dv; /* command line option: enhanced=1, basic=0 */ + u8 bus_reset; /* 1 to allow bus reset */ u8 rsvd[1]; -} ScsiCfgData; +}SpiCfgData; typedef struct _SasCfgData { u8 ptClear; /* 1 to automatically clear the @@ -486,6 +487,12 @@ typedef struct _SasCfgData { */ }SasCfgData; +typedef struct _RaidCfgData { + IOCPage2_t *pIocPg2; /* table of Raid Volumes */ + IOCPage3_t *pIocPg3; /* table of physical disks */ + int isRaid; /* bit field, 1 if RAID */ +}RaidCfgData; + /* * Adapter Structure - pci_dev specific. Maximum: MPT_MAX_ADAPTERS */ @@ -546,7 +553,8 @@ typedef struct _MPT_ADAPTER struct pci_dev *pcidev; /* struct pci_dev pointer */ u8 __iomem *memmap; /* mmap address */ struct Scsi_Host *sh; /* Scsi Host pointer */ - ScsiCfgData spi_data; /* Scsi config. data */ + SpiCfgData spi_data; /* Scsi config. data */ + RaidCfgData raid_data; /* Raid config. data */ SasCfgData sas_data; /* Sas config. data */ MPT_IOCTL *ioctl; /* ioctl data pointer */ struct proc_dir_entry *ioc_dentry; diff --git a/drivers/message/fusion/mptctl.c b/drivers/message/fusion/mptctl.c index 7577c2417e2..cb2d59d5f5a 100644 --- a/drivers/message/fusion/mptctl.c +++ b/drivers/message/fusion/mptctl.c @@ -1326,7 +1326,7 @@ mptctl_gettargetinfo (unsigned long arg) */ if (hd && hd->Targets) { mpt_findImVolumes(ioc); - pIoc2 = ioc->spi_data.pIocPg2; + pIoc2 = ioc->raid_data.pIocPg2; for ( id = 0; id <= max_id; ) { if ( pIoc2 && pIoc2->NumActiveVolumes ) { if ( id == pIoc2->RaidVolume[0].VolumeID ) { @@ -1348,7 +1348,7 @@ mptctl_gettargetinfo (unsigned long arg) --maxWordsLeft; goto next_id; } else { - pIoc3 = ioc->spi_data.pIocPg3; + pIoc3 = ioc->raid_data.pIocPg3; for ( jj = 0; jj < pIoc3->NumPhysDisks; jj++ ) { if ( pIoc3->PhysDisk[jj].PhysDiskID == id ) goto next_id; diff --git a/drivers/message/fusion/mptfc.c b/drivers/message/fusion/mptfc.c index 13771abea13..a628be9bbba 100644 --- a/drivers/message/fusion/mptfc.c +++ b/drivers/message/fusion/mptfc.c @@ -189,7 +189,7 @@ mptfc_probe(struct pci_dev *pdev, const struct pci_device_id *id) printk(MYIOC_s_WARN_FMT "Skipping ioc=%p because SCSI Initiator mode is NOT enabled!\n", ioc->name, ioc); - return -ENODEV; + return 0; } sh = scsi_host_alloc(&mptfc_driver_template, sizeof(MPT_SCSI_HOST)); diff --git a/drivers/message/fusion/mptlan.c b/drivers/message/fusion/mptlan.c index 52794be5a95..ed3c891e388 100644 --- a/drivers/message/fusion/mptlan.c +++ b/drivers/message/fusion/mptlan.c @@ -312,7 +312,12 @@ static int mpt_lan_ioc_reset(MPT_ADAPTER *ioc, int reset_phase) { struct net_device *dev = ioc->netdev; - struct mpt_lan_priv *priv = netdev_priv(dev); + struct mpt_lan_priv *priv; + + if (dev == NULL) + return(1); + else + priv = netdev_priv(dev); dlprintk((KERN_INFO MYNAM ": IOC %s_reset routed to LAN driver!\n", reset_phase==MPT_IOC_SETUP_RESET ? "setup" : ( diff --git a/drivers/message/fusion/mptsas.c b/drivers/message/fusion/mptsas.c index 0d9a192e1bd..429820e48c6 100644 --- a/drivers/message/fusion/mptsas.c +++ b/drivers/message/fusion/mptsas.c @@ -980,7 +980,7 @@ mptsas_probe(struct pci_dev *pdev, const struct pci_device_id *id) printk(MYIOC_s_WARN_FMT "Skipping ioc=%p because SCSI Initiator mode " "is NOT enabled!\n", ioc->name, ioc); - return -ENODEV; + return 0; } sh = scsi_host_alloc(&mptsas_driver_template, sizeof(MPT_SCSI_HOST)); diff --git a/drivers/message/fusion/mptscsih.c b/drivers/message/fusion/mptscsih.c index 58b5fdee009..8dd25aac535 100644 --- a/drivers/message/fusion/mptscsih.c +++ b/drivers/message/fusion/mptscsih.c @@ -93,8 +93,9 @@ typedef struct _BIG_SENSE_BUF { #define MPT_ICFLAG_BUF_CAP 0x01 /* ReadBuffer Read Capacity format */ #define MPT_ICFLAG_ECHO 0x02 /* ReadBuffer Echo buffer format */ -#define MPT_ICFLAG_PHYS_DISK 0x04 /* Any SCSI IO but do Phys Disk Format */ -#define MPT_ICFLAG_TAGGED_CMD 0x08 /* Do tagged IO */ +#define MPT_ICFLAG_EBOS 0x04 /* ReadBuffer Echo buffer has EBOS */ +#define MPT_ICFLAG_PHYS_DISK 0x08 /* Any SCSI IO but do Phys Disk Format */ +#define MPT_ICFLAG_TAGGED_CMD 0x10 /* Do tagged IO */ #define MPT_ICFLAG_DID_RESET 0x20 /* Bus Reset occurred with this command */ #define MPT_ICFLAG_RESERVED 0x40 /* Reserved has been issued */ @@ -159,6 +160,8 @@ int mptscsih_scandv_complete(MPT_ADAPTER *ioc, MPT_FRAME_HDR *mf, MPT_FRAME_HDR static int mptscsih_do_cmd(MPT_SCSI_HOST *hd, INTERNAL_CMD *iocmd); static int mptscsih_synchronize_cache(MPT_SCSI_HOST *hd, int portnum); +static struct work_struct mptscsih_persistTask; + #ifdef MPTSCSIH_ENABLE_DOMAIN_VALIDATION static int mptscsih_do_raid(MPT_SCSI_HOST *hd, u8 action, INTERNAL_CMD *io); static void mptscsih_domainValidation(void *hd); @@ -167,6 +170,7 @@ static void mptscsih_qas_check(MPT_SCSI_HOST *hd, int id); static int mptscsih_doDv(MPT_SCSI_HOST *hd, int channel, int target); static void mptscsih_dv_parms(MPT_SCSI_HOST *hd, DVPARAMETERS *dv,void *pPage); static void mptscsih_fillbuf(char *buffer, int size, int index, int width); +static void mptscsih_set_dvflags_raid(MPT_SCSI_HOST *hd, int id); #endif void mptscsih_remove(struct pci_dev *); @@ -606,11 +610,24 @@ mptscsih_io_done(MPT_ADAPTER *ioc, MPT_FRAME_HDR *mf, MPT_FRAME_HDR *mr) xfer_cnt = le32_to_cpu(pScsiReply->TransferCount); sc->resid = sc->request_bufflen - xfer_cnt; + /* + * if we get a data underrun indication, yet no data was + * transferred and the SCSI status indicates that the + * command was never started, change the data underrun + * to success + */ + if (status == MPI_IOCSTATUS_SCSI_DATA_UNDERRUN && xfer_cnt == 0 && + (scsi_status == MPI_SCSI_STATUS_BUSY || + scsi_status == MPI_SCSI_STATUS_RESERVATION_CONFLICT || + scsi_status == MPI_SCSI_STATUS_TASK_SET_FULL)) { + status = MPI_IOCSTATUS_SUCCESS; + } + dreplyprintk((KERN_NOTICE "Reply ha=%d id=%d lun=%d:\n" "IOCStatus=%04xh SCSIState=%02xh SCSIStatus=%02xh\n" "resid=%d bufflen=%d xfer_cnt=%d\n", ioc->id, pScsiReq->TargetID, pScsiReq->LUN[1], - status, scsi_state, scsi_status, sc->resid, + status, scsi_state, scsi_status, sc->resid, sc->request_bufflen, xfer_cnt)); if (scsi_state & MPI_SCSI_STATE_AUTOSENSE_VALID) @@ -619,8 +636,11 @@ mptscsih_io_done(MPT_ADAPTER *ioc, MPT_FRAME_HDR *mf, MPT_FRAME_HDR *mr) /* * Look for + dump FCP ResponseInfo[]! */ - if (scsi_state & MPI_SCSI_STATE_RESPONSE_INFO_VALID) { - printk(KERN_NOTICE " FCP_ResponseInfo=%08xh\n", + if (scsi_state & MPI_SCSI_STATE_RESPONSE_INFO_VALID && + pScsiReply->ResponseInfo) { + printk(KERN_NOTICE "ha=%d id=%d lun=%d: " + "FCP_ResponseInfo=%08xh\n", + ioc->id, pScsiReq->TargetID, pScsiReq->LUN[1], le32_to_cpu(pScsiReply->ResponseInfo)); } @@ -661,23 +681,13 @@ mptscsih_io_done(MPT_ADAPTER *ioc, MPT_FRAME_HDR *mf, MPT_FRAME_HDR *mr) break; case MPI_IOCSTATUS_SCSI_RESIDUAL_MISMATCH: /* 0x0049 */ - if ( xfer_cnt >= sc->underflow ) { - /* Sufficient data transfer occurred */ + sc->resid = sc->request_bufflen - xfer_cnt; + if((xfer_cnt==0)||(sc->underflow > xfer_cnt)) + sc->result=DID_SOFT_ERROR << 16; + else /* Sufficient data transfer occurred */ sc->result = (DID_OK << 16) | scsi_status; - } else if ( xfer_cnt == 0 ) { - /* A CRC Error causes this condition; retry */ - sc->result = (DRIVER_SENSE << 24) | (DID_OK << 16) | - (CHECK_CONDITION << 1); - sc->sense_buffer[0] = 0x70; - sc->sense_buffer[2] = NO_SENSE; - sc->sense_buffer[12] = 0; - sc->sense_buffer[13] = 0; - } else { - sc->result = DID_SOFT_ERROR << 16; - } - dreplyprintk((KERN_NOTICE - "RESIDUAL_MISMATCH: result=%x on id=%d\n", - sc->result, sc->device->id)); + dreplyprintk((KERN_NOTICE + "RESIDUAL_MISMATCH: result=%x on id=%d\n", sc->result, sc->device->id)); break; case MPI_IOCSTATUS_SCSI_DATA_UNDERRUN: /* 0x0045 */ @@ -692,7 +702,10 @@ mptscsih_io_done(MPT_ADAPTER *ioc, MPT_FRAME_HDR *mf, MPT_FRAME_HDR *mr) ; } else { if (xfer_cnt < sc->underflow) { - sc->result = DID_SOFT_ERROR << 16; + if (scsi_status == SAM_STAT_BUSY) + sc->result = SAM_STAT_BUSY; + else + sc->result = DID_SOFT_ERROR << 16; } if (scsi_state & (MPI_SCSI_STATE_AUTOSENSE_FAILED | MPI_SCSI_STATE_NO_SCSI_STATUS)) { /* What to do? @@ -717,8 +730,10 @@ mptscsih_io_done(MPT_ADAPTER *ioc, MPT_FRAME_HDR *mf, MPT_FRAME_HDR *mr) case MPI_IOCSTATUS_SCSI_RECOVERED_ERROR: /* 0x0040 */ case MPI_IOCSTATUS_SUCCESS: /* 0x0000 */ - scsi_status = pScsiReply->SCSIStatus; - sc->result = (DID_OK << 16) | scsi_status; + if (scsi_status == MPI_SCSI_STATUS_BUSY) + sc->result = (DID_BUS_BUSY << 16) | scsi_status; + else + sc->result = (DID_OK << 16) | scsi_status; if (scsi_state == 0) { ; } else if (scsi_state & MPI_SCSI_STATE_AUTOSENSE_VALID) { @@ -890,12 +905,13 @@ mptscsih_search_running_cmds(MPT_SCSI_HOST *hd, uint target, uint lun) SCSIIORequest_t *mf = NULL; int ii; int max = hd->ioc->req_depth; + struct scsi_cmnd *sc; dsprintk((KERN_INFO MYNAM ": search_running target %d lun %d max %d\n", target, lun, max)); for (ii=0; ii < max; ii++) { - if (hd->ScsiLookup[ii] != NULL) { + if ((sc = hd->ScsiLookup[ii]) != NULL) { mf = (SCSIIORequest_t *)MPT_INDEX_2_MFPTR(hd->ioc, ii); @@ -910,9 +926,22 @@ mptscsih_search_running_cmds(MPT_SCSI_HOST *hd, uint target, uint lun) hd->ScsiLookup[ii] = NULL; mptscsih_freeChainBuffers(hd->ioc, ii); mpt_free_msg_frame(hd->ioc, (MPT_FRAME_HDR *)mf); + if (sc->use_sg) { + pci_unmap_sg(hd->ioc->pcidev, + (struct scatterlist *) sc->request_buffer, + sc->use_sg, + sc->sc_data_direction); + } else if (sc->request_bufflen) { + pci_unmap_single(hd->ioc->pcidev, + sc->SCp.dma_handle, + sc->request_bufflen, + sc->sc_data_direction); + } + sc->host_scribble = NULL; + sc->result = DID_NO_CONNECT << 16; + sc->scsi_done(sc); } } - return; } @@ -967,8 +996,10 @@ mptscsih_remove(struct pci_dev *pdev) unsigned long flags; int sz1; - if(!host) + if(!host) { + mpt_detach(pdev); return; + } scsi_remove_host(host); @@ -1422,6 +1453,7 @@ mptscsih_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *)) return 0; fail: + hd->ScsiLookup[my_idx] = NULL; mptscsih_freeChainBuffers(hd->ioc, my_idx); mpt_free_msg_frame(hd->ioc, mf); return SCSI_MLQUEUE_HOST_BUSY; @@ -1709,24 +1741,23 @@ mptscsih_abort(struct scsi_cmnd * SCpnt) MPT_FRAME_HDR *mf; u32 ctx2abort; int scpnt_idx; + int retval; /* If we can't locate our host adapter structure, return FAILED status. */ if ((hd = (MPT_SCSI_HOST *) SCpnt->device->host->hostdata) == NULL) { SCpnt->result = DID_RESET << 16; SCpnt->scsi_done(SCpnt); - dfailprintk((KERN_WARNING MYNAM ": mptscsih_abort: " + dfailprintk((KERN_INFO MYNAM ": mptscsih_abort: " "Can't locate host! (sc=%p)\n", SCpnt)); return FAILED; } ioc = hd->ioc; - if (hd->resetPending) + if (hd->resetPending) { return FAILED; - - printk(KERN_WARNING MYNAM ": %s: >> Attempting task abort! (sc=%p)\n", - hd->ioc->name, SCpnt); + } if (hd->timeouts < -1) hd->timeouts++; @@ -1734,16 +1765,20 @@ mptscsih_abort(struct scsi_cmnd * SCpnt) /* Find this command */ if ((scpnt_idx = SCPNT_TO_LOOKUP_IDX(SCpnt)) < 0) { - /* Cmd not found in ScsiLookup. + /* Cmd not found in ScsiLookup. * Do OS callback. */ SCpnt->result = DID_RESET << 16; - dtmprintk((KERN_WARNING MYNAM ": %s: mptscsih_abort: " + dtmprintk((KERN_INFO MYNAM ": %s: mptscsih_abort: " "Command not in the active list! (sc=%p)\n", hd->ioc->name, SCpnt)); return SUCCESS; } + printk(KERN_WARNING MYNAM ": %s: attempting task abort! (sc=%p)\n", + hd->ioc->name, SCpnt); + scsi_print_command(SCpnt); + /* Most important! Set TaskMsgContext to SCpnt's MsgContext! * (the IO to be ABORT'd) * @@ -1756,38 +1791,22 @@ mptscsih_abort(struct scsi_cmnd * SCpnt) hd->abortSCpnt = SCpnt; - if (mptscsih_TMHandler(hd, MPI_SCSITASKMGMT_TASKTYPE_ABORT_TASK, + retval = mptscsih_TMHandler(hd, MPI_SCSITASKMGMT_TASKTYPE_ABORT_TASK, SCpnt->device->channel, SCpnt->device->id, SCpnt->device->lun, - ctx2abort, 2 /* 2 second timeout */) - < 0) { + ctx2abort, 2 /* 2 second timeout */); - /* The TM request failed and the subsequent FW-reload failed! - * Fatal error case. - */ - printk(MYIOC_s_WARN_FMT "Error issuing abort task! (sc=%p)\n", - hd->ioc->name, SCpnt); + printk (KERN_WARNING MYNAM ": %s: task abort: %s (sc=%p)\n", + hd->ioc->name, + ((retval == 0) ? "SUCCESS" : "FAILED" ), SCpnt); - /* We must clear our pending flag before clearing our state. - */ + if (retval == 0) + return SUCCESS; + + if(retval != FAILED ) { hd->tmPending = 0; hd->tmState = TM_STATE_NONE; - - /* Unmap the DMA buffers, if any. */ - if (SCpnt->use_sg) { - pci_unmap_sg(ioc->pcidev, (struct scatterlist *) SCpnt->request_buffer, - SCpnt->use_sg, SCpnt->sc_data_direction); - } else if (SCpnt->request_bufflen) { - pci_unmap_single(ioc->pcidev, SCpnt->SCp.dma_handle, - SCpnt->request_bufflen, SCpnt->sc_data_direction); - } - hd->ScsiLookup[scpnt_idx] = NULL; - SCpnt->result = DID_RESET << 16; - SCpnt->scsi_done(SCpnt); /* Issue the command callback */ - mptscsih_freeChainBuffers(ioc, scpnt_idx); - mpt_free_msg_frame(ioc, mf); - return FAILED; } - return SUCCESS; + return FAILED; } /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ @@ -1803,11 +1822,12 @@ int mptscsih_dev_reset(struct scsi_cmnd * SCpnt) { MPT_SCSI_HOST *hd; + int retval; /* If we can't locate our host adapter structure, return FAILED status. */ if ((hd = (MPT_SCSI_HOST *) SCpnt->device->host->hostdata) == NULL){ - dtmprintk((KERN_WARNING MYNAM ": mptscsih_dev_reset: " + dtmprintk((KERN_INFO MYNAM ": mptscsih_dev_reset: " "Can't locate host! (sc=%p)\n", SCpnt)); return FAILED; @@ -1816,24 +1836,26 @@ mptscsih_dev_reset(struct scsi_cmnd * SCpnt) if (hd->resetPending) return FAILED; - printk(KERN_WARNING MYNAM ": %s: >> Attempting target reset! (sc=%p)\n", + printk(KERN_WARNING MYNAM ": %s: attempting target reset! (sc=%p)\n", hd->ioc->name, SCpnt); + scsi_print_command(SCpnt); - if (mptscsih_TMHandler(hd, MPI_SCSITASKMGMT_TASKTYPE_TARGET_RESET, + retval = mptscsih_TMHandler(hd, MPI_SCSITASKMGMT_TASKTYPE_TARGET_RESET, SCpnt->device->channel, SCpnt->device->id, - 0, 0, 5 /* 5 second timeout */) - < 0){ - /* The TM request failed and the subsequent FW-reload failed! - * Fatal error case. - */ - printk(MYIOC_s_WARN_FMT "Error processing TaskMgmt request (sc=%p)\n", - hd->ioc->name, SCpnt); + 0, 0, 5 /* 5 second timeout */); + + printk (KERN_WARNING MYNAM ": %s: target reset: %s (sc=%p)\n", + hd->ioc->name, + ((retval == 0) ? "SUCCESS" : "FAILED" ), SCpnt); + + if (retval == 0) + return SUCCESS; + + if(retval != FAILED ) { hd->tmPending = 0; hd->tmState = TM_STATE_NONE; - return FAILED; } - - return SUCCESS; + return FAILED; } /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ @@ -1849,41 +1871,39 @@ int mptscsih_bus_reset(struct scsi_cmnd * SCpnt) { MPT_SCSI_HOST *hd; - spinlock_t *host_lock = SCpnt->device->host->host_lock; + int retval; /* If we can't locate our host adapter structure, return FAILED status. */ if ((hd = (MPT_SCSI_HOST *) SCpnt->device->host->hostdata) == NULL){ - dtmprintk((KERN_WARNING MYNAM ": mptscsih_bus_reset: " + dtmprintk((KERN_INFO MYNAM ": mptscsih_bus_reset: " "Can't locate host! (sc=%p)\n", SCpnt ) ); return FAILED; } - printk(KERN_WARNING MYNAM ": %s: >> Attempting bus reset! (sc=%p)\n", + printk(KERN_WARNING MYNAM ": %s: attempting bus reset! (sc=%p)\n", hd->ioc->name, SCpnt); + scsi_print_command(SCpnt); if (hd->timeouts < -1) hd->timeouts++; - /* We are now ready to execute the task management request. */ - if (mptscsih_TMHandler(hd, MPI_SCSITASKMGMT_TASKTYPE_RESET_BUS, - SCpnt->device->channel, 0, 0, 0, 5 /* 5 second timeout */) - < 0){ + retval = mptscsih_TMHandler(hd, MPI_SCSITASKMGMT_TASKTYPE_RESET_BUS, + SCpnt->device->channel, 0, 0, 0, 5 /* 5 second timeout */); - /* The TM request failed and the subsequent FW-reload failed! - * Fatal error case. - */ - printk(MYIOC_s_WARN_FMT - "Error processing TaskMgmt request (sc=%p)\n", - hd->ioc->name, SCpnt); + printk (KERN_WARNING MYNAM ": %s: bus reset: %s (sc=%p)\n", + hd->ioc->name, + ((retval == 0) ? "SUCCESS" : "FAILED" ), SCpnt); + + if (retval == 0) + return SUCCESS; + + if(retval != FAILED ) { hd->tmPending = 0; hd->tmState = TM_STATE_NONE; - spin_lock_irq(host_lock); - return FAILED; } - - return SUCCESS; + return FAILED; } /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ @@ -2165,7 +2185,7 @@ mptscsih_slave_alloc(struct scsi_device *device) vdev->raidVolume = 0; hd->Targets[device->id] = vdev; if (hd->ioc->bus_type == SCSI) { - if (hd->ioc->spi_data.isRaid & (1 << device->id)) { + if (hd->ioc->raid_data.isRaid & (1 << device->id)) { vdev->raidVolume = 1; ddvtprintk((KERN_INFO "RAID Volume @ id %d\n", device->id)); @@ -2180,22 +2200,6 @@ mptscsih_slave_alloc(struct scsi_device *device) return 0; } -static int -mptscsih_is_raid_volume(MPT_SCSI_HOST *hd, uint id) -{ - int i; - - if (!hd->ioc->spi_data.isRaid || !hd->ioc->spi_data.pIocPg3) - return 0; - - for (i = 0; i < hd->ioc->spi_data.pIocPg3->NumPhysDisks; i++) { - if (id == hd->ioc->spi_data.pIocPg3->PhysDisk[i].PhysDiskID) - return 1; - } - - return 0; -} - /* * OS entry point to allow for host driver to free allocated memory * Called if no device present or device being unloaded @@ -2223,7 +2227,7 @@ mptscsih_slave_destroy(struct scsi_device *device) hd->Targets[target] = NULL; if (hd->ioc->bus_type == SCSI) { - if (mptscsih_is_raid_volume(hd, target)) { + if (mptscsih_is_phys_disk(hd->ioc, target)) { hd->ioc->spi_data.forceDv |= MPT_SCSICFG_RELOAD_IOC_PG3; } else { hd->ioc->spi_data.dvStatus[target] = @@ -2436,6 +2440,7 @@ mptscsih_ioc_reset(MPT_ADAPTER *ioc, int reset_phase) { MPT_SCSI_HOST *hd; unsigned long flags; + int ii; dtmprintk((KERN_WARNING MYNAM ": IOC %s_reset routed to SCSI host driver!\n", @@ -2493,11 +2498,8 @@ mptscsih_ioc_reset(MPT_ADAPTER *ioc, int reset_phase) /* ScsiLookup initialization */ - { - int ii; - for (ii=0; ii < hd->ioc->req_depth; ii++) - hd->ScsiLookup[ii] = NULL; - } + for (ii=0; ii < hd->ioc->req_depth; ii++) + hd->ScsiLookup[ii] = NULL; /* 2. Chain Buffer initialization */ @@ -2545,6 +2547,16 @@ mptscsih_ioc_reset(MPT_ADAPTER *ioc, int reset_phase) return 1; /* currently means nothing really */ } +/*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ +/* work queue thread to clear the persitency table */ +static void +mptscsih_sas_persist_clear_table(void * arg) +{ + MPT_ADAPTER *ioc = (MPT_ADAPTER *)arg; + + mptbase_sas_persist_operation(ioc, MPI_SAS_OP_CLEAR_NOT_PRESENT); +} + /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ int mptscsih_event_process(MPT_ADAPTER *ioc, EventNotificationReply_t *pEvReply) @@ -2555,18 +2567,18 @@ mptscsih_event_process(MPT_ADAPTER *ioc, EventNotificationReply_t *pEvReply) devtprintk((MYIOC_s_INFO_FMT "MPT event (=%02Xh) routed to SCSI host driver!\n", ioc->name, event)); + if (ioc->sh == NULL || + ((hd = (MPT_SCSI_HOST *)ioc->sh->hostdata) == NULL)) + return 1; + switch (event) { case MPI_EVENT_UNIT_ATTENTION: /* 03 */ /* FIXME! */ break; case MPI_EVENT_IOC_BUS_RESET: /* 04 */ case MPI_EVENT_EXT_BUS_RESET: /* 05 */ - hd = NULL; - if (ioc->sh) { - hd = (MPT_SCSI_HOST *) ioc->sh->hostdata; - if (hd && (ioc->bus_type == SCSI) && (hd->soft_resets < -1)) - hd->soft_resets++; - } + if (hd && (ioc->bus_type == SCSI) && (hd->soft_resets < -1)) + hd->soft_resets++; break; case MPI_EVENT_LOGOUT: /* 09 */ /* FIXME! */ @@ -2585,69 +2597,24 @@ mptscsih_event_process(MPT_ADAPTER *ioc, EventNotificationReply_t *pEvReply) break; case MPI_EVENT_INTEGRATED_RAID: /* 0B */ + { + pMpiEventDataRaid_t pRaidEventData = + (pMpiEventDataRaid_t) pEvReply->Data; #ifdef MPTSCSIH_ENABLE_DOMAIN_VALIDATION - /* negoNvram set to 0 if DV enabled and to USE_NVRAM if - * if DV disabled. Need to check for target mode. - */ - hd = NULL; - if (ioc->sh) - hd = (MPT_SCSI_HOST *) ioc->sh->hostdata; - - if (hd && (ioc->bus_type == SCSI) && (hd->negoNvram == 0)) { - ScsiCfgData *pSpi; - Ioc3PhysDisk_t *pPDisk; - int numPDisk; - u8 reason; - u8 physDiskNum; - - reason = (le32_to_cpu(pEvReply->Data[0]) & 0x00FF0000) >> 16; - if (reason == MPI_EVENT_RAID_RC_DOMAIN_VAL_NEEDED) { - /* New or replaced disk. - * Set DV flag and schedule DV. - */ - pSpi = &ioc->spi_data; - physDiskNum = (le32_to_cpu(pEvReply->Data[0]) & 0xFF000000) >> 24; - ddvtprintk(("DV requested for phys disk id %d\n", physDiskNum)); - if (pSpi->pIocPg3) { - pPDisk = pSpi->pIocPg3->PhysDisk; - numPDisk =pSpi->pIocPg3->NumPhysDisks; - - while (numPDisk) { - if (physDiskNum == pPDisk->PhysDiskNum) { - pSpi->dvStatus[pPDisk->PhysDiskID] = (MPT_SCSICFG_NEED_DV | MPT_SCSICFG_DV_NOT_DONE); - pSpi->forceDv = MPT_SCSICFG_NEED_DV; - ddvtprintk(("NEED_DV set for phys disk id %d\n", pPDisk->PhysDiskID)); - break; - } - pPDisk++; - numPDisk--; - } - - if (numPDisk == 0) { - /* The physical disk that needs DV was not found - * in the stored IOC Page 3. The driver must reload - * this page. DV routine will set the NEED_DV flag for - * all phys disks that have DV_NOT_DONE set. - */ - pSpi->forceDv = MPT_SCSICFG_NEED_DV | MPT_SCSICFG_RELOAD_IOC_PG3; - ddvtprintk(("phys disk %d not found. Setting reload IOC Pg3 Flag\n", physDiskNum)); - } - } - } - } + /* Domain Validation Needed */ + if (ioc->bus_type == SCSI && + pRaidEventData->ReasonCode == + MPI_EVENT_RAID_RC_DOMAIN_VAL_NEEDED) + mptscsih_set_dvflags_raid(hd, pRaidEventData->PhysDiskNum); #endif + break; + } -#if defined(MPT_DEBUG_DV) || defined(MPT_DEBUG_DV_TINY) - printk("Raid Event RF: "); - { - u32 *m = (u32 *)pEvReply; - int ii; - int n = (int)pEvReply->MsgLength; - for (ii=6; ii < n; ii++) - printk(" %08x", le32_to_cpu(m[ii])); - printk("\n"); - } -#endif + /* Persistent table is full. */ + case MPI_EVENT_PERSISTENT_TABLE_FULL: + INIT_WORK(&mptscsih_persistTask, + mptscsih_sas_persist_clear_table,(void *)ioc); + schedule_work(&mptscsih_persistTask); break; case MPI_EVENT_NONE: /* 00 */ @@ -2684,7 +2651,7 @@ mptscsih_initTarget(MPT_SCSI_HOST *hd, int bus_id, int target_id, u8 lun, char * { int indexed_lun, lun_index; VirtDevice *vdev; - ScsiCfgData *pSpi; + SpiCfgData *pSpi; char data_56; dinitprintk((MYIOC_s_INFO_FMT "initTarget bus=%d id=%d lun=%d hd=%p\n", @@ -2791,7 +2758,7 @@ mptscsih_initTarget(MPT_SCSI_HOST *hd, int bus_id, int target_id, u8 lun, char * static void mptscsih_setTargetNegoParms(MPT_SCSI_HOST *hd, VirtDevice *target, char byte56) { - ScsiCfgData *pspi_data = &hd->ioc->spi_data; + SpiCfgData *pspi_data = &hd->ioc->spi_data; int id = (int) target->target_id; int nvram; VirtDevice *vdev; @@ -2970,11 +2937,13 @@ mptscsih_setTargetNegoParms(MPT_SCSI_HOST *hd, VirtDevice *target, char byte56) static void mptscsih_set_dvflags(MPT_SCSI_HOST *hd, SCSIIORequest_t *pReq) { + MPT_ADAPTER *ioc = hd->ioc; u8 cmd; - ScsiCfgData *pSpi; + SpiCfgData *pSpi; - ddvtprintk((" set_dvflags: id=%d lun=%d negoNvram=%x cmd=%x\n", - pReq->TargetID, pReq->LUN[1], hd->negoNvram, pReq->CDB[0])); + ddvtprintk((MYIOC_s_NOTE_FMT + " set_dvflags: id=%d lun=%d negoNvram=%x cmd=%x\n", + hd->ioc->name, pReq->TargetID, pReq->LUN[1], hd->negoNvram, pReq->CDB[0])); if ((pReq->LUN[1] != 0) || (hd->negoNvram != 0)) return; @@ -2982,12 +2951,12 @@ mptscsih_set_dvflags(MPT_SCSI_HOST *hd, SCSIIORequest_t *pReq) cmd = pReq->CDB[0]; if ((cmd == READ_CAPACITY) || (cmd == MODE_SENSE)) { - pSpi = &hd->ioc->spi_data; - if ((pSpi->isRaid & (1 << pReq->TargetID)) && pSpi->pIocPg3) { + pSpi = &ioc->spi_data; + if ((ioc->raid_data.isRaid & (1 << pReq->TargetID)) && ioc->raid_data.pIocPg3) { /* Set NEED_DV for all hidden disks */ - Ioc3PhysDisk_t *pPDisk = pSpi->pIocPg3->PhysDisk; - int numPDisk = pSpi->pIocPg3->NumPhysDisks; + Ioc3PhysDisk_t *pPDisk = ioc->raid_data.pIocPg3->PhysDisk; + int numPDisk = ioc->raid_data.pIocPg3->NumPhysDisks; while (numPDisk) { pSpi->dvStatus[pPDisk->PhysDiskID] |= MPT_SCSICFG_NEED_DV; @@ -3001,6 +2970,50 @@ mptscsih_set_dvflags(MPT_SCSI_HOST *hd, SCSIIORequest_t *pReq) } } +/* mptscsih_raid_set_dv_flags() + * + * New or replaced disk. Set DV flag and schedule DV. + */ +static void +mptscsih_set_dvflags_raid(MPT_SCSI_HOST *hd, int id) +{ + MPT_ADAPTER *ioc = hd->ioc; + SpiCfgData *pSpi = &ioc->spi_data; + Ioc3PhysDisk_t *pPDisk; + int numPDisk; + + if (hd->negoNvram != 0) + return; + + ddvtprintk(("DV requested for phys disk id %d\n", id)); + if (ioc->raid_data.pIocPg3) { + pPDisk = ioc->raid_data.pIocPg3->PhysDisk; + numPDisk = ioc->raid_data.pIocPg3->NumPhysDisks; + while (numPDisk) { + if (id == pPDisk->PhysDiskNum) { + pSpi->dvStatus[pPDisk->PhysDiskID] = + (MPT_SCSICFG_NEED_DV | MPT_SCSICFG_DV_NOT_DONE); + pSpi->forceDv = MPT_SCSICFG_NEED_DV; + ddvtprintk(("NEED_DV set for phys disk id %d\n", + pPDisk->PhysDiskID)); + break; + } + pPDisk++; + numPDisk--; + } + + if (numPDisk == 0) { + /* The physical disk that needs DV was not found + * in the stored IOC Page 3. The driver must reload + * this page. DV routine will set the NEED_DV flag for + * all phys disks that have DV_NOT_DONE set. + */ + pSpi->forceDv = MPT_SCSICFG_NEED_DV | MPT_SCSICFG_RELOAD_IOC_PG3; + ddvtprintk(("phys disk %d not found. Setting reload IOC Pg3 Flag\n",id)); + } + } +} + /*=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=*/ /* * If no Target, bus reset on 1st I/O. Set the flag to @@ -3088,7 +3101,7 @@ mptscsih_writeSDP1(MPT_SCSI_HOST *hd, int portnum, int target_id, int flags) MPT_ADAPTER *ioc = hd->ioc; Config_t *pReq; SCSIDevicePage1_t *pData; - VirtDevice *pTarget; + VirtDevice *pTarget=NULL; MPT_FRAME_HDR *mf; dma_addr_t dataDma; u16 req_idx; @@ -3187,7 +3200,7 @@ mptscsih_writeSDP1(MPT_SCSI_HOST *hd, int portnum, int target_id, int flags) #endif if (flags & MPT_SCSICFG_BLK_NEGO) - negoFlags = MPT_TARGET_NO_NEGO_WIDE | MPT_TARGET_NO_NEGO_SYNC; + negoFlags |= MPT_TARGET_NO_NEGO_WIDE | MPT_TARGET_NO_NEGO_SYNC; mptscsih_setDevicePage1Flags(width, factor, offset, &requested, &configuration, negoFlags); @@ -4008,7 +4021,7 @@ mptscsih_synchronize_cache(MPT_SCSI_HOST *hd, int portnum) /* If target Ptr NULL or if this target is NOT a disk, skip. */ - if ((pTarget) && (pTarget->tflags & MPT_TARGET_FLAGS_Q_YES)){ + if ((pTarget) && (pTarget->inq_data[0] == TYPE_DISK)){ for (lun=0; lun <= MPT_LAST_LUN; lun++) { /* If LUN present, issue the command */ @@ -4103,9 +4116,9 @@ mptscsih_domainValidation(void *arg) if ((ioc->spi_data.forceDv & MPT_SCSICFG_RELOAD_IOC_PG3) != 0) { mpt_read_ioc_pg_3(ioc); - if (ioc->spi_data.pIocPg3) { - Ioc3PhysDisk_t *pPDisk = ioc->spi_data.pIocPg3->PhysDisk; - int numPDisk = ioc->spi_data.pIocPg3->NumPhysDisks; + if (ioc->raid_data.pIocPg3) { + Ioc3PhysDisk_t *pPDisk = ioc->raid_data.pIocPg3->PhysDisk; + int numPDisk = ioc->raid_data.pIocPg3->NumPhysDisks; while (numPDisk) { if (ioc->spi_data.dvStatus[pPDisk->PhysDiskID] & MPT_SCSICFG_DV_NOT_DONE) @@ -4144,7 +4157,7 @@ mptscsih_domainValidation(void *arg) isPhysDisk = mptscsih_is_phys_disk(ioc, id); if (isPhysDisk) { for (ii=0; ii < MPT_MAX_SCSI_DEVICES; ii++) { - if (hd->ioc->spi_data.isRaid & (1 << ii)) { + if (hd->ioc->raid_data.isRaid & (1 << ii)) { hd->ioc->spi_data.dvStatus[ii] |= MPT_SCSICFG_DV_PENDING; } } @@ -4163,7 +4176,7 @@ mptscsih_domainValidation(void *arg) if (isPhysDisk) { for (ii=0; ii < MPT_MAX_SCSI_DEVICES; ii++) { - if (hd->ioc->spi_data.isRaid & (1 << ii)) { + if (hd->ioc->raid_data.isRaid & (1 << ii)) { hd->ioc->spi_data.dvStatus[ii] &= ~MPT_SCSICFG_DV_PENDING; } } @@ -4185,21 +4198,21 @@ mptscsih_domainValidation(void *arg) /* Search IOC page 3 to determine if this is hidden physical disk */ -static int +/* Search IOC page 3 to determine if this is hidden physical disk + */ +static int mptscsih_is_phys_disk(MPT_ADAPTER *ioc, int id) { - if (ioc->spi_data.pIocPg3) { - Ioc3PhysDisk_t *pPDisk = ioc->spi_data.pIocPg3->PhysDisk; - int numPDisk = ioc->spi_data.pIocPg3->NumPhysDisks; + int i; - while (numPDisk) { - if (pPDisk->PhysDiskID == id) { - return 1; - } - pPDisk++; - numPDisk--; - } + if (!ioc->raid_data.isRaid || !ioc->raid_data.pIocPg3) + return 0; + + for (i = 0; i < ioc->raid_data.pIocPg3->NumPhysDisks; i++) { + if (id == ioc->raid_data.pIocPg3->PhysDisk[i].PhysDiskID) + return 1; } + return 0; } @@ -4405,7 +4418,7 @@ mptscsih_doDv(MPT_SCSI_HOST *hd, int bus_number, int id) /* Skip this ID? Set cfg.cfghdr.hdr to force config page write */ { - ScsiCfgData *pspi_data = &hd->ioc->spi_data; + SpiCfgData *pspi_data = &hd->ioc->spi_data; if (pspi_data->nvram && (pspi_data->nvram[id] != MPT_HOST_NVRAM_INVALID)) { /* Set the factor from nvram */ nfactor = (pspi_data->nvram[id] & MPT_NVRAM_SYNC_MASK) >> 8; @@ -4435,11 +4448,11 @@ mptscsih_doDv(MPT_SCSI_HOST *hd, int bus_number, int id) } /* Finish iocmd inititialization - hidden or visible disk? */ - if (ioc->spi_data.pIocPg3) { + if (ioc->raid_data.pIocPg3) { /* Search IOC page 3 for matching id */ - Ioc3PhysDisk_t *pPDisk = ioc->spi_data.pIocPg3->PhysDisk; - int numPDisk = ioc->spi_data.pIocPg3->NumPhysDisks; + Ioc3PhysDisk_t *pPDisk = ioc->raid_data.pIocPg3->PhysDisk; + int numPDisk = ioc->raid_data.pIocPg3->NumPhysDisks; while (numPDisk) { if (pPDisk->PhysDiskID == id) { @@ -4463,7 +4476,7 @@ mptscsih_doDv(MPT_SCSI_HOST *hd, int bus_number, int id) /* RAID Volume ID's may double for a physical device. If RAID but * not a physical ID as well, skip DV. */ - if ((hd->ioc->spi_data.isRaid & (1 << id)) && !(iocmd.flags & MPT_ICFLAG_PHYS_DISK)) + if ((hd->ioc->raid_data.isRaid & (1 << id)) && !(iocmd.flags & MPT_ICFLAG_PHYS_DISK)) goto target_done; @@ -4812,6 +4825,8 @@ mptscsih_doDv(MPT_SCSI_HOST *hd, int bus_number, int id) notDone = 0; if (iocmd.flags & MPT_ICFLAG_ECHO) { bufsize = ((pbuf1[2] & 0x1F) <<8) | pbuf1[3]; + if (pbuf1[0] & 0x01) + iocmd.flags |= MPT_ICFLAG_EBOS; } else { bufsize = pbuf1[1]<<16 | pbuf1[2]<<8 | pbuf1[3]; } @@ -4908,6 +4923,9 @@ mptscsih_doDv(MPT_SCSI_HOST *hd, int bus_number, int id) } iocmd.flags &= ~MPT_ICFLAG_DID_RESET; + if (iocmd.flags & MPT_ICFLAG_EBOS) + goto skip_Reserve; + repeat = 5; while (repeat && (!(iocmd.flags & MPT_ICFLAG_RESERVED))) { iocmd.cmd = RESERVE; @@ -4951,6 +4969,7 @@ mptscsih_doDv(MPT_SCSI_HOST *hd, int bus_number, int id) } } +skip_Reserve: mptscsih_fillbuf(pbuf1, sz, patt, 1); iocmd.cmd = WRITE_BUFFER; iocmd.data_dma = buf1_dma; @@ -5195,11 +5214,12 @@ mptscsih_dv_parms(MPT_SCSI_HOST *hd, DVPARAMETERS *dv,void *pPage) * If not an LVD bus, the adapter minSyncFactor has been * already throttled back. */ + negoFlags = hd->ioc->spi_data.noQas; if ((hd->Targets)&&((pTarget = hd->Targets[(int)id]) != NULL) && !pTarget->raidVolume) { width = pTarget->maxWidth; offset = pTarget->maxOffset; factor = pTarget->minSyncFactor; - negoFlags = pTarget->negoFlags; + negoFlags |= pTarget->negoFlags; } else { if (hd->ioc->spi_data.nvram && (hd->ioc->spi_data.nvram[id] != MPT_HOST_NVRAM_INVALID)) { data = hd->ioc->spi_data.nvram[id]; @@ -5220,7 +5240,6 @@ mptscsih_dv_parms(MPT_SCSI_HOST *hd, DVPARAMETERS *dv,void *pPage) } /* Set the negotiation flags */ - negoFlags = hd->ioc->spi_data.noQas; if (!width) negoFlags |= MPT_TARGET_NO_NEGO_WIDE; diff --git a/drivers/message/fusion/mptscsih.h b/drivers/message/fusion/mptscsih.h index 51c0255ac16..09389af9845 100644 --- a/drivers/message/fusion/mptscsih.h +++ b/drivers/message/fusion/mptscsih.h @@ -1,5 +1,5 @@ /* - * linux/drivers/message/fusion/mptscsi.h + * linux/drivers/message/fusion/mptscsih.h * High performance SCSI / Fibre Channel SCSI Host device driver. * For use with PCI chip/adapter(s): * LSIFC9xx/LSI409xx Fibre Channel @@ -53,8 +53,8 @@ * SCSI Public stuff... */ -#define MPT_SCSI_CMD_PER_DEV_HIGH 31 -#define MPT_SCSI_CMD_PER_DEV_LOW 7 +#define MPT_SCSI_CMD_PER_DEV_HIGH 64 +#define MPT_SCSI_CMD_PER_DEV_LOW 32 #define MPT_SCSI_CMD_PER_LUN 7 @@ -77,6 +77,7 @@ #define MPTSCSIH_MAX_WIDTH 1 #define MPTSCSIH_MIN_SYNC 0x08 #define MPTSCSIH_SAF_TE 0 +#define MPTSCSIH_PT_CLEAR 0 #endif @@ -105,3 +106,4 @@ extern int mptscsih_event_process(MPT_ADAPTER *ioc, EventNotificationReply_t *pE extern int mptscsih_ioc_reset(MPT_ADAPTER *ioc, int post_reset); extern int mptscsih_change_queue_depth(struct scsi_device *sdev, int qdepth); extern void mptscsih_timer_expired(unsigned long data); +extern void scsi_print_command(struct scsi_cmnd *cmd); diff --git a/drivers/message/fusion/mptspi.c b/drivers/message/fusion/mptspi.c index 587d1274fd7..5c0e307d1d5 100644 --- a/drivers/message/fusion/mptspi.c +++ b/drivers/message/fusion/mptspi.c @@ -199,7 +199,7 @@ mptspi_probe(struct pci_dev *pdev, const struct pci_device_id *id) printk(MYIOC_s_WARN_FMT "Skipping ioc=%p because SCSI Initiator mode is NOT enabled!\n", ioc->name, ioc); - return -ENODEV; + return 0; } sh = scsi_host_alloc(&mptspi_driver_template, sizeof(MPT_SCSI_HOST)); From e0fc15bef0e8c6b5abad6e10cfe3d42e278ae8e8 Mon Sep 17 00:00:00 2001 From: "Moore, Eric Dean" Date: Thu, 15 Sep 2005 13:17:14 -0600 Subject: [PATCH 035/204] [SCSI] fusion SAS support (mptsas driver) minor fix On Thursday, September 15, 2005 6:22 AM, Christoph Hellwig wrote: > Looks good to me, except for the spurious scsi_print_command prototype > in mptscsih.h. The attached patch addresses that concern. Signed-off-by: Eric Moore Signed-off-by: James Bottomley --- drivers/message/fusion/mptscsih.c | 1 + drivers/message/fusion/mptscsih.h | 1 - 2 files changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/message/fusion/mptscsih.c b/drivers/message/fusion/mptscsih.c index 8dd25aac535..5cb07eb224d 100644 --- a/drivers/message/fusion/mptscsih.c +++ b/drivers/message/fusion/mptscsih.c @@ -62,6 +62,7 @@ #include #include #include +#include #include "mptbase.h" #include "mptscsih.h" diff --git a/drivers/message/fusion/mptscsih.h b/drivers/message/fusion/mptscsih.h index 09389af9845..971fda4b8b5 100644 --- a/drivers/message/fusion/mptscsih.h +++ b/drivers/message/fusion/mptscsih.h @@ -106,4 +106,3 @@ extern int mptscsih_event_process(MPT_ADAPTER *ioc, EventNotificationReply_t *pE extern int mptscsih_ioc_reset(MPT_ADAPTER *ioc, int post_reset); extern int mptscsih_change_queue_depth(struct scsi_device *sdev, int qdepth); extern void mptscsih_timer_expired(unsigned long data); -extern void scsi_print_command(struct scsi_cmnd *cmd); From 3734d24b2e8d85796de70c13705cfb7cbb1d77df Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Tue, 13 Sep 2005 21:47:11 +0200 Subject: [PATCH 036/204] [SCSI] zfcp: fix race conditions when accessing erp_action lists o always use locking when changing erp_action lists, o avoid escalation to ERP_ACTION_REOPEN_PORT_FORCED if erp_action is still in use for ERP_ACTION_REOPEN_PORT Signed-off-by: Andreas Herrmann Signed-off-by: James Bottomley --- drivers/s390/scsi/zfcp_erp.c | 24 +++++++++++++++++------- 1 file changed, 17 insertions(+), 7 deletions(-) diff --git a/drivers/s390/scsi/zfcp_erp.c b/drivers/s390/scsi/zfcp_erp.c index cb4f612550b..376cb0f6cb7 100644 --- a/drivers/s390/scsi/zfcp_erp.c +++ b/drivers/s390/scsi/zfcp_erp.c @@ -886,7 +886,7 @@ static int zfcp_erp_strategy_check_fsfreq(struct zfcp_erp_action *erp_action) { int retval = 0; - struct zfcp_fsf_req *fsf_req; + struct zfcp_fsf_req *fsf_req = NULL; struct zfcp_adapter *adapter = erp_action->adapter; if (erp_action->fsf_req) { @@ -896,7 +896,7 @@ zfcp_erp_strategy_check_fsfreq(struct zfcp_erp_action *erp_action) list_for_each_entry(fsf_req, &adapter->fsf_req_list_head, list) if (fsf_req == erp_action->fsf_req) break; - if (fsf_req == erp_action->fsf_req) { + if (fsf_req && (fsf_req->erp_action == erp_action)) { /* fsf_req still exists */ debug_text_event(adapter->erp_dbf, 3, "a_ca_req"); debug_event(adapter->erp_dbf, 3, &fsf_req, @@ -2291,7 +2291,9 @@ zfcp_erp_adapter_strategy_open_fsf_xconfig(struct zfcp_erp_action *erp_action) atomic_clear_mask(ZFCP_STATUS_ADAPTER_HOST_CON_INIT, &adapter->status); ZFCP_LOG_DEBUG("Doing exchange config data\n"); + write_lock(&adapter->erp_lock); zfcp_erp_action_to_running(erp_action); + write_unlock(&adapter->erp_lock); zfcp_erp_timeout_init(erp_action); if (zfcp_fsf_exchange_config_data(erp_action)) { retval = ZFCP_ERP_FAILED; @@ -3194,11 +3196,19 @@ zfcp_erp_action_enqueue(int action, /* fall through !!! */ case ZFCP_ERP_ACTION_REOPEN_PORT_FORCED: - if (atomic_test_mask - (ZFCP_STATUS_COMMON_ERP_INUSE, &port->status) - && port->erp_action.action == - ZFCP_ERP_ACTION_REOPEN_PORT_FORCED) { - debug_text_event(adapter->erp_dbf, 4, "pf_actenq_drp"); + if (atomic_test_mask(ZFCP_STATUS_COMMON_ERP_INUSE, + &port->status)) { + if (port->erp_action.action != + ZFCP_ERP_ACTION_REOPEN_PORT_FORCED) { + ZFCP_LOG_INFO("dropped erp action %i (port " + "0x%016Lx, action in use: %i)\n", + action, port->wwpn, + port->erp_action.action); + debug_text_event(adapter->erp_dbf, 4, + "pf_actenq_drp"); + } else + debug_text_event(adapter->erp_dbf, 4, + "pf_actenq_drpcp"); debug_event(adapter->erp_dbf, 4, &port->wwpn, sizeof (wwn_t)); goto out; From 059c97d0434834d291eff94669ca2dd3eaac9d28 Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Tue, 13 Sep 2005 21:47:52 +0200 Subject: [PATCH 037/204] [SCSI] zfcp: remove union zfcp_req_data, use unit refcount for FCP commands o union zfcp_req_data removed o increment unit refcount when processing FCP commands (This fixes a theoretical race: When all scsi commands of a unit are aborted and the scsi_device is removed then the unit could be removed before all fsf_requests of that unit are completely processed.) Signed-off-by: Andreas Herrmann Signed-off-by: James Bottomley --- drivers/s390/scsi/zfcp_aux.c | 8 +- drivers/s390/scsi/zfcp_def.h | 69 +------------ drivers/s390/scsi/zfcp_fsf.c | 101 ++++++++++--------- drivers/s390/scsi/zfcp_scsi.c | 176 +++++----------------------------- 4 files changed, 80 insertions(+), 274 deletions(-) diff --git a/drivers/s390/scsi/zfcp_aux.c b/drivers/s390/scsi/zfcp_aux.c index bfe3ba73bc0..17f9989238f 100644 --- a/drivers/s390/scsi/zfcp_aux.c +++ b/drivers/s390/scsi/zfcp_aux.c @@ -141,7 +141,7 @@ zfcp_cmd_dbf_event_fsf(const char *text, struct zfcp_fsf_req *fsf_req, spin_lock_irqsave(&adapter->dbf_lock, flags); if (zfcp_fsf_req_is_scsi_cmnd(fsf_req)) { - scsi_cmnd = fsf_req->data.send_fcp_command_task.scsi_cmnd; + scsi_cmnd = (struct scsi_cmnd*) fsf_req->data; debug_text_event(adapter->cmd_dbf, level, "fsferror"); debug_text_event(adapter->cmd_dbf, level, text); debug_event(adapter->cmd_dbf, level, &fsf_req, @@ -167,14 +167,12 @@ void zfcp_cmd_dbf_event_scsi(const char *text, struct scsi_cmnd *scsi_cmnd) { struct zfcp_adapter *adapter; - union zfcp_req_data *req_data; struct zfcp_fsf_req *fsf_req; int level = ((host_byte(scsi_cmnd->result) != 0) ? 1 : 5); unsigned long flags; adapter = (struct zfcp_adapter *) scsi_cmnd->device->host->hostdata[0]; - req_data = (union zfcp_req_data *) scsi_cmnd->host_scribble; - fsf_req = (req_data ? req_data->send_fcp_command_task.fsf_req : NULL); + fsf_req = (struct zfcp_fsf_req *) scsi_cmnd->host_scribble; spin_lock_irqsave(&adapter->dbf_lock, flags); debug_text_event(adapter->cmd_dbf, level, "hostbyte"); debug_text_event(adapter->cmd_dbf, level, text); @@ -1609,7 +1607,7 @@ zfcp_fsf_incoming_els(struct zfcp_fsf_req *fsf_req) u32 els_type; struct zfcp_adapter *adapter; - status_buffer = fsf_req->data.status_read.buffer; + status_buffer = (struct fsf_status_read_buffer *) fsf_req->data; els_type = *(u32 *) (status_buffer->payload); adapter = fsf_req->adapter; diff --git a/drivers/s390/scsi/zfcp_def.h b/drivers/s390/scsi/zfcp_def.h index 455e902533a..9160e68c4a2 100644 --- a/drivers/s390/scsi/zfcp_def.h +++ b/drivers/s390/scsi/zfcp_def.h @@ -635,45 +635,6 @@ struct zfcp_adapter_mempool { mempool_t *data_gid_pn; }; -struct zfcp_exchange_config_data{ -}; - -struct zfcp_open_port { - struct zfcp_port *port; -}; - -struct zfcp_close_port { - struct zfcp_port *port; -}; - -struct zfcp_open_unit { - struct zfcp_unit *unit; -}; - -struct zfcp_close_unit { - struct zfcp_unit *unit; -}; - -struct zfcp_close_physical_port { - struct zfcp_port *port; -}; - -struct zfcp_send_fcp_command_task { - struct zfcp_fsf_req *fsf_req; - struct zfcp_unit *unit; - struct scsi_cmnd *scsi_cmnd; - unsigned long start_jiffies; -}; - -struct zfcp_send_fcp_command_task_management { - struct zfcp_unit *unit; -}; - -struct zfcp_abort_fcp_command { - struct zfcp_fsf_req *fsf_req; - struct zfcp_unit *unit; -}; - /* * header for CT_IU */ @@ -781,33 +742,6 @@ struct zfcp_send_els { int status; }; -struct zfcp_status_read { - struct fsf_status_read_buffer *buffer; -}; - -struct zfcp_fsf_done { - struct completion *complete; - int status; -}; - -/* request specific data */ -union zfcp_req_data { - struct zfcp_exchange_config_data exchange_config_data; - struct zfcp_open_port open_port; - struct zfcp_close_port close_port; - struct zfcp_open_unit open_unit; - struct zfcp_close_unit close_unit; - struct zfcp_close_physical_port close_physical_port; - struct zfcp_send_fcp_command_task send_fcp_command_task; - struct zfcp_send_fcp_command_task_management - send_fcp_command_task_management; - struct zfcp_abort_fcp_command abort_fcp_command; - struct zfcp_send_ct *send_ct; - struct zfcp_send_els *send_els; - struct zfcp_status_read status_read; - struct fsf_qtcb_bottom_port *port_data; -}; - struct zfcp_qdio_queue { struct qdio_buffer *buffer[QDIO_MAX_BUFFERS_PER_Q]; /* SBALs */ u8 free_index; /* index of next free bfr @@ -963,11 +897,12 @@ struct zfcp_fsf_req { u32 fsf_command; /* FSF Command copy */ struct fsf_qtcb *qtcb; /* address of associated QTCB */ u32 seq_no; /* Sequence number of request */ - union zfcp_req_data data; /* Info fields of request */ + unsigned long data; /* private data of request */ struct zfcp_erp_action *erp_action; /* used if this request is issued on behalf of erp */ mempool_t *pool; /* used if request was alloacted from emergency pool */ + struct zfcp_unit *unit; }; typedef void zfcp_fsf_req_handler_t(struct zfcp_fsf_req*); diff --git a/drivers/s390/scsi/zfcp_fsf.c b/drivers/s390/scsi/zfcp_fsf.c index c007b6424e7..64e2c348018 100644 --- a/drivers/s390/scsi/zfcp_fsf.c +++ b/drivers/s390/scsi/zfcp_fsf.c @@ -821,7 +821,7 @@ zfcp_fsf_status_read(struct zfcp_adapter *adapter, int req_flags) goto failed_buf; } memset(status_buffer, 0, sizeof (struct fsf_status_read_buffer)); - fsf_req->data.status_read.buffer = status_buffer; + fsf_req->data = (unsigned long) status_buffer; /* insert pointer to respective buffer */ sbale = zfcp_qdio_sbale_curr(fsf_req); @@ -859,7 +859,7 @@ zfcp_fsf_status_read_port_closed(struct zfcp_fsf_req *fsf_req) struct zfcp_port *port; unsigned long flags; - status_buffer = fsf_req->data.status_read.buffer; + status_buffer = (struct fsf_status_read_buffer *) fsf_req->data; adapter = fsf_req->adapter; read_lock_irqsave(&zfcp_data.config_lock, flags); @@ -918,7 +918,7 @@ zfcp_fsf_status_read_handler(struct zfcp_fsf_req *fsf_req) int retval = 0; struct zfcp_adapter *adapter = fsf_req->adapter; struct fsf_status_read_buffer *status_buffer = - fsf_req->data.status_read.buffer; + (struct fsf_status_read_buffer *) fsf_req->data; if (fsf_req->status & ZFCP_STATUS_FSFREQ_DISMISSED) { mempool_free(status_buffer, adapter->pool.data_status_read); @@ -1093,7 +1093,7 @@ zfcp_fsf_abort_fcp_command(unsigned long old_req_id, sbale[0].flags |= SBAL_FLAGS0_TYPE_READ; sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY; - fsf_req->data.abort_fcp_command.unit = unit; + fsf_req->data = (unsigned long) unit; /* set handles of unit and its parent port in QTCB */ fsf_req->qtcb->header.lun_handle = unit->handle; @@ -1139,7 +1139,7 @@ static int zfcp_fsf_abort_fcp_command_handler(struct zfcp_fsf_req *new_fsf_req) { int retval = -EINVAL; - struct zfcp_unit *unit = new_fsf_req->data.abort_fcp_command.unit; + struct zfcp_unit *unit; unsigned char status_qual = new_fsf_req->qtcb->header.fsf_status_qual.word[0]; @@ -1150,6 +1150,8 @@ zfcp_fsf_abort_fcp_command_handler(struct zfcp_fsf_req *new_fsf_req) goto skip_fsfstatus; } + unit = (struct zfcp_unit *) new_fsf_req->data; + /* evaluate FSF status in QTCB */ switch (new_fsf_req->qtcb->header.fsf_status) { @@ -1414,7 +1416,7 @@ zfcp_fsf_send_ct(struct zfcp_send_ct *ct, mempool_t *pool, fsf_req->qtcb->header.port_handle = port->handle; fsf_req->qtcb->bottom.support.service_class = adapter->fc_service_class; fsf_req->qtcb->bottom.support.timeout = ct->timeout; - fsf_req->data.send_ct = ct; + fsf_req->data = (unsigned long) ct; /* start QDIO request for this FSF request */ ret = zfcp_fsf_req_send(fsf_req, ct->timer); @@ -1445,10 +1447,10 @@ zfcp_fsf_send_ct(struct zfcp_send_ct *ct, mempool_t *pool, * zfcp_fsf_send_ct_handler - handler for Generic Service requests * @fsf_req: pointer to struct zfcp_fsf_req * - * Data specific for the Generic Service request is passed by - * fsf_req->data.send_ct - * Usually a specific handler for the request is called via - * fsf_req->data.send_ct->handler at end of this function. + * Data specific for the Generic Service request is passed using + * fsf_req->data. There we find the pointer to struct zfcp_send_ct. + * Usually a specific handler for the CT request is called which is + * found in this structure. */ static int zfcp_fsf_send_ct_handler(struct zfcp_fsf_req *fsf_req) @@ -1462,7 +1464,7 @@ zfcp_fsf_send_ct_handler(struct zfcp_fsf_req *fsf_req) u16 subtable, rule, counter; adapter = fsf_req->adapter; - send_ct = fsf_req->data.send_ct; + send_ct = (struct zfcp_send_ct *) fsf_req->data; port = send_ct->port; header = &fsf_req->qtcb->header; bottom = &fsf_req->qtcb->bottom.support; @@ -1714,7 +1716,7 @@ zfcp_fsf_send_els(struct zfcp_send_els *els) fsf_req->qtcb->bottom.support.d_id = d_id; fsf_req->qtcb->bottom.support.service_class = adapter->fc_service_class; fsf_req->qtcb->bottom.support.timeout = ZFCP_ELS_TIMEOUT; - fsf_req->data.send_els = els; + fsf_req->data = (unsigned long) els; sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0); @@ -1746,10 +1748,10 @@ zfcp_fsf_send_els(struct zfcp_send_els *els) * zfcp_fsf_send_els_handler - handler for ELS commands * @fsf_req: pointer to struct zfcp_fsf_req * - * Data specific for the ELS command is passed by - * fsf_req->data.send_els - * Usually a specific handler for the command is called via - * fsf_req->data.send_els->handler at end of this function. + * Data specific for the ELS command is passed using + * fsf_req->data. There we find the pointer to struct zfcp_send_els. + * Usually a specific handler for the ELS command is called which is + * found in this structure. */ static int zfcp_fsf_send_els_handler(struct zfcp_fsf_req *fsf_req) { @@ -1762,7 +1764,7 @@ static int zfcp_fsf_send_els_handler(struct zfcp_fsf_req *fsf_req) int retval = -EINVAL; u16 subtable, rule, counter; - send_els = fsf_req->data.send_els; + send_els = (struct zfcp_send_els *) fsf_req->data; adapter = send_els->adapter; port = send_els->port; d_id = send_els->d_id; @@ -2211,12 +2213,12 @@ zfcp_fsf_exchange_port_data(struct zfcp_adapter *adapter, goto out; } + fsf_req->data = (unsigned long) data; + sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0); sbale[0].flags |= SBAL_FLAGS0_TYPE_READ; sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY; - fsf_req->data.port_data = data; - init_timer(timer); timer->function = zfcp_fsf_request_timeout_handler; timer->data = (unsigned long) adapter; @@ -2257,7 +2259,9 @@ static void zfcp_fsf_exchange_port_data_handler(struct zfcp_fsf_req *fsf_req) { struct fsf_qtcb_bottom_port *bottom; - struct fsf_qtcb_bottom_port *data = fsf_req->data.port_data; + struct fsf_qtcb_bottom_port *data; + + data = (struct fsf_qtcb_bottom_port*) fsf_req->data; if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) return; @@ -2312,7 +2316,7 @@ zfcp_fsf_open_port(struct zfcp_erp_action *erp_action) erp_action->fsf_req->qtcb->bottom.support.d_id = erp_action->port->d_id; atomic_set_mask(ZFCP_STATUS_COMMON_OPENING, &erp_action->port->status); - erp_action->fsf_req->data.open_port.port = erp_action->port; + erp_action->fsf_req->data = (unsigned long) erp_action->port; erp_action->fsf_req->erp_action = erp_action; /* start QDIO request for this FSF request */ @@ -2353,7 +2357,7 @@ zfcp_fsf_open_port_handler(struct zfcp_fsf_req *fsf_req) struct fsf_qtcb_header *header; u16 subtable, rule, counter; - port = fsf_req->data.open_port.port; + port = (struct zfcp_port *) fsf_req->data; header = &fsf_req->qtcb->header; if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) { @@ -2566,7 +2570,7 @@ zfcp_fsf_close_port(struct zfcp_erp_action *erp_action) sbale[1].flags |= SBAL_FLAGS_LAST_ENTRY; atomic_set_mask(ZFCP_STATUS_COMMON_CLOSING, &erp_action->port->status); - erp_action->fsf_req->data.close_port.port = erp_action->port; + erp_action->fsf_req->data = (unsigned long) erp_action->port; erp_action->fsf_req->erp_action = erp_action; erp_action->fsf_req->qtcb->header.port_handle = erp_action->port->handle; @@ -2606,7 +2610,7 @@ zfcp_fsf_close_port_handler(struct zfcp_fsf_req *fsf_req) int retval = -EINVAL; struct zfcp_port *port; - port = fsf_req->data.close_port.port; + port = (struct zfcp_port *) fsf_req->data; if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) { /* don't change port status in our bookkeeping */ @@ -2703,7 +2707,7 @@ zfcp_fsf_close_physical_port(struct zfcp_erp_action *erp_action) atomic_set_mask(ZFCP_STATUS_PORT_PHYS_CLOSING, &erp_action->port->status); /* save a pointer to this port */ - erp_action->fsf_req->data.close_physical_port.port = erp_action->port; + erp_action->fsf_req->data = (unsigned long) erp_action->port; /* port to be closeed */ erp_action->fsf_req->qtcb->header.port_handle = erp_action->port->handle; @@ -2747,7 +2751,7 @@ zfcp_fsf_close_physical_port_handler(struct zfcp_fsf_req *fsf_req) struct fsf_qtcb_header *header; u16 subtable, rule, counter; - port = fsf_req->data.close_physical_port.port; + port = (struct zfcp_port *) fsf_req->data; header = &fsf_req->qtcb->header; if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) { @@ -2911,7 +2915,7 @@ zfcp_fsf_open_unit(struct zfcp_erp_action *erp_action) erp_action->fsf_req->qtcb->bottom.support.option = FSF_OPEN_LUN_SUPPRESS_BOXING; atomic_set_mask(ZFCP_STATUS_COMMON_OPENING, &erp_action->unit->status); - erp_action->fsf_req->data.open_unit.unit = erp_action->unit; + erp_action->fsf_req->data = (unsigned long) erp_action->unit; erp_action->fsf_req->erp_action = erp_action; /* start QDIO request for this FSF request */ @@ -2957,7 +2961,7 @@ zfcp_fsf_open_unit_handler(struct zfcp_fsf_req *fsf_req) u16 subtable, rule, counter; u32 allowed, exclusive, readwrite; - unit = fsf_req->data.open_unit.unit; + unit = (struct zfcp_unit *) fsf_req->data; if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) { /* don't change unit status in our bookkeeping */ @@ -3242,7 +3246,7 @@ zfcp_fsf_close_unit(struct zfcp_erp_action *erp_action) erp_action->port->handle; erp_action->fsf_req->qtcb->header.lun_handle = erp_action->unit->handle; atomic_set_mask(ZFCP_STATUS_COMMON_CLOSING, &erp_action->unit->status); - erp_action->fsf_req->data.close_unit.unit = erp_action->unit; + erp_action->fsf_req->data = (unsigned long) erp_action->unit; erp_action->fsf_req->erp_action = erp_action; /* start QDIO request for this FSF request */ @@ -3281,7 +3285,7 @@ zfcp_fsf_close_unit_handler(struct zfcp_fsf_req *fsf_req) int retval = -EINVAL; struct zfcp_unit *unit; - unit = fsf_req->data.close_unit.unit; /* restore unit */ + unit = (struct zfcp_unit *) fsf_req->data; if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) { /* don't change unit status in our bookkeeping */ @@ -3436,21 +3440,14 @@ zfcp_fsf_send_fcp_command_task(struct zfcp_adapter *adapter, goto failed_req_create; } - /* - * associate FSF request with SCSI request - * (need this for look up on abort) - */ - fsf_req->data.send_fcp_command_task.fsf_req = fsf_req; - scsi_cmnd->host_scribble = (char *) &(fsf_req->data); + zfcp_unit_get(unit); + fsf_req->unit = unit; - /* - * associate SCSI command with FSF request - * (need this for look up on normal command completion) - */ - fsf_req->data.send_fcp_command_task.scsi_cmnd = scsi_cmnd; - fsf_req->data.send_fcp_command_task.start_jiffies = jiffies; - fsf_req->data.send_fcp_command_task.unit = unit; - ZFCP_LOG_DEBUG("unit=%p, fcp_lun=0x%016Lx\n", unit, unit->fcp_lun); + /* associate FSF request with SCSI request (for look up on abort) */ + scsi_cmnd->host_scribble = (char *) fsf_req; + + /* associate SCSI command with FSF request */ + fsf_req->data = (unsigned long) scsi_cmnd; /* set handles of unit and its parent port in QTCB */ fsf_req->qtcb->header.lun_handle = unit->handle; @@ -3584,6 +3581,7 @@ zfcp_fsf_send_fcp_command_task(struct zfcp_adapter *adapter, send_failed: no_fit: failed_scsi_cmnd: + zfcp_unit_put(unit); zfcp_fsf_req_free(fsf_req); fsf_req = NULL; scsi_cmnd->host_scribble = NULL; @@ -3640,7 +3638,7 @@ zfcp_fsf_send_fcp_command_task_management(struct zfcp_adapter *adapter, * hold a pointer to the unit being target of this * task management request */ - fsf_req->data.send_fcp_command_task_management.unit = unit; + fsf_req->data = (unsigned long) unit; /* set FSF related fields in QTCB */ fsf_req->qtcb->header.lun_handle = unit->handle; @@ -3706,9 +3704,9 @@ zfcp_fsf_send_fcp_command_handler(struct zfcp_fsf_req *fsf_req) header = &fsf_req->qtcb->header; if (unlikely(fsf_req->status & ZFCP_STATUS_FSFREQ_TASK_MANAGEMENT)) - unit = fsf_req->data.send_fcp_command_task_management.unit; + unit = (struct zfcp_unit *) fsf_req->data; else - unit = fsf_req->data.send_fcp_command_task.unit; + unit = fsf_req->unit; if (unlikely(fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR)) { /* go directly to calls of special handlers */ @@ -3947,6 +3945,8 @@ zfcp_fsf_send_fcp_command_handler(struct zfcp_fsf_req *fsf_req) zfcp_fsf_send_fcp_command_task_management_handler(fsf_req); } else { retval = zfcp_fsf_send_fcp_command_task_handler(fsf_req); + fsf_req->unit = NULL; + zfcp_unit_put(unit); } return retval; } @@ -3970,10 +3970,10 @@ zfcp_fsf_send_fcp_command_task_handler(struct zfcp_fsf_req *fsf_req) u32 sns_len; char *fcp_rsp_info = zfcp_get_fcp_rsp_info_ptr(fcp_rsp_iu); unsigned long flags; - struct zfcp_unit *unit = fsf_req->data.send_fcp_command_task.unit; + struct zfcp_unit *unit = fsf_req->unit; read_lock_irqsave(&fsf_req->adapter->abort_lock, flags); - scpnt = fsf_req->data.send_fcp_command_task.scsi_cmnd; + scpnt = (struct scsi_cmnd *) fsf_req->data; if (unlikely(!scpnt)) { ZFCP_LOG_DEBUG ("Command with fsf_req %p is not associated to " @@ -4198,8 +4198,7 @@ zfcp_fsf_send_fcp_command_task_management_handler(struct zfcp_fsf_req *fsf_req) struct fcp_rsp_iu *fcp_rsp_iu = (struct fcp_rsp_iu *) &(fsf_req->qtcb->bottom.io.fcp_rsp); char *fcp_rsp_info = zfcp_get_fcp_rsp_info_ptr(fcp_rsp_iu); - struct zfcp_unit *unit = - fsf_req->data.send_fcp_command_task_management.unit; + struct zfcp_unit *unit = (struct zfcp_unit *) fsf_req->data; del_timer(&fsf_req->adapter->scsi_er_timer); if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) { diff --git a/drivers/s390/scsi/zfcp_scsi.c b/drivers/s390/scsi/zfcp_scsi.c index 31a76065cf2..fffd1239901 100644 --- a/drivers/s390/scsi/zfcp_scsi.c +++ b/drivers/s390/scsi/zfcp_scsi.c @@ -414,67 +414,37 @@ zfcp_port_lookup(struct zfcp_adapter *adapter, int channel, scsi_id_t id) return (struct zfcp_port *) NULL; } -/* - * function: zfcp_scsi_eh_abort_handler +/** + * zfcp_scsi_eh_abort_handler - abort the specified SCSI command + * @scpnt: pointer to scsi_cmnd to be aborted + * Return: SUCCESS - command has been aborted and cleaned up in internal + * bookkeeping, SCSI stack won't be called for aborted command + * FAILED - otherwise * - * purpose: tries to abort the specified (timed out) SCSI command - * - * note: We do not need to care for a SCSI command which completes - * normally but late during this abort routine runs. - * We are allowed to return late commands to the SCSI stack. - * It tracks the state of commands and will handle late commands. - * (Usually, the normal completion of late commands is ignored with - * respect to the running abort operation. Grep for 'done_late' - * in the SCSI stacks sources.) - * - * returns: SUCCESS - command has been aborted and cleaned up in internal - * bookkeeping, - * SCSI stack won't be called for aborted command - * FAILED - otherwise + * We do not need to care for a SCSI command which completes normally + * but late during this abort routine runs. We are allowed to return + * late commands to the SCSI stack. It tracks the state of commands and + * will handle late commands. (Usually, the normal completion of late + * commands is ignored with respect to the running abort operation.) */ int -__zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt) +zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt) { + struct Scsi_Host *scsi_host; + struct zfcp_adapter *adapter; + struct zfcp_unit *unit; int retval = SUCCESS; struct zfcp_fsf_req *new_fsf_req, *old_fsf_req; - struct zfcp_adapter *adapter = (struct zfcp_adapter *) scpnt->device->host->hostdata[0]; - struct zfcp_unit *unit = (struct zfcp_unit *) scpnt->device->hostdata; - struct zfcp_port *port = unit->port; - struct Scsi_Host *scsi_host = scpnt->device->host; - union zfcp_req_data *req_data = NULL; unsigned long flags; - u32 status = 0; - /* the components of a abort_dbf record (fixed size record) */ - u64 dbf_scsi_cmnd = (unsigned long) scpnt; - char dbf_opcode[ZFCP_ABORT_DBF_LENGTH]; - wwn_t dbf_wwn = port->wwpn; - fcp_lun_t dbf_fcp_lun = unit->fcp_lun; - u64 dbf_retries = scpnt->retries; - u64 dbf_allowed = scpnt->allowed; - u64 dbf_timeout = 0; - u64 dbf_fsf_req = 0; - u64 dbf_fsf_status = 0; - u64 dbf_fsf_qual[2] = { 0, 0 }; - char dbf_result[ZFCP_ABORT_DBF_LENGTH] = "##undef"; - - memset(dbf_opcode, 0, ZFCP_ABORT_DBF_LENGTH); - memcpy(dbf_opcode, - scpnt->cmnd, - min(scpnt->cmd_len, (unsigned char) ZFCP_ABORT_DBF_LENGTH)); + scsi_host = scpnt->device->host; + adapter = (struct zfcp_adapter *) scsi_host->hostdata[0]; + unit = (struct zfcp_unit *) scpnt->device->hostdata; ZFCP_LOG_INFO("aborting scsi_cmnd=%p on adapter %s\n", scpnt, zfcp_get_busid_by_adapter(adapter)); - spin_unlock_irq(scsi_host->host_lock); - - /* - * Race condition between normal (late) completion and abort has - * to be avoided. - * The entirity of all accesses to scsi_req have to be atomic. - * scsi_req is usually part of the fsf_req and thus we block the - * release of fsf_req as long as we need to access scsi_req. - */ + /* avoid race condition between late normal completion and abort */ write_lock_irqsave(&adapter->abort_lock, flags); /* @@ -484,144 +454,48 @@ __zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt) * this routine returns. (scpnt is parameter passed to this routine * and must not disappear during abort even on late completion.) */ - req_data = (union zfcp_req_data *) scpnt->host_scribble; - /* DEBUG */ - ZFCP_LOG_DEBUG("req_data=%p\n", req_data); - if (!req_data) { - ZFCP_LOG_DEBUG("late command completion overtook abort\n"); - /* - * That's it. - * Do not initiate abort but return SUCCESS. - */ - write_unlock_irqrestore(&adapter->abort_lock, flags); - retval = SUCCESS; - strncpy(dbf_result, "##late1", ZFCP_ABORT_DBF_LENGTH); - goto out; - } - - /* Figure out which fsf_req needs to be aborted. */ - old_fsf_req = req_data->send_fcp_command_task.fsf_req; - - dbf_fsf_req = (unsigned long) old_fsf_req; - dbf_timeout = - (jiffies - req_data->send_fcp_command_task.start_jiffies) / HZ; - - ZFCP_LOG_DEBUG("old_fsf_req=%p\n", old_fsf_req); + old_fsf_req = (struct zfcp_fsf_req *) scpnt->host_scribble; if (!old_fsf_req) { write_unlock_irqrestore(&adapter->abort_lock, flags); ZFCP_LOG_NORMAL("bug: no old fsf request found\n"); - ZFCP_LOG_NORMAL("req_data:\n"); - ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_NORMAL, - (char *) req_data, sizeof (union zfcp_req_data)); ZFCP_LOG_NORMAL("scsi_cmnd:\n"); ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_NORMAL, (char *) scpnt, sizeof (struct scsi_cmnd)); retval = FAILED; - strncpy(dbf_result, "##bug:r", ZFCP_ABORT_DBF_LENGTH); goto out; } - old_fsf_req->data.send_fcp_command_task.scsi_cmnd = NULL; - /* mark old request as being aborted */ + old_fsf_req->data = 0; old_fsf_req->status |= ZFCP_STATUS_FSFREQ_ABORTING; - /* - * We have to collect all information (e.g. unit) needed by - * zfcp_fsf_abort_fcp_command before calling that routine - * since that routine is not allowed to access - * fsf_req which it is going to abort. - * This is because of we need to release fsf_req_list_lock - * before calling zfcp_fsf_abort_fcp_command. - * Since this lock will not be held, fsf_req may complete - * late and may be released meanwhile. - */ - ZFCP_LOG_DEBUG("unit 0x%016Lx (%p)\n", unit->fcp_lun, unit); - /* - * We block (call schedule) - * That's why we must release the lock and enable the - * interrupts before. - * On the other hand we do not need the lock anymore since - * all critical accesses to scsi_req are done. - */ + /* don't access old_fsf_req after releasing the abort_lock */ write_unlock_irqrestore(&adapter->abort_lock, flags); /* call FSF routine which does the abort */ new_fsf_req = zfcp_fsf_abort_fcp_command((unsigned long) old_fsf_req, adapter, unit, 0); - ZFCP_LOG_DEBUG("new_fsf_req=%p\n", new_fsf_req); if (!new_fsf_req) { retval = FAILED; ZFCP_LOG_NORMAL("error: initiation of Abort FCP Cmnd " "failed\n"); - strncpy(dbf_result, "##nores", ZFCP_ABORT_DBF_LENGTH); goto out; } /* wait for completion of abort */ - ZFCP_LOG_DEBUG("waiting for cleanup...\n"); -#if 1 - /* - * FIXME: - * copying zfcp_fsf_req_wait_and_cleanup code is not really nice - */ __wait_event(new_fsf_req->completion_wq, new_fsf_req->status & ZFCP_STATUS_FSFREQ_COMPLETED); - status = new_fsf_req->status; - dbf_fsf_status = new_fsf_req->qtcb->header.fsf_status; - /* - * Ralphs special debug load provides timestamps in the FSF - * status qualifier. This might be specified later if being - * useful for debugging aborts. - */ - dbf_fsf_qual[0] = - *(u64 *) & new_fsf_req->qtcb->header.fsf_status_qual.word[0]; - dbf_fsf_qual[1] = - *(u64 *) & new_fsf_req->qtcb->header.fsf_status_qual.word[2]; zfcp_fsf_req_free(new_fsf_req); -#else - retval = zfcp_fsf_req_wait_and_cleanup(new_fsf_req, - ZFCP_UNINTERRUPTIBLE, &status); -#endif - ZFCP_LOG_DEBUG("Waiting for cleanup complete, status=0x%x\n", status); + /* status should be valid since signals were not permitted */ - if (status & ZFCP_STATUS_FSFREQ_ABORTSUCCEEDED) { + if (new_fsf_req->status & ZFCP_STATUS_FSFREQ_ABORTSUCCEEDED) { retval = SUCCESS; - strncpy(dbf_result, "##succ", ZFCP_ABORT_DBF_LENGTH); - } else if (status & ZFCP_STATUS_FSFREQ_ABORTNOTNEEDED) { + } else if (new_fsf_req->status & ZFCP_STATUS_FSFREQ_ABORTNOTNEEDED) { retval = SUCCESS; - strncpy(dbf_result, "##late2", ZFCP_ABORT_DBF_LENGTH); } else { retval = FAILED; - strncpy(dbf_result, "##fail", ZFCP_ABORT_DBF_LENGTH); } - out: - debug_event(adapter->abort_dbf, 1, &dbf_scsi_cmnd, sizeof (u64)); - debug_event(adapter->abort_dbf, 1, &dbf_opcode, ZFCP_ABORT_DBF_LENGTH); - debug_event(adapter->abort_dbf, 1, &dbf_wwn, sizeof (wwn_t)); - debug_event(adapter->abort_dbf, 1, &dbf_fcp_lun, sizeof (fcp_lun_t)); - debug_event(adapter->abort_dbf, 1, &dbf_retries, sizeof (u64)); - debug_event(adapter->abort_dbf, 1, &dbf_allowed, sizeof (u64)); - debug_event(adapter->abort_dbf, 1, &dbf_timeout, sizeof (u64)); - debug_event(adapter->abort_dbf, 1, &dbf_fsf_req, sizeof (u64)); - debug_event(adapter->abort_dbf, 1, &dbf_fsf_status, sizeof (u64)); - debug_event(adapter->abort_dbf, 1, &dbf_fsf_qual[0], sizeof (u64)); - debug_event(adapter->abort_dbf, 1, &dbf_fsf_qual[1], sizeof (u64)); - debug_text_event(adapter->abort_dbf, 1, dbf_result); - - spin_lock_irq(scsi_host->host_lock); return retval; } -int -zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt) -{ - int rc; - struct Scsi_Host *scsi_host = scpnt->device->host; - spin_lock_irq(scsi_host->host_lock); - rc = __zfcp_scsi_eh_abort_handler(scpnt); - spin_unlock_irq(scsi_host->host_lock); - return rc; -} - /* * function: zfcp_scsi_eh_device_reset_handler * From 77eb1699c76177af2f3d65c8ae7934cf304e0254 Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Tue, 13 Sep 2005 21:48:33 +0200 Subject: [PATCH 038/204] [SCSI] zfcp: remove function zfcp_fsf_req_wait_and_cleanup Signed-off-by: Andreas Herrmann Signed-off-by: James Bottomley --- drivers/s390/scsi/zfcp_ext.h | 1 - drivers/s390/scsi/zfcp_fsf.c | 46 ----------------------------------- drivers/s390/scsi/zfcp_scsi.c | 21 +++++++--------- 3 files changed, 9 insertions(+), 59 deletions(-) diff --git a/drivers/s390/scsi/zfcp_ext.h b/drivers/s390/scsi/zfcp_ext.h index cd98a2de9f8..c400e3b9de9 100644 --- a/drivers/s390/scsi/zfcp_ext.h +++ b/drivers/s390/scsi/zfcp_ext.h @@ -109,7 +109,6 @@ extern int zfcp_fsf_req_create(struct zfcp_adapter *, u32, int, mempool_t *, extern int zfcp_fsf_send_ct(struct zfcp_send_ct *, mempool_t *, struct zfcp_erp_action *); extern int zfcp_fsf_send_els(struct zfcp_send_els *); -extern int zfcp_fsf_req_wait_and_cleanup(struct zfcp_fsf_req *, int, u32 *); extern int zfcp_fsf_send_fcp_command_task(struct zfcp_adapter *, struct zfcp_unit *, struct scsi_cmnd *, diff --git a/drivers/s390/scsi/zfcp_fsf.c b/drivers/s390/scsi/zfcp_fsf.c index 64e2c348018..929f6c9cad3 100644 --- a/drivers/s390/scsi/zfcp_fsf.c +++ b/drivers/s390/scsi/zfcp_fsf.c @@ -4548,52 +4548,6 @@ skip_fsfstatus: return retval; } - -/* - * function: zfcp_fsf_req_wait_and_cleanup - * - * purpose: - * - * FIXME(design): signal seems to be <0 !!! - * returns: 0 - request completed (*status is valid), cleanup succ. - * <0 - request completed (*status is valid), cleanup failed - * >0 - signal which interrupted waiting (*status invalid), - * request not completed, no cleanup - * - * *status is a copy of status of completed fsf_req - */ -int -zfcp_fsf_req_wait_and_cleanup(struct zfcp_fsf_req *fsf_req, - int interruptible, u32 * status) -{ - int retval = 0; - int signal = 0; - - if (interruptible) { - __wait_event_interruptible(fsf_req->completion_wq, - fsf_req->status & - ZFCP_STATUS_FSFREQ_COMPLETED, - signal); - if (signal) { - ZFCP_LOG_DEBUG("Caught signal %i while waiting for the " - "completion of the request at %p\n", - signal, fsf_req); - retval = signal; - goto out; - } - } else { - __wait_event(fsf_req->completion_wq, - fsf_req->status & ZFCP_STATUS_FSFREQ_COMPLETED); - } - - *status = fsf_req->status; - - /* cleanup request */ - zfcp_fsf_req_free(fsf_req); - out: - return retval; -} - static inline int zfcp_fsf_req_sbal_check(unsigned long *flags, struct zfcp_qdio_queue *queue, int needed) diff --git a/drivers/s390/scsi/zfcp_scsi.c b/drivers/s390/scsi/zfcp_scsi.c index fffd1239901..3b5d48c6153 100644 --- a/drivers/s390/scsi/zfcp_scsi.c +++ b/drivers/s390/scsi/zfcp_scsi.c @@ -558,9 +558,8 @@ static int zfcp_task_management_function(struct zfcp_unit *unit, u8 tm_flags) { struct zfcp_adapter *adapter = unit->port->adapter; - int retval; - int status; struct zfcp_fsf_req *fsf_req; + int retval = 0; /* issue task management function */ fsf_req = zfcp_fsf_send_fcp_command_task_management @@ -574,18 +573,16 @@ zfcp_task_management_function(struct zfcp_unit *unit, u8 tm_flags) goto out; } - retval = zfcp_fsf_req_wait_and_cleanup(fsf_req, - ZFCP_UNINTERRUPTIBLE, &status); - /* - * check completion status of task management function - * (status should always be valid since no signals permitted) - */ - if (status & ZFCP_STATUS_FSFREQ_TMFUNCFAILED) + __wait_event(fsf_req->completion_wq, + fsf_req->status & ZFCP_STATUS_FSFREQ_COMPLETED); + + /* check completion status of task management function */ + if (fsf_req->status & ZFCP_STATUS_FSFREQ_TMFUNCFAILED) retval = -EIO; - else if (status & ZFCP_STATUS_FSFREQ_TMFUNCNOTSUPP) + else if (fsf_req->status & ZFCP_STATUS_FSFREQ_TMFUNCNOTSUPP) retval = -ENOTSUPP; - else - retval = 0; + + zfcp_fsf_req_free(fsf_req); out: return retval; } From 810f1e3ea5cc0a812816af97020a27c73441f8e9 Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Tue, 13 Sep 2005 21:49:52 +0200 Subject: [PATCH 039/204] [SCSI] zfcp: shorten eh_bus_reset and eh_host_reset handlers Signed-off-by: Andreas Herrmann Signed-off-by: James Bottomley --- drivers/s390/scsi/zfcp_scsi.c | 40 ++++++++++++----------------------- 1 file changed, 14 insertions(+), 26 deletions(-) diff --git a/drivers/s390/scsi/zfcp_scsi.c b/drivers/s390/scsi/zfcp_scsi.c index 3b5d48c6153..eeb5152854e 100644 --- a/drivers/s390/scsi/zfcp_scsi.c +++ b/drivers/s390/scsi/zfcp_scsi.c @@ -587,50 +587,38 @@ zfcp_task_management_function(struct zfcp_unit *unit, u8 tm_flags) return retval; } -/* - * function: zfcp_scsi_eh_bus_reset_handler - * - * purpose: - * - * returns: +/** + * zfcp_scsi_eh_bus_reset_handler - reset bus (reopen adapter) */ int zfcp_scsi_eh_bus_reset_handler(struct scsi_cmnd *scpnt) { - int retval = 0; - struct zfcp_unit *unit; + struct zfcp_unit *unit = (struct zfcp_unit*) scpnt->device->hostdata; + struct zfcp_adapter *adapter = unit->port->adapter; - unit = (struct zfcp_unit *) scpnt->device->hostdata; ZFCP_LOG_NORMAL("bus reset because of problems with " "unit 0x%016Lx\n", unit->fcp_lun); - zfcp_erp_adapter_reopen(unit->port->adapter, 0); - zfcp_erp_wait(unit->port->adapter); - retval = SUCCESS; + zfcp_erp_adapter_reopen(adapter, 0); + zfcp_erp_wait(adapter); - return retval; + return SUCCESS; } -/* - * function: zfcp_scsi_eh_host_reset_handler - * - * purpose: - * - * returns: +/** + * zfcp_scsi_eh_host_reset_handler - reset host (reopen adapter) */ int zfcp_scsi_eh_host_reset_handler(struct scsi_cmnd *scpnt) { - int retval = 0; - struct zfcp_unit *unit; + struct zfcp_unit *unit = (struct zfcp_unit*) scpnt->device->hostdata; + struct zfcp_adapter *adapter = unit->port->adapter; - unit = (struct zfcp_unit *) scpnt->device->hostdata; ZFCP_LOG_NORMAL("host reset because of problems with " "unit 0x%016Lx\n", unit->fcp_lun); - zfcp_erp_adapter_reopen(unit->port->adapter, 0); - zfcp_erp_wait(unit->port->adapter); - retval = SUCCESS; + zfcp_erp_adapter_reopen(adapter, 0); + zfcp_erp_wait(adapter); - return retval; + return SUCCESS; } /* From 8a36e4532ea10471f0a8605207d071361d7be2c3 Mon Sep 17 00:00:00 2001 From: Maxim Shchetynin Date: Tue, 13 Sep 2005 21:50:38 +0200 Subject: [PATCH 040/204] [SCSI] zfcp: enhancement of zfcp debug features Debug features (DBFs) els_dbf, cmd_dbf and abt_dbf were removed and san_dbf, hba_dbf and scsi_dbf were introduced. The erp_dbf did not change. The new traces improve debugging of problems with zfcp, scsi-stack, multipath and hardware in the SAN. san_dbf traces things like ELS and CT commands, hba_dbf saves HBA specific information of requests, and scsi_dbf saves FCP and SCSI specific information of requests. Common to all new DBFs is that they provide a so called structured view. This significantly improves readability of the traces. Signed-off-by: Andreas Herrmann Signed-off-by: James Bottomley --- drivers/s390/scsi/Makefile | 2 +- drivers/s390/scsi/zfcp_aux.c | 180 +------ drivers/s390/scsi/zfcp_dbf.c | 988 ++++++++++++++++++++++++++++++++++ drivers/s390/scsi/zfcp_def.h | 213 ++++++-- drivers/s390/scsi/zfcp_ext.h | 26 +- drivers/s390/scsi/zfcp_fsf.c | 241 +++------ drivers/s390/scsi/zfcp_fsf.h | 3 + drivers/s390/scsi/zfcp_qdio.c | 30 +- drivers/s390/scsi/zfcp_scsi.c | 49 +- 9 files changed, 1297 insertions(+), 435 deletions(-) create mode 100644 drivers/s390/scsi/zfcp_dbf.c diff --git a/drivers/s390/scsi/Makefile b/drivers/s390/scsi/Makefile index fc145307a7d..d6a78f1a2f1 100644 --- a/drivers/s390/scsi/Makefile +++ b/drivers/s390/scsi/Makefile @@ -3,7 +3,7 @@ # zfcp-objs := zfcp_aux.o zfcp_ccw.o zfcp_scsi.o zfcp_erp.o zfcp_qdio.o \ - zfcp_fsf.o zfcp_sysfs_adapter.o zfcp_sysfs_port.o \ + zfcp_fsf.o zfcp_dbf.o zfcp_sysfs_adapter.o zfcp_sysfs_port.o \ zfcp_sysfs_unit.o zfcp_sysfs_driver.o obj-$(CONFIG_ZFCP) += zfcp.o diff --git a/drivers/s390/scsi/zfcp_aux.c b/drivers/s390/scsi/zfcp_aux.c index 17f9989238f..0b5087f7cab 100644 --- a/drivers/s390/scsi/zfcp_aux.c +++ b/drivers/s390/scsi/zfcp_aux.c @@ -122,93 +122,6 @@ _zfcp_hex_dump(char *addr, int count) #define ZFCP_LOG_AREA ZFCP_LOG_AREA_OTHER -static inline int -zfcp_fsf_req_is_scsi_cmnd(struct zfcp_fsf_req *fsf_req) -{ - return ((fsf_req->fsf_command == FSF_QTCB_FCP_CMND) && - !(fsf_req->status & ZFCP_STATUS_FSFREQ_TASK_MANAGEMENT)); -} - -void -zfcp_cmd_dbf_event_fsf(const char *text, struct zfcp_fsf_req *fsf_req, - void *add_data, int add_length) -{ - struct zfcp_adapter *adapter = fsf_req->adapter; - struct scsi_cmnd *scsi_cmnd; - int level = 3; - int i; - unsigned long flags; - - spin_lock_irqsave(&adapter->dbf_lock, flags); - if (zfcp_fsf_req_is_scsi_cmnd(fsf_req)) { - scsi_cmnd = (struct scsi_cmnd*) fsf_req->data; - debug_text_event(adapter->cmd_dbf, level, "fsferror"); - debug_text_event(adapter->cmd_dbf, level, text); - debug_event(adapter->cmd_dbf, level, &fsf_req, - sizeof (unsigned long)); - debug_event(adapter->cmd_dbf, level, &fsf_req->seq_no, - sizeof (u32)); - debug_event(adapter->cmd_dbf, level, &scsi_cmnd, - sizeof (unsigned long)); - debug_event(adapter->cmd_dbf, level, &scsi_cmnd->cmnd, - min(ZFCP_CMD_DBF_LENGTH, (int)scsi_cmnd->cmd_len)); - for (i = 0; i < add_length; i += ZFCP_CMD_DBF_LENGTH) - debug_event(adapter->cmd_dbf, - level, - (char *) add_data + i, - min(ZFCP_CMD_DBF_LENGTH, add_length - i)); - } - spin_unlock_irqrestore(&adapter->dbf_lock, flags); -} - -/* XXX additionally log unit if available */ -/* ---> introduce new parameter for unit, see 2.4 code */ -void -zfcp_cmd_dbf_event_scsi(const char *text, struct scsi_cmnd *scsi_cmnd) -{ - struct zfcp_adapter *adapter; - struct zfcp_fsf_req *fsf_req; - int level = ((host_byte(scsi_cmnd->result) != 0) ? 1 : 5); - unsigned long flags; - - adapter = (struct zfcp_adapter *) scsi_cmnd->device->host->hostdata[0]; - fsf_req = (struct zfcp_fsf_req *) scsi_cmnd->host_scribble; - spin_lock_irqsave(&adapter->dbf_lock, flags); - debug_text_event(adapter->cmd_dbf, level, "hostbyte"); - debug_text_event(adapter->cmd_dbf, level, text); - debug_event(adapter->cmd_dbf, level, &scsi_cmnd->result, sizeof (u32)); - debug_event(adapter->cmd_dbf, level, &scsi_cmnd, - sizeof (unsigned long)); - debug_event(adapter->cmd_dbf, level, &scsi_cmnd->cmnd, - min(ZFCP_CMD_DBF_LENGTH, (int)scsi_cmnd->cmd_len)); - if (likely(fsf_req)) { - debug_event(adapter->cmd_dbf, level, &fsf_req, - sizeof (unsigned long)); - debug_event(adapter->cmd_dbf, level, &fsf_req->seq_no, - sizeof (u32)); - } else { - debug_text_event(adapter->cmd_dbf, level, ""); - debug_text_event(adapter->cmd_dbf, level, ""); - } - spin_unlock_irqrestore(&adapter->dbf_lock, flags); -} - -void -zfcp_in_els_dbf_event(struct zfcp_adapter *adapter, const char *text, - struct fsf_status_read_buffer *status_buffer, int length) -{ - int level = 1; - int i; - - debug_text_event(adapter->in_els_dbf, level, text); - debug_event(adapter->in_els_dbf, level, &status_buffer->d_id, 8); - for (i = 0; i < length; i += ZFCP_IN_ELS_DBF_LENGTH) - debug_event(adapter->in_els_dbf, - level, - (char *) status_buffer->payload + i, - min(ZFCP_IN_ELS_DBF_LENGTH, length - i)); -} - /** * zfcp_device_setup - setup function * @str: pointer to parameter string @@ -1015,81 +928,6 @@ zfcp_free_low_mem_buffers(struct zfcp_adapter *adapter) mempool_destroy(adapter->pool.data_gid_pn); } -/** - * zfcp_adapter_debug_register - registers debug feature for an adapter - * @adapter: pointer to adapter for which debug features should be registered - * return: -ENOMEM on error, 0 otherwise - */ -int -zfcp_adapter_debug_register(struct zfcp_adapter *adapter) -{ - char dbf_name[20]; - - /* debug feature area which records SCSI command failures (hostbyte) */ - spin_lock_init(&adapter->dbf_lock); - - sprintf(dbf_name, ZFCP_CMD_DBF_NAME "%s", - zfcp_get_busid_by_adapter(adapter)); - adapter->cmd_dbf = debug_register(dbf_name, ZFCP_CMD_DBF_INDEX, - ZFCP_CMD_DBF_AREAS, - ZFCP_CMD_DBF_LENGTH); - debug_register_view(adapter->cmd_dbf, &debug_hex_ascii_view); - debug_set_level(adapter->cmd_dbf, ZFCP_CMD_DBF_LEVEL); - - /* debug feature area which records SCSI command aborts */ - sprintf(dbf_name, ZFCP_ABORT_DBF_NAME "%s", - zfcp_get_busid_by_adapter(adapter)); - adapter->abort_dbf = debug_register(dbf_name, ZFCP_ABORT_DBF_INDEX, - ZFCP_ABORT_DBF_AREAS, - ZFCP_ABORT_DBF_LENGTH); - debug_register_view(adapter->abort_dbf, &debug_hex_ascii_view); - debug_set_level(adapter->abort_dbf, ZFCP_ABORT_DBF_LEVEL); - - /* debug feature area which records incoming ELS commands */ - sprintf(dbf_name, ZFCP_IN_ELS_DBF_NAME "%s", - zfcp_get_busid_by_adapter(adapter)); - adapter->in_els_dbf = debug_register(dbf_name, ZFCP_IN_ELS_DBF_INDEX, - ZFCP_IN_ELS_DBF_AREAS, - ZFCP_IN_ELS_DBF_LENGTH); - debug_register_view(adapter->in_els_dbf, &debug_hex_ascii_view); - debug_set_level(adapter->in_els_dbf, ZFCP_IN_ELS_DBF_LEVEL); - - /* debug feature area which records erp events */ - sprintf(dbf_name, ZFCP_ERP_DBF_NAME "%s", - zfcp_get_busid_by_adapter(adapter)); - adapter->erp_dbf = debug_register(dbf_name, ZFCP_ERP_DBF_INDEX, - ZFCP_ERP_DBF_AREAS, - ZFCP_ERP_DBF_LENGTH); - debug_register_view(adapter->erp_dbf, &debug_hex_ascii_view); - debug_set_level(adapter->erp_dbf, ZFCP_ERP_DBF_LEVEL); - - if (!(adapter->cmd_dbf && adapter->abort_dbf && - adapter->in_els_dbf && adapter->erp_dbf)) { - zfcp_adapter_debug_unregister(adapter); - return -ENOMEM; - } - - return 0; - -} - -/** - * zfcp_adapter_debug_unregister - unregisters debug feature for an adapter - * @adapter: pointer to adapter for which debug features should be unregistered - */ -void -zfcp_adapter_debug_unregister(struct zfcp_adapter *adapter) -{ - debug_unregister(adapter->abort_dbf); - debug_unregister(adapter->cmd_dbf); - debug_unregister(adapter->erp_dbf); - debug_unregister(adapter->in_els_dbf); - adapter->abort_dbf = NULL; - adapter->cmd_dbf = NULL; - adapter->erp_dbf = NULL; - adapter->in_els_dbf = NULL; -} - void zfcp_dummy_release(struct device *dev) { @@ -1460,10 +1298,6 @@ zfcp_fsf_incoming_els_rscn(struct zfcp_adapter *adapter, /* see FC-FS */ no_entries = (fcp_rscn_head->payload_len / 4); - zfcp_in_els_dbf_event(adapter, "##rscn", status_buffer, - fcp_rscn_head->payload_len); - - debug_text_event(adapter->erp_dbf, 1, "unsol_els_rscn:"); for (i = 1; i < no_entries; i++) { /* skip head and start with 1st element */ fcp_rscn_element++; @@ -1495,8 +1329,6 @@ zfcp_fsf_incoming_els_rscn(struct zfcp_adapter *adapter, (ZFCP_STATUS_PORT_DID_DID, &port->status)) { ZFCP_LOG_INFO("incoming RSCN, trying to open " "port 0x%016Lx\n", port->wwpn); - debug_text_event(adapter->erp_dbf, 1, - "unsol_els_rscnu:"); zfcp_erp_port_reopen(port, ZFCP_STATUS_COMMON_ERP_FAILED); continue; @@ -1522,8 +1354,6 @@ zfcp_fsf_incoming_els_rscn(struct zfcp_adapter *adapter, */ ZFCP_LOG_INFO("incoming RSCN, trying to open " "port 0x%016Lx\n", port->wwpn); - debug_text_event(adapter->erp_dbf, 1, - "unsol_els_rscnk:"); zfcp_test_link(port); } } @@ -1539,8 +1369,6 @@ zfcp_fsf_incoming_els_plogi(struct zfcp_adapter *adapter, struct zfcp_port *port; unsigned long flags; - zfcp_in_els_dbf_event(adapter, "##plogi", status_buffer, 28); - read_lock_irqsave(&zfcp_data.config_lock, flags); list_for_each_entry(port, &adapter->port_list_head, list) { if (port->wwpn == (*(wwn_t *) & els_logi->nport_wwn)) @@ -1554,8 +1382,6 @@ zfcp_fsf_incoming_els_plogi(struct zfcp_adapter *adapter, status_buffer->d_id, zfcp_get_busid_by_adapter(adapter)); } else { - debug_text_event(adapter->erp_dbf, 1, "unsol_els_plogi:"); - debug_event(adapter->erp_dbf, 1, &els_logi->nport_wwn, 8); zfcp_erp_port_forced_reopen(port, 0); } } @@ -1568,8 +1394,6 @@ zfcp_fsf_incoming_els_logo(struct zfcp_adapter *adapter, struct zfcp_port *port; unsigned long flags; - zfcp_in_els_dbf_event(adapter, "##logo", status_buffer, 16); - read_lock_irqsave(&zfcp_data.config_lock, flags); list_for_each_entry(port, &adapter->port_list_head, list) { if (port->wwpn == els_logo->nport_wwpn) @@ -1583,8 +1407,6 @@ zfcp_fsf_incoming_els_logo(struct zfcp_adapter *adapter, status_buffer->d_id, zfcp_get_busid_by_adapter(adapter)); } else { - debug_text_event(adapter->erp_dbf, 1, "unsol_els_logo:"); - debug_event(adapter->erp_dbf, 1, &els_logo->nport_wwpn, 8); zfcp_erp_port_forced_reopen(port, 0); } } @@ -1593,7 +1415,6 @@ static void zfcp_fsf_incoming_els_unknown(struct zfcp_adapter *adapter, struct fsf_status_read_buffer *status_buffer) { - zfcp_in_els_dbf_event(adapter, "##undef", status_buffer, 24); ZFCP_LOG_NORMAL("warning: unknown incoming ELS 0x%08x " "for adapter %s\n", *(u32 *) (status_buffer->payload), zfcp_get_busid_by_adapter(adapter)); @@ -1611,6 +1432,7 @@ zfcp_fsf_incoming_els(struct zfcp_fsf_req *fsf_req) els_type = *(u32 *) (status_buffer->payload); adapter = fsf_req->adapter; + zfcp_san_dbf_event_incoming_els(fsf_req); if (els_type == LS_PLOGI) zfcp_fsf_incoming_els_plogi(adapter, status_buffer); else if (els_type == LS_LOGO) diff --git a/drivers/s390/scsi/zfcp_dbf.c b/drivers/s390/scsi/zfcp_dbf.c new file mode 100644 index 00000000000..fff1537335c --- /dev/null +++ b/drivers/s390/scsi/zfcp_dbf.c @@ -0,0 +1,988 @@ +/* + * + * linux/drivers/s390/scsi/zfcp_dbf.c + * + * FCP adapter driver for IBM eServer zSeries + * + * Debugging facilities + * + * (C) Copyright IBM Corp. 2005 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2, or (at your option) + * any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + */ + +#define ZFCP_DBF_REVISION "$Revision$" + +#include +#include +#include "zfcp_ext.h" + +static u32 dbfsize = 4; + +module_param(dbfsize, uint, 0400); +MODULE_PARM_DESC(dbfsize, + "number of pages for each debug feature area (default 4)"); + +#define ZFCP_LOG_AREA ZFCP_LOG_AREA_OTHER + +static inline int +zfcp_dbf_stck(char *out_buf, const char *label, unsigned long long stck) +{ + unsigned long long sec; + struct timespec xtime; + int len = 0; + + stck -= 0x8126d60e46000000LL - (0x3c26700LL * 1000000 * 4096); + sec = stck >> 12; + do_div(sec, 1000000); + xtime.tv_sec = sec; + stck -= (sec * 1000000) << 12; + xtime.tv_nsec = ((stck * 1000) >> 12); + len += sprintf(out_buf + len, "%-24s%011lu:%06lu\n", + label, xtime.tv_sec, xtime.tv_nsec); + + return len; +} + +static int zfcp_dbf_tag(char *out_buf, const char *label, const char *tag) +{ + int len = 0, i; + + len += sprintf(out_buf + len, "%-24s", label); + for (i = 0; i < ZFCP_DBF_TAG_SIZE; i++) + len += sprintf(out_buf + len, "%c", tag[i]); + len += sprintf(out_buf + len, "\n"); + + return len; +} + +static int +zfcp_dbf_view(char *out_buf, const char *label, const char *format, ...) +{ + va_list arg; + int len = 0; + + len += sprintf(out_buf + len, "%-24s", label); + va_start(arg, format); + len += vsprintf(out_buf + len, format, arg); + va_end(arg); + len += sprintf(out_buf + len, "\n"); + + return len; +} + +static int +zfcp_dbf_view_dump(char *out_buf, const char *label, + char *buffer, int buflen, int offset, int total_size) +{ + int len = 0; + + if (offset == 0) + len += sprintf(out_buf + len, "%-24s ", label); + + while (buflen--) { + if (offset > 0) { + if ((offset % 32) == 0) + len += sprintf(out_buf + len, "\n%-24c ", ' '); + else if ((offset % 4) == 0) + len += sprintf(out_buf + len, " "); + } + len += sprintf(out_buf + len, "%02x", *buffer++); + if (++offset == total_size) { + len += sprintf(out_buf + len, "\n"); + break; + } + } + + if (total_size == 0) + len += sprintf(out_buf + len, "\n"); + + return len; +} + +static inline int +zfcp_dbf_view_header(debug_info_t * id, struct debug_view *view, int area, + debug_entry_t * entry, char *out_buf) +{ + struct zfcp_dbf_dump *dump = (struct zfcp_dbf_dump *)DEBUG_DATA(entry); + int len = 0; + + if (strncmp(dump->tag, "dump", ZFCP_DBF_TAG_SIZE) != 0) { + len += zfcp_dbf_stck(out_buf + len, "timestamp", + entry->id.stck); + len += zfcp_dbf_view(out_buf + len, "cpu", "%02i", + entry->id.fields.cpuid); + } else { + len += zfcp_dbf_view_dump(out_buf + len, NULL, + dump->data, + dump->size, + dump->offset, dump->total_size); + if ((dump->offset + dump->size) == dump->total_size) + len += sprintf(out_buf + len, "\n"); + } + + return len; +} + +inline void zfcp_hba_dbf_event_fsf_response(struct zfcp_fsf_req *fsf_req) +{ + struct zfcp_adapter *adapter = fsf_req->adapter; + struct fsf_qtcb *qtcb = fsf_req->qtcb; + union fsf_prot_status_qual *prot_status_qual = + &qtcb->prefix.prot_status_qual; + union fsf_status_qual *fsf_status_qual = &qtcb->header.fsf_status_qual; + struct scsi_cmnd *scsi_cmnd; + struct zfcp_port *port; + struct zfcp_unit *unit; + struct zfcp_send_els *send_els; + struct zfcp_hba_dbf_record *rec = &adapter->hba_dbf_buf; + struct zfcp_hba_dbf_record_response *response = &rec->type.response; + int level; + unsigned long flags; + + spin_lock_irqsave(&adapter->hba_dbf_lock, flags); + memset(rec, 0, sizeof(struct zfcp_hba_dbf_record)); + strncpy(rec->tag, "resp", ZFCP_DBF_TAG_SIZE); + + if ((qtcb->prefix.prot_status != FSF_PROT_GOOD) && + (qtcb->prefix.prot_status != FSF_PROT_FSF_STATUS_PRESENTED)) { + strncpy(rec->tag2, "perr", ZFCP_DBF_TAG_SIZE); + level = 1; + } else if (qtcb->header.fsf_status != FSF_GOOD) { + strncpy(rec->tag2, "ferr", ZFCP_DBF_TAG_SIZE); + level = 1; + } else if ((fsf_req->fsf_command == FSF_QTCB_OPEN_PORT_WITH_DID) || + (fsf_req->fsf_command == FSF_QTCB_OPEN_LUN)) { + strncpy(rec->tag2, "open", ZFCP_DBF_TAG_SIZE); + level = 4; + } else if ((prot_status_qual->doubleword[0] != 0) || + (prot_status_qual->doubleword[1] != 0) || + (fsf_status_qual->doubleword[0] != 0) || + (fsf_status_qual->doubleword[1] != 0)) { + strncpy(rec->tag2, "qual", ZFCP_DBF_TAG_SIZE); + level = 3; + } else { + strncpy(rec->tag2, "norm", ZFCP_DBF_TAG_SIZE); + level = 6; + } + + response->fsf_command = fsf_req->fsf_command; + response->fsf_reqid = (unsigned long)fsf_req; + response->fsf_seqno = fsf_req->seq_no; + response->fsf_issued = fsf_req->issued; + response->fsf_prot_status = qtcb->prefix.prot_status; + response->fsf_status = qtcb->header.fsf_status; + memcpy(response->fsf_prot_status_qual, + prot_status_qual, FSF_PROT_STATUS_QUAL_SIZE); + memcpy(response->fsf_status_qual, + fsf_status_qual, FSF_STATUS_QUALIFIER_SIZE); + response->fsf_req_status = fsf_req->status; + response->sbal_first = fsf_req->sbal_first; + response->sbal_curr = fsf_req->sbal_curr; + response->sbal_last = fsf_req->sbal_last; + response->pool = fsf_req->pool != NULL; + response->erp_action = (unsigned long)fsf_req->erp_action; + + switch (fsf_req->fsf_command) { + case FSF_QTCB_FCP_CMND: + if (fsf_req->status & ZFCP_STATUS_FSFREQ_TASK_MANAGEMENT) + break; + scsi_cmnd = (struct scsi_cmnd *)fsf_req->data; + if (scsi_cmnd != NULL) { + response->data.send_fcp.scsi_cmnd + = (unsigned long)scsi_cmnd; + response->data.send_fcp.scsi_serial + = scsi_cmnd->serial_number; + } + break; + + case FSF_QTCB_OPEN_PORT_WITH_DID: + case FSF_QTCB_CLOSE_PORT: + case FSF_QTCB_CLOSE_PHYSICAL_PORT: + port = (struct zfcp_port *)fsf_req->data; + response->data.port.wwpn = port->wwpn; + response->data.port.d_id = port->d_id; + response->data.port.port_handle = qtcb->header.port_handle; + break; + + case FSF_QTCB_OPEN_LUN: + case FSF_QTCB_CLOSE_LUN: + unit = (struct zfcp_unit *)fsf_req->data; + port = unit->port; + response->data.unit.wwpn = port->wwpn; + response->data.unit.fcp_lun = unit->fcp_lun; + response->data.unit.port_handle = qtcb->header.port_handle; + response->data.unit.lun_handle = qtcb->header.lun_handle; + break; + + case FSF_QTCB_SEND_ELS: + send_els = (struct zfcp_send_els *)fsf_req->data; + response->data.send_els.d_id = qtcb->bottom.support.d_id; + response->data.send_els.ls_code = send_els->ls_code >> 24; + break; + + case FSF_QTCB_ABORT_FCP_CMND: + case FSF_QTCB_SEND_GENERIC: + case FSF_QTCB_EXCHANGE_CONFIG_DATA: + case FSF_QTCB_EXCHANGE_PORT_DATA: + case FSF_QTCB_DOWNLOAD_CONTROL_FILE: + case FSF_QTCB_UPLOAD_CONTROL_FILE: + break; + } + + debug_event(adapter->hba_dbf, level, + rec, sizeof(struct zfcp_hba_dbf_record)); + spin_unlock_irqrestore(&adapter->hba_dbf_lock, flags); +} + +inline void +zfcp_hba_dbf_event_fsf_unsol(const char *tag, struct zfcp_adapter *adapter, + struct fsf_status_read_buffer *status_buffer) +{ + struct zfcp_hba_dbf_record *rec = &adapter->hba_dbf_buf; + unsigned long flags; + + spin_lock_irqsave(&adapter->hba_dbf_lock, flags); + memset(rec, 0, sizeof(struct zfcp_hba_dbf_record)); + strncpy(rec->tag, "stat", ZFCP_DBF_TAG_SIZE); + strncpy(rec->tag2, tag, ZFCP_DBF_TAG_SIZE); + + rec->type.status.failed = adapter->status_read_failed; + if (status_buffer != NULL) { + rec->type.status.status_type = status_buffer->status_type; + rec->type.status.status_subtype = status_buffer->status_subtype; + memcpy(&rec->type.status.queue_designator, + &status_buffer->queue_designator, + sizeof(struct fsf_queue_designator)); + + switch (status_buffer->status_type) { + case FSF_STATUS_READ_SENSE_DATA_AVAIL: + rec->type.status.payload_size = + ZFCP_DBF_UNSOL_PAYLOAD_SENSE_DATA_AVAIL; + break; + + case FSF_STATUS_READ_BIT_ERROR_THRESHOLD: + rec->type.status.payload_size = + ZFCP_DBF_UNSOL_PAYLOAD_BIT_ERROR_THRESHOLD; + break; + + case FSF_STATUS_READ_LINK_DOWN: + rec->type.status.payload_size = sizeof(u64); + break; + + } + memcpy(&rec->type.status.payload, + &status_buffer->payload, rec->type.status.payload_size); + } + + debug_event(adapter->hba_dbf, 2, + rec, sizeof(struct zfcp_hba_dbf_record)); + spin_unlock_irqrestore(&adapter->hba_dbf_lock, flags); +} + +inline void +zfcp_hba_dbf_event_qdio(struct zfcp_adapter *adapter, unsigned int status, + unsigned int qdio_error, unsigned int siga_error, + int sbal_index, int sbal_count) +{ + struct zfcp_hba_dbf_record *rec = &adapter->hba_dbf_buf; + unsigned long flags; + + spin_lock_irqsave(&adapter->hba_dbf_lock, flags); + memset(rec, 0, sizeof(struct zfcp_hba_dbf_record)); + strncpy(rec->tag, "qdio", ZFCP_DBF_TAG_SIZE); + rec->type.qdio.status = status; + rec->type.qdio.qdio_error = qdio_error; + rec->type.qdio.siga_error = siga_error; + rec->type.qdio.sbal_index = sbal_index; + rec->type.qdio.sbal_count = sbal_count; + debug_event(adapter->hba_dbf, 0, + rec, sizeof(struct zfcp_hba_dbf_record)); + spin_unlock_irqrestore(&adapter->hba_dbf_lock, flags); +} + +static inline int +zfcp_hba_dbf_view_response(char *out_buf, + struct zfcp_hba_dbf_record_response *rec) +{ + int len = 0; + + len += zfcp_dbf_view(out_buf + len, "fsf_command", "0x%08x", + rec->fsf_command); + len += zfcp_dbf_view(out_buf + len, "fsf_reqid", "0x%0Lx", + rec->fsf_reqid); + len += zfcp_dbf_view(out_buf + len, "fsf_seqno", "0x%08x", + rec->fsf_seqno); + len += zfcp_dbf_stck(out_buf + len, "fsf_issued", rec->fsf_issued); + len += zfcp_dbf_view(out_buf + len, "fsf_prot_status", "0x%08x", + rec->fsf_prot_status); + len += zfcp_dbf_view(out_buf + len, "fsf_status", "0x%08x", + rec->fsf_status); + len += zfcp_dbf_view_dump(out_buf + len, "fsf_prot_status_qual", + rec->fsf_prot_status_qual, + FSF_PROT_STATUS_QUAL_SIZE, + 0, FSF_PROT_STATUS_QUAL_SIZE); + len += zfcp_dbf_view_dump(out_buf + len, "fsf_status_qual", + rec->fsf_status_qual, + FSF_STATUS_QUALIFIER_SIZE, + 0, FSF_STATUS_QUALIFIER_SIZE); + len += zfcp_dbf_view(out_buf + len, "fsf_req_status", "0x%08x", + rec->fsf_req_status); + len += zfcp_dbf_view(out_buf + len, "sbal_first", "0x%02x", + rec->sbal_first); + len += zfcp_dbf_view(out_buf + len, "sbal_curr", "0x%02x", + rec->sbal_curr); + len += zfcp_dbf_view(out_buf + len, "sbal_last", "0x%02x", + rec->sbal_last); + len += zfcp_dbf_view(out_buf + len, "pool", "0x%02x", rec->pool); + + switch (rec->fsf_command) { + case FSF_QTCB_FCP_CMND: + if (rec->fsf_req_status & ZFCP_STATUS_FSFREQ_TASK_MANAGEMENT) + break; + len += zfcp_dbf_view(out_buf + len, "scsi_cmnd", "0x%0Lx", + rec->data.send_fcp.scsi_cmnd); + len += zfcp_dbf_view(out_buf + len, "scsi_serial", "0x%016Lx", + rec->data.send_fcp.scsi_serial); + break; + + case FSF_QTCB_OPEN_PORT_WITH_DID: + case FSF_QTCB_CLOSE_PORT: + case FSF_QTCB_CLOSE_PHYSICAL_PORT: + len += zfcp_dbf_view(out_buf + len, "wwpn", "0x%016Lx", + rec->data.port.wwpn); + len += zfcp_dbf_view(out_buf + len, "d_id", "0x%06x", + rec->data.port.d_id); + len += zfcp_dbf_view(out_buf + len, "port_handle", "0x%08x", + rec->data.port.port_handle); + break; + + case FSF_QTCB_OPEN_LUN: + case FSF_QTCB_CLOSE_LUN: + len += zfcp_dbf_view(out_buf + len, "wwpn", "0x%016Lx", + rec->data.unit.wwpn); + len += zfcp_dbf_view(out_buf + len, "fcp_lun", "0x%016Lx", + rec->data.unit.fcp_lun); + len += zfcp_dbf_view(out_buf + len, "port_handle", "0x%08x", + rec->data.unit.port_handle); + len += zfcp_dbf_view(out_buf + len, "lun_handle", "0x%08x", + rec->data.unit.lun_handle); + break; + + case FSF_QTCB_SEND_ELS: + len += zfcp_dbf_view(out_buf + len, "d_id", "0x%06x", + rec->data.send_els.d_id); + len += zfcp_dbf_view(out_buf + len, "ls_code", "0x%02x", + rec->data.send_els.ls_code); + break; + + case FSF_QTCB_ABORT_FCP_CMND: + case FSF_QTCB_SEND_GENERIC: + case FSF_QTCB_EXCHANGE_CONFIG_DATA: + case FSF_QTCB_EXCHANGE_PORT_DATA: + case FSF_QTCB_DOWNLOAD_CONTROL_FILE: + case FSF_QTCB_UPLOAD_CONTROL_FILE: + break; + } + + return len; +} + +static inline int +zfcp_hba_dbf_view_status(char *out_buf, struct zfcp_hba_dbf_record_status *rec) +{ + int len = 0; + + len += zfcp_dbf_view(out_buf + len, "failed", "0x%02x", rec->failed); + len += zfcp_dbf_view(out_buf + len, "status_type", "0x%08x", + rec->status_type); + len += zfcp_dbf_view(out_buf + len, "status_subtype", "0x%08x", + rec->status_subtype); + len += zfcp_dbf_view_dump(out_buf + len, "queue_designator", + (char *)&rec->queue_designator, + sizeof(struct fsf_queue_designator), + 0, sizeof(struct fsf_queue_designator)); + len += zfcp_dbf_view_dump(out_buf + len, "payload", + (char *)&rec->payload, + rec->payload_size, 0, rec->payload_size); + + return len; +} + +static inline int +zfcp_hba_dbf_view_qdio(char *out_buf, struct zfcp_hba_dbf_record_qdio *rec) +{ + int len = 0; + + len += zfcp_dbf_view(out_buf + len, "status", "0x%08x", rec->status); + len += zfcp_dbf_view(out_buf + len, "qdio_error", "0x%08x", + rec->qdio_error); + len += zfcp_dbf_view(out_buf + len, "siga_error", "0x%08x", + rec->siga_error); + len += zfcp_dbf_view(out_buf + len, "sbal_index", "0x%02x", + rec->sbal_index); + len += zfcp_dbf_view(out_buf + len, "sbal_count", "0x%02x", + rec->sbal_count); + + return len; +} + +static int +zfcp_hba_dbf_view_format(debug_info_t * id, struct debug_view *view, + char *out_buf, const char *in_buf) +{ + struct zfcp_hba_dbf_record *rec = (struct zfcp_hba_dbf_record *)in_buf; + int len = 0; + + if (strncmp(rec->tag, "dump", ZFCP_DBF_TAG_SIZE) == 0) + return 0; + + len += zfcp_dbf_tag(out_buf + len, "tag", rec->tag); + if (isalpha(rec->tag2[0])) + len += zfcp_dbf_tag(out_buf + len, "tag2", rec->tag2); + if (strncmp(rec->tag, "resp", ZFCP_DBF_TAG_SIZE) == 0) + len += zfcp_hba_dbf_view_response(out_buf + len, + &rec->type.response); + else if (strncmp(rec->tag, "stat", ZFCP_DBF_TAG_SIZE) == 0) + len += zfcp_hba_dbf_view_status(out_buf + len, + &rec->type.status); + else if (strncmp(rec->tag, "qdio", ZFCP_DBF_TAG_SIZE) == 0) + len += zfcp_hba_dbf_view_qdio(out_buf + len, &rec->type.qdio); + + len += sprintf(out_buf + len, "\n"); + + return len; +} + +struct debug_view zfcp_hba_dbf_view = { + "structured", + NULL, + &zfcp_dbf_view_header, + &zfcp_hba_dbf_view_format, + NULL, + NULL +}; + +inline void +_zfcp_san_dbf_event_common_ct(const char *tag, struct zfcp_fsf_req *fsf_req, + fc_id_t s_id, fc_id_t d_id, + void *buffer, int buflen) +{ + struct zfcp_send_ct *send_ct = (struct zfcp_send_ct *)fsf_req->data; + struct zfcp_port *port = send_ct->port; + struct zfcp_adapter *adapter = port->adapter; + struct ct_hdr *header = (struct ct_hdr *)buffer; + struct zfcp_san_dbf_record *rec = &adapter->san_dbf_buf; + struct zfcp_san_dbf_record_ct *ct = &rec->type.ct; + unsigned long flags; + + spin_lock_irqsave(&adapter->san_dbf_lock, flags); + memset(rec, 0, sizeof(struct zfcp_san_dbf_record)); + strncpy(rec->tag, tag, ZFCP_DBF_TAG_SIZE); + rec->fsf_reqid = (unsigned long)fsf_req; + rec->fsf_seqno = fsf_req->seq_no; + rec->s_id = s_id; + rec->d_id = d_id; + if (strncmp(tag, "octc", ZFCP_DBF_TAG_SIZE) == 0) { + ct->type.request.cmd_req_code = header->cmd_rsp_code; + ct->type.request.revision = header->revision; + ct->type.request.gs_type = header->gs_type; + ct->type.request.gs_subtype = header->gs_subtype; + ct->type.request.options = header->options; + ct->type.request.max_res_size = header->max_res_size; + } else if (strncmp(tag, "rctc", ZFCP_DBF_TAG_SIZE) == 0) { + ct->type.response.cmd_rsp_code = header->cmd_rsp_code; + ct->type.response.revision = header->revision; + ct->type.response.reason_code = header->reason_code; + ct->type.response.reason_code_expl = header->reason_code_expl; + ct->type.response.vendor_unique = header->vendor_unique; + } + ct->payload_size = + min(buflen - (int)sizeof(struct ct_hdr), ZFCP_DBF_CT_PAYLOAD); + memcpy(ct->payload, buffer + sizeof(struct ct_hdr), ct->payload_size); + debug_event(adapter->san_dbf, 3, + rec, sizeof(struct zfcp_san_dbf_record)); + spin_unlock_irqrestore(&adapter->san_dbf_lock, flags); +} + +inline void zfcp_san_dbf_event_ct_request(struct zfcp_fsf_req *fsf_req) +{ + struct zfcp_send_ct *ct = (struct zfcp_send_ct *)fsf_req->data; + struct zfcp_port *port = ct->port; + struct zfcp_adapter *adapter = port->adapter; + + _zfcp_san_dbf_event_common_ct("octc", + fsf_req, adapter->s_id, port->d_id, + zfcp_sg_to_address(ct->req), + ct->req->length); +} + +inline void zfcp_san_dbf_event_ct_response(struct zfcp_fsf_req *fsf_req) +{ + struct zfcp_send_ct *ct = (struct zfcp_send_ct *)fsf_req->data; + struct zfcp_port *port = ct->port; + struct zfcp_adapter *adapter = port->adapter; + + _zfcp_san_dbf_event_common_ct("rctc", + fsf_req, port->d_id, adapter->s_id, + zfcp_sg_to_address(ct->resp), + ct->resp->length); +} + +static inline void +_zfcp_san_dbf_event_common_els(const char *tag, int level, + struct zfcp_fsf_req *fsf_req, + fc_id_t s_id, fc_id_t d_id, u8 ls_code, + void *buffer, int buflen) +{ + struct zfcp_adapter *adapter = fsf_req->adapter; + struct zfcp_san_dbf_record *rec = &adapter->san_dbf_buf; + struct zfcp_dbf_dump *dump = (struct zfcp_dbf_dump *)rec; + unsigned long flags; + int offset = 0; + + spin_lock_irqsave(&adapter->san_dbf_lock, flags); + do { + memset(rec, 0, sizeof(struct zfcp_san_dbf_record)); + if (offset == 0) { + strncpy(rec->tag, tag, ZFCP_DBF_TAG_SIZE); + rec->fsf_reqid = (unsigned long)fsf_req; + rec->fsf_seqno = fsf_req->seq_no; + rec->s_id = s_id; + rec->d_id = d_id; + rec->type.els.ls_code = ls_code; + buflen = min(buflen, ZFCP_DBF_ELS_MAX_PAYLOAD); + rec->type.els.payload_size = buflen; + memcpy(rec->type.els.payload, + buffer, min(buflen, ZFCP_DBF_ELS_PAYLOAD)); + offset += min(buflen, ZFCP_DBF_ELS_PAYLOAD); + } else { + strncpy(dump->tag, "dump", ZFCP_DBF_TAG_SIZE); + dump->total_size = buflen; + dump->offset = offset; + dump->size = min(buflen - offset, + (int)sizeof(struct zfcp_san_dbf_record) + - (int)sizeof(struct zfcp_dbf_dump)); + memcpy(dump->data, buffer + offset, dump->size); + offset += dump->size; + } + debug_event(adapter->san_dbf, level, + rec, sizeof(struct zfcp_san_dbf_record)); + } while (offset < buflen); + spin_unlock_irqrestore(&adapter->san_dbf_lock, flags); +} + +inline void zfcp_san_dbf_event_els_request(struct zfcp_fsf_req *fsf_req) +{ + struct zfcp_send_els *els = (struct zfcp_send_els *)fsf_req->data; + + _zfcp_san_dbf_event_common_els("oels", 2, + fsf_req, els->adapter->s_id, els->d_id, + *(u8 *) zfcp_sg_to_address(els->req), + zfcp_sg_to_address(els->req), + els->req->length); +} + +inline void zfcp_san_dbf_event_els_response(struct zfcp_fsf_req *fsf_req) +{ + struct zfcp_send_els *els = (struct zfcp_send_els *)fsf_req->data; + + _zfcp_san_dbf_event_common_els("rels", 2, + fsf_req, els->d_id, els->adapter->s_id, + *(u8 *) zfcp_sg_to_address(els->req), + zfcp_sg_to_address(els->resp), + els->resp->length); +} + +inline void zfcp_san_dbf_event_incoming_els(struct zfcp_fsf_req *fsf_req) +{ + struct zfcp_adapter *adapter = fsf_req->adapter; + struct fsf_status_read_buffer *status_buffer = + (struct fsf_status_read_buffer *)fsf_req->data; + int length = (int)status_buffer->length - + (int)((void *)&status_buffer->payload - (void *)status_buffer); + + _zfcp_san_dbf_event_common_els("iels", 1, + fsf_req, status_buffer->d_id, + adapter->s_id, + *(u8 *) status_buffer->payload, + (void *)status_buffer->payload, length); +} + +static int +zfcp_san_dbf_view_format(debug_info_t * id, struct debug_view *view, + char *out_buf, const char *in_buf) +{ + struct zfcp_san_dbf_record *rec = (struct zfcp_san_dbf_record *)in_buf; + char *buffer = NULL; + int buflen = 0, total = 0; + int len = 0; + + if (strncmp(rec->tag, "dump", ZFCP_DBF_TAG_SIZE) == 0) + return 0; + + len += zfcp_dbf_tag(out_buf + len, "tag", rec->tag); + len += zfcp_dbf_view(out_buf + len, "fsf_reqid", "0x%0Lx", + rec->fsf_reqid); + len += zfcp_dbf_view(out_buf + len, "fsf_seqno", "0x%08x", + rec->fsf_seqno); + len += zfcp_dbf_view(out_buf + len, "s_id", "0x%06x", rec->s_id); + len += zfcp_dbf_view(out_buf + len, "d_id", "0x%06x", rec->d_id); + + if (strncmp(rec->tag, "octc", ZFCP_DBF_TAG_SIZE) == 0) { + len += zfcp_dbf_view(out_buf + len, "cmd_req_code", "0x%04x", + rec->type.ct.type.request.cmd_req_code); + len += zfcp_dbf_view(out_buf + len, "revision", "0x%02x", + rec->type.ct.type.request.revision); + len += zfcp_dbf_view(out_buf + len, "gs_type", "0x%02x", + rec->type.ct.type.request.gs_type); + len += zfcp_dbf_view(out_buf + len, "gs_subtype", "0x%02x", + rec->type.ct.type.request.gs_subtype); + len += zfcp_dbf_view(out_buf + len, "options", "0x%02x", + rec->type.ct.type.request.options); + len += zfcp_dbf_view(out_buf + len, "max_res_size", "0x%04x", + rec->type.ct.type.request.max_res_size); + total = rec->type.ct.payload_size; + buffer = rec->type.ct.payload; + buflen = min(total, ZFCP_DBF_CT_PAYLOAD); + } else if (strncmp(rec->tag, "rctc", ZFCP_DBF_TAG_SIZE) == 0) { + len += zfcp_dbf_view(out_buf + len, "cmd_rsp_code", "0x%04x", + rec->type.ct.type.response.cmd_rsp_code); + len += zfcp_dbf_view(out_buf + len, "revision", "0x%02x", + rec->type.ct.type.response.revision); + len += zfcp_dbf_view(out_buf + len, "reason_code", "0x%02x", + rec->type.ct.type.response.reason_code); + len += + zfcp_dbf_view(out_buf + len, "reason_code_expl", "0x%02x", + rec->type.ct.type.response.reason_code_expl); + len += + zfcp_dbf_view(out_buf + len, "vendor_unique", "0x%02x", + rec->type.ct.type.response.vendor_unique); + total = rec->type.ct.payload_size; + buffer = rec->type.ct.payload; + buflen = min(total, ZFCP_DBF_CT_PAYLOAD); + } else if (strncmp(rec->tag, "oels", ZFCP_DBF_TAG_SIZE) == 0 || + strncmp(rec->tag, "rels", ZFCP_DBF_TAG_SIZE) == 0 || + strncmp(rec->tag, "iels", ZFCP_DBF_TAG_SIZE) == 0) { + len += zfcp_dbf_view(out_buf + len, "ls_code", "0x%02x", + rec->type.els.ls_code); + total = rec->type.els.payload_size; + buffer = rec->type.els.payload; + buflen = min(total, ZFCP_DBF_ELS_PAYLOAD); + } + + len += zfcp_dbf_view_dump(out_buf + len, "payload", + buffer, buflen, 0, total); + + if (buflen == total) + len += sprintf(out_buf + len, "\n"); + + return len; +} + +struct debug_view zfcp_san_dbf_view = { + "structured", + NULL, + &zfcp_dbf_view_header, + &zfcp_san_dbf_view_format, + NULL, + NULL +}; + +static inline void +_zfcp_scsi_dbf_event_common(const char *tag, const char *tag2, int level, + struct zfcp_adapter *adapter, + struct scsi_cmnd *scsi_cmnd, + struct zfcp_fsf_req *new_fsf_req) +{ + struct zfcp_fsf_req *fsf_req = + (struct zfcp_fsf_req *)scsi_cmnd->host_scribble; + struct zfcp_scsi_dbf_record *rec = &adapter->scsi_dbf_buf; + struct zfcp_dbf_dump *dump = (struct zfcp_dbf_dump *)rec; + unsigned long flags; + struct fcp_rsp_iu *fcp_rsp; + char *fcp_rsp_info = NULL, *fcp_sns_info = NULL; + int offset = 0, buflen = 0; + + spin_lock_irqsave(&adapter->scsi_dbf_lock, flags); + do { + memset(rec, 0, sizeof(struct zfcp_scsi_dbf_record)); + if (offset == 0) { + strncpy(rec->tag, tag, ZFCP_DBF_TAG_SIZE); + strncpy(rec->tag2, tag2, ZFCP_DBF_TAG_SIZE); + if (scsi_cmnd->device) { + rec->scsi_id = scsi_cmnd->device->id; + rec->scsi_lun = scsi_cmnd->device->lun; + } + rec->scsi_result = scsi_cmnd->result; + rec->scsi_cmnd = (unsigned long)scsi_cmnd; + rec->scsi_serial = scsi_cmnd->serial_number; + memcpy(rec->scsi_opcode, + &scsi_cmnd->cmnd, + min((int)scsi_cmnd->cmd_len, + ZFCP_DBF_SCSI_OPCODE)); + rec->scsi_retries = scsi_cmnd->retries; + rec->scsi_allowed = scsi_cmnd->allowed; + if (fsf_req != NULL) { + fcp_rsp = (struct fcp_rsp_iu *) + &(fsf_req->qtcb->bottom.io.fcp_rsp); + fcp_rsp_info = + zfcp_get_fcp_rsp_info_ptr(fcp_rsp); + fcp_sns_info = + zfcp_get_fcp_sns_info_ptr(fcp_rsp); + + rec->type.fcp.rsp_validity = + fcp_rsp->validity.value; + rec->type.fcp.rsp_scsi_status = + fcp_rsp->scsi_status; + rec->type.fcp.rsp_resid = fcp_rsp->fcp_resid; + if (fcp_rsp->validity.bits.fcp_rsp_len_valid) + rec->type.fcp.rsp_code = + *(fcp_rsp_info + 3); + if (fcp_rsp->validity.bits.fcp_sns_len_valid) { + buflen = min((int)fcp_rsp->fcp_sns_len, + ZFCP_DBF_SCSI_MAX_FCP_SNS_INFO); + rec->type.fcp.sns_info_len = buflen; + memcpy(rec->type.fcp.sns_info, + fcp_sns_info, + min(buflen, + ZFCP_DBF_SCSI_FCP_SNS_INFO)); + offset += min(buflen, + ZFCP_DBF_SCSI_FCP_SNS_INFO); + } + + rec->fsf_reqid = (unsigned long)fsf_req; + rec->fsf_seqno = fsf_req->seq_no; + rec->fsf_issued = fsf_req->issued; + } + if (new_fsf_req != NULL) { + rec->type.new_fsf_req.fsf_reqid = + (unsigned long) + new_fsf_req; + rec->type.new_fsf_req.fsf_seqno = + new_fsf_req->seq_no; + rec->type.new_fsf_req.fsf_issued = + new_fsf_req->issued; + } + } else { + strncpy(dump->tag, "dump", ZFCP_DBF_TAG_SIZE); + dump->total_size = buflen; + dump->offset = offset; + dump->size = min(buflen - offset, + (int)sizeof(struct + zfcp_scsi_dbf_record) - + (int)sizeof(struct zfcp_dbf_dump)); + memcpy(dump->data, fcp_sns_info + offset, dump->size); + offset += dump->size; + } + debug_event(adapter->scsi_dbf, level, + rec, sizeof(struct zfcp_scsi_dbf_record)); + } while (offset < buflen); + spin_unlock_irqrestore(&adapter->scsi_dbf_lock, flags); +} + +inline void +zfcp_scsi_dbf_event_result(const char *tag, int level, + struct zfcp_adapter *adapter, + struct scsi_cmnd *scsi_cmnd) +{ + _zfcp_scsi_dbf_event_common("rslt", + tag, level, adapter, scsi_cmnd, NULL); +} + +inline void +zfcp_scsi_dbf_event_abort(const char *tag, struct zfcp_adapter *adapter, + struct scsi_cmnd *scsi_cmnd, + struct zfcp_fsf_req *new_fsf_req) +{ + _zfcp_scsi_dbf_event_common("abrt", + tag, 1, adapter, scsi_cmnd, new_fsf_req); +} + +inline void +zfcp_scsi_dbf_event_devreset(const char *tag, u8 flag, struct zfcp_unit *unit, + struct scsi_cmnd *scsi_cmnd) +{ + struct zfcp_adapter *adapter = unit->port->adapter; + + _zfcp_scsi_dbf_event_common(flag == FCP_TARGET_RESET ? "trst" : "lrst", + tag, 1, adapter, scsi_cmnd, NULL); +} + +static int +zfcp_scsi_dbf_view_format(debug_info_t * id, struct debug_view *view, + char *out_buf, const char *in_buf) +{ + struct zfcp_scsi_dbf_record *rec = + (struct zfcp_scsi_dbf_record *)in_buf; + int len = 0; + + if (strncmp(rec->tag, "dump", ZFCP_DBF_TAG_SIZE) == 0) + return 0; + + len += zfcp_dbf_tag(out_buf + len, "tag", rec->tag); + len += zfcp_dbf_tag(out_buf + len, "tag2", rec->tag2); + len += zfcp_dbf_view(out_buf + len, "scsi_id", "0x%08x", rec->scsi_id); + len += zfcp_dbf_view(out_buf + len, "scsi_lun", "0x%08x", + rec->scsi_lun); + len += zfcp_dbf_view(out_buf + len, "scsi_result", "0x%08x", + rec->scsi_result); + len += zfcp_dbf_view(out_buf + len, "scsi_cmnd", "0x%0Lx", + rec->scsi_cmnd); + len += zfcp_dbf_view(out_buf + len, "scsi_serial", "0x%016Lx", + rec->scsi_serial); + len += zfcp_dbf_view_dump(out_buf + len, "scsi_opcode", + rec->scsi_opcode, + ZFCP_DBF_SCSI_OPCODE, + 0, ZFCP_DBF_SCSI_OPCODE); + len += zfcp_dbf_view(out_buf + len, "scsi_retries", "0x%02x", + rec->scsi_retries); + len += zfcp_dbf_view(out_buf + len, "scsi_allowed", "0x%02x", + rec->scsi_allowed); + len += zfcp_dbf_view(out_buf + len, "fsf_reqid", "0x%0Lx", + rec->fsf_reqid); + len += zfcp_dbf_view(out_buf + len, "fsf_seqno", "0x%08x", + rec->fsf_seqno); + len += zfcp_dbf_stck(out_buf + len, "fsf_issued", rec->fsf_issued); + if (strncmp(rec->tag, "rslt", ZFCP_DBF_TAG_SIZE) == 0) { + len += + zfcp_dbf_view(out_buf + len, "fcp_rsp_validity", "0x%02x", + rec->type.fcp.rsp_validity); + len += + zfcp_dbf_view(out_buf + len, "fcp_rsp_scsi_status", + "0x%02x", rec->type.fcp.rsp_scsi_status); + len += + zfcp_dbf_view(out_buf + len, "fcp_rsp_resid", "0x%08x", + rec->type.fcp.rsp_resid); + len += + zfcp_dbf_view(out_buf + len, "fcp_rsp_code", "0x%08x", + rec->type.fcp.rsp_code); + len += + zfcp_dbf_view(out_buf + len, "fcp_sns_info_len", "0x%08x", + rec->type.fcp.sns_info_len); + len += + zfcp_dbf_view_dump(out_buf + len, "fcp_sns_info", + rec->type.fcp.sns_info, + min((int)rec->type.fcp.sns_info_len, + ZFCP_DBF_SCSI_FCP_SNS_INFO), 0, + rec->type.fcp.sns_info_len); + } else if (strncmp(rec->tag, "abrt", ZFCP_DBF_TAG_SIZE) == 0) { + len += zfcp_dbf_view(out_buf + len, "fsf_reqid_abort", "0x%0Lx", + rec->type.new_fsf_req.fsf_reqid); + len += zfcp_dbf_view(out_buf + len, "fsf_seqno_abort", "0x%08x", + rec->type.new_fsf_req.fsf_seqno); + len += zfcp_dbf_stck(out_buf + len, "fsf_issued", + rec->type.new_fsf_req.fsf_issued); + } else if ((strncmp(rec->tag, "trst", ZFCP_DBF_TAG_SIZE) == 0) || + (strncmp(rec->tag, "lrst", ZFCP_DBF_TAG_SIZE) == 0)) { + len += zfcp_dbf_view(out_buf + len, "fsf_reqid_reset", "0x%0Lx", + rec->type.new_fsf_req.fsf_reqid); + len += zfcp_dbf_view(out_buf + len, "fsf_seqno_reset", "0x%08x", + rec->type.new_fsf_req.fsf_seqno); + len += zfcp_dbf_stck(out_buf + len, "fsf_issued", + rec->type.new_fsf_req.fsf_issued); + } + + len += sprintf(out_buf + len, "\n"); + + return len; +} + +struct debug_view zfcp_scsi_dbf_view = { + "structured", + NULL, + &zfcp_dbf_view_header, + &zfcp_scsi_dbf_view_format, + NULL, + NULL +}; + +/** + * zfcp_adapter_debug_register - registers debug feature for an adapter + * @adapter: pointer to adapter for which debug features should be registered + * return: -ENOMEM on error, 0 otherwise + */ +int zfcp_adapter_debug_register(struct zfcp_adapter *adapter) +{ + char dbf_name[DEBUG_MAX_NAME_LEN]; + + /* debug feature area which records recovery activity */ + spin_lock_init(&adapter->erp_dbf_lock); + sprintf(dbf_name, "zfcp_%s_erp", zfcp_get_busid_by_adapter(adapter)); + adapter->erp_dbf = debug_register(dbf_name, dbfsize, 2, + sizeof(struct zfcp_erp_dbf_record)); + if (!adapter->erp_dbf) + goto failed; + debug_register_view(adapter->erp_dbf, &debug_hex_ascii_view); + debug_set_level(adapter->erp_dbf, 3); + + /* debug feature area which records HBA (FSF and QDIO) conditions */ + spin_lock_init(&adapter->hba_dbf_lock); + sprintf(dbf_name, "zfcp_%s_hba", zfcp_get_busid_by_adapter(adapter)); + adapter->hba_dbf = debug_register(dbf_name, dbfsize, 1, + sizeof(struct zfcp_hba_dbf_record)); + if (!adapter->hba_dbf) + goto failed; + debug_register_view(adapter->hba_dbf, &debug_hex_ascii_view); + debug_register_view(adapter->hba_dbf, &zfcp_hba_dbf_view); + debug_set_level(adapter->hba_dbf, 3); + + /* debug feature area which records SAN command failures and recovery */ + spin_lock_init(&adapter->san_dbf_lock); + sprintf(dbf_name, "zfcp_%s_san", zfcp_get_busid_by_adapter(adapter)); + adapter->san_dbf = debug_register(dbf_name, dbfsize, 1, + sizeof(struct zfcp_san_dbf_record)); + if (!adapter->san_dbf) + goto failed; + debug_register_view(adapter->san_dbf, &debug_hex_ascii_view); + debug_register_view(adapter->san_dbf, &zfcp_san_dbf_view); + debug_set_level(adapter->san_dbf, 6); + + /* debug feature area which records SCSI command failures and recovery */ + spin_lock_init(&adapter->scsi_dbf_lock); + sprintf(dbf_name, "zfcp_%s_scsi", zfcp_get_busid_by_adapter(adapter)); + adapter->scsi_dbf = debug_register(dbf_name, dbfsize, 1, + sizeof(struct zfcp_scsi_dbf_record)); + if (!adapter->scsi_dbf) + goto failed; + debug_register_view(adapter->scsi_dbf, &debug_hex_ascii_view); + debug_register_view(adapter->scsi_dbf, &zfcp_scsi_dbf_view); + debug_set_level(adapter->scsi_dbf, 3); + + return 0; + + failed: + zfcp_adapter_debug_unregister(adapter); + + return -ENOMEM; +} + +/** + * zfcp_adapter_debug_unregister - unregisters debug feature for an adapter + * @adapter: pointer to adapter for which debug features should be unregistered + */ +void zfcp_adapter_debug_unregister(struct zfcp_adapter *adapter) +{ + debug_unregister(adapter->scsi_dbf); + debug_unregister(adapter->san_dbf); + debug_unregister(adapter->hba_dbf); + debug_unregister(adapter->erp_dbf); + adapter->scsi_dbf = NULL; + adapter->san_dbf = NULL; + adapter->hba_dbf = NULL; + adapter->erp_dbf = NULL; +} + +#undef ZFCP_LOG_AREA diff --git a/drivers/s390/scsi/zfcp_def.h b/drivers/s390/scsi/zfcp_def.h index 9160e68c4a2..fc5bb6f3180 100644 --- a/drivers/s390/scsi/zfcp_def.h +++ b/drivers/s390/scsi/zfcp_def.h @@ -66,7 +66,7 @@ /********************* GENERAL DEFINES *********************************/ /* zfcp version number, it consists of major, minor, and patch-level number */ -#define ZFCP_VERSION "4.3.0" +#define ZFCP_VERSION "4.4.0" /** * zfcp_sg_to_address - determine kernel address from struct scatterlist @@ -280,6 +280,171 @@ struct fcp_logo { wwn_t nport_wwpn; } __attribute__((packed)); +/* + * DBF stuff + */ +#define ZFCP_DBF_TAG_SIZE 4 + +struct zfcp_dbf_dump { + u8 tag[ZFCP_DBF_TAG_SIZE]; + u32 total_size; /* size of total dump data */ + u32 offset; /* how much data has being already dumped */ + u32 size; /* how much data comes with this record */ + u8 data[]; /* dump data */ +} __attribute__ ((packed)); + +/* FIXME: to be inflated when reworking the erp dbf */ +struct zfcp_erp_dbf_record { + u8 dummy[16]; +} __attribute__ ((packed)); + +struct zfcp_hba_dbf_record_response { + u32 fsf_command; + u64 fsf_reqid; + u32 fsf_seqno; + u64 fsf_issued; + u32 fsf_prot_status; + u32 fsf_status; + u8 fsf_prot_status_qual[FSF_PROT_STATUS_QUAL_SIZE]; + u8 fsf_status_qual[FSF_STATUS_QUALIFIER_SIZE]; + u32 fsf_req_status; + u8 sbal_first; + u8 sbal_curr; + u8 sbal_last; + u8 pool; + u64 erp_action; + union { + struct { + u64 scsi_cmnd; + u64 scsi_serial; + } send_fcp; + struct { + u64 wwpn; + u32 d_id; + u32 port_handle; + } port; + struct { + u64 wwpn; + u64 fcp_lun; + u32 port_handle; + u32 lun_handle; + } unit; + struct { + u32 d_id; + u8 ls_code; + } send_els; + } data; +} __attribute__ ((packed)); + +struct zfcp_hba_dbf_record_status { + u8 failed; + u32 status_type; + u32 status_subtype; + struct fsf_queue_designator + queue_designator; + u32 payload_size; +#define ZFCP_DBF_UNSOL_PAYLOAD 80 +#define ZFCP_DBF_UNSOL_PAYLOAD_SENSE_DATA_AVAIL 32 +#define ZFCP_DBF_UNSOL_PAYLOAD_BIT_ERROR_THRESHOLD 56 +#define ZFCP_DBF_UNSOL_PAYLOAD_FEATURE_UPDATE_ALERT 2 * sizeof(u32) + u8 payload[ZFCP_DBF_UNSOL_PAYLOAD]; +} __attribute__ ((packed)); + +struct zfcp_hba_dbf_record_qdio { + u32 status; + u32 qdio_error; + u32 siga_error; + u8 sbal_index; + u8 sbal_count; +} __attribute__ ((packed)); + +struct zfcp_hba_dbf_record { + u8 tag[ZFCP_DBF_TAG_SIZE]; + u8 tag2[ZFCP_DBF_TAG_SIZE]; + union { + struct zfcp_hba_dbf_record_response response; + struct zfcp_hba_dbf_record_status status; + struct zfcp_hba_dbf_record_qdio qdio; + } type; +} __attribute__ ((packed)); + +struct zfcp_san_dbf_record_ct { + union { + struct { + u16 cmd_req_code; + u8 revision; + u8 gs_type; + u8 gs_subtype; + u8 options; + u16 max_res_size; + } request; + struct { + u16 cmd_rsp_code; + u8 revision; + u8 reason_code; + u8 reason_code_expl; + u8 vendor_unique; + } response; + } type; + u32 payload_size; +#define ZFCP_DBF_CT_PAYLOAD 24 + u8 payload[ZFCP_DBF_CT_PAYLOAD]; +} __attribute__ ((packed)); + +struct zfcp_san_dbf_record_els { + u8 ls_code; + u32 payload_size; +#define ZFCP_DBF_ELS_PAYLOAD 32 +#define ZFCP_DBF_ELS_MAX_PAYLOAD 1024 + u8 payload[ZFCP_DBF_ELS_PAYLOAD]; +} __attribute__ ((packed)); + +struct zfcp_san_dbf_record { + u8 tag[ZFCP_DBF_TAG_SIZE]; + u64 fsf_reqid; + u32 fsf_seqno; + u32 s_id; + u32 d_id; + union { + struct zfcp_san_dbf_record_ct ct; + struct zfcp_san_dbf_record_els els; + } type; +} __attribute__ ((packed)); + +struct zfcp_scsi_dbf_record { + u8 tag[ZFCP_DBF_TAG_SIZE]; + u8 tag2[ZFCP_DBF_TAG_SIZE]; + u32 scsi_id; + u32 scsi_lun; + u32 scsi_result; + u64 scsi_cmnd; + u64 scsi_serial; +#define ZFCP_DBF_SCSI_OPCODE 16 + u8 scsi_opcode[ZFCP_DBF_SCSI_OPCODE]; + u8 scsi_retries; + u8 scsi_allowed; + u64 fsf_reqid; + u32 fsf_seqno; + u64 fsf_issued; + union { + struct { + u64 fsf_reqid; + u32 fsf_seqno; + u64 fsf_issued; + } new_fsf_req; + struct { + u8 rsp_validity; + u8 rsp_scsi_status; + u32 rsp_resid; + u8 rsp_code; +#define ZFCP_DBF_SCSI_FCP_SNS_INFO 16 +#define ZFCP_DBF_SCSI_MAX_FCP_SNS_INFO 256 + u32 sns_info_len; + u8 sns_info[ZFCP_DBF_SCSI_FCP_SNS_INFO]; + } fcp; + } type; +} __attribute__ ((packed)); + /* * FC-FS stuff */ @@ -339,34 +504,6 @@ struct zfcp_rc_entry { */ #define ZFCP_CT_TIMEOUT (3 * R_A_TOV) - -/***************** S390 DEBUG FEATURE SPECIFIC DEFINES ***********************/ - -/* debug feature entries per adapter */ -#define ZFCP_ERP_DBF_INDEX 1 -#define ZFCP_ERP_DBF_AREAS 2 -#define ZFCP_ERP_DBF_LENGTH 16 -#define ZFCP_ERP_DBF_LEVEL 3 -#define ZFCP_ERP_DBF_NAME "zfcperp" - -#define ZFCP_CMD_DBF_INDEX 2 -#define ZFCP_CMD_DBF_AREAS 1 -#define ZFCP_CMD_DBF_LENGTH 8 -#define ZFCP_CMD_DBF_LEVEL 3 -#define ZFCP_CMD_DBF_NAME "zfcpcmd" - -#define ZFCP_ABORT_DBF_INDEX 2 -#define ZFCP_ABORT_DBF_AREAS 1 -#define ZFCP_ABORT_DBF_LENGTH 8 -#define ZFCP_ABORT_DBF_LEVEL 6 -#define ZFCP_ABORT_DBF_NAME "zfcpabt" - -#define ZFCP_IN_ELS_DBF_INDEX 2 -#define ZFCP_IN_ELS_DBF_AREAS 1 -#define ZFCP_IN_ELS_DBF_LENGTH 8 -#define ZFCP_IN_ELS_DBF_LEVEL 6 -#define ZFCP_IN_ELS_DBF_NAME "zfcpels" - /******************** LOGGING MACROS AND DEFINES *****************************/ /* @@ -823,11 +960,18 @@ struct zfcp_adapter { u32 erp_low_mem_count; /* nr of erp actions waiting for memory */ struct zfcp_port *nameserver_port; /* adapter's nameserver */ - debug_info_t *erp_dbf; /* S/390 debug features */ - debug_info_t *abort_dbf; - debug_info_t *in_els_dbf; - debug_info_t *cmd_dbf; - spinlock_t dbf_lock; + debug_info_t *erp_dbf; + debug_info_t *hba_dbf; + debug_info_t *san_dbf; /* debug feature areas */ + debug_info_t *scsi_dbf; + spinlock_t erp_dbf_lock; + spinlock_t hba_dbf_lock; + spinlock_t san_dbf_lock; + spinlock_t scsi_dbf_lock; + struct zfcp_erp_dbf_record erp_dbf_buf; + struct zfcp_hba_dbf_record hba_dbf_buf; + struct zfcp_san_dbf_record san_dbf_buf; + struct zfcp_scsi_dbf_record scsi_dbf_buf; struct zfcp_adapter_mempool pool; /* Adapter memory pools */ struct qdio_initialize qdio_init_data; /* for qdio_establish */ struct device generic_services; /* directory for WKA ports */ @@ -902,6 +1046,7 @@ struct zfcp_fsf_req { issued on behalf of erp */ mempool_t *pool; /* used if request was alloacted from emergency pool */ + unsigned long long issued; /* request sent time (STCK) */ struct zfcp_unit *unit; }; diff --git a/drivers/s390/scsi/zfcp_ext.h b/drivers/s390/scsi/zfcp_ext.h index c400e3b9de9..104b7423fd4 100644 --- a/drivers/s390/scsi/zfcp_ext.h +++ b/drivers/s390/scsi/zfcp_ext.h @@ -181,9 +181,25 @@ extern void zfcp_erp_port_access_changed(struct zfcp_port *); extern void zfcp_erp_unit_access_changed(struct zfcp_unit *); /******************************** AUX ****************************************/ -extern void zfcp_cmd_dbf_event_fsf(const char *, struct zfcp_fsf_req *, - void *, int); -extern void zfcp_cmd_dbf_event_scsi(const char *, struct scsi_cmnd *); -extern void zfcp_in_els_dbf_event(struct zfcp_adapter *, const char *, - struct fsf_status_read_buffer *, int); +extern void zfcp_hba_dbf_event_fsf_response(struct zfcp_fsf_req *); +extern void zfcp_hba_dbf_event_fsf_unsol(const char *, struct zfcp_adapter *, + struct fsf_status_read_buffer *); +extern void zfcp_hba_dbf_event_qdio(struct zfcp_adapter *, + unsigned int, unsigned int, unsigned int, + int, int); + +extern void zfcp_san_dbf_event_ct_request(struct zfcp_fsf_req *); +extern void zfcp_san_dbf_event_ct_response(struct zfcp_fsf_req *); +extern void zfcp_san_dbf_event_els_request(struct zfcp_fsf_req *); +extern void zfcp_san_dbf_event_els_response(struct zfcp_fsf_req *); +extern void zfcp_san_dbf_event_incoming_els(struct zfcp_fsf_req *); + +extern void zfcp_scsi_dbf_event_result(const char *, int, struct zfcp_adapter *, + struct scsi_cmnd *); +extern void zfcp_scsi_dbf_event_abort(const char *, struct zfcp_adapter *, + struct scsi_cmnd *, + struct zfcp_fsf_req *); +extern void zfcp_scsi_dbf_event_devreset(const char *, u8, struct zfcp_unit *, + struct scsi_cmnd *); + #endif /* ZFCP_EXT_H */ diff --git a/drivers/s390/scsi/zfcp_fsf.c b/drivers/s390/scsi/zfcp_fsf.c index 929f6c9cad3..80ab721ddff 100644 --- a/drivers/s390/scsi/zfcp_fsf.c +++ b/drivers/s390/scsi/zfcp_fsf.c @@ -285,51 +285,51 @@ zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *fsf_req) { int retval = 0; struct zfcp_adapter *adapter = fsf_req->adapter; + struct fsf_qtcb *qtcb = fsf_req->qtcb; + union fsf_prot_status_qual *prot_status_qual = + &qtcb->prefix.prot_status_qual; - ZFCP_LOG_DEBUG("QTCB is at %p\n", fsf_req->qtcb); + zfcp_hba_dbf_event_fsf_response(fsf_req); if (fsf_req->status & ZFCP_STATUS_FSFREQ_DISMISSED) { ZFCP_LOG_DEBUG("fsf_req 0x%lx has been dismissed\n", (unsigned long) fsf_req); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR | ZFCP_STATUS_FSFREQ_RETRY; /* only for SCSI cmnds. */ - zfcp_cmd_dbf_event_fsf("dismiss", fsf_req, NULL, 0); goto skip_protstatus; } /* log additional information provided by FSF (if any) */ - if (unlikely(fsf_req->qtcb->header.log_length)) { + if (unlikely(qtcb->header.log_length)) { /* do not trust them ;-) */ - if (fsf_req->qtcb->header.log_start > sizeof(struct fsf_qtcb)) { + if (qtcb->header.log_start > sizeof(struct fsf_qtcb)) { ZFCP_LOG_NORMAL ("bug: ULP (FSF logging) log data starts " "beyond end of packet header. Ignored. " "(start=%i, size=%li)\n", - fsf_req->qtcb->header.log_start, + qtcb->header.log_start, sizeof(struct fsf_qtcb)); goto forget_log; } - if ((size_t) (fsf_req->qtcb->header.log_start + - fsf_req->qtcb->header.log_length) + if ((size_t) (qtcb->header.log_start + qtcb->header.log_length) > sizeof(struct fsf_qtcb)) { ZFCP_LOG_NORMAL("bug: ULP (FSF logging) log data ends " "beyond end of packet header. Ignored. " "(start=%i, length=%i, size=%li)\n", - fsf_req->qtcb->header.log_start, - fsf_req->qtcb->header.log_length, + qtcb->header.log_start, + qtcb->header.log_length, sizeof(struct fsf_qtcb)); goto forget_log; } ZFCP_LOG_TRACE("ULP log data: \n"); ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_TRACE, - (char *) fsf_req->qtcb + - fsf_req->qtcb->header.log_start, - fsf_req->qtcb->header.log_length); + (char *) qtcb + qtcb->header.log_start, + qtcb->header.log_length); } forget_log: /* evaluate FSF Protocol Status */ - switch (fsf_req->qtcb->prefix.prot_status) { + switch (qtcb->prefix.prot_status) { case FSF_PROT_GOOD: case FSF_PROT_FSF_STATUS_PRESENTED: @@ -340,14 +340,9 @@ zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *fsf_req) "microcode of version 0x%x, the device driver " "only supports 0x%x. Aborting.\n", zfcp_get_busid_by_adapter(adapter), - fsf_req->qtcb->prefix.prot_status_qual. - version_error.fsf_version, ZFCP_QTCB_VERSION); - /* stop operation for this adapter */ - debug_text_exception(adapter->erp_dbf, 0, "prot_ver_err"); + prot_status_qual->version_error.fsf_version, + ZFCP_QTCB_VERSION); zfcp_erp_adapter_shutdown(adapter, 0); - zfcp_cmd_dbf_event_fsf("qverserr", fsf_req, - &fsf_req->qtcb->prefix.prot_status_qual, - sizeof (union fsf_prot_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -355,16 +350,10 @@ zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *fsf_req) ZFCP_LOG_NORMAL("bug: Sequence number mismatch between " "driver (0x%x) and adapter %s (0x%x). " "Restarting all operations on this adapter.\n", - fsf_req->qtcb->prefix.req_seq_no, + qtcb->prefix.req_seq_no, zfcp_get_busid_by_adapter(adapter), - fsf_req->qtcb->prefix.prot_status_qual. - sequence_error.exp_req_seq_no); - debug_text_exception(adapter->erp_dbf, 0, "prot_seq_err"); - /* restart operation on this adapter */ + prot_status_qual->sequence_error.exp_req_seq_no); zfcp_erp_adapter_reopen(adapter, 0); - zfcp_cmd_dbf_event_fsf("seqnoerr", fsf_req, - &fsf_req->qtcb->prefix.prot_status_qual, - sizeof (union fsf_prot_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_RETRY; fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -375,22 +364,14 @@ zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *fsf_req) "that used on adapter %s. " "Stopping all operations on this adapter.\n", zfcp_get_busid_by_adapter(adapter)); - debug_text_exception(adapter->erp_dbf, 0, "prot_unsup_qtcb"); zfcp_erp_adapter_shutdown(adapter, 0); - zfcp_cmd_dbf_event_fsf("unsqtcbt", fsf_req, - &fsf_req->qtcb->prefix.prot_status_qual, - sizeof (union fsf_prot_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; case FSF_PROT_HOST_CONNECTION_INITIALIZING: - zfcp_cmd_dbf_event_fsf("hconinit", fsf_req, - &fsf_req->qtcb->prefix.prot_status_qual, - sizeof (union fsf_prot_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; atomic_set_mask(ZFCP_STATUS_ADAPTER_HOST_CON_INIT, &(adapter->status)); - debug_text_event(adapter->erp_dbf, 3, "prot_con_init"); break; case FSF_PROT_DUPLICATE_REQUEST_ID: @@ -413,11 +394,7 @@ zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *fsf_req) fsf_req, zfcp_get_busid_by_adapter(adapter)); } - debug_text_exception(adapter->erp_dbf, 0, "prot_dup_id"); zfcp_erp_adapter_shutdown(adapter, 0); - zfcp_cmd_dbf_event_fsf("dupreqid", fsf_req, - &fsf_req->qtcb->prefix.prot_status_qual, - sizeof (union fsf_prot_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -483,8 +460,7 @@ zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *fsf_req) break; case FSF_PROT_REEST_QUEUE: - debug_text_event(adapter->erp_dbf, 1, "prot_reest_queue"); - ZFCP_LOG_INFO("The local link to adapter with " + ZFCP_LOG_NORMAL("The local link to adapter with " "%s was re-plugged. " "Re-starting operations on this adapter.\n", zfcp_get_busid_by_adapter(adapter)); @@ -495,9 +471,6 @@ zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *fsf_req) zfcp_erp_adapter_reopen(adapter, ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED | ZFCP_STATUS_COMMON_ERP_FAILED); - zfcp_cmd_dbf_event_fsf("reestque", fsf_req, - &fsf_req->qtcb->prefix.prot_status_qual, - sizeof (union fsf_prot_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -507,12 +480,7 @@ zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *fsf_req) "Restarting all operations on this " "adapter.\n", zfcp_get_busid_by_adapter(adapter)); - debug_text_event(adapter->erp_dbf, 0, "prot_err_sta"); - /* restart operation on this adapter */ zfcp_erp_adapter_reopen(adapter, 0); - zfcp_cmd_dbf_event_fsf("proterrs", fsf_req, - &fsf_req->qtcb->prefix.prot_status_qual, - sizeof (union fsf_prot_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_RETRY; fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -524,11 +492,7 @@ zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *fsf_req) "Stopping all operations on this adapter. " "(debug info 0x%x).\n", zfcp_get_busid_by_adapter(adapter), - fsf_req->qtcb->prefix.prot_status); - debug_text_event(adapter->erp_dbf, 0, "prot_inval:"); - debug_exception(adapter->erp_dbf, 0, - &fsf_req->qtcb->prefix.prot_status, - sizeof (u32)); + qtcb->prefix.prot_status); zfcp_erp_adapter_shutdown(adapter, 0); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; } @@ -568,28 +532,18 @@ zfcp_fsf_fsfstatus_eval(struct zfcp_fsf_req *fsf_req) "(debug info 0x%x).\n", zfcp_get_busid_by_adapter(fsf_req->adapter), fsf_req->qtcb->header.fsf_command); - debug_text_exception(fsf_req->adapter->erp_dbf, 0, - "fsf_s_unknown"); zfcp_erp_adapter_shutdown(fsf_req->adapter, 0); - zfcp_cmd_dbf_event_fsf("unknownc", fsf_req, - &fsf_req->qtcb->header.fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; case FSF_FCP_RSP_AVAILABLE: ZFCP_LOG_DEBUG("FCP Sense data will be presented to the " "SCSI stack.\n"); - debug_text_event(fsf_req->adapter->erp_dbf, 3, "fsf_s_rsp"); break; case FSF_ADAPTER_STATUS_AVAILABLE: - debug_text_event(fsf_req->adapter->erp_dbf, 2, "fsf_s_astatus"); zfcp_fsf_fsfstatus_qual_eval(fsf_req); break; - - default: - break; } skip_fsfstatus: @@ -617,44 +571,28 @@ zfcp_fsf_fsfstatus_qual_eval(struct zfcp_fsf_req *fsf_req) switch (fsf_req->qtcb->header.fsf_status_qual.word[0]) { case FSF_SQ_FCP_RSP_AVAILABLE: - debug_text_event(fsf_req->adapter->erp_dbf, 4, "fsf_sq_rsp"); break; case FSF_SQ_RETRY_IF_POSSIBLE: /* The SCSI-stack may now issue retries or escalate */ - debug_text_event(fsf_req->adapter->erp_dbf, 2, "fsf_sq_retry"); - zfcp_cmd_dbf_event_fsf("sqretry", fsf_req, - &fsf_req->qtcb->header.fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; case FSF_SQ_COMMAND_ABORTED: /* Carry the aborted state on to upper layer */ - debug_text_event(fsf_req->adapter->erp_dbf, 2, "fsf_sq_abort"); - zfcp_cmd_dbf_event_fsf("sqabort", fsf_req, - &fsf_req->qtcb->header.fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ABORTED; fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; case FSF_SQ_NO_RECOM: - debug_text_exception(fsf_req->adapter->erp_dbf, 0, - "fsf_sq_no_rec"); ZFCP_LOG_NORMAL("bug: No recommendation could be given for a" "problem on the adapter %s " "Stopping all operations on this adapter. ", zfcp_get_busid_by_adapter(fsf_req->adapter)); zfcp_erp_adapter_shutdown(fsf_req->adapter, 0); - zfcp_cmd_dbf_event_fsf("sqnrecom", fsf_req, - &fsf_req->qtcb->header.fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; case FSF_SQ_ULP_PROGRAMMING_ERROR: ZFCP_LOG_NORMAL("error: not enough SBALs for data transfer " "(adapter %s)\n", zfcp_get_busid_by_adapter(fsf_req->adapter)); - debug_text_exception(fsf_req->adapter->erp_dbf, 0, - "fsf_sq_ulp_err"); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; case FSF_SQ_INVOKE_LINK_TEST_PROCEDURE: @@ -668,13 +606,6 @@ zfcp_fsf_fsfstatus_qual_eval(struct zfcp_fsf_req *fsf_req) ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_NORMAL, (char *) &fsf_req->qtcb->header.fsf_status_qual, sizeof (union fsf_status_qual)); - debug_text_event(fsf_req->adapter->erp_dbf, 0, "fsf_sq_inval:"); - debug_exception(fsf_req->adapter->erp_dbf, 0, - &fsf_req->qtcb->header.fsf_status_qual.word[0], - sizeof (u32)); - zfcp_cmd_dbf_event_fsf("squndef", fsf_req, - &fsf_req->qtcb->header.fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; } @@ -696,11 +627,6 @@ zfcp_fsf_req_dispatch(struct zfcp_fsf_req *fsf_req) struct zfcp_adapter *adapter = fsf_req->adapter; int retval = 0; - if (unlikely(fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR)) { - ZFCP_LOG_TRACE("fsf_req=%p, QTCB=%p\n", fsf_req, fsf_req->qtcb); - ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_TRACE, - (char *) fsf_req->qtcb, sizeof(struct fsf_qtcb)); - } switch (fsf_req->fsf_command) { @@ -760,13 +686,13 @@ zfcp_fsf_req_dispatch(struct zfcp_fsf_req *fsf_req) fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; ZFCP_LOG_NORMAL("bug: Command issued by the device driver is " "not supported by the adapter %s\n", - zfcp_get_busid_by_adapter(fsf_req->adapter)); + zfcp_get_busid_by_adapter(adapter)); if (fsf_req->fsf_command != fsf_req->qtcb->header.fsf_command) ZFCP_LOG_NORMAL ("bug: Command issued by the device driver differs " "from the command returned by the adapter %s " "(debug info 0x%x, 0x%x).\n", - zfcp_get_busid_by_adapter(fsf_req->adapter), + zfcp_get_busid_by_adapter(adapter), fsf_req->fsf_command, fsf_req->qtcb->header.fsf_command); } @@ -774,8 +700,6 @@ zfcp_fsf_req_dispatch(struct zfcp_fsf_req *fsf_req) if (!erp_action) return retval; - debug_text_event(adapter->erp_dbf, 3, "a_frh"); - debug_event(adapter->erp_dbf, 3, &erp_action->action, sizeof (int)); zfcp_erp_async_handler(erp_action, 0); return retval; @@ -846,6 +770,7 @@ zfcp_fsf_status_read(struct zfcp_adapter *adapter, int req_flags) failed_buf: zfcp_fsf_req_free(fsf_req); failed_req_create: + zfcp_hba_dbf_event_fsf_unsol("fail", adapter, NULL); out: write_unlock_irqrestore(&adapter->request_queue.queue_lock, lock_flags); return retval; @@ -921,35 +846,30 @@ zfcp_fsf_status_read_handler(struct zfcp_fsf_req *fsf_req) (struct fsf_status_read_buffer *) fsf_req->data; if (fsf_req->status & ZFCP_STATUS_FSFREQ_DISMISSED) { + zfcp_hba_dbf_event_fsf_unsol("dism", adapter, status_buffer); mempool_free(status_buffer, adapter->pool.data_status_read); zfcp_fsf_req_free(fsf_req); goto out; } + zfcp_hba_dbf_event_fsf_unsol("read", adapter, status_buffer); + switch (status_buffer->status_type) { case FSF_STATUS_READ_PORT_CLOSED: - debug_text_event(adapter->erp_dbf, 3, "unsol_pclosed:"); - debug_event(adapter->erp_dbf, 3, - &status_buffer->d_id, sizeof (u32)); zfcp_fsf_status_read_port_closed(fsf_req); break; case FSF_STATUS_READ_INCOMING_ELS: - debug_text_event(adapter->erp_dbf, 3, "unsol_els:"); zfcp_fsf_incoming_els(fsf_req); break; case FSF_STATUS_READ_SENSE_DATA_AVAIL: - debug_text_event(adapter->erp_dbf, 3, "unsol_sense:"); ZFCP_LOG_INFO("unsolicited sense data received (adapter %s)\n", zfcp_get_busid_by_adapter(adapter)); - ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_NORMAL, (char *) status_buffer, - sizeof(struct fsf_status_read_buffer)); break; case FSF_STATUS_READ_BIT_ERROR_THRESHOLD: - debug_text_event(adapter->erp_dbf, 3, "unsol_bit_err:"); ZFCP_LOG_NORMAL("Bit error threshold data received:\n"); ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_NORMAL, (char *) status_buffer, @@ -980,35 +900,30 @@ zfcp_fsf_status_read_handler(struct zfcp_fsf_req *fsf_req) break; case FSF_STATUS_READ_CFDC_UPDATED: - debug_text_event(adapter->erp_dbf, 2, "unsol_cfdc_update:"); - ZFCP_LOG_INFO("CFDC has been updated on the adapter %s\n", + ZFCP_LOG_NORMAL("CFDC has been updated on the adapter %s\n", zfcp_get_busid_by_adapter(adapter)); zfcp_erp_adapter_access_changed(adapter); break; case FSF_STATUS_READ_CFDC_HARDENED: - debug_text_event(adapter->erp_dbf, 2, "unsol_cfdc_harden:"); switch (status_buffer->status_subtype) { case FSF_STATUS_READ_SUB_CFDC_HARDENED_ON_SE: - ZFCP_LOG_INFO("CFDC of adapter %s saved on SE\n", + ZFCP_LOG_NORMAL("CFDC of adapter %s saved on SE\n", zfcp_get_busid_by_adapter(adapter)); break; case FSF_STATUS_READ_SUB_CFDC_HARDENED_ON_SE2: - ZFCP_LOG_INFO("CFDC of adapter %s has been copied " + ZFCP_LOG_NORMAL("CFDC of adapter %s has been copied " "to the secondary SE\n", zfcp_get_busid_by_adapter(adapter)); break; default: - ZFCP_LOG_INFO("CFDC of adapter %s has been hardened\n", + ZFCP_LOG_NORMAL("CFDC of adapter %s has been hardened\n", zfcp_get_busid_by_adapter(adapter)); } break; default: - debug_text_event(adapter->erp_dbf, 0, "unsol_unknown:"); - debug_exception(adapter->erp_dbf, 0, - &status_buffer->status_type, sizeof (u32)); - ZFCP_LOG_NORMAL("bug: An unsolicited status packet of unknown " + ZFCP_LOG_NORMAL("warning: An unsolicited status packet of unknown " "type was received (debug info 0x%x)\n", status_buffer->status_type); ZFCP_LOG_DEBUG("Dump of status_read_buffer %p:\n", @@ -1418,6 +1333,8 @@ zfcp_fsf_send_ct(struct zfcp_send_ct *ct, mempool_t *pool, fsf_req->qtcb->bottom.support.timeout = ct->timeout; fsf_req->data = (unsigned long) ct; + zfcp_san_dbf_event_ct_request(fsf_req); + /* start QDIO request for this FSF request */ ret = zfcp_fsf_req_send(fsf_req, ct->timer); if (ret) { @@ -1476,6 +1393,7 @@ zfcp_fsf_send_ct_handler(struct zfcp_fsf_req *fsf_req) switch (header->fsf_status) { case FSF_GOOD: + zfcp_san_dbf_event_ct_response(fsf_req); retval = 0; break; @@ -1720,6 +1638,8 @@ zfcp_fsf_send_els(struct zfcp_send_els *els) sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0); + zfcp_san_dbf_event_els_request(fsf_req); + /* start QDIO request for this FSF request */ ret = zfcp_fsf_req_send(fsf_req, els->timer); if (ret) { @@ -1777,6 +1697,7 @@ static int zfcp_fsf_send_els_handler(struct zfcp_fsf_req *fsf_req) switch (header->fsf_status) { case FSF_GOOD: + zfcp_san_dbf_event_els_response(fsf_req); retval = 0; break; @@ -3309,9 +3230,6 @@ zfcp_fsf_close_unit_handler(struct zfcp_fsf_req *fsf_req) debug_text_event(fsf_req->adapter->erp_dbf, 1, "fsf_s_phand_nv"); zfcp_erp_adapter_reopen(unit->port->adapter, 0); - zfcp_cmd_dbf_event_fsf("porthinv", fsf_req, - &fsf_req->qtcb->header.fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -3330,9 +3248,6 @@ zfcp_fsf_close_unit_handler(struct zfcp_fsf_req *fsf_req) debug_text_event(fsf_req->adapter->erp_dbf, 1, "fsf_s_lhand_nv"); zfcp_erp_port_reopen(unit->port, 0); - zfcp_cmd_dbf_event_fsf("lunhinv", fsf_req, - &fsf_req->qtcb->header.fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -3763,10 +3678,6 @@ zfcp_fsf_send_fcp_command_handler(struct zfcp_fsf_req *fsf_req) debug_text_event(fsf_req->adapter->erp_dbf, 1, "fsf_s_hand_mis"); zfcp_erp_adapter_reopen(unit->port->adapter, 0); - zfcp_cmd_dbf_event_fsf("handmism", - fsf_req, - &header->fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -3787,10 +3698,6 @@ zfcp_fsf_send_fcp_command_handler(struct zfcp_fsf_req *fsf_req) debug_text_exception(fsf_req->adapter->erp_dbf, 0, "fsf_s_class_nsup"); zfcp_erp_adapter_shutdown(unit->port->adapter, 0); - zfcp_cmd_dbf_event_fsf("unsclass", - fsf_req, - &header->fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -3809,10 +3716,6 @@ zfcp_fsf_send_fcp_command_handler(struct zfcp_fsf_req *fsf_req) debug_text_event(fsf_req->adapter->erp_dbf, 1, "fsf_s_fcp_lun_nv"); zfcp_erp_port_reopen(unit->port, 0); - zfcp_cmd_dbf_event_fsf("fluninv", - fsf_req, - &header->fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -3851,10 +3754,6 @@ zfcp_fsf_send_fcp_command_handler(struct zfcp_fsf_req *fsf_req) debug_text_event(fsf_req->adapter->erp_dbf, 0, "fsf_s_dir_ind_nv"); zfcp_erp_adapter_shutdown(unit->port->adapter, 0); - zfcp_cmd_dbf_event_fsf("dirinv", - fsf_req, - &header->fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -3870,10 +3769,6 @@ zfcp_fsf_send_fcp_command_handler(struct zfcp_fsf_req *fsf_req) debug_text_event(fsf_req->adapter->erp_dbf, 0, "fsf_s_cmd_len_nv"); zfcp_erp_adapter_shutdown(unit->port->adapter, 0); - zfcp_cmd_dbf_event_fsf("cleninv", - fsf_req, - &header->fsf_status_qual, - sizeof (union fsf_status_qual)); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -4043,7 +3938,6 @@ zfcp_fsf_send_fcp_command_task_handler(struct zfcp_fsf_req *fsf_req) ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG, (char *) &fsf_req->qtcb-> bottom.io.fcp_cmnd, FSF_FCP_CMND_SIZE); - zfcp_cmd_dbf_event_fsf("clenmis", fsf_req, NULL, 0); set_host_byte(&scpnt->result, DID_ERROR); goto skip_fsfstatus; case RSP_CODE_FIELD_INVALID: @@ -4062,7 +3956,6 @@ zfcp_fsf_send_fcp_command_task_handler(struct zfcp_fsf_req *fsf_req) (char *) &fsf_req->qtcb-> bottom.io.fcp_cmnd, FSF_FCP_CMND_SIZE); set_host_byte(&scpnt->result, DID_ERROR); - zfcp_cmd_dbf_event_fsf("codeinv", fsf_req, NULL, 0); goto skip_fsfstatus; case RSP_CODE_RO_MISMATCH: /* hardware bug */ @@ -4079,7 +3972,6 @@ zfcp_fsf_send_fcp_command_task_handler(struct zfcp_fsf_req *fsf_req) ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG, (char *) &fsf_req->qtcb-> bottom.io.fcp_cmnd, FSF_FCP_CMND_SIZE); - zfcp_cmd_dbf_event_fsf("codemism", fsf_req, NULL, 0); set_host_byte(&scpnt->result, DID_ERROR); goto skip_fsfstatus; default: @@ -4096,7 +3988,6 @@ zfcp_fsf_send_fcp_command_task_handler(struct zfcp_fsf_req *fsf_req) ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_DEBUG, (char *) &fsf_req->qtcb-> bottom.io.fcp_cmnd, FSF_FCP_CMND_SIZE); - zfcp_cmd_dbf_event_fsf("undeffcp", fsf_req, NULL, 0); set_host_byte(&scpnt->result, DID_ERROR); goto skip_fsfstatus; } @@ -4158,19 +4049,17 @@ zfcp_fsf_send_fcp_command_task_handler(struct zfcp_fsf_req *fsf_req) skip_fsfstatus: ZFCP_LOG_DEBUG("scpnt->result =0x%x\n", scpnt->result); - zfcp_cmd_dbf_event_scsi("response", scpnt); + if (scpnt->result != 0) + zfcp_scsi_dbf_event_result("erro", 3, fsf_req->adapter, scpnt); + else if (scpnt->retries > 0) + zfcp_scsi_dbf_event_result("retr", 4, fsf_req->adapter, scpnt); + else + zfcp_scsi_dbf_event_result("norm", 6, fsf_req->adapter, scpnt); /* cleanup pointer (need this especially for abort) */ scpnt->host_scribble = NULL; - /* - * NOTE: - * according to the outcome of a discussion on linux-scsi we - * don't need to grab the io_request_lock here since we use - * the new eh - */ /* always call back */ - (scpnt->scsi_done) (scpnt); /* @@ -4563,15 +4452,16 @@ zfcp_fsf_req_sbal_check(unsigned long *flags, * set qtcb pointer in fsf_req and initialize QTCB */ static inline void -zfcp_fsf_req_qtcb_init(struct zfcp_fsf_req *fsf_req, u32 fsf_cmd) +zfcp_fsf_req_qtcb_init(struct zfcp_fsf_req *fsf_req) { if (likely(fsf_req->qtcb != NULL)) { + fsf_req->qtcb->prefix.req_seq_no = fsf_req->adapter->fsf_req_seq_no; fsf_req->qtcb->prefix.req_id = (unsigned long)fsf_req; fsf_req->qtcb->prefix.ulp_info = ZFCP_ULP_INFO_VERSION; - fsf_req->qtcb->prefix.qtcb_type = fsf_qtcb_type[fsf_cmd]; + fsf_req->qtcb->prefix.qtcb_type = fsf_qtcb_type[fsf_req->fsf_command]; fsf_req->qtcb->prefix.qtcb_version = ZFCP_QTCB_VERSION; fsf_req->qtcb->header.req_handle = (unsigned long)fsf_req; - fsf_req->qtcb->header.fsf_command = fsf_cmd; + fsf_req->qtcb->header.fsf_command = fsf_req->fsf_command; } } @@ -4639,7 +4529,10 @@ zfcp_fsf_req_create(struct zfcp_adapter *adapter, u32 fsf_cmd, int req_flags, goto failed_fsf_req; } - zfcp_fsf_req_qtcb_init(fsf_req, fsf_cmd); + fsf_req->adapter = adapter; + fsf_req->fsf_command = fsf_cmd; + + zfcp_fsf_req_qtcb_init(fsf_req); /* initialize waitqueue which may be used to wait on this request completion */ @@ -4661,8 +4554,10 @@ zfcp_fsf_req_create(struct zfcp_adapter *adapter, u32 fsf_cmd, int req_flags, goto failed_sbals; } - fsf_req->adapter = adapter; /* pointer to "parent" adapter */ - fsf_req->fsf_command = fsf_cmd; + if (fsf_req->qtcb) { + fsf_req->seq_no = adapter->fsf_req_seq_no; + fsf_req->qtcb->prefix.req_seq_no = adapter->fsf_req_seq_no; + } fsf_req->sbal_number = 1; fsf_req->sbal_first = req_queue->free_index; fsf_req->sbal_curr = req_queue->free_index; @@ -4713,9 +4608,9 @@ zfcp_fsf_req_send(struct zfcp_fsf_req *fsf_req, struct timer_list *timer) struct zfcp_adapter *adapter; struct zfcp_qdio_queue *req_queue; volatile struct qdio_buffer_element *sbale; + int inc_seq_no; int new_distance_from_int; unsigned long flags; - int inc_seq_no = 1; int retval = 0; adapter = fsf_req->adapter; @@ -4729,23 +4624,13 @@ zfcp_fsf_req_send(struct zfcp_fsf_req *fsf_req, struct timer_list *timer) ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_TRACE, (char *) sbale[1].addr, sbale[1].length); - /* set sequence counter in QTCB */ - if (likely(fsf_req->qtcb)) { - fsf_req->qtcb->prefix.req_seq_no = adapter->fsf_req_seq_no; - fsf_req->seq_no = adapter->fsf_req_seq_no; - ZFCP_LOG_TRACE("FSF request %p of adapter %s gets " - "FSF sequence counter value of %i\n", - fsf_req, - zfcp_get_busid_by_adapter(adapter), - fsf_req->qtcb->prefix.req_seq_no); - } else - inc_seq_no = 0; - /* put allocated FSF request at list tail */ spin_lock_irqsave(&adapter->fsf_req_list_lock, flags); list_add_tail(&fsf_req->list, &adapter->fsf_req_list_head); spin_unlock_irqrestore(&adapter->fsf_req_list_lock, flags); + inc_seq_no = (fsf_req->qtcb != NULL); + /* figure out expiration time of timeout and start timeout */ if (unlikely(timer)) { timer->expires += jiffies; @@ -4775,6 +4660,8 @@ zfcp_fsf_req_send(struct zfcp_fsf_req *fsf_req, struct timer_list *timer) req_queue->free_index %= QDIO_MAX_BUFFERS_PER_Q; /* wrap if needed */ new_distance_from_int = zfcp_qdio_determine_pci(req_queue, fsf_req); + fsf_req->issued = get_clock(); + retval = do_QDIO(adapter->ccw_device, QDIO_FLAG_SYNC_OUTPUT, 0, fsf_req->sbal_first, fsf_req->sbal_number, NULL); @@ -4813,15 +4700,11 @@ zfcp_fsf_req_send(struct zfcp_fsf_req *fsf_req, struct timer_list *timer) * routines resulting in missing sequence counter values * otherwise, */ + /* Don't increase for unsolicited status */ - if (likely(inc_seq_no)) { + if (inc_seq_no) adapter->fsf_req_seq_no++; - ZFCP_LOG_TRACE - ("FSF sequence counter value of adapter %s " - "increased to %i\n", - zfcp_get_busid_by_adapter(adapter), - adapter->fsf_req_seq_no); - } + /* count FSF requests pending */ atomic_inc(&adapter->fsf_reqs_active); } diff --git a/drivers/s390/scsi/zfcp_fsf.h b/drivers/s390/scsi/zfcp_fsf.h index 07140dfda2a..57ce0706007 100644 --- a/drivers/s390/scsi/zfcp_fsf.h +++ b/drivers/s390/scsi/zfcp_fsf.h @@ -116,6 +116,7 @@ #define FSF_INVALID_COMMAND_OPTION 0x000000E5 /* #define FSF_ERROR 0x000000FF */ +#define FSF_PROT_STATUS_QUAL_SIZE 16 #define FSF_STATUS_QUALIFIER_SIZE 16 /* FSF status qualifier, recommendations */ @@ -311,6 +312,7 @@ struct fsf_qual_locallink_error { } __attribute__ ((packed)); union fsf_prot_status_qual { + u64 doubleword[FSF_PROT_STATUS_QUAL_SIZE / sizeof(u64)]; struct fsf_qual_version_error version_error; struct fsf_qual_sequence_error sequence_error; struct fsf_qual_locallink_error locallink_error; @@ -331,6 +333,7 @@ union fsf_status_qual { u8 byte[FSF_STATUS_QUALIFIER_SIZE]; u16 halfword[FSF_STATUS_QUALIFIER_SIZE / sizeof (u16)]; u32 word[FSF_STATUS_QUALIFIER_SIZE / sizeof (u32)]; + u64 doubleword[FSF_STATUS_QUALIFIER_SIZE / sizeof(u64)]; struct fsf_queue_designator fsf_queue_designator; } __attribute__ ((packed)); diff --git a/drivers/s390/scsi/zfcp_qdio.c b/drivers/s390/scsi/zfcp_qdio.c index 24e16ec331d..d719f66a29a 100644 --- a/drivers/s390/scsi/zfcp_qdio.c +++ b/drivers/s390/scsi/zfcp_qdio.c @@ -54,8 +54,7 @@ static inline int zfcp_qdio_sbals_from_buffer static qdio_handler_t zfcp_qdio_request_handler; static qdio_handler_t zfcp_qdio_response_handler; static int zfcp_qdio_handler_error_check(struct zfcp_adapter *, - unsigned int, - unsigned int, unsigned int); + unsigned int, unsigned int, unsigned int, int, int); #define ZFCP_LOG_AREA ZFCP_LOG_AREA_QDIO @@ -214,22 +213,12 @@ zfcp_qdio_allocate(struct zfcp_adapter *adapter) * */ static inline int -zfcp_qdio_handler_error_check(struct zfcp_adapter *adapter, - unsigned int status, - unsigned int qdio_error, unsigned int siga_error) +zfcp_qdio_handler_error_check(struct zfcp_adapter *adapter, unsigned int status, + unsigned int qdio_error, unsigned int siga_error, + int first_element, int elements_processed) { int retval = 0; - if (ZFCP_LOG_CHECK(ZFCP_LOG_LEVEL_TRACE)) { - if (status & QDIO_STATUS_INBOUND_INT) { - ZFCP_LOG_TRACE("status is" - " QDIO_STATUS_INBOUND_INT \n"); - } - if (status & QDIO_STATUS_OUTBOUND_INT) { - ZFCP_LOG_TRACE("status is" - " QDIO_STATUS_OUTBOUND_INT \n"); - } - } if (unlikely(status & QDIO_STATUS_LOOK_FOR_ERROR)) { retval = -EIO; @@ -237,9 +226,10 @@ zfcp_qdio_handler_error_check(struct zfcp_adapter *adapter, "qdio_error=0x%x, siga_error=0x%x)\n", status, qdio_error, siga_error); - /* Restarting IO on the failed adapter from scratch */ - debug_text_event(adapter->erp_dbf, 1, "qdio_err"); + zfcp_hba_dbf_event_qdio(adapter, status, qdio_error, siga_error, + first_element, elements_processed); /* + * Restarting IO on the failed adapter from scratch. * Since we have been using this adapter, it is save to assume * that it is not failed but recoverable. The card seems to * report link-up events by self-initiated queue shutdown. @@ -282,7 +272,8 @@ zfcp_qdio_request_handler(struct ccw_device *ccw_device, first_element, elements_processed); if (unlikely(zfcp_qdio_handler_error_check(adapter, status, qdio_error, - siga_error))) + siga_error, first_element, + elements_processed))) goto out; /* * we stored address of struct zfcp_adapter data structure @@ -334,7 +325,8 @@ zfcp_qdio_response_handler(struct ccw_device *ccw_device, queue = &adapter->response_queue; if (unlikely(zfcp_qdio_handler_error_check(adapter, status, qdio_error, - siga_error))) + siga_error, first_element, + elements_processed))) goto out; /* diff --git a/drivers/s390/scsi/zfcp_scsi.c b/drivers/s390/scsi/zfcp_scsi.c index eeb5152854e..c1403a23174 100644 --- a/drivers/s390/scsi/zfcp_scsi.c +++ b/drivers/s390/scsi/zfcp_scsi.c @@ -44,7 +44,8 @@ static int zfcp_scsi_eh_abort_handler(struct scsi_cmnd *); static int zfcp_scsi_eh_device_reset_handler(struct scsi_cmnd *); static int zfcp_scsi_eh_bus_reset_handler(struct scsi_cmnd *); static int zfcp_scsi_eh_host_reset_handler(struct scsi_cmnd *); -static int zfcp_task_management_function(struct zfcp_unit *, u8); +static int zfcp_task_management_function(struct zfcp_unit *, u8, + struct scsi_cmnd *); static struct zfcp_unit *zfcp_unit_lookup(struct zfcp_adapter *, int, scsi_id_t, scsi_lun_t); @@ -242,7 +243,10 @@ static void zfcp_scsi_command_fail(struct scsi_cmnd *scpnt, int result) { set_host_byte(&scpnt->result, result); - zfcp_cmd_dbf_event_scsi("failing", scpnt); + if ((scpnt->device != NULL) && (scpnt->device->host != NULL)) + zfcp_scsi_dbf_event_result("fail", 4, + (struct zfcp_adapter*) scpnt->device->host->hostdata[0], + scpnt); /* return directly */ scpnt->scsi_done(scpnt); } @@ -434,7 +438,8 @@ zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt) struct zfcp_adapter *adapter; struct zfcp_unit *unit; int retval = SUCCESS; - struct zfcp_fsf_req *new_fsf_req, *old_fsf_req; + struct zfcp_fsf_req *new_fsf_req = NULL; + struct zfcp_fsf_req *old_fsf_req; unsigned long flags; scsi_host = scpnt->device->host; @@ -457,11 +462,8 @@ zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt) old_fsf_req = (struct zfcp_fsf_req *) scpnt->host_scribble; if (!old_fsf_req) { write_unlock_irqrestore(&adapter->abort_lock, flags); - ZFCP_LOG_NORMAL("bug: no old fsf request found\n"); - ZFCP_LOG_NORMAL("scsi_cmnd:\n"); - ZFCP_HEX_DUMP(ZFCP_LOG_LEVEL_NORMAL, - (char *) scpnt, sizeof (struct scsi_cmnd)); - retval = FAILED; + zfcp_scsi_dbf_event_abort("lte1", adapter, scpnt, new_fsf_req); + retval = SUCCESS; goto out; } old_fsf_req->data = 0; @@ -473,25 +475,27 @@ zfcp_scsi_eh_abort_handler(struct scsi_cmnd *scpnt) new_fsf_req = zfcp_fsf_abort_fcp_command((unsigned long) old_fsf_req, adapter, unit, 0); if (!new_fsf_req) { + ZFCP_LOG_INFO("error: initiation of Abort FCP Cmnd failed\n"); retval = FAILED; - ZFCP_LOG_NORMAL("error: initiation of Abort FCP Cmnd " - "failed\n"); goto out; } /* wait for completion of abort */ __wait_event(new_fsf_req->completion_wq, new_fsf_req->status & ZFCP_STATUS_FSFREQ_COMPLETED); - zfcp_fsf_req_free(new_fsf_req); /* status should be valid since signals were not permitted */ if (new_fsf_req->status & ZFCP_STATUS_FSFREQ_ABORTSUCCEEDED) { + zfcp_scsi_dbf_event_abort("okay", adapter, scpnt, new_fsf_req); retval = SUCCESS; } else if (new_fsf_req->status & ZFCP_STATUS_FSFREQ_ABORTNOTNEEDED) { + zfcp_scsi_dbf_event_abort("lte2", adapter, scpnt, new_fsf_req); retval = SUCCESS; } else { + zfcp_scsi_dbf_event_abort("fail", adapter, scpnt, new_fsf_req); retval = FAILED; } + zfcp_fsf_req_free(new_fsf_req); out: return retval; } @@ -525,8 +529,9 @@ zfcp_scsi_eh_device_reset_handler(struct scsi_cmnd *scpnt) */ if (!atomic_test_mask(ZFCP_STATUS_UNIT_NOTSUPPUNITRESET, &unit->status)) { - retval = - zfcp_task_management_function(unit, FCP_LOGICAL_UNIT_RESET); + retval = zfcp_task_management_function(unit, + FCP_LOGICAL_UNIT_RESET, + scpnt); if (retval) { ZFCP_LOG_DEBUG("unit reset failed (unit=%p)\n", unit); if (retval == -ENOTSUPP) @@ -542,7 +547,7 @@ zfcp_scsi_eh_device_reset_handler(struct scsi_cmnd *scpnt) goto out; } } - retval = zfcp_task_management_function(unit, FCP_TARGET_RESET); + retval = zfcp_task_management_function(unit, FCP_TARGET_RESET, scpnt); if (retval) { ZFCP_LOG_DEBUG("target reset failed (unit=%p)\n", unit); retval = FAILED; @@ -555,7 +560,8 @@ zfcp_scsi_eh_device_reset_handler(struct scsi_cmnd *scpnt) } static int -zfcp_task_management_function(struct zfcp_unit *unit, u8 tm_flags) +zfcp_task_management_function(struct zfcp_unit *unit, u8 tm_flags, + struct scsi_cmnd *scpnt) { struct zfcp_adapter *adapter = unit->port->adapter; struct zfcp_fsf_req *fsf_req; @@ -569,6 +575,7 @@ zfcp_task_management_function(struct zfcp_unit *unit, u8 tm_flags) "failed for unit 0x%016Lx on port 0x%016Lx on " "adapter %s\n", unit->fcp_lun, unit->port->wwpn, zfcp_get_busid_by_adapter(adapter)); + zfcp_scsi_dbf_event_devreset("nres", tm_flags, unit, scpnt); retval = -ENOMEM; goto out; } @@ -576,11 +583,17 @@ zfcp_task_management_function(struct zfcp_unit *unit, u8 tm_flags) __wait_event(fsf_req->completion_wq, fsf_req->status & ZFCP_STATUS_FSFREQ_COMPLETED); - /* check completion status of task management function */ - if (fsf_req->status & ZFCP_STATUS_FSFREQ_TMFUNCFAILED) + /* + * check completion status of task management function + */ + if (fsf_req->status & ZFCP_STATUS_FSFREQ_TMFUNCFAILED) { + zfcp_scsi_dbf_event_devreset("fail", tm_flags, unit, scpnt); retval = -EIO; - else if (fsf_req->status & ZFCP_STATUS_FSFREQ_TMFUNCNOTSUPP) + } else if (fsf_req->status & ZFCP_STATUS_FSFREQ_TMFUNCNOTSUPP) { + zfcp_scsi_dbf_event_devreset("nsup", tm_flags, unit, scpnt); retval = -ENOTSUPP; + } else + zfcp_scsi_dbf_event_devreset("okay", tm_flags, unit, scpnt); zfcp_fsf_req_free(fsf_req); out: From aef4a983090fa590481a86d9690dc3fa6bb121fa Mon Sep 17 00:00:00 2001 From: Maxim Shchetynin Date: Tue, 13 Sep 2005 21:51:16 +0200 Subject: [PATCH 041/204] [SCSI] zfcp: provide support for NPIV N_Port ID Virtualization (NPIV) allows a single FCP port to appear as multiple, distinct ports providing separate port identification. NPIV is supported by FC HBAs on System z9. zfcp was adapted to support this new feature. Signed-off-by: Andreas Herrmann Signed-off-by: James Bottomley --- drivers/s390/scsi/zfcp_dbf.c | 11 +- drivers/s390/scsi/zfcp_def.h | 13 +- drivers/s390/scsi/zfcp_erp.c | 95 ++++++- drivers/s390/scsi/zfcp_ext.h | 3 +- drivers/s390/scsi/zfcp_fsf.c | 330 ++++++++++++++++--------- drivers/s390/scsi/zfcp_fsf.h | 51 +++- drivers/s390/scsi/zfcp_sysfs_adapter.c | 4 + 7 files changed, 367 insertions(+), 140 deletions(-) diff --git a/drivers/s390/scsi/zfcp_dbf.c b/drivers/s390/scsi/zfcp_dbf.c index fff1537335c..64d9b90373f 100644 --- a/drivers/s390/scsi/zfcp_dbf.c +++ b/drivers/s390/scsi/zfcp_dbf.c @@ -279,9 +279,18 @@ zfcp_hba_dbf_event_fsf_unsol(const char *tag, struct zfcp_adapter *adapter, break; case FSF_STATUS_READ_LINK_DOWN: - rec->type.status.payload_size = sizeof(u64); + switch (status_buffer->status_subtype) { + case FSF_STATUS_READ_SUB_NO_PHYSICAL_LINK: + case FSF_STATUS_READ_SUB_FDISC_FAILED: + rec->type.status.payload_size = + sizeof(struct fsf_link_down_info); + } break; + case FSF_STATUS_READ_FEATURE_UPDATE_ALERT: + rec->type.status.payload_size = + ZFCP_DBF_UNSOL_PAYLOAD_FEATURE_UPDATE_ALERT; + break; } memcpy(&rec->type.status.payload, &status_buffer->payload, rec->type.status.payload_size); diff --git a/drivers/s390/scsi/zfcp_def.h b/drivers/s390/scsi/zfcp_def.h index fc5bb6f3180..4b6af8e07e8 100644 --- a/drivers/s390/scsi/zfcp_def.h +++ b/drivers/s390/scsi/zfcp_def.h @@ -66,7 +66,7 @@ /********************* GENERAL DEFINES *********************************/ /* zfcp version number, it consists of major, minor, and patch-level number */ -#define ZFCP_VERSION "4.4.0" +#define ZFCP_VERSION "4.5.0" /** * zfcp_sg_to_address - determine kernel address from struct scatterlist @@ -154,6 +154,11 @@ typedef u32 scsi_lun_t; #define ZFCP_EXCHANGE_CONFIG_DATA_FIRST_SLEEP 100 #define ZFCP_EXCHANGE_CONFIG_DATA_RETRIES 7 +/* Retry 5 times every 2 second, then every minute */ +#define ZFCP_EXCHANGE_PORT_DATA_SHORT_RETRIES 5 +#define ZFCP_EXCHANGE_PORT_DATA_SHORT_SLEEP 200 +#define ZFCP_EXCHANGE_PORT_DATA_LONG_SLEEP 6000 + /* timeout value for "default timer" for fsf requests */ #define ZFCP_FSF_REQUEST_TIMEOUT (60*HZ); @@ -638,6 +643,7 @@ do { \ #define ZFCP_STATUS_ADAPTER_ERP_THREAD_KILL 0x00000080 #define ZFCP_STATUS_ADAPTER_ERP_PENDING 0x00000100 #define ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED 0x00000200 +#define ZFCP_STATUS_ADAPTER_XPORT_OK 0x00000800 #define ZFCP_STATUS_ADAPTER_SCSI_UP \ (ZFCP_STATUS_COMMON_UNBLOCKED | \ @@ -915,13 +921,16 @@ struct zfcp_adapter { wwn_t peer_wwnn; /* P2P peer WWNN */ wwn_t peer_wwpn; /* P2P peer WWPN */ fc_id_t peer_d_id; /* P2P peer D_ID */ + wwn_t physical_wwpn; /* WWPN of physical port */ + fc_id_t physical_s_id; /* local FC port ID */ struct ccw_device *ccw_device; /* S/390 ccw device */ u8 fc_service_class; u32 fc_topology; /* FC topology */ u32 fc_link_speed; /* FC interface speed */ u32 hydra_version; /* Hydra version */ u32 fsf_lic_version; - u32 supported_features;/* of FCP channel */ + u32 adapter_features; /* FCP channel features */ + u32 connection_features; /* host connection features */ u32 hardware_version; /* of FCP channel */ u8 serial_number[32]; /* of hardware */ struct Scsi_Host *scsi_host; /* Pointer to mid-layer */ diff --git a/drivers/s390/scsi/zfcp_erp.c b/drivers/s390/scsi/zfcp_erp.c index 376cb0f6cb7..c4a6799aed4 100644 --- a/drivers/s390/scsi/zfcp_erp.c +++ b/drivers/s390/scsi/zfcp_erp.c @@ -82,6 +82,7 @@ static int zfcp_erp_adapter_strategy_open(struct zfcp_erp_action *); static int zfcp_erp_adapter_strategy_open_qdio(struct zfcp_erp_action *); static int zfcp_erp_adapter_strategy_open_fsf(struct zfcp_erp_action *); static int zfcp_erp_adapter_strategy_open_fsf_xconfig(struct zfcp_erp_action *); +static int zfcp_erp_adapter_strategy_open_fsf_xport(struct zfcp_erp_action *); static int zfcp_erp_adapter_strategy_open_fsf_statusread( struct zfcp_erp_action *); @@ -2258,16 +2259,21 @@ zfcp_erp_adapter_strategy_close_qdio(struct zfcp_erp_action *erp_action) static int zfcp_erp_adapter_strategy_open_fsf(struct zfcp_erp_action *erp_action) { - int retval; + int xconfig, xport; - /* do 'exchange configuration data' */ - retval = zfcp_erp_adapter_strategy_open_fsf_xconfig(erp_action); - if (retval == ZFCP_ERP_FAILED) - return retval; + if (atomic_test_mask(ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED, + &erp_action->adapter->status)) { + zfcp_erp_adapter_strategy_open_fsf_xport(erp_action); + atomic_set(&erp_action->adapter->erp_counter, 0); + return ZFCP_ERP_FAILED; + } - /* start the desired number of Status Reads */ - retval = zfcp_erp_adapter_strategy_open_fsf_statusread(erp_action); - return retval; + xconfig = zfcp_erp_adapter_strategy_open_fsf_xconfig(erp_action); + xport = zfcp_erp_adapter_strategy_open_fsf_xport(erp_action); + if ((xconfig == ZFCP_ERP_FAILED) || (xport == ZFCP_ERP_FAILED)) + return ZFCP_ERP_FAILED; + + return zfcp_erp_adapter_strategy_open_fsf_statusread(erp_action); } /* @@ -2350,6 +2356,76 @@ zfcp_erp_adapter_strategy_open_fsf_xconfig(struct zfcp_erp_action *erp_action) return retval; } +static int +zfcp_erp_adapter_strategy_open_fsf_xport(struct zfcp_erp_action *erp_action) +{ + int retval = ZFCP_ERP_SUCCEEDED; + int retries; + int sleep; + struct zfcp_adapter *adapter = erp_action->adapter; + + atomic_clear_mask(ZFCP_STATUS_ADAPTER_XPORT_OK, &adapter->status); + + for (retries = 0; ; retries++) { + ZFCP_LOG_DEBUG("Doing exchange port data\n"); + zfcp_erp_action_to_running(erp_action); + zfcp_erp_timeout_init(erp_action); + if (zfcp_fsf_exchange_port_data(erp_action, adapter, NULL)) { + retval = ZFCP_ERP_FAILED; + debug_text_event(adapter->erp_dbf, 5, "a_fstx_xf"); + ZFCP_LOG_INFO("error: initiation of exchange of " + "port data failed for adapter %s\n", + zfcp_get_busid_by_adapter(adapter)); + break; + } + debug_text_event(adapter->erp_dbf, 6, "a_fstx_xok"); + ZFCP_LOG_DEBUG("Xchange underway\n"); + + /* + * Why this works: + * Both the normal completion handler as well as the timeout + * handler will do an 'up' when the 'exchange port data' + * request completes or times out. Thus, the signal to go on + * won't be lost utilizing this semaphore. + * Furthermore, this 'adapter_reopen' action is + * guaranteed to be the only action being there (highest action + * which prevents other actions from being created). + * Resulting from that, the wake signal recognized here + * _must_ be the one belonging to the 'exchange port + * data' request. + */ + down(&adapter->erp_ready_sem); + if (erp_action->status & ZFCP_STATUS_ERP_TIMEDOUT) { + ZFCP_LOG_INFO("error: exchange of port data " + "for adapter %s timed out\n", + zfcp_get_busid_by_adapter(adapter)); + break; + } + + if (!atomic_test_mask(ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED, + &adapter->status)) + break; + + ZFCP_LOG_DEBUG("host connection still initialising... " + "waiting and retrying...\n"); + /* sleep a little bit before retry */ + sleep = retries < ZFCP_EXCHANGE_PORT_DATA_SHORT_RETRIES ? + ZFCP_EXCHANGE_PORT_DATA_SHORT_SLEEP : + ZFCP_EXCHANGE_PORT_DATA_LONG_SLEEP; + msleep(jiffies_to_msecs(sleep)); + } + + if (atomic_test_mask(ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED, + &adapter->status)) { + ZFCP_LOG_INFO("error: exchange of port data for " + "adapter %s failed\n", + zfcp_get_busid_by_adapter(adapter)); + retval = ZFCP_ERP_FAILED; + } + + return retval; +} + /* * function: * @@ -3599,6 +3675,9 @@ zfcp_erp_adapter_access_changed(struct zfcp_adapter *adapter) struct zfcp_port *port; unsigned long flags; + if (adapter->connection_features & FSF_FEATURE_NPIV_MODE) + return; + debug_text_event(adapter->erp_dbf, 3, "a_access_recover"); debug_event(adapter->erp_dbf, 3, &adapter->name, 8); diff --git a/drivers/s390/scsi/zfcp_ext.h b/drivers/s390/scsi/zfcp_ext.h index 104b7423fd4..c3782261cb5 100644 --- a/drivers/s390/scsi/zfcp_ext.h +++ b/drivers/s390/scsi/zfcp_ext.h @@ -96,7 +96,8 @@ extern int zfcp_fsf_open_unit(struct zfcp_erp_action *); extern int zfcp_fsf_close_unit(struct zfcp_erp_action *); extern int zfcp_fsf_exchange_config_data(struct zfcp_erp_action *); -extern int zfcp_fsf_exchange_port_data(struct zfcp_adapter *, +extern int zfcp_fsf_exchange_port_data(struct zfcp_erp_action *, + struct zfcp_adapter *, struct fsf_qtcb_bottom_port *); extern int zfcp_fsf_control_file(struct zfcp_adapter *, struct zfcp_fsf_req **, u32, u32, struct zfcp_sg_list *); diff --git a/drivers/s390/scsi/zfcp_fsf.c b/drivers/s390/scsi/zfcp_fsf.c index 80ab721ddff..f994f453e2d 100644 --- a/drivers/s390/scsi/zfcp_fsf.c +++ b/drivers/s390/scsi/zfcp_fsf.c @@ -59,6 +59,8 @@ static int zfcp_fsf_req_send(struct zfcp_fsf_req *, struct timer_list *); static int zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *); static int zfcp_fsf_fsfstatus_eval(struct zfcp_fsf_req *); static int zfcp_fsf_fsfstatus_qual_eval(struct zfcp_fsf_req *); +static void zfcp_fsf_link_down_info_eval(struct zfcp_adapter *, + struct fsf_link_down_info *); static int zfcp_fsf_req_dispatch(struct zfcp_fsf_req *); static void zfcp_fsf_req_dismiss(struct zfcp_fsf_req *); @@ -375,87 +377,19 @@ zfcp_fsf_protstatus_eval(struct zfcp_fsf_req *fsf_req) break; case FSF_PROT_DUPLICATE_REQUEST_ID: - if (fsf_req->qtcb) { ZFCP_LOG_NORMAL("bug: The request identifier 0x%Lx " "to the adapter %s is ambiguous. " - "Stopping all operations on this " - "adapter.\n", - *(unsigned long long *) - (&fsf_req->qtcb->bottom.support. - req_handle), + "Stopping all operations on this adapter.\n", + *(unsigned long long*) + (&qtcb->bottom.support.req_handle), zfcp_get_busid_by_adapter(adapter)); - } else { - ZFCP_LOG_NORMAL("bug: The request identifier %p " - "to the adapter %s is ambiguous. " - "Stopping all operations on this " - "adapter. " - "(bug: got this for an unsolicited " - "status read request)\n", - fsf_req, - zfcp_get_busid_by_adapter(adapter)); - } zfcp_erp_adapter_shutdown(adapter, 0); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; case FSF_PROT_LINK_DOWN: - /* - * 'test and set' is not atomic here - - * it's ok as long as calls to our response queue handler - * (and thus execution of this code here) are serialized - * by the qdio module - */ - if (!atomic_test_mask(ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED, - &adapter->status)) { - switch (fsf_req->qtcb->prefix.prot_status_qual. - locallink_error.code) { - case FSF_PSQ_LINK_NOLIGHT: - ZFCP_LOG_INFO("The local link to adapter %s " - "is down (no light detected).\n", - zfcp_get_busid_by_adapter( - adapter)); - break; - case FSF_PSQ_LINK_WRAPPLUG: - ZFCP_LOG_INFO("The local link to adapter %s " - "is down (wrap plug detected).\n", - zfcp_get_busid_by_adapter( - adapter)); - break; - case FSF_PSQ_LINK_NOFCP: - ZFCP_LOG_INFO("The local link to adapter %s " - "is down (adjacent node on " - "link does not support FCP).\n", - zfcp_get_busid_by_adapter( - adapter)); - break; - default: - ZFCP_LOG_INFO("The local link to adapter %s " - "is down " - "(warning: unknown reason " - "code).\n", - zfcp_get_busid_by_adapter( - adapter)); - break; - - } - /* - * Due to the 'erp failed' flag the adapter won't - * be recovered but will be just set to 'blocked' - * state. All subordinary devices will have state - * 'blocked' and 'erp failed', too. - * Thus the adapter is still able to provide - * 'link up' status without being flooded with - * requests. - * (note: even 'close port' is not permitted) - */ - ZFCP_LOG_INFO("Stopping all operations for adapter " - "%s.\n", - zfcp_get_busid_by_adapter(adapter)); - atomic_set_mask(ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED | - ZFCP_STATUS_COMMON_ERP_FAILED, - &adapter->status); - zfcp_erp_adapter_reopen(adapter, 0); - } + zfcp_fsf_link_down_info_eval(adapter, + &prot_status_qual->link_down_info); fsf_req->status |= ZFCP_STATUS_FSFREQ_ERROR; break; @@ -613,6 +547,110 @@ zfcp_fsf_fsfstatus_qual_eval(struct zfcp_fsf_req *fsf_req) return retval; } +/** + * zfcp_fsf_link_down_info_eval - evaluate link down information block + */ +static void +zfcp_fsf_link_down_info_eval(struct zfcp_adapter *adapter, + struct fsf_link_down_info *link_down) +{ + switch (link_down->error_code) { + case FSF_PSQ_LINK_NO_LIGHT: + ZFCP_LOG_NORMAL("The local link to adapter %s is down " + "(no light detected)\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_PSQ_LINK_WRAP_PLUG: + ZFCP_LOG_NORMAL("The local link to adapter %s is down " + "(wrap plug detected)\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_PSQ_LINK_NO_FCP: + ZFCP_LOG_NORMAL("The local link to adapter %s is down " + "(adjacent node on link does not support FCP)\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_PSQ_LINK_FIRMWARE_UPDATE: + ZFCP_LOG_NORMAL("The local link to adapter %s is down " + "(firmware update in progress)\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_PSQ_LINK_INVALID_WWPN: + ZFCP_LOG_NORMAL("The local link to adapter %s is down " + "(duplicate or invalid WWPN detected)\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_PSQ_LINK_NO_NPIV_SUPPORT: + ZFCP_LOG_NORMAL("The local link to adapter %s is down " + "(no support for NPIV by Fabric)\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_PSQ_LINK_NO_FCP_RESOURCES: + ZFCP_LOG_NORMAL("The local link to adapter %s is down " + "(out of resource in FCP daughtercard)\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_PSQ_LINK_NO_FABRIC_RESOURCES: + ZFCP_LOG_NORMAL("The local link to adapter %s is down " + "(out of resource in Fabric)\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_PSQ_LINK_FABRIC_LOGIN_UNABLE: + ZFCP_LOG_NORMAL("The local link to adapter %s is down " + "(unable to Fabric login)\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_PSQ_LINK_WWPN_ASSIGNMENT_CORRUPTED: + ZFCP_LOG_NORMAL("WWPN assignment file corrupted on adapter %s\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_PSQ_LINK_MODE_TABLE_CURRUPTED: + ZFCP_LOG_NORMAL("Mode table corrupted on adapter %s\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_PSQ_LINK_NO_WWPN_ASSIGNMENT: + ZFCP_LOG_NORMAL("No WWPN for assignment table on adapter %s\n", + zfcp_get_busid_by_adapter(adapter)); + break; + default: + ZFCP_LOG_NORMAL("The local link to adapter %s is down " + "(warning: unknown reason code %d)\n", + zfcp_get_busid_by_adapter(adapter), + link_down->error_code); + } + + if (adapter->connection_features & FSF_FEATURE_NPIV_MODE) + ZFCP_LOG_DEBUG("Debug information to link down: " + "primary_status=0x%02x " + "ioerr_code=0x%02x " + "action_code=0x%02x " + "reason_code=0x%02x " + "explanation_code=0x%02x " + "vendor_specific_code=0x%02x\n", + link_down->primary_status, + link_down->ioerr_code, + link_down->action_code, + link_down->reason_code, + link_down->explanation_code, + link_down->vendor_specific_code); + + if (!atomic_test_mask(ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED, + &adapter->status)) { + atomic_set_mask(ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED, + &adapter->status); + switch (link_down->error_code) { + case FSF_PSQ_LINK_NO_LIGHT: + case FSF_PSQ_LINK_WRAP_PLUG: + case FSF_PSQ_LINK_NO_FCP: + case FSF_PSQ_LINK_FIRMWARE_UPDATE: + zfcp_erp_adapter_reopen(adapter, 0); + break; + default: + zfcp_erp_adapter_failed(adapter); + } + } +} + /* * function: zfcp_fsf_req_dispatch * @@ -877,17 +915,32 @@ zfcp_fsf_status_read_handler(struct zfcp_fsf_req *fsf_req) break; case FSF_STATUS_READ_LINK_DOWN: - debug_text_event(adapter->erp_dbf, 0, "unsol_link_down:"); - ZFCP_LOG_INFO("Local link to adapter %s is down\n", + switch (status_buffer->status_subtype) { + case FSF_STATUS_READ_SUB_NO_PHYSICAL_LINK: + ZFCP_LOG_INFO("Physical link to adapter %s is down\n", + zfcp_get_busid_by_adapter(adapter)); + break; + case FSF_STATUS_READ_SUB_FDISC_FAILED: + ZFCP_LOG_INFO("Local link to adapter %s is down " + "due to failed FDISC login\n", zfcp_get_busid_by_adapter(adapter)); - atomic_set_mask(ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED, - &adapter->status); - zfcp_erp_adapter_failed(adapter); + break; + case FSF_STATUS_READ_SUB_FIRMWARE_UPDATE: + ZFCP_LOG_INFO("Local link to adapter %s is down " + "due to firmware update on adapter\n", + zfcp_get_busid_by_adapter(adapter)); + break; + default: + ZFCP_LOG_INFO("Local link to adapter %s is down " + "due to unknown reason\n", + zfcp_get_busid_by_adapter(adapter)); + }; + zfcp_fsf_link_down_info_eval(adapter, + (struct fsf_link_down_info *) &status_buffer->payload); break; case FSF_STATUS_READ_LINK_UP: - debug_text_event(adapter->erp_dbf, 2, "unsol_link_up:"); - ZFCP_LOG_INFO("Local link to adapter %s was replugged. " + ZFCP_LOG_NORMAL("Local link to adapter %s was replugged. " "Restarting operations on this adapter\n", zfcp_get_busid_by_adapter(adapter)); /* All ports should be marked as ready to run again */ @@ -922,6 +975,16 @@ zfcp_fsf_status_read_handler(struct zfcp_fsf_req *fsf_req) } break; + case FSF_STATUS_READ_FEATURE_UPDATE_ALERT: + debug_text_event(adapter->erp_dbf, 2, "unsol_features:"); + ZFCP_LOG_INFO("List of supported features on adapter %s has " + "been changed from 0x%08X to 0x%08X\n", + zfcp_get_busid_by_adapter(adapter), + *(u32*) (status_buffer->payload + 4), + *(u32*) (status_buffer->payload)); + adapter->adapter_features = *(u32*) status_buffer->payload; + break; + default: ZFCP_LOG_NORMAL("warning: An unsolicited status packet of unknown " "type was received (debug info 0x%x)\n", @@ -1281,7 +1344,7 @@ zfcp_fsf_send_ct(struct zfcp_send_ct *ct, mempool_t *pool, sbale[3].addr = zfcp_sg_to_address(&ct->resp[0]); sbale[3].length = ct->resp[0].length; sbale[3].flags |= SBAL_FLAGS_LAST_ENTRY; - } else if (adapter->supported_features & + } else if (adapter->adapter_features & FSF_FEATURE_ELS_CT_CHAINED_SBALS) { /* try to use chained SBALs */ bytes = zfcp_qdio_sbals_from_sg(fsf_req, @@ -1584,7 +1647,7 @@ zfcp_fsf_send_els(struct zfcp_send_els *els) sbale[3].addr = zfcp_sg_to_address(&els->resp[0]); sbale[3].length = els->resp[0].length; sbale[3].flags |= SBAL_FLAGS_LAST_ENTRY; - } else if (adapter->supported_features & + } else if (adapter->adapter_features & FSF_FEATURE_ELS_CT_CHAINED_SBALS) { /* try to use chained SBALs */ bytes = zfcp_qdio_sbals_from_sg(fsf_req, @@ -1877,7 +1940,9 @@ zfcp_fsf_exchange_config_data(struct zfcp_erp_action *erp_action) erp_action->fsf_req->erp_action = erp_action; erp_action->fsf_req->qtcb->bottom.config.feature_selection = - (FSF_FEATURE_CFDC | FSF_FEATURE_LUN_SHARING); + FSF_FEATURE_CFDC | + FSF_FEATURE_LUN_SHARING | + FSF_FEATURE_UPDATE_ALERT; /* start QDIO request for this FSF request */ retval = zfcp_fsf_req_send(erp_action->fsf_req, &erp_action->timer); @@ -1918,7 +1983,8 @@ zfcp_fsf_exchange_config_evaluate(struct zfcp_fsf_req *fsf_req, int xchg_ok) ZFCP_LOG_DEBUG("low/high QTCB version 0x%x/0x%x of FSF\n", bottom->low_qtcb_version, bottom->high_qtcb_version); adapter->fsf_lic_version = bottom->lic_version; - adapter->supported_features = bottom->supported_features; + adapter->adapter_features = bottom->adapter_features; + adapter->connection_features = bottom->connection_features; adapter->peer_wwpn = 0; adapter->peer_wwnn = 0; adapter->peer_d_id = 0; @@ -1930,6 +1996,10 @@ zfcp_fsf_exchange_config_evaluate(struct zfcp_fsf_req *fsf_req, int xchg_ok) adapter->fc_topology = bottom->fc_topology; adapter->fc_link_speed = bottom->fc_link_speed; adapter->hydra_version = bottom->adapter_type; + if (adapter->physical_wwpn == 0) + adapter->physical_wwpn = adapter->wwpn; + if (adapter->physical_s_id == 0) + adapter->physical_s_id = adapter->s_id; } else { adapter->wwnn = 0; adapter->wwpn = 0; @@ -1945,7 +2015,7 @@ zfcp_fsf_exchange_config_evaluate(struct zfcp_fsf_req *fsf_req, int xchg_ok) adapter->peer_wwnn = bottom->plogi_payload.wwnn; } - if(adapter->supported_features & FSF_FEATURE_HBAAPI_MANAGEMENT){ + if (adapter->adapter_features & FSF_FEATURE_HBAAPI_MANAGEMENT) { adapter->hardware_version = bottom->hardware_version; memcpy(adapter->serial_number, bottom->serial_number, 17); EBCASC(adapter->serial_number, sizeof(adapter->serial_number)); @@ -2001,11 +2071,12 @@ zfcp_fsf_exchange_config_data_handler(struct zfcp_fsf_req *fsf_req) { struct fsf_qtcb_bottom_config *bottom; struct zfcp_adapter *adapter = fsf_req->adapter; + struct fsf_qtcb *qtcb = fsf_req->qtcb; if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) return -EIO; - switch (fsf_req->qtcb->header.fsf_status) { + switch (qtcb->header.fsf_status) { case FSF_GOOD: if (zfcp_fsf_exchange_config_evaluate(fsf_req, 1)) @@ -2035,7 +2106,7 @@ zfcp_fsf_exchange_config_data_handler(struct zfcp_fsf_req *fsf_req) zfcp_erp_adapter_shutdown(adapter, 0); return -EIO; case FSF_TOPO_FABRIC: - ZFCP_LOG_INFO("Switched fabric fibrechannel " + ZFCP_LOG_NORMAL("Switched fabric fibrechannel " "network detected at adapter %s.\n", zfcp_get_busid_by_adapter(adapter)); break; @@ -2053,7 +2124,7 @@ zfcp_fsf_exchange_config_data_handler(struct zfcp_fsf_req *fsf_req) zfcp_erp_adapter_shutdown(adapter, 0); return -EIO; } - bottom = &fsf_req->qtcb->bottom.config; + bottom = &qtcb->bottom.config; if (bottom->max_qtcb_size < sizeof(struct fsf_qtcb)) { ZFCP_LOG_NORMAL("bug: Maximum QTCB size (%d bytes) " "allowed by the adapter %s " @@ -2078,12 +2149,10 @@ zfcp_fsf_exchange_config_data_handler(struct zfcp_fsf_req *fsf_req) if (zfcp_fsf_exchange_config_evaluate(fsf_req, 0)) return -EIO; - ZFCP_LOG_INFO("Local link to adapter %s is down\n", - zfcp_get_busid_by_adapter(adapter)); - atomic_set_mask(ZFCP_STATUS_ADAPTER_XCONFIG_OK | - ZFCP_STATUS_ADAPTER_LINK_UNPLUGGED, - &adapter->status); - zfcp_erp_adapter_failed(adapter); + atomic_set_mask(ZFCP_STATUS_ADAPTER_XCONFIG_OK, &adapter->status); + + zfcp_fsf_link_down_info_eval(adapter, + &qtcb->header.fsf_status_qual.link_down_info); break; default: debug_text_event(fsf_req->adapter->erp_dbf, 0, "fsf-stat-ng"); @@ -2097,11 +2166,13 @@ zfcp_fsf_exchange_config_data_handler(struct zfcp_fsf_req *fsf_req) /** * zfcp_fsf_exchange_port_data - request information about local port + * @erp_action: ERP action for the adapter for which port data is requested * @adapter: for which port data is requested * @data: response to exchange port data request */ int -zfcp_fsf_exchange_port_data(struct zfcp_adapter *adapter, +zfcp_fsf_exchange_port_data(struct zfcp_erp_action *erp_action, + struct zfcp_adapter *adapter, struct fsf_qtcb_bottom_port *data) { volatile struct qdio_buffer_element *sbale; @@ -2110,7 +2181,7 @@ zfcp_fsf_exchange_port_data(struct zfcp_adapter *adapter, struct zfcp_fsf_req *fsf_req; struct timer_list *timer; - if(!(adapter->supported_features & FSF_FEATURE_HBAAPI_MANAGEMENT)){ + if (!(adapter->adapter_features & FSF_FEATURE_HBAAPI_MANAGEMENT)) { ZFCP_LOG_INFO("error: exchange port data " "command not supported by adapter %s\n", zfcp_get_busid_by_adapter(adapter)); @@ -2134,6 +2205,12 @@ zfcp_fsf_exchange_port_data(struct zfcp_adapter *adapter, goto out; } + if (erp_action) { + erp_action->fsf_req = fsf_req; + fsf_req->erp_action = erp_action; + } + + if (data) fsf_req->data = (unsigned long) data; sbale = zfcp_qdio_sbale_req(fsf_req, fsf_req->sbal_curr, 0); @@ -2151,6 +2228,8 @@ zfcp_fsf_exchange_port_data(struct zfcp_adapter *adapter, "command on the adapter %s\n", zfcp_get_busid_by_adapter(adapter)); zfcp_fsf_req_free(fsf_req); + if (erp_action) + erp_action->fsf_req = NULL; write_unlock_irqrestore(&adapter->request_queue.queue_lock, lock_flags); goto out; @@ -2179,23 +2258,40 @@ zfcp_fsf_exchange_port_data(struct zfcp_adapter *adapter, static void zfcp_fsf_exchange_port_data_handler(struct zfcp_fsf_req *fsf_req) { - struct fsf_qtcb_bottom_port *bottom; - struct fsf_qtcb_bottom_port *data; - - data = (struct fsf_qtcb_bottom_port*) fsf_req->data; + struct zfcp_adapter *adapter = fsf_req->adapter; + struct fsf_qtcb *qtcb = fsf_req->qtcb; + struct fsf_qtcb_bottom_port *bottom, *data; if (fsf_req->status & ZFCP_STATUS_FSFREQ_ERROR) return; - switch (fsf_req->qtcb->header.fsf_status) { + switch (qtcb->header.fsf_status) { case FSF_GOOD: - bottom = &fsf_req->qtcb->bottom.port; - memcpy(data, bottom, sizeof(*data)); + atomic_set_mask(ZFCP_STATUS_ADAPTER_XPORT_OK, &adapter->status); + + bottom = &qtcb->bottom.port; + data = (struct fsf_qtcb_bottom_port*) fsf_req->data; + if (data) + memcpy(data, bottom, sizeof(struct fsf_qtcb_bottom_port)); + if (adapter->connection_features & FSF_FEATURE_NPIV_MODE) { + adapter->physical_wwpn = bottom->wwpn; + adapter->physical_s_id = bottom->fc_port_id; + } else { + adapter->physical_wwpn = adapter->wwpn; + adapter->physical_s_id = adapter->s_id; + } + break; + + case FSF_EXCHANGE_CONFIG_DATA_INCOMPLETE: + atomic_set_mask(ZFCP_STATUS_ADAPTER_XPORT_OK, &adapter->status); + + zfcp_fsf_link_down_info_eval(adapter, + &qtcb->header.fsf_status_qual.link_down_info); break; default: - debug_text_event(fsf_req->adapter->erp_dbf, 0, "xchg-port-ng"); - debug_event(fsf_req->adapter->erp_dbf, 0, + debug_text_event(adapter->erp_dbf, 0, "xchg-port-ng"); + debug_event(adapter->erp_dbf, 0, &fsf_req->qtcb->header.fsf_status, sizeof(u32)); } } @@ -2629,7 +2725,7 @@ zfcp_fsf_close_physical_port(struct zfcp_erp_action *erp_action) &erp_action->port->status); /* save a pointer to this port */ erp_action->fsf_req->data = (unsigned long) erp_action->port; - /* port to be closeed */ + /* port to be closed */ erp_action->fsf_req->qtcb->header.port_handle = erp_action->port->handle; erp_action->fsf_req->erp_action = erp_action; @@ -2833,6 +2929,7 @@ zfcp_fsf_open_unit(struct zfcp_erp_action *erp_action) erp_action->port->handle; erp_action->fsf_req->qtcb->bottom.support.fcp_lun = erp_action->unit->fcp_lun; + if (!(erp_action->adapter->connection_features & FSF_FEATURE_NPIV_MODE)) erp_action->fsf_req->qtcb->bottom.support.option = FSF_OPEN_LUN_SUPPRESS_BOXING; atomic_set_mask(ZFCP_STATUS_COMMON_OPENING, &erp_action->unit->status); @@ -2880,7 +2977,7 @@ zfcp_fsf_open_unit_handler(struct zfcp_fsf_req *fsf_req) struct fsf_qtcb_bottom_support *bottom; struct fsf_queue_designator *queue_designator; u16 subtable, rule, counter; - u32 allowed, exclusive, readwrite; + int exclusive, readwrite; unit = (struct zfcp_unit *) fsf_req->data; @@ -2894,10 +2991,6 @@ zfcp_fsf_open_unit_handler(struct zfcp_fsf_req *fsf_req) bottom = &fsf_req->qtcb->bottom.support; queue_designator = &header->fsf_status_qual.fsf_queue_designator; - allowed = bottom->lun_access_info & FSF_UNIT_ACCESS_OPEN_LUN_ALLOWED; - exclusive = bottom->lun_access_info & FSF_UNIT_ACCESS_EXCLUSIVE; - readwrite = bottom->lun_access_info & FSF_UNIT_ACCESS_OUTBOUND_TRANSFER; - atomic_clear_mask(ZFCP_STATUS_COMMON_ACCESS_DENIED | ZFCP_STATUS_UNIT_SHARED | ZFCP_STATUS_UNIT_READONLY, @@ -3071,10 +3164,15 @@ zfcp_fsf_open_unit_handler(struct zfcp_fsf_req *fsf_req) unit->handle); /* mark unit as open */ atomic_set_mask(ZFCP_STATUS_COMMON_OPEN, &unit->status); - atomic_clear_mask(ZFCP_STATUS_COMMON_ACCESS_DENIED | - ZFCP_STATUS_COMMON_ACCESS_BOXED, - &unit->status); - if (adapter->supported_features & FSF_FEATURE_LUN_SHARING){ + + if (!(adapter->connection_features & FSF_FEATURE_NPIV_MODE) && + (adapter->adapter_features & FSF_FEATURE_LUN_SHARING) && + (adapter->ccw_device->id.dev_model != ZFCP_DEVICE_MODEL_PRIV)) { + exclusive = (bottom->lun_access_info & + FSF_UNIT_ACCESS_EXCLUSIVE); + readwrite = (bottom->lun_access_info & + FSF_UNIT_ACCESS_OUTBOUND_TRANSFER); + if (!exclusive) atomic_set_mask(ZFCP_STATUS_UNIT_SHARED, &unit->status); @@ -4164,7 +4262,7 @@ zfcp_fsf_control_file(struct zfcp_adapter *adapter, int direction; int retval = 0; - if (!(adapter->supported_features & FSF_FEATURE_CFDC)) { + if (!(adapter->adapter_features & FSF_FEATURE_CFDC)) { ZFCP_LOG_INFO("cfdc not supported (adapter %s)\n", zfcp_get_busid_by_adapter(adapter)); retval = -EOPNOTSUPP; diff --git a/drivers/s390/scsi/zfcp_fsf.h b/drivers/s390/scsi/zfcp_fsf.h index 57ce0706007..48719f05595 100644 --- a/drivers/s390/scsi/zfcp_fsf.h +++ b/drivers/s390/scsi/zfcp_fsf.h @@ -140,9 +140,18 @@ #define FSF_SQ_CFDC_SUBTABLE_LUN 0x0004 /* FSF status qualifier (most significant 4 bytes), local link down */ -#define FSF_PSQ_LINK_NOLIGHT 0x00000004 -#define FSF_PSQ_LINK_WRAPPLUG 0x00000008 -#define FSF_PSQ_LINK_NOFCP 0x00000010 +#define FSF_PSQ_LINK_NO_LIGHT 0x00000004 +#define FSF_PSQ_LINK_WRAP_PLUG 0x00000008 +#define FSF_PSQ_LINK_NO_FCP 0x00000010 +#define FSF_PSQ_LINK_FIRMWARE_UPDATE 0x00000020 +#define FSF_PSQ_LINK_INVALID_WWPN 0x00000100 +#define FSF_PSQ_LINK_NO_NPIV_SUPPORT 0x00000200 +#define FSF_PSQ_LINK_NO_FCP_RESOURCES 0x00000400 +#define FSF_PSQ_LINK_NO_FABRIC_RESOURCES 0x00000800 +#define FSF_PSQ_LINK_FABRIC_LOGIN_UNABLE 0x00001000 +#define FSF_PSQ_LINK_WWPN_ASSIGNMENT_CORRUPTED 0x00002000 +#define FSF_PSQ_LINK_MODE_TABLE_CURRUPTED 0x00004000 +#define FSF_PSQ_LINK_NO_WWPN_ASSIGNMENT 0x00008000 /* payload size in status read buffer */ #define FSF_STATUS_READ_PAYLOAD_SIZE 4032 @@ -155,15 +164,21 @@ #define FSF_STATUS_READ_INCOMING_ELS 0x00000002 #define FSF_STATUS_READ_SENSE_DATA_AVAIL 0x00000003 #define FSF_STATUS_READ_BIT_ERROR_THRESHOLD 0x00000004 -#define FSF_STATUS_READ_LINK_DOWN 0x00000005 /* FIXME: really? */ +#define FSF_STATUS_READ_LINK_DOWN 0x00000005 #define FSF_STATUS_READ_LINK_UP 0x00000006 #define FSF_STATUS_READ_CFDC_UPDATED 0x0000000A #define FSF_STATUS_READ_CFDC_HARDENED 0x0000000B +#define FSF_STATUS_READ_FEATURE_UPDATE_ALERT 0x0000000C /* status subtypes in status read buffer */ #define FSF_STATUS_READ_SUB_CLOSE_PHYS_PORT 0x00000001 #define FSF_STATUS_READ_SUB_ERROR_PORT 0x00000002 +/* status subtypes for link down */ +#define FSF_STATUS_READ_SUB_NO_PHYSICAL_LINK 0x00000000 +#define FSF_STATUS_READ_SUB_FDISC_FAILED 0x00000001 +#define FSF_STATUS_READ_SUB_FIRMWARE_UPDATE 0x00000002 + /* status subtypes for CFDC */ #define FSF_STATUS_READ_SUB_CFDC_HARDENED_ON_SE 0x00000002 #define FSF_STATUS_READ_SUB_CFDC_HARDENED_ON_SE2 0x0000000F @@ -194,11 +209,15 @@ #define FSF_QTCB_LOG_SIZE 1024 /* channel features */ -#define FSF_FEATURE_QTCB_SUPPRESSION 0x00000001 #define FSF_FEATURE_CFDC 0x00000002 #define FSF_FEATURE_LUN_SHARING 0x00000004 #define FSF_FEATURE_HBAAPI_MANAGEMENT 0x00000010 #define FSF_FEATURE_ELS_CT_CHAINED_SBALS 0x00000020 +#define FSF_FEATURE_UPDATE_ALERT 0x00000100 + +/* host connection features */ +#define FSF_FEATURE_NPIV_MODE 0x00000001 +#define FSF_FEATURE_VM_ASSIGNED_WWPN 0x00000002 /* option */ #define FSF_OPEN_LUN_SUPPRESS_BOXING 0x00000001 @@ -306,16 +325,23 @@ struct fsf_qual_sequence_error { u32 res1[3]; } __attribute__ ((packed)); -struct fsf_qual_locallink_error { - u32 code; - u32 res1[3]; +struct fsf_link_down_info { + u32 error_code; + u32 res1; + u8 res2[2]; + u8 primary_status; + u8 ioerr_code; + u8 action_code; + u8 reason_code; + u8 explanation_code; + u8 vendor_specific_code; } __attribute__ ((packed)); union fsf_prot_status_qual { u64 doubleword[FSF_PROT_STATUS_QUAL_SIZE / sizeof(u64)]; struct fsf_qual_version_error version_error; struct fsf_qual_sequence_error sequence_error; - struct fsf_qual_locallink_error locallink_error; + struct fsf_link_down_info link_down_info; } __attribute__ ((packed)); struct fsf_qtcb_prefix { @@ -335,6 +361,7 @@ union fsf_status_qual { u32 word[FSF_STATUS_QUALIFIER_SIZE / sizeof (u32)]; u64 doubleword[FSF_STATUS_QUALIFIER_SIZE / sizeof(u64)]; struct fsf_queue_designator fsf_queue_designator; + struct fsf_link_down_info link_down_info; } __attribute__ ((packed)); struct fsf_qtcb_header { @@ -409,8 +436,8 @@ struct fsf_qtcb_bottom_config { u32 low_qtcb_version; u32 max_qtcb_size; u32 max_data_transfer_size; - u32 supported_features; - u8 res1[4]; + u32 adapter_features; + u32 connection_features; u32 fc_topology; u32 fc_link_speed; u32 adapter_type; @@ -428,7 +455,7 @@ struct fsf_qtcb_bottom_config { } __attribute__ ((packed)); struct fsf_qtcb_bottom_port { - u8 res1[8]; + u64 wwpn; u32 fc_port_id; u32 port_type; u32 port_state; diff --git a/drivers/s390/scsi/zfcp_sysfs_adapter.c b/drivers/s390/scsi/zfcp_sysfs_adapter.c index e7345a74800..c406ea5c5e9 100644 --- a/drivers/s390/scsi/zfcp_sysfs_adapter.c +++ b/drivers/s390/scsi/zfcp_sysfs_adapter.c @@ -68,6 +68,8 @@ ZFCP_DEFINE_ADAPTER_ATTR(s_id, "0x%06x\n", adapter->s_id); ZFCP_DEFINE_ADAPTER_ATTR(peer_wwnn, "0x%016llx\n", adapter->peer_wwnn); ZFCP_DEFINE_ADAPTER_ATTR(peer_wwpn, "0x%016llx\n", adapter->peer_wwpn); ZFCP_DEFINE_ADAPTER_ATTR(peer_d_id, "0x%06x\n", adapter->peer_d_id); +ZFCP_DEFINE_ADAPTER_ATTR(physical_wwpn, "0x%016llx\n", adapter->physical_wwpn); +ZFCP_DEFINE_ADAPTER_ATTR(physical_s_id, "0x%06x\n", adapter->physical_s_id); ZFCP_DEFINE_ADAPTER_ATTR(card_version, "0x%04x\n", adapter->hydra_version); ZFCP_DEFINE_ADAPTER_ATTR(lic_version, "0x%08x\n", adapter->fsf_lic_version); ZFCP_DEFINE_ADAPTER_ATTR(fc_link_speed, "%d Gb/s\n", adapter->fc_link_speed); @@ -261,6 +263,8 @@ static struct attribute *zfcp_adapter_attrs[] = { &dev_attr_peer_wwnn.attr, &dev_attr_peer_wwpn.attr, &dev_attr_peer_d_id.attr, + &dev_attr_physical_wwpn.attr, + &dev_attr_physical_s_id.attr, &dev_attr_card_version.attr, &dev_attr_lic_version.attr, &dev_attr_fc_link_speed.attr, From 13e1e1f08c1c098c7574c1fa72bd8c67792dc89b Mon Sep 17 00:00:00 2001 From: Andreas Herrmann Date: Mon, 19 Sep 2005 16:56:17 +0200 Subject: [PATCH 042/204] [SCSI] zfcp: add additional fc_host attributes this patch adds some fc host attributes and removes its equivalents from the zfcp_adapter structure and zfcp specific sysfs subtree. Furthermore it removes superfluous calls to fc_remort_port_delete when an adapter is set offline because rports will be removed by fc_remove_host anyway. Signed-off-by: Andreas Herrmann Signed-off-by: James Bottomley --- drivers/s390/scsi/zfcp_ccw.c | 10 ---- drivers/s390/scsi/zfcp_dbf.c | 32 ++++++------- drivers/s390/scsi/zfcp_def.h | 16 ++----- drivers/s390/scsi/zfcp_erp.c | 16 +++---- drivers/s390/scsi/zfcp_fsf.c | 65 ++++++++++++++------------ drivers/s390/scsi/zfcp_scsi.c | 21 ++++----- drivers/s390/scsi/zfcp_sysfs_adapter.c | 10 ---- 7 files changed, 72 insertions(+), 98 deletions(-) diff --git a/drivers/s390/scsi/zfcp_ccw.c b/drivers/s390/scsi/zfcp_ccw.c index b30abab77da..0fc46381fc2 100644 --- a/drivers/s390/scsi/zfcp_ccw.c +++ b/drivers/s390/scsi/zfcp_ccw.c @@ -202,19 +202,9 @@ static int zfcp_ccw_set_offline(struct ccw_device *ccw_device) { struct zfcp_adapter *adapter; - struct zfcp_port *port; - struct fc_rport *rport; down(&zfcp_data.config_sema); adapter = dev_get_drvdata(&ccw_device->dev); - /* might be racy, but we cannot take config_lock due to the fact that - fc_remote_port_delete might sleep */ - list_for_each_entry(port, &adapter->port_list_head, list) - if (port->rport) { - rport = port->rport; - port->rport = NULL; - fc_remote_port_delete(rport); - } zfcp_erp_adapter_shutdown(adapter, 0); zfcp_erp_wait(adapter); zfcp_adapter_scsi_unregister(adapter); diff --git a/drivers/s390/scsi/zfcp_dbf.c b/drivers/s390/scsi/zfcp_dbf.c index 64d9b90373f..826fb3b0060 100644 --- a/drivers/s390/scsi/zfcp_dbf.c +++ b/drivers/s390/scsi/zfcp_dbf.c @@ -486,8 +486,7 @@ struct debug_view zfcp_hba_dbf_view = { inline void _zfcp_san_dbf_event_common_ct(const char *tag, struct zfcp_fsf_req *fsf_req, - fc_id_t s_id, fc_id_t d_id, - void *buffer, int buflen) + u32 s_id, u32 d_id, void *buffer, int buflen) { struct zfcp_send_ct *send_ct = (struct zfcp_send_ct *)fsf_req->data; struct zfcp_port *port = send_ct->port; @@ -532,9 +531,9 @@ inline void zfcp_san_dbf_event_ct_request(struct zfcp_fsf_req *fsf_req) struct zfcp_port *port = ct->port; struct zfcp_adapter *adapter = port->adapter; - _zfcp_san_dbf_event_common_ct("octc", - fsf_req, adapter->s_id, port->d_id, - zfcp_sg_to_address(ct->req), + _zfcp_san_dbf_event_common_ct("octc", fsf_req, + fc_host_port_id(adapter->scsi_host), + port->d_id, zfcp_sg_to_address(ct->req), ct->req->length); } @@ -544,17 +543,16 @@ inline void zfcp_san_dbf_event_ct_response(struct zfcp_fsf_req *fsf_req) struct zfcp_port *port = ct->port; struct zfcp_adapter *adapter = port->adapter; - _zfcp_san_dbf_event_common_ct("rctc", - fsf_req, port->d_id, adapter->s_id, + _zfcp_san_dbf_event_common_ct("rctc", fsf_req, port->d_id, + fc_host_port_id(adapter->scsi_host), zfcp_sg_to_address(ct->resp), ct->resp->length); } static inline void _zfcp_san_dbf_event_common_els(const char *tag, int level, - struct zfcp_fsf_req *fsf_req, - fc_id_t s_id, fc_id_t d_id, u8 ls_code, - void *buffer, int buflen) + struct zfcp_fsf_req *fsf_req, u32 s_id, + u32 d_id, u8 ls_code, void *buffer, int buflen) { struct zfcp_adapter *adapter = fsf_req->adapter; struct zfcp_san_dbf_record *rec = &adapter->san_dbf_buf; @@ -597,8 +595,9 @@ inline void zfcp_san_dbf_event_els_request(struct zfcp_fsf_req *fsf_req) { struct zfcp_send_els *els = (struct zfcp_send_els *)fsf_req->data; - _zfcp_san_dbf_event_common_els("oels", 2, - fsf_req, els->adapter->s_id, els->d_id, + _zfcp_san_dbf_event_common_els("oels", 2, fsf_req, + fc_host_port_id(els->adapter->scsi_host), + els->d_id, *(u8 *) zfcp_sg_to_address(els->req), zfcp_sg_to_address(els->req), els->req->length); @@ -608,8 +607,8 @@ inline void zfcp_san_dbf_event_els_response(struct zfcp_fsf_req *fsf_req) { struct zfcp_send_els *els = (struct zfcp_send_els *)fsf_req->data; - _zfcp_san_dbf_event_common_els("rels", 2, - fsf_req, els->d_id, els->adapter->s_id, + _zfcp_san_dbf_event_common_els("rels", 2, fsf_req, els->d_id, + fc_host_port_id(els->adapter->scsi_host), *(u8 *) zfcp_sg_to_address(els->req), zfcp_sg_to_address(els->resp), els->resp->length); @@ -623,9 +622,8 @@ inline void zfcp_san_dbf_event_incoming_els(struct zfcp_fsf_req *fsf_req) int length = (int)status_buffer->length - (int)((void *)&status_buffer->payload - (void *)status_buffer); - _zfcp_san_dbf_event_common_els("iels", 1, - fsf_req, status_buffer->d_id, - adapter->s_id, + _zfcp_san_dbf_event_common_els("iels", 1, fsf_req, status_buffer->d_id, + fc_host_port_id(adapter->scsi_host), *(u8 *) status_buffer->payload, (void *)status_buffer->payload, length); } diff --git a/drivers/s390/scsi/zfcp_def.h b/drivers/s390/scsi/zfcp_def.h index 4b6af8e07e8..d81b737d68c 100644 --- a/drivers/s390/scsi/zfcp_def.h +++ b/drivers/s390/scsi/zfcp_def.h @@ -165,7 +165,6 @@ typedef u32 scsi_lun_t; /*************** FIBRE CHANNEL PROTOCOL SPECIFIC DEFINES ********************/ typedef unsigned long long wwn_t; -typedef unsigned int fc_id_t; typedef unsigned long long fcp_lun_t; /* data length field may be at variable position in FCP-2 FCP_CMND IU */ typedef unsigned int fcp_dl_t; @@ -806,7 +805,7 @@ struct ct_iu_gid_pn_req { /* FS_ACC IU and data unit for GID_PN nameserver request */ struct ct_iu_gid_pn_resp { struct ct_hdr header; - fc_id_t d_id; + u32 d_id; } __attribute__ ((packed)); typedef void (*zfcp_send_ct_handler_t)(unsigned long); @@ -872,7 +871,7 @@ typedef void (*zfcp_send_els_handler_t)(unsigned long); struct zfcp_send_els { struct zfcp_adapter *adapter; struct zfcp_port *port; - fc_id_t d_id; + u32 d_id; struct scatterlist *req; struct scatterlist *resp; unsigned int req_count; @@ -915,24 +914,19 @@ struct zfcp_adapter { atomic_t refcount; /* reference count */ wait_queue_head_t remove_wq; /* can be used to wait for refcount drop to zero */ - wwn_t wwnn; /* WWNN */ - wwn_t wwpn; /* WWPN */ - fc_id_t s_id; /* N_Port ID */ wwn_t peer_wwnn; /* P2P peer WWNN */ wwn_t peer_wwpn; /* P2P peer WWPN */ - fc_id_t peer_d_id; /* P2P peer D_ID */ + u32 peer_d_id; /* P2P peer D_ID */ wwn_t physical_wwpn; /* WWPN of physical port */ - fc_id_t physical_s_id; /* local FC port ID */ + u32 physical_s_id; /* local FC port ID */ struct ccw_device *ccw_device; /* S/390 ccw device */ u8 fc_service_class; u32 fc_topology; /* FC topology */ - u32 fc_link_speed; /* FC interface speed */ u32 hydra_version; /* Hydra version */ u32 fsf_lic_version; u32 adapter_features; /* FCP channel features */ u32 connection_features; /* host connection features */ u32 hardware_version; /* of FCP channel */ - u8 serial_number[32]; /* of hardware */ struct Scsi_Host *scsi_host; /* Pointer to mid-layer */ unsigned short scsi_host_no; /* Assigned host number */ unsigned char name[9]; @@ -1006,7 +1000,7 @@ struct zfcp_port { atomic_t status; /* status of this remote port */ wwn_t wwnn; /* WWNN if known */ wwn_t wwpn; /* WWPN */ - fc_id_t d_id; /* D_ID */ + u32 d_id; /* D_ID */ u32 handle; /* handle assigned by FSF */ struct zfcp_erp_action erp_action; /* pending error recovery */ atomic_t erp_counter; diff --git a/drivers/s390/scsi/zfcp_erp.c b/drivers/s390/scsi/zfcp_erp.c index c4a6799aed4..023f4e558ae 100644 --- a/drivers/s390/scsi/zfcp_erp.c +++ b/drivers/s390/scsi/zfcp_erp.c @@ -346,13 +346,13 @@ zfcp_erp_adisc(struct zfcp_port *port) /* acc. to FC-FS, hard_nport_id in ADISC should not be set for ports without FC-AL-2 capability, so we don't set it */ - adisc->wwpn = adapter->wwpn; - adisc->wwnn = adapter->wwnn; - adisc->nport_id = adapter->s_id; + adisc->wwpn = fc_host_port_name(adapter->scsi_host); + adisc->wwnn = fc_host_node_name(adapter->scsi_host); + adisc->nport_id = fc_host_port_id(adapter->scsi_host); ZFCP_LOG_INFO("ADISC request from s_id 0x%08x to d_id 0x%08x " "(wwpn=0x%016Lx, wwnn=0x%016Lx, " "hard_nport_id=0x%08x, nport_id=0x%08x)\n", - adapter->s_id, send_els->d_id, (wwn_t) adisc->wwpn, + adisc->nport_id, send_els->d_id, (wwn_t) adisc->wwpn, (wwn_t) adisc->wwnn, adisc->hard_nport_id, adisc->nport_id); @@ -405,7 +405,7 @@ zfcp_erp_adisc_handler(unsigned long data) struct zfcp_send_els *send_els; struct zfcp_port *port; struct zfcp_adapter *adapter; - fc_id_t d_id; + u32 d_id; struct zfcp_ls_adisc_acc *adisc; send_els = (struct zfcp_send_els *) data; @@ -436,9 +436,9 @@ zfcp_erp_adisc_handler(unsigned long data) ZFCP_LOG_INFO("ADISC response from d_id 0x%08x to s_id " "0x%08x (wwpn=0x%016Lx, wwnn=0x%016Lx, " "hard_nport_id=0x%08x, nport_id=0x%08x)\n", - d_id, adapter->s_id, (wwn_t) adisc->wwpn, - (wwn_t) adisc->wwnn, adisc->hard_nport_id, - adisc->nport_id); + d_id, fc_host_port_id(adapter->scsi_host), + (wwn_t) adisc->wwpn, (wwn_t) adisc->wwnn, + adisc->hard_nport_id, adisc->nport_id); /* set wwnn for port */ if (port->wwnn == 0) diff --git a/drivers/s390/scsi/zfcp_fsf.c b/drivers/s390/scsi/zfcp_fsf.c index f994f453e2d..3b0fc1163f5 100644 --- a/drivers/s390/scsi/zfcp_fsf.c +++ b/drivers/s390/scsi/zfcp_fsf.c @@ -1617,7 +1617,7 @@ zfcp_fsf_send_els(struct zfcp_send_els *els) { volatile struct qdio_buffer_element *sbale; struct zfcp_fsf_req *fsf_req; - fc_id_t d_id; + u32 d_id; struct zfcp_adapter *adapter; unsigned long lock_flags; int bytes; @@ -1740,7 +1740,7 @@ static int zfcp_fsf_send_els_handler(struct zfcp_fsf_req *fsf_req) { struct zfcp_adapter *adapter; struct zfcp_port *port; - fc_id_t d_id; + u32 d_id; struct fsf_qtcb_header *header; struct fsf_qtcb_bottom_support *bottom; struct zfcp_send_els *send_els; @@ -1978,6 +1978,7 @@ zfcp_fsf_exchange_config_evaluate(struct zfcp_fsf_req *fsf_req, int xchg_ok) { struct fsf_qtcb_bottom_config *bottom; struct zfcp_adapter *adapter = fsf_req->adapter; + struct Scsi_Host *shost = adapter->scsi_host; bottom = &fsf_req->qtcb->bottom.config; ZFCP_LOG_DEBUG("low/high QTCB version 0x%x/0x%x of FSF\n", @@ -1990,22 +1991,23 @@ zfcp_fsf_exchange_config_evaluate(struct zfcp_fsf_req *fsf_req, int xchg_ok) adapter->peer_d_id = 0; if (xchg_ok) { - adapter->wwnn = bottom->nport_serv_param.wwnn; - adapter->wwpn = bottom->nport_serv_param.wwpn; - adapter->s_id = bottom->s_id & ZFCP_DID_MASK; + fc_host_node_name(shost) = bottom->nport_serv_param.wwnn; + fc_host_port_name(shost) = bottom->nport_serv_param.wwpn; + fc_host_port_id(shost) = bottom->s_id & ZFCP_DID_MASK; + fc_host_speed(shost) = bottom->fc_link_speed; + fc_host_supported_classes(shost) = FC_COS_CLASS2 | FC_COS_CLASS3; adapter->fc_topology = bottom->fc_topology; - adapter->fc_link_speed = bottom->fc_link_speed; adapter->hydra_version = bottom->adapter_type; if (adapter->physical_wwpn == 0) - adapter->physical_wwpn = adapter->wwpn; + adapter->physical_wwpn = fc_host_port_name(shost); if (adapter->physical_s_id == 0) - adapter->physical_s_id = adapter->s_id; + adapter->physical_s_id = fc_host_port_id(shost); } else { - adapter->wwnn = 0; - adapter->wwpn = 0; - adapter->s_id = 0; + fc_host_node_name(shost) = 0; + fc_host_port_name(shost) = 0; + fc_host_port_id(shost) = 0; + fc_host_speed(shost) = FC_PORTSPEED_UNKNOWN; adapter->fc_topology = 0; - adapter->fc_link_speed = 0; adapter->hydra_version = 0; } @@ -2017,24 +2019,26 @@ zfcp_fsf_exchange_config_evaluate(struct zfcp_fsf_req *fsf_req, int xchg_ok) if (adapter->adapter_features & FSF_FEATURE_HBAAPI_MANAGEMENT) { adapter->hardware_version = bottom->hardware_version; - memcpy(adapter->serial_number, bottom->serial_number, 17); - EBCASC(adapter->serial_number, sizeof(adapter->serial_number)); + memcpy(fc_host_serial_number(shost), bottom->serial_number, + min(FC_SERIAL_NUMBER_SIZE, 17)); + EBCASC(fc_host_serial_number(shost), + min(FC_SERIAL_NUMBER_SIZE, 17)); } ZFCP_LOG_NORMAL("The adapter %s reported the following characteristics:\n" - "WWNN 0x%016Lx, " - "WWPN 0x%016Lx, " - "S_ID 0x%08x,\n" - "adapter version 0x%x, " - "LIC version 0x%x, " - "FC link speed %d Gb/s\n", - zfcp_get_busid_by_adapter(adapter), - adapter->wwnn, - adapter->wwpn, - (unsigned int) adapter->s_id, - adapter->hydra_version, - adapter->fsf_lic_version, - adapter->fc_link_speed); + "WWNN 0x%016Lx, " + "WWPN 0x%016Lx, " + "S_ID 0x%08x,\n" + "adapter version 0x%x, " + "LIC version 0x%x, " + "FC link speed %d Gb/s\n", + zfcp_get_busid_by_adapter(adapter), + (wwn_t) fc_host_node_name(shost), + (wwn_t) fc_host_port_name(shost), + fc_host_port_id(shost), + adapter->hydra_version, + adapter->fsf_lic_version, + fc_host_speed(shost)); if (ZFCP_QTCB_VERSION < bottom->low_qtcb_version) { ZFCP_LOG_NORMAL("error: the adapter %s " "only supports newer control block " @@ -2055,7 +2059,6 @@ zfcp_fsf_exchange_config_evaluate(struct zfcp_fsf_req *fsf_req, int xchg_ok) zfcp_erp_adapter_shutdown(adapter, 0); return -EIO; } - zfcp_set_fc_host_attrs(adapter); return 0; } @@ -2259,6 +2262,7 @@ static void zfcp_fsf_exchange_port_data_handler(struct zfcp_fsf_req *fsf_req) { struct zfcp_adapter *adapter = fsf_req->adapter; + struct Scsi_Host *shost = adapter->scsi_host; struct fsf_qtcb *qtcb = fsf_req->qtcb; struct fsf_qtcb_bottom_port *bottom, *data; @@ -2277,9 +2281,10 @@ zfcp_fsf_exchange_port_data_handler(struct zfcp_fsf_req *fsf_req) adapter->physical_wwpn = bottom->wwpn; adapter->physical_s_id = bottom->fc_port_id; } else { - adapter->physical_wwpn = adapter->wwpn; - adapter->physical_s_id = adapter->s_id; + adapter->physical_wwpn = fc_host_port_name(shost); + adapter->physical_s_id = fc_host_port_id(shost); } + fc_host_maxframe_size(shost) = bottom->maximum_frame_size; break; case FSF_EXCHANGE_CONFIG_DATA_INCOMPLETE: diff --git a/drivers/s390/scsi/zfcp_scsi.c b/drivers/s390/scsi/zfcp_scsi.c index c1403a23174..3dcd1bfba3b 100644 --- a/drivers/s390/scsi/zfcp_scsi.c +++ b/drivers/s390/scsi/zfcp_scsi.c @@ -698,10 +698,16 @@ void zfcp_adapter_scsi_unregister(struct zfcp_adapter *adapter) { struct Scsi_Host *shost; + struct zfcp_port *port; shost = adapter->scsi_host; if (!shost) return; + read_lock_irq(&zfcp_data.config_lock); + list_for_each_entry(port, &adapter->port_list_head, list) + if (port->rport) + port->rport = NULL; + read_unlock_irq(&zfcp_data.config_lock); fc_remove_host(shost); scsi_remove_host(shost); scsi_host_put(shost); @@ -776,18 +782,6 @@ zfcp_get_node_name(struct scsi_target *starget) read_unlock_irqrestore(&zfcp_data.config_lock, flags); } -void -zfcp_set_fc_host_attrs(struct zfcp_adapter *adapter) -{ - struct Scsi_Host *shost = adapter->scsi_host; - - fc_host_node_name(shost) = adapter->wwnn; - fc_host_port_name(shost) = adapter->wwpn; - strncpy(fc_host_serial_number(shost), adapter->serial_number, - min(FC_SERIAL_NUMBER_SIZE, 32)); - fc_host_supported_classes(shost) = FC_COS_CLASS2 | FC_COS_CLASS3; -} - struct fc_function_template zfcp_transport_functions = { .get_starget_port_id = zfcp_get_port_id, .get_starget_port_name = zfcp_get_port_name, @@ -799,7 +793,10 @@ struct fc_function_template zfcp_transport_functions = { .show_host_node_name = 1, .show_host_port_name = 1, .show_host_supported_classes = 1, + .show_host_maxframe_size = 1, .show_host_serial_number = 1, + .show_host_speed = 1, + .show_host_port_id = 1, }; /** diff --git a/drivers/s390/scsi/zfcp_sysfs_adapter.c b/drivers/s390/scsi/zfcp_sysfs_adapter.c index c406ea5c5e9..0cd435280e7 100644 --- a/drivers/s390/scsi/zfcp_sysfs_adapter.c +++ b/drivers/s390/scsi/zfcp_sysfs_adapter.c @@ -62,9 +62,6 @@ static ssize_t zfcp_sysfs_adapter_##_name##_show(struct device *dev, struct devi static DEVICE_ATTR(_name, S_IRUGO, zfcp_sysfs_adapter_##_name##_show, NULL); ZFCP_DEFINE_ADAPTER_ATTR(status, "0x%08x\n", atomic_read(&adapter->status)); -ZFCP_DEFINE_ADAPTER_ATTR(wwnn, "0x%016llx\n", adapter->wwnn); -ZFCP_DEFINE_ADAPTER_ATTR(wwpn, "0x%016llx\n", adapter->wwpn); -ZFCP_DEFINE_ADAPTER_ATTR(s_id, "0x%06x\n", adapter->s_id); ZFCP_DEFINE_ADAPTER_ATTR(peer_wwnn, "0x%016llx\n", adapter->peer_wwnn); ZFCP_DEFINE_ADAPTER_ATTR(peer_wwpn, "0x%016llx\n", adapter->peer_wwpn); ZFCP_DEFINE_ADAPTER_ATTR(peer_d_id, "0x%06x\n", adapter->peer_d_id); @@ -72,13 +69,11 @@ ZFCP_DEFINE_ADAPTER_ATTR(physical_wwpn, "0x%016llx\n", adapter->physical_wwpn); ZFCP_DEFINE_ADAPTER_ATTR(physical_s_id, "0x%06x\n", adapter->physical_s_id); ZFCP_DEFINE_ADAPTER_ATTR(card_version, "0x%04x\n", adapter->hydra_version); ZFCP_DEFINE_ADAPTER_ATTR(lic_version, "0x%08x\n", adapter->fsf_lic_version); -ZFCP_DEFINE_ADAPTER_ATTR(fc_link_speed, "%d Gb/s\n", adapter->fc_link_speed); ZFCP_DEFINE_ADAPTER_ATTR(fc_service_class, "%d\n", adapter->fc_service_class); ZFCP_DEFINE_ADAPTER_ATTR(fc_topology, "%s\n", fc_topologies[adapter->fc_topology]); ZFCP_DEFINE_ADAPTER_ATTR(hardware_version, "0x%08x\n", adapter->hardware_version); -ZFCP_DEFINE_ADAPTER_ATTR(serial_number, "%17s\n", adapter->serial_number); ZFCP_DEFINE_ADAPTER_ATTR(scsi_host_no, "0x%x\n", adapter->scsi_host_no); ZFCP_DEFINE_ADAPTER_ATTR(in_recovery, "%d\n", atomic_test_mask (ZFCP_STATUS_COMMON_ERP_INUSE, &adapter->status)); @@ -257,9 +252,6 @@ static struct attribute *zfcp_adapter_attrs[] = { &dev_attr_in_recovery.attr, &dev_attr_port_remove.attr, &dev_attr_port_add.attr, - &dev_attr_wwnn.attr, - &dev_attr_wwpn.attr, - &dev_attr_s_id.attr, &dev_attr_peer_wwnn.attr, &dev_attr_peer_wwpn.attr, &dev_attr_peer_d_id.attr, @@ -267,13 +259,11 @@ static struct attribute *zfcp_adapter_attrs[] = { &dev_attr_physical_s_id.attr, &dev_attr_card_version.attr, &dev_attr_lic_version.attr, - &dev_attr_fc_link_speed.attr, &dev_attr_fc_service_class.attr, &dev_attr_fc_topology.attr, &dev_attr_scsi_host_no.attr, &dev_attr_status.attr, &dev_attr_hardware_version.attr, - &dev_attr_serial_number.attr, NULL }; From eff4c654b1a4a5e5493fbdc3affa6dd48765c085 Mon Sep 17 00:00:00 2001 From: Hal Rosenstock Date: Mon, 19 Sep 2005 13:51:01 -0700 Subject: [PATCH 043/204] [PATCH] IB: Fix data length for RMPP SA sends We need to subtract off the header length from our payload length when sending multi-packet SA messages. Signed-off-by: Hal Rosenstock Signed-off-by: Roland Dreier --- drivers/infiniband/core/user_mad.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c index 7c2f03057dd..a64d6b4dcc1 100644 --- a/drivers/infiniband/core/user_mad.c +++ b/drivers/infiniband/core/user_mad.c @@ -334,10 +334,11 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf, ret = -EINVAL; goto err_ah; } - /* Validate that management class can support RMPP */ + + /* Validate that the management class can support RMPP */ if (rmpp_mad->mad_hdr.mgmt_class == IB_MGMT_CLASS_SUBN_ADM) { hdr_len = offsetof(struct ib_sa_mad, data); - data_len = length; + data_len = length - hdr_len; } else if ((rmpp_mad->mad_hdr.mgmt_class >= IB_MGMT_CLASS_VENDOR_RANGE2_START) && (rmpp_mad->mad_hdr.mgmt_class <= IB_MGMT_CLASS_VENDOR_RANGE2_END)) { hdr_len = offsetof(struct ib_vendor_mad, data); From 1d25cd6cc2528e4af12ab18e84fe95ed78f3f21a Mon Sep 17 00:00:00 2001 From: Robert Olsson Date: Mon, 19 Sep 2005 15:29:52 -0700 Subject: [PATCH 044/204] [IPV4]: fib_trie tnode stats refinements * Prints the route tnode and set the stats level deepth as before. Signed-off-by: Robert Olsson Signed-off-by: David S. Miller --- net/ipv4/fib_trie.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c index 1b63b482416..3dd70787081 100644 --- a/net/ipv4/fib_trie.c +++ b/net/ipv4/fib_trie.c @@ -2029,7 +2029,7 @@ static struct node *fib_trie_get_first(struct fib_trie_iter *iter, iter->tnode = (struct tnode *) n; iter->trie = t; iter->index = 0; - iter->depth = 0; + iter->depth = 1; return n; } return NULL; @@ -2274,11 +2274,12 @@ static int fib_trie_seq_show(struct seq_file *seq, void *v) seq_puts(seq, ":\n"); else seq_puts(seq, "
:\n"); - } else { - seq_indent(seq, iter->depth-1); - seq_printf(seq, " +-- %d.%d.%d.%d/%d\n", - NIPQUAD(prf), tn->pos); - } + } + seq_indent(seq, iter->depth-1); + seq_printf(seq, " +-- %d.%d.%d.%d/%d %d %d %d\n", + NIPQUAD(prf), tn->pos, tn->bits, tn->full_children, + tn->empty_children); + } else { struct leaf *l = (struct leaf *) n; int i; From 772cb712b1373d335ef2874ea357ec681edc754b Mon Sep 17 00:00:00 2001 From: Robert Olsson Date: Mon, 19 Sep 2005 15:31:18 -0700 Subject: [PATCH 045/204] [IPV4]: fib_trie RCU refinements * This patch is from Paul McKenney's RCU reviewing. Signed-off-by: Robert Olsson Signed-off-by: David S. Miller --- net/ipv4/fib_trie.c | 21 ++++++++++----------- 1 file changed, 10 insertions(+), 11 deletions(-) diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c index 3dd70787081..90ae70870a1 100644 --- a/net/ipv4/fib_trie.c +++ b/net/ipv4/fib_trie.c @@ -43,7 +43,7 @@ * 2 of the License, or (at your option) any later version. */ -#define VERSION "0.403" +#define VERSION "0.404" #include #include @@ -224,7 +224,7 @@ static inline int tkey_mismatch(t_key a, int offset, t_key b) Consider a node 'n' and its parent 'tp'. If n is a leaf, every bit in its key is significant. Its presence is - necessitaded by path compression, since during a tree traversal (when + necessitated by path compression, since during a tree traversal (when searching for a leaf - unless we are doing an insertion) we will completely ignore all skipped bits we encounter. Thus we need to verify, at the end of a potentially successful search, that we have indeed been walking the @@ -836,11 +836,12 @@ static void trie_init(struct trie *t) #endif } -/* readside most use rcu_read_lock currently dump routines +/* readside must use rcu_read_lock currently dump routines via get_fa_head and dump */ -static struct leaf_info *find_leaf_info(struct hlist_head *head, int plen) +static struct leaf_info *find_leaf_info(struct leaf *l, int plen) { + struct hlist_head *head = &l->list; struct hlist_node *node; struct leaf_info *li; @@ -853,7 +854,7 @@ static struct leaf_info *find_leaf_info(struct hlist_head *head, int plen) static inline struct list_head * get_fa_head(struct leaf *l, int plen) { - struct leaf_info *li = find_leaf_info(&l->list, plen); + struct leaf_info *li = find_leaf_info(l, plen); if (!li) return NULL; @@ -1248,7 +1249,7 @@ err: } -/* should be clalled with rcu_read_lock */ +/* should be called with rcu_read_lock */ static inline int check_leaf(struct trie *t, struct leaf *l, t_key key, int *plen, const struct flowi *flp, struct fib_result *res) @@ -1590,7 +1591,7 @@ fn_trie_delete(struct fib_table *tb, struct rtmsg *r, struct kern_rta *rta, rtmsg_fib(RTM_DELROUTE, htonl(key), fa, plen, tb->tb_id, nlhdr, req); l = fib_find_node(t, key); - li = find_leaf_info(&l->list, plen); + li = find_leaf_info(l, plen); list_del_rcu(&fa->fa_list); @@ -1714,7 +1715,6 @@ static int fn_trie_flush(struct fib_table *tb) t->revision++; - rcu_read_lock(); for (h = 0; (l = nextleaf(t, l)) != NULL; h++) { found += trie_flush_leaf(t, l); @@ -1722,7 +1722,6 @@ static int fn_trie_flush(struct fib_table *tb) trie_leaf_remove(t, ll->key); ll = l; } - rcu_read_unlock(); if (ll && hlist_empty(&ll->list)) trie_leaf_remove(t, ll->key); @@ -2288,7 +2287,7 @@ static int fib_trie_seq_show(struct seq_file *seq, void *v) seq_indent(seq, iter->depth); seq_printf(seq, " |-- %d.%d.%d.%d\n", NIPQUAD(val)); for (i = 32; i >= 0; i--) { - struct leaf_info *li = find_leaf_info(&l->list, i); + struct leaf_info *li = find_leaf_info(l, i); if (li) { struct fib_alias *fa; list_for_each_entry_rcu(fa, &li->falh, fa_list) { @@ -2384,7 +2383,7 @@ static int fib_route_seq_show(struct seq_file *seq, void *v) return 0; for (i=32; i>=0; i--) { - struct leaf_info *li = find_leaf_info(&l->list, i); + struct leaf_info *li = find_leaf_info(l, i); struct fib_alias *fa; u32 mask, prefix; From 926b50f92a30090da2c1a8675de954c2d9b09732 Mon Sep 17 00:00:00 2001 From: Harald Welte Date: Mon, 19 Sep 2005 15:33:08 -0700 Subject: [PATCH 046/204] [NETFILTER]: Add new PPTP conntrack and NAT helper This new "version 3" PPTP conntrack/nat helper is finally ready for mainline inclusion. Special thanks to lots of last-minute bugfixing by Patric McHardy. Signed-off-by: Harald Welte Signed-off-by: David S. Miller --- include/linux/netfilter_ipv4/ip_conntrack.h | 12 + .../linux/netfilter_ipv4/ip_conntrack_pptp.h | 332 ++++++++ .../netfilter_ipv4/ip_conntrack_proto_gre.h | 114 +++ .../linux/netfilter_ipv4/ip_conntrack_tuple.h | 7 + include/linux/netfilter_ipv4/ip_nat_pptp.h | 11 + net/ipv4/netfilter/Kconfig | 22 + net/ipv4/netfilter/Makefile | 5 + net/ipv4/netfilter/ip_conntrack_helper_pptp.c | 805 ++++++++++++++++++ net/ipv4/netfilter/ip_conntrack_proto_gre.c | 327 +++++++ net/ipv4/netfilter/ip_nat_helper_pptp.c | 401 +++++++++ net/ipv4/netfilter/ip_nat_proto_gre.c | 214 +++++ 11 files changed, 2250 insertions(+) create mode 100644 include/linux/netfilter_ipv4/ip_conntrack_pptp.h create mode 100644 include/linux/netfilter_ipv4/ip_conntrack_proto_gre.h create mode 100644 include/linux/netfilter_ipv4/ip_nat_pptp.h create mode 100644 net/ipv4/netfilter/ip_conntrack_helper_pptp.c create mode 100644 net/ipv4/netfilter/ip_conntrack_proto_gre.c create mode 100644 net/ipv4/netfilter/ip_nat_helper_pptp.c create mode 100644 net/ipv4/netfilter/ip_nat_proto_gre.c diff --git a/include/linux/netfilter_ipv4/ip_conntrack.h b/include/linux/netfilter_ipv4/ip_conntrack.h index 7e033e9271a..2df446c952e 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack.h +++ b/include/linux/netfilter_ipv4/ip_conntrack.h @@ -133,11 +133,13 @@ enum ip_conntrack_expect_events { #include #include +#include #include /* per conntrack: protocol private data */ union ip_conntrack_proto { /* insert conntrack proto private data here */ + struct ip_ct_gre gre; struct ip_ct_sctp sctp; struct ip_ct_tcp tcp; struct ip_ct_icmp icmp; @@ -148,6 +150,7 @@ union ip_conntrack_expect_proto { }; /* Add protocol helper include file here */ +#include #include #include #include @@ -155,12 +158,20 @@ union ip_conntrack_expect_proto { /* per conntrack: application helper private data */ union ip_conntrack_help { /* insert conntrack helper private data (master) here */ + struct ip_ct_pptp_master ct_pptp_info; struct ip_ct_ftp_master ct_ftp_info; struct ip_ct_irc_master ct_irc_info; }; #ifdef CONFIG_IP_NF_NAT_NEEDED #include +#include + +/* per conntrack: nat application helper private data */ +union ip_conntrack_nat_help { + /* insert nat helper private data here */ + struct ip_nat_pptp nat_pptp_info; +}; #endif #include @@ -223,6 +234,7 @@ struct ip_conntrack #ifdef CONFIG_IP_NF_NAT_NEEDED struct { struct ip_nat_info info; + union ip_conntrack_nat_help help; #if defined(CONFIG_IP_NF_TARGET_MASQUERADE) || \ defined(CONFIG_IP_NF_TARGET_MASQUERADE_MODULE) int masq_index; diff --git a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h new file mode 100644 index 00000000000..389e3851d52 --- /dev/null +++ b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h @@ -0,0 +1,332 @@ +/* PPTP constants and structs */ +#ifndef _CONNTRACK_PPTP_H +#define _CONNTRACK_PPTP_H + +/* state of the control session */ +enum pptp_ctrlsess_state { + PPTP_SESSION_NONE, /* no session present */ + PPTP_SESSION_ERROR, /* some session error */ + PPTP_SESSION_STOPREQ, /* stop_sess request seen */ + PPTP_SESSION_REQUESTED, /* start_sess request seen */ + PPTP_SESSION_CONFIRMED, /* session established */ +}; + +/* state of the call inside the control session */ +enum pptp_ctrlcall_state { + PPTP_CALL_NONE, + PPTP_CALL_ERROR, + PPTP_CALL_OUT_REQ, + PPTP_CALL_OUT_CONF, + PPTP_CALL_IN_REQ, + PPTP_CALL_IN_REP, + PPTP_CALL_IN_CONF, + PPTP_CALL_CLEAR_REQ, +}; + + +/* conntrack private data */ +struct ip_ct_pptp_master { + enum pptp_ctrlsess_state sstate; /* session state */ + + /* everything below is going to be per-expectation in newnat, + * since there could be more than one call within one session */ + enum pptp_ctrlcall_state cstate; /* call state */ + u_int16_t pac_call_id; /* call id of PAC, host byte order */ + u_int16_t pns_call_id; /* call id of PNS, host byte order */ + + /* in pre-2.6.11 this used to be per-expect. Now it is per-conntrack + * and therefore imposes a fixed limit on the number of maps */ + struct ip_ct_gre_keymap *keymap_orig, *keymap_reply; +}; + +/* conntrack_expect private member */ +struct ip_ct_pptp_expect { + enum pptp_ctrlcall_state cstate; /* call state */ + u_int16_t pac_call_id; /* call id of PAC */ + u_int16_t pns_call_id; /* call id of PNS */ +}; + + +#ifdef __KERNEL__ + +#define IP_CONNTR_PPTP PPTP_CONTROL_PORT + +#define PPTP_CONTROL_PORT 1723 + +#define PPTP_PACKET_CONTROL 1 +#define PPTP_PACKET_MGMT 2 + +#define PPTP_MAGIC_COOKIE 0x1a2b3c4d + +struct pptp_pkt_hdr { + __u16 packetLength; + __u16 packetType; + __u32 magicCookie; +}; + +/* PptpControlMessageType values */ +#define PPTP_START_SESSION_REQUEST 1 +#define PPTP_START_SESSION_REPLY 2 +#define PPTP_STOP_SESSION_REQUEST 3 +#define PPTP_STOP_SESSION_REPLY 4 +#define PPTP_ECHO_REQUEST 5 +#define PPTP_ECHO_REPLY 6 +#define PPTP_OUT_CALL_REQUEST 7 +#define PPTP_OUT_CALL_REPLY 8 +#define PPTP_IN_CALL_REQUEST 9 +#define PPTP_IN_CALL_REPLY 10 +#define PPTP_IN_CALL_CONNECT 11 +#define PPTP_CALL_CLEAR_REQUEST 12 +#define PPTP_CALL_DISCONNECT_NOTIFY 13 +#define PPTP_WAN_ERROR_NOTIFY 14 +#define PPTP_SET_LINK_INFO 15 + +#define PPTP_MSG_MAX 15 + +/* PptpGeneralError values */ +#define PPTP_ERROR_CODE_NONE 0 +#define PPTP_NOT_CONNECTED 1 +#define PPTP_BAD_FORMAT 2 +#define PPTP_BAD_VALUE 3 +#define PPTP_NO_RESOURCE 4 +#define PPTP_BAD_CALLID 5 +#define PPTP_REMOVE_DEVICE_ERROR 6 + +struct PptpControlHeader { + __u16 messageType; + __u16 reserved; +}; + +/* FramingCapability Bitmap Values */ +#define PPTP_FRAME_CAP_ASYNC 0x1 +#define PPTP_FRAME_CAP_SYNC 0x2 + +/* BearerCapability Bitmap Values */ +#define PPTP_BEARER_CAP_ANALOG 0x1 +#define PPTP_BEARER_CAP_DIGITAL 0x2 + +struct PptpStartSessionRequest { + __u16 protocolVersion; + __u8 reserved1; + __u8 reserved2; + __u32 framingCapability; + __u32 bearerCapability; + __u16 maxChannels; + __u16 firmwareRevision; + __u8 hostName[64]; + __u8 vendorString[64]; +}; + +/* PptpStartSessionResultCode Values */ +#define PPTP_START_OK 1 +#define PPTP_START_GENERAL_ERROR 2 +#define PPTP_START_ALREADY_CONNECTED 3 +#define PPTP_START_NOT_AUTHORIZED 4 +#define PPTP_START_UNKNOWN_PROTOCOL 5 + +struct PptpStartSessionReply { + __u16 protocolVersion; + __u8 resultCode; + __u8 generalErrorCode; + __u32 framingCapability; + __u32 bearerCapability; + __u16 maxChannels; + __u16 firmwareRevision; + __u8 hostName[64]; + __u8 vendorString[64]; +}; + +/* PptpStopReasons */ +#define PPTP_STOP_NONE 1 +#define PPTP_STOP_PROTOCOL 2 +#define PPTP_STOP_LOCAL_SHUTDOWN 3 + +struct PptpStopSessionRequest { + __u8 reason; +}; + +/* PptpStopSessionResultCode */ +#define PPTP_STOP_OK 1 +#define PPTP_STOP_GENERAL_ERROR 2 + +struct PptpStopSessionReply { + __u8 resultCode; + __u8 generalErrorCode; +}; + +struct PptpEchoRequest { + __u32 identNumber; +}; + +/* PptpEchoReplyResultCode */ +#define PPTP_ECHO_OK 1 +#define PPTP_ECHO_GENERAL_ERROR 2 + +struct PptpEchoReply { + __u32 identNumber; + __u8 resultCode; + __u8 generalErrorCode; + __u16 reserved; +}; + +/* PptpFramingType */ +#define PPTP_ASYNC_FRAMING 1 +#define PPTP_SYNC_FRAMING 2 +#define PPTP_DONT_CARE_FRAMING 3 + +/* PptpCallBearerType */ +#define PPTP_ANALOG_TYPE 1 +#define PPTP_DIGITAL_TYPE 2 +#define PPTP_DONT_CARE_BEARER_TYPE 3 + +struct PptpOutCallRequest { + __u16 callID; + __u16 callSerialNumber; + __u32 minBPS; + __u32 maxBPS; + __u32 bearerType; + __u32 framingType; + __u16 packetWindow; + __u16 packetProcDelay; + __u16 reserved1; + __u16 phoneNumberLength; + __u16 reserved2; + __u8 phoneNumber[64]; + __u8 subAddress[64]; +}; + +/* PptpCallResultCode */ +#define PPTP_OUTCALL_CONNECT 1 +#define PPTP_OUTCALL_GENERAL_ERROR 2 +#define PPTP_OUTCALL_NO_CARRIER 3 +#define PPTP_OUTCALL_BUSY 4 +#define PPTP_OUTCALL_NO_DIAL_TONE 5 +#define PPTP_OUTCALL_TIMEOUT 6 +#define PPTP_OUTCALL_DONT_ACCEPT 7 + +struct PptpOutCallReply { + __u16 callID; + __u16 peersCallID; + __u8 resultCode; + __u8 generalErrorCode; + __u16 causeCode; + __u32 connectSpeed; + __u16 packetWindow; + __u16 packetProcDelay; + __u32 physChannelID; +}; + +struct PptpInCallRequest { + __u16 callID; + __u16 callSerialNumber; + __u32 callBearerType; + __u32 physChannelID; + __u16 dialedNumberLength; + __u16 dialingNumberLength; + __u8 dialedNumber[64]; + __u8 dialingNumber[64]; + __u8 subAddress[64]; +}; + +/* PptpInCallResultCode */ +#define PPTP_INCALL_ACCEPT 1 +#define PPTP_INCALL_GENERAL_ERROR 2 +#define PPTP_INCALL_DONT_ACCEPT 3 + +struct PptpInCallReply { + __u16 callID; + __u16 peersCallID; + __u8 resultCode; + __u8 generalErrorCode; + __u16 packetWindow; + __u16 packetProcDelay; + __u16 reserved; +}; + +struct PptpInCallConnected { + __u16 peersCallID; + __u16 reserved; + __u32 connectSpeed; + __u16 packetWindow; + __u16 packetProcDelay; + __u32 callFramingType; +}; + +struct PptpClearCallRequest { + __u16 callID; + __u16 reserved; +}; + +struct PptpCallDisconnectNotify { + __u16 callID; + __u8 resultCode; + __u8 generalErrorCode; + __u16 causeCode; + __u16 reserved; + __u8 callStatistics[128]; +}; + +struct PptpWanErrorNotify { + __u16 peersCallID; + __u16 reserved; + __u32 crcErrors; + __u32 framingErrors; + __u32 hardwareOverRuns; + __u32 bufferOverRuns; + __u32 timeoutErrors; + __u32 alignmentErrors; +}; + +struct PptpSetLinkInfo { + __u16 peersCallID; + __u16 reserved; + __u32 sendAccm; + __u32 recvAccm; +}; + + +struct pptp_priv_data { + __u16 call_id; + __u16 mcall_id; + __u16 pcall_id; +}; + +union pptp_ctrl_union { + struct PptpStartSessionRequest sreq; + struct PptpStartSessionReply srep; + struct PptpStopSessionRequest streq; + struct PptpStopSessionReply strep; + struct PptpOutCallRequest ocreq; + struct PptpOutCallReply ocack; + struct PptpInCallRequest icreq; + struct PptpInCallReply icack; + struct PptpInCallConnected iccon; + struct PptpClearCallRequest clrreq; + struct PptpCallDisconnectNotify disc; + struct PptpWanErrorNotify wanerr; + struct PptpSetLinkInfo setlink; +}; + +extern int +(*ip_nat_pptp_hook_outbound)(struct sk_buff **pskb, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + struct PptpControlHeader *ctlh, + union pptp_ctrl_union *pptpReq); + +extern int +(*ip_nat_pptp_hook_inbound)(struct sk_buff **pskb, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + struct PptpControlHeader *ctlh, + union pptp_ctrl_union *pptpReq); + +extern int +(*ip_nat_pptp_hook_exp_gre)(struct ip_conntrack_expect *exp_orig, + struct ip_conntrack_expect *exp_reply); + +extern void +(*ip_nat_pptp_hook_expectfn)(struct ip_conntrack *ct, + struct ip_conntrack_expect *exp); +#endif /* __KERNEL__ */ +#endif /* _CONNTRACK_PPTP_H */ diff --git a/include/linux/netfilter_ipv4/ip_conntrack_proto_gre.h b/include/linux/netfilter_ipv4/ip_conntrack_proto_gre.h new file mode 100644 index 00000000000..8d090ef82f5 --- /dev/null +++ b/include/linux/netfilter_ipv4/ip_conntrack_proto_gre.h @@ -0,0 +1,114 @@ +#ifndef _CONNTRACK_PROTO_GRE_H +#define _CONNTRACK_PROTO_GRE_H +#include + +/* GRE PROTOCOL HEADER */ + +/* GRE Version field */ +#define GRE_VERSION_1701 0x0 +#define GRE_VERSION_PPTP 0x1 + +/* GRE Protocol field */ +#define GRE_PROTOCOL_PPTP 0x880B + +/* GRE Flags */ +#define GRE_FLAG_C 0x80 +#define GRE_FLAG_R 0x40 +#define GRE_FLAG_K 0x20 +#define GRE_FLAG_S 0x10 +#define GRE_FLAG_A 0x80 + +#define GRE_IS_C(f) ((f)&GRE_FLAG_C) +#define GRE_IS_R(f) ((f)&GRE_FLAG_R) +#define GRE_IS_K(f) ((f)&GRE_FLAG_K) +#define GRE_IS_S(f) ((f)&GRE_FLAG_S) +#define GRE_IS_A(f) ((f)&GRE_FLAG_A) + +/* GRE is a mess: Four different standards */ +struct gre_hdr { +#if defined(__LITTLE_ENDIAN_BITFIELD) + __u16 rec:3, + srr:1, + seq:1, + key:1, + routing:1, + csum:1, + version:3, + reserved:4, + ack:1; +#elif defined(__BIG_ENDIAN_BITFIELD) + __u16 csum:1, + routing:1, + key:1, + seq:1, + srr:1, + rec:3, + ack:1, + reserved:4, + version:3; +#else +#error "Adjust your defines" +#endif + __u16 protocol; +}; + +/* modified GRE header for PPTP */ +struct gre_hdr_pptp { + __u8 flags; /* bitfield */ + __u8 version; /* should be GRE_VERSION_PPTP */ + __u16 protocol; /* should be GRE_PROTOCOL_PPTP */ + __u16 payload_len; /* size of ppp payload, not inc. gre header */ + __u16 call_id; /* peer's call_id for this session */ + __u32 seq; /* sequence number. Present if S==1 */ + __u32 ack; /* seq number of highest packet recieved by */ + /* sender in this session */ +}; + + +/* this is part of ip_conntrack */ +struct ip_ct_gre { + unsigned int stream_timeout; + unsigned int timeout; +}; + +#ifdef __KERNEL__ +struct ip_conntrack_expect; +struct ip_conntrack; + +/* structure for original <-> reply keymap */ +struct ip_ct_gre_keymap { + struct list_head list; + + struct ip_conntrack_tuple tuple; +}; + +/* add new tuple->key_reply pair to keymap */ +int ip_ct_gre_keymap_add(struct ip_conntrack *ct, + struct ip_conntrack_tuple *t, + int reply); + +/* delete keymap entries */ +void ip_ct_gre_keymap_destroy(struct ip_conntrack *ct); + + +/* get pointer to gre key, if present */ +static inline u_int32_t *gre_key(struct gre_hdr *greh) +{ + if (!greh->key) + return NULL; + if (greh->csum || greh->routing) + return (u_int32_t *) (greh+sizeof(*greh)+4); + return (u_int32_t *) (greh+sizeof(*greh)); +} + +/* get pointer ot gre csum, if present */ +static inline u_int16_t *gre_csum(struct gre_hdr *greh) +{ + if (!greh->csum) + return NULL; + return (u_int16_t *) (greh+sizeof(*greh)); +} + +#endif /* __KERNEL__ */ + +#endif /* _CONNTRACK_PROTO_GRE_H */ diff --git a/include/linux/netfilter_ipv4/ip_conntrack_tuple.h b/include/linux/netfilter_ipv4/ip_conntrack_tuple.h index c33f0b5e0d0..14dc0f7b655 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack_tuple.h +++ b/include/linux/netfilter_ipv4/ip_conntrack_tuple.h @@ -28,6 +28,9 @@ union ip_conntrack_manip_proto struct { u_int16_t port; } sctp; + struct { + u_int16_t key; /* key is 32bit, pptp only uses 16 */ + } gre; }; /* The manipulable part of the tuple. */ @@ -61,6 +64,10 @@ struct ip_conntrack_tuple struct { u_int16_t port; } sctp; + struct { + u_int16_t key; /* key is 32bit, + * pptp only uses 16 */ + } gre; } u; /* The protocol. */ diff --git a/include/linux/netfilter_ipv4/ip_nat_pptp.h b/include/linux/netfilter_ipv4/ip_nat_pptp.h new file mode 100644 index 00000000000..eaf66c2e8f9 --- /dev/null +++ b/include/linux/netfilter_ipv4/ip_nat_pptp.h @@ -0,0 +1,11 @@ +/* PPTP constants and structs */ +#ifndef _NAT_PPTP_H +#define _NAT_PPTP_H + +/* conntrack private data */ +struct ip_nat_pptp { + u_int16_t pns_call_id; /* NAT'ed PNS call id */ + u_int16_t pac_call_id; /* NAT'ed PAC call id */ +}; + +#endif /* _NAT_PPTP_H */ diff --git a/net/ipv4/netfilter/Kconfig b/net/ipv4/netfilter/Kconfig index e2162d27007..3cf9b451675 100644 --- a/net/ipv4/netfilter/Kconfig +++ b/net/ipv4/netfilter/Kconfig @@ -137,6 +137,22 @@ config IP_NF_AMANDA To compile it as a module, choose M here. If unsure, say Y. +config IP_NF_PPTP + tristate 'PPTP protocol support' + help + This module adds support for PPTP (Point to Point Tunnelling + Protocol, RFC2637) conncection tracking and NAT. + + If you are running PPTP sessions over a stateful firewall or NAT + box, you may want to enable this feature. + + Please note that not all PPTP modes of operation are supported yet. + For more info, read top of the file + net/ipv4/netfilter/ip_conntrack_pptp.c + + If you want to compile it as a module, say M here and read + Documentation/modules.txt. If unsure, say `N'. + config IP_NF_QUEUE tristate "IP Userspace queueing via NETLINK (OBSOLETE)" help @@ -621,6 +637,12 @@ config IP_NF_NAT_AMANDA default IP_NF_NAT if IP_NF_AMANDA=y default m if IP_NF_AMANDA=m +config IP_NF_NAT_PPTP + tristate + depends on IP_NF_NAT!=n && IP_NF_PPTP!=n + default IP_NF_NAT if IP_NF_PPTP=y + default m if IP_NF_PPTP=m + # mangle + specific targets config IP_NF_MANGLE tristate "Packet mangling" diff --git a/net/ipv4/netfilter/Makefile b/net/ipv4/netfilter/Makefile index 1ba0db74681..3d45d3c0283 100644 --- a/net/ipv4/netfilter/Makefile +++ b/net/ipv4/netfilter/Makefile @@ -6,6 +6,9 @@ ip_conntrack-objs := ip_conntrack_standalone.o ip_conntrack_core.o ip_conntrack_proto_generic.o ip_conntrack_proto_tcp.o ip_conntrack_proto_udp.o ip_conntrack_proto_icmp.o iptable_nat-objs := ip_nat_standalone.o ip_nat_rule.o ip_nat_core.o ip_nat_helper.o ip_nat_proto_unknown.o ip_nat_proto_tcp.o ip_nat_proto_udp.o ip_nat_proto_icmp.o +ip_conntrack_pptp-objs := ip_conntrack_helper_pptp.o ip_conntrack_proto_gre.o +ip_nat_pptp-objs := ip_nat_helper_pptp.o ip_nat_proto_gre.o + # connection tracking obj-$(CONFIG_IP_NF_CONNTRACK) += ip_conntrack.o @@ -17,6 +20,7 @@ obj-$(CONFIG_IP_NF_CONNTRACK_NETLINK) += ip_conntrack_netlink.o obj-$(CONFIG_IP_NF_CT_PROTO_SCTP) += ip_conntrack_proto_sctp.o # connection tracking helpers +obj-$(CONFIG_IP_NF_PPTP) += ip_conntrack_pptp.o obj-$(CONFIG_IP_NF_AMANDA) += ip_conntrack_amanda.o obj-$(CONFIG_IP_NF_TFTP) += ip_conntrack_tftp.o obj-$(CONFIG_IP_NF_FTP) += ip_conntrack_ftp.o @@ -24,6 +28,7 @@ obj-$(CONFIG_IP_NF_IRC) += ip_conntrack_irc.o obj-$(CONFIG_IP_NF_NETBIOS_NS) += ip_conntrack_netbios_ns.o # NAT helpers +obj-$(CONFIG_IP_NF_NAT_PPTP) += ip_nat_pptp.o obj-$(CONFIG_IP_NF_NAT_AMANDA) += ip_nat_amanda.o obj-$(CONFIG_IP_NF_NAT_TFTP) += ip_nat_tftp.o obj-$(CONFIG_IP_NF_NAT_FTP) += ip_nat_ftp.o diff --git a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c new file mode 100644 index 00000000000..79db5b70d5f --- /dev/null +++ b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c @@ -0,0 +1,805 @@ +/* + * ip_conntrack_pptp.c - Version 3.0 + * + * Connection tracking support for PPTP (Point to Point Tunneling Protocol). + * PPTP is a a protocol for creating virtual private networks. + * It is a specification defined by Microsoft and some vendors + * working with Microsoft. PPTP is built on top of a modified + * version of the Internet Generic Routing Encapsulation Protocol. + * GRE is defined in RFC 1701 and RFC 1702. Documentation of + * PPTP can be found in RFC 2637 + * + * (C) 2000-2005 by Harald Welte + * + * Development of this code funded by Astaro AG (http://www.astaro.com/) + * + * Limitations: + * - We blindly assume that control connections are always + * established in PNS->PAC direction. This is a violation + * of RFFC2673 + * - We can only support one single call within each session + * + * TODO: + * - testing of incoming PPTP calls + * + * Changes: + * 2002-02-05 - Version 1.3 + * - Call ip_conntrack_unexpect_related() from + * pptp_destroy_siblings() to destroy expectations in case + * CALL_DISCONNECT_NOTIFY or tcp fin packet was seen + * (Philip Craig ) + * - Add Version information at module loadtime + * 2002-02-10 - Version 1.6 + * - move to C99 style initializers + * - remove second expectation if first arrives + * 2004-10-22 - Version 2.0 + * - merge Mandrake's 2.6.x port with recent 2.6.x API changes + * - fix lots of linear skb assumptions from Mandrake's port + * 2005-06-10 - Version 2.1 + * - use ip_conntrack_expect_free() instead of kfree() on the + * expect's (which are from the slab for quite some time) + * 2005-06-10 - Version 3.0 + * - port helper to post-2.6.11 API changes, + * funded by Oxcoda NetBox Blue (http://www.netboxblue.com/) + * 2005-07-30 - Version 3.1 + * - port helper to 2.6.13 API changes + * + */ + +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +#define IP_CT_PPTP_VERSION "3.1" + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Harald Welte "); +MODULE_DESCRIPTION("Netfilter connection tracking helper module for PPTP"); + +static DEFINE_SPINLOCK(ip_pptp_lock); + +int +(*ip_nat_pptp_hook_outbound)(struct sk_buff **pskb, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + struct PptpControlHeader *ctlh, + union pptp_ctrl_union *pptpReq); + +int +(*ip_nat_pptp_hook_inbound)(struct sk_buff **pskb, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + struct PptpControlHeader *ctlh, + union pptp_ctrl_union *pptpReq); + +int +(*ip_nat_pptp_hook_exp_gre)(struct ip_conntrack_expect *expect_orig, + struct ip_conntrack_expect *expect_reply); + +void +(*ip_nat_pptp_hook_expectfn)(struct ip_conntrack *ct, + struct ip_conntrack_expect *exp); + +#if 0 +/* PptpControlMessageType names */ +const char *pptp_msg_name[] = { + "UNKNOWN_MESSAGE", + "START_SESSION_REQUEST", + "START_SESSION_REPLY", + "STOP_SESSION_REQUEST", + "STOP_SESSION_REPLY", + "ECHO_REQUEST", + "ECHO_REPLY", + "OUT_CALL_REQUEST", + "OUT_CALL_REPLY", + "IN_CALL_REQUEST", + "IN_CALL_REPLY", + "IN_CALL_CONNECT", + "CALL_CLEAR_REQUEST", + "CALL_DISCONNECT_NOTIFY", + "WAN_ERROR_NOTIFY", + "SET_LINK_INFO" +}; +EXPORT_SYMBOL(pptp_msg_name); +#define DEBUGP(format, args...) printk(KERN_DEBUG "%s:%s: " format, __FILE__, __FUNCTION__, ## args) +#else +#define DEBUGP(format, args...) +#endif + +#define SECS *HZ +#define MINS * 60 SECS +#define HOURS * 60 MINS + +#define PPTP_GRE_TIMEOUT (10 MINS) +#define PPTP_GRE_STREAM_TIMEOUT (5 HOURS) + +static void pptp_expectfn(struct ip_conntrack *ct, + struct ip_conntrack_expect *exp) +{ + DEBUGP("increasing timeouts\n"); + + /* increase timeout of GRE data channel conntrack entry */ + ct->proto.gre.timeout = PPTP_GRE_TIMEOUT; + ct->proto.gre.stream_timeout = PPTP_GRE_STREAM_TIMEOUT; + + /* Can you see how rusty this code is, compared with the pre-2.6.11 + * one? That's what happened to my shiny newnat of 2002 ;( -HW */ + + if (!ip_nat_pptp_hook_expectfn) { + struct ip_conntrack_tuple inv_t; + struct ip_conntrack_expect *exp_other; + + /* obviously this tuple inversion only works until you do NAT */ + invert_tuplepr(&inv_t, &exp->tuple); + DEBUGP("trying to unexpect other dir: "); + DUMP_TUPLE(&inv_t); + + exp_other = ip_conntrack_expect_find(&inv_t); + if (exp_other) { + /* delete other expectation. */ + DEBUGP("found\n"); + ip_conntrack_unexpect_related(exp_other); + ip_conntrack_expect_put(exp_other); + } else { + DEBUGP("not found\n"); + } + } else { + /* we need more than simple inversion */ + ip_nat_pptp_hook_expectfn(ct, exp); + } +} + +static int destroy_sibling_or_exp(const struct ip_conntrack_tuple *t) +{ + struct ip_conntrack_tuple_hash *h; + struct ip_conntrack_expect *exp; + + DEBUGP("trying to timeout ct or exp for tuple "); + DUMP_TUPLE(t); + + h = ip_conntrack_find_get(t, NULL); + if (h) { + struct ip_conntrack *sibling = tuplehash_to_ctrack(h); + DEBUGP("setting timeout of conntrack %p to 0\n", sibling); + sibling->proto.gre.timeout = 0; + sibling->proto.gre.stream_timeout = 0; + /* refresh_acct will not modify counters if skb == NULL */ + if (del_timer(&sibling->timeout)) + sibling->timeout.function((unsigned long)sibling); + ip_conntrack_put(sibling); + return 1; + } else { + exp = ip_conntrack_expect_find(t); + if (exp) { + DEBUGP("unexpect_related of expect %p\n", exp); + ip_conntrack_unexpect_related(exp); + ip_conntrack_expect_put(exp); + return 1; + } + } + + return 0; +} + + +/* timeout GRE data connections */ +static void pptp_destroy_siblings(struct ip_conntrack *ct) +{ + struct ip_conntrack_tuple t; + + /* Since ct->sibling_list has literally rusted away in 2.6.11, + * we now need another way to find out about our sibling + * contrack and expects... -HW */ + + /* try original (pns->pac) tuple */ + memcpy(&t, &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple, sizeof(t)); + t.dst.protonum = IPPROTO_GRE; + t.src.u.gre.key = htons(ct->help.ct_pptp_info.pns_call_id); + t.dst.u.gre.key = htons(ct->help.ct_pptp_info.pac_call_id); + + if (!destroy_sibling_or_exp(&t)) + DEBUGP("failed to timeout original pns->pac ct/exp\n"); + + /* try reply (pac->pns) tuple */ + memcpy(&t, &ct->tuplehash[IP_CT_DIR_REPLY].tuple, sizeof(t)); + t.dst.protonum = IPPROTO_GRE; + t.src.u.gre.key = htons(ct->help.ct_pptp_info.pac_call_id); + t.dst.u.gre.key = htons(ct->help.ct_pptp_info.pns_call_id); + + if (!destroy_sibling_or_exp(&t)) + DEBUGP("failed to timeout reply pac->pns ct/exp\n"); +} + +/* expect GRE connections (PNS->PAC and PAC->PNS direction) */ +static inline int +exp_gre(struct ip_conntrack *master, + u_int32_t seq, + u_int16_t callid, + u_int16_t peer_callid) +{ + struct ip_conntrack_tuple inv_tuple; + struct ip_conntrack_tuple exp_tuples[] = { + /* tuple in original direction, PNS->PAC */ + { .src = { .ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip, + .u = { .gre = { .key = peer_callid } } + }, + .dst = { .ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.ip, + .u = { .gre = { .key = callid } }, + .protonum = IPPROTO_GRE + }, + }, + /* tuple in reply direction, PAC->PNS */ + { .src = { .ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip, + .u = { .gre = { .key = callid } } + }, + .dst = { .ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip, + .u = { .gre = { .key = peer_callid } }, + .protonum = IPPROTO_GRE + }, + } + }; + struct ip_conntrack_expect *exp_orig, *exp_reply; + int ret = 1; + + exp_orig = ip_conntrack_expect_alloc(master); + if (exp_orig == NULL) + goto out; + + exp_reply = ip_conntrack_expect_alloc(master); + if (exp_reply == NULL) + goto out_put_orig; + + memcpy(&exp_orig->tuple, &exp_tuples[0], sizeof(exp_orig->tuple)); + + exp_orig->mask.src.ip = 0xffffffff; + exp_orig->mask.src.u.all = 0; + exp_orig->mask.dst.u.all = 0; + exp_orig->mask.dst.u.gre.key = 0xffff; + exp_orig->mask.dst.ip = 0xffffffff; + exp_orig->mask.dst.protonum = 0xff; + + exp_orig->master = master; + exp_orig->expectfn = pptp_expectfn; + exp_orig->flags = 0; + + exp_orig->dir = IP_CT_DIR_ORIGINAL; + + /* both expectations are identical apart from tuple */ + memcpy(exp_reply, exp_orig, sizeof(*exp_reply)); + memcpy(&exp_reply->tuple, &exp_tuples[1], sizeof(exp_reply->tuple)); + + exp_reply->dir = !exp_orig->dir; + + if (ip_nat_pptp_hook_exp_gre) + ret = ip_nat_pptp_hook_exp_gre(exp_orig, exp_reply); + else { + + DEBUGP("calling expect_related PNS->PAC"); + DUMP_TUPLE(&exp_orig->tuple); + + if (ip_conntrack_expect_related(exp_orig) != 0) { + DEBUGP("cannot expect_related()\n"); + goto out_put_both; + } + + DEBUGP("calling expect_related PAC->PNS"); + DUMP_TUPLE(&exp_reply->tuple); + + if (ip_conntrack_expect_related(exp_reply) != 0) { + DEBUGP("cannot expect_related()\n"); + goto out_unexpect_orig; + } + + /* Add GRE keymap entries */ + if (ip_ct_gre_keymap_add(master, &exp_reply->tuple, 0) != 0) { + DEBUGP("cannot keymap_add() exp\n"); + goto out_unexpect_both; + } + + invert_tuplepr(&inv_tuple, &exp_reply->tuple); + if (ip_ct_gre_keymap_add(master, &inv_tuple, 1) != 0) { + ip_ct_gre_keymap_destroy(master); + DEBUGP("cannot keymap_add() exp_inv\n"); + goto out_unexpect_both; + } + ret = 0; + } + +out_put_both: + ip_conntrack_expect_put(exp_reply); +out_put_orig: + ip_conntrack_expect_put(exp_orig); +out: + return ret; + +out_unexpect_both: + ip_conntrack_unexpect_related(exp_reply); +out_unexpect_orig: + ip_conntrack_unexpect_related(exp_orig); + goto out_put_both; +} + +static inline int +pptp_inbound_pkt(struct sk_buff **pskb, + struct tcphdr *tcph, + unsigned int nexthdr_off, + unsigned int datalen, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo) +{ + struct PptpControlHeader _ctlh, *ctlh; + unsigned int reqlen; + union pptp_ctrl_union _pptpReq, *pptpReq; + struct ip_ct_pptp_master *info = &ct->help.ct_pptp_info; + u_int16_t msg, *cid, *pcid; + u_int32_t seq; + + ctlh = skb_header_pointer(*pskb, nexthdr_off, sizeof(_ctlh), &_ctlh); + if (!ctlh) { + DEBUGP("error during skb_header_pointer\n"); + return NF_ACCEPT; + } + nexthdr_off += sizeof(_ctlh); + datalen -= sizeof(_ctlh); + + reqlen = datalen; + if (reqlen > sizeof(*pptpReq)) + reqlen = sizeof(*pptpReq); + pptpReq = skb_header_pointer(*pskb, nexthdr_off, reqlen, &_pptpReq); + if (!pptpReq) { + DEBUGP("error during skb_header_pointer\n"); + return NF_ACCEPT; + } + + msg = ntohs(ctlh->messageType); + DEBUGP("inbound control message %s\n", pptp_msg_name[msg]); + + switch (msg) { + case PPTP_START_SESSION_REPLY: + if (reqlen < sizeof(_pptpReq.srep)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server confirms new control session */ + if (info->sstate < PPTP_SESSION_REQUESTED) { + DEBUGP("%s without START_SESS_REQUEST\n", + pptp_msg_name[msg]); + break; + } + if (pptpReq->srep.resultCode == PPTP_START_OK) + info->sstate = PPTP_SESSION_CONFIRMED; + else + info->sstate = PPTP_SESSION_ERROR; + break; + + case PPTP_STOP_SESSION_REPLY: + if (reqlen < sizeof(_pptpReq.strep)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server confirms end of control session */ + if (info->sstate > PPTP_SESSION_STOPREQ) { + DEBUGP("%s without STOP_SESS_REQUEST\n", + pptp_msg_name[msg]); + break; + } + if (pptpReq->strep.resultCode == PPTP_STOP_OK) + info->sstate = PPTP_SESSION_NONE; + else + info->sstate = PPTP_SESSION_ERROR; + break; + + case PPTP_OUT_CALL_REPLY: + if (reqlen < sizeof(_pptpReq.ocack)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server accepted call, we now expect GRE frames */ + if (info->sstate != PPTP_SESSION_CONFIRMED) { + DEBUGP("%s but no session\n", pptp_msg_name[msg]); + break; + } + if (info->cstate != PPTP_CALL_OUT_REQ && + info->cstate != PPTP_CALL_OUT_CONF) { + DEBUGP("%s without OUTCALL_REQ\n", pptp_msg_name[msg]); + break; + } + if (pptpReq->ocack.resultCode != PPTP_OUTCALL_CONNECT) { + info->cstate = PPTP_CALL_NONE; + break; + } + + cid = &pptpReq->ocack.callID; + pcid = &pptpReq->ocack.peersCallID; + + info->pac_call_id = ntohs(*cid); + + if (htons(info->pns_call_id) != *pcid) { + DEBUGP("%s for unknown callid %u\n", + pptp_msg_name[msg], ntohs(*pcid)); + break; + } + + DEBUGP("%s, CID=%X, PCID=%X\n", pptp_msg_name[msg], + ntohs(*cid), ntohs(*pcid)); + + info->cstate = PPTP_CALL_OUT_CONF; + + seq = ntohl(tcph->seq) + sizeof(struct pptp_pkt_hdr) + + sizeof(struct PptpControlHeader) + + ((void *)pcid - (void *)pptpReq); + + if (exp_gre(ct, seq, *cid, *pcid) != 0) + printk("ip_conntrack_pptp: error during exp_gre\n"); + break; + + case PPTP_IN_CALL_REQUEST: + if (reqlen < sizeof(_pptpReq.icack)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server tells us about incoming call request */ + if (info->sstate != PPTP_SESSION_CONFIRMED) { + DEBUGP("%s but no session\n", pptp_msg_name[msg]); + break; + } + pcid = &pptpReq->icack.peersCallID; + DEBUGP("%s, PCID=%X\n", pptp_msg_name[msg], ntohs(*pcid)); + info->cstate = PPTP_CALL_IN_REQ; + info->pac_call_id = ntohs(*pcid); + break; + + case PPTP_IN_CALL_CONNECT: + if (reqlen < sizeof(_pptpReq.iccon)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server tells us about incoming call established */ + if (info->sstate != PPTP_SESSION_CONFIRMED) { + DEBUGP("%s but no session\n", pptp_msg_name[msg]); + break; + } + if (info->sstate != PPTP_CALL_IN_REP + && info->sstate != PPTP_CALL_IN_CONF) { + DEBUGP("%s but never sent IN_CALL_REPLY\n", + pptp_msg_name[msg]); + break; + } + + pcid = &pptpReq->iccon.peersCallID; + cid = &info->pac_call_id; + + if (info->pns_call_id != ntohs(*pcid)) { + DEBUGP("%s for unknown CallID %u\n", + pptp_msg_name[msg], ntohs(*cid)); + break; + } + + DEBUGP("%s, PCID=%X\n", pptp_msg_name[msg], ntohs(*pcid)); + info->cstate = PPTP_CALL_IN_CONF; + + /* we expect a GRE connection from PAC to PNS */ + seq = ntohl(tcph->seq) + sizeof(struct pptp_pkt_hdr) + + sizeof(struct PptpControlHeader) + + ((void *)pcid - (void *)pptpReq); + + if (exp_gre(ct, seq, *cid, *pcid) != 0) + printk("ip_conntrack_pptp: error during exp_gre\n"); + + break; + + case PPTP_CALL_DISCONNECT_NOTIFY: + if (reqlen < sizeof(_pptpReq.disc)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* server confirms disconnect */ + cid = &pptpReq->disc.callID; + DEBUGP("%s, CID=%X\n", pptp_msg_name[msg], ntohs(*cid)); + info->cstate = PPTP_CALL_NONE; + + /* untrack this call id, unexpect GRE packets */ + pptp_destroy_siblings(ct); + break; + + case PPTP_WAN_ERROR_NOTIFY: + break; + + case PPTP_ECHO_REQUEST: + case PPTP_ECHO_REPLY: + /* I don't have to explain these ;) */ + break; + default: + DEBUGP("invalid %s (TY=%d)\n", (msg <= PPTP_MSG_MAX) + ? pptp_msg_name[msg]:pptp_msg_name[0], msg); + break; + } + + + if (ip_nat_pptp_hook_inbound) + return ip_nat_pptp_hook_inbound(pskb, ct, ctinfo, ctlh, + pptpReq); + + return NF_ACCEPT; + +} + +static inline int +pptp_outbound_pkt(struct sk_buff **pskb, + struct tcphdr *tcph, + unsigned int nexthdr_off, + unsigned int datalen, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo) +{ + struct PptpControlHeader _ctlh, *ctlh; + unsigned int reqlen; + union pptp_ctrl_union _pptpReq, *pptpReq; + struct ip_ct_pptp_master *info = &ct->help.ct_pptp_info; + u_int16_t msg, *cid, *pcid; + + ctlh = skb_header_pointer(*pskb, nexthdr_off, sizeof(_ctlh), &_ctlh); + if (!ctlh) + return NF_ACCEPT; + nexthdr_off += sizeof(_ctlh); + datalen -= sizeof(_ctlh); + + reqlen = datalen; + if (reqlen > sizeof(*pptpReq)) + reqlen = sizeof(*pptpReq); + pptpReq = skb_header_pointer(*pskb, nexthdr_off, reqlen, &_pptpReq); + if (!pptpReq) + return NF_ACCEPT; + + msg = ntohs(ctlh->messageType); + DEBUGP("outbound control message %s\n", pptp_msg_name[msg]); + + switch (msg) { + case PPTP_START_SESSION_REQUEST: + /* client requests for new control session */ + if (info->sstate != PPTP_SESSION_NONE) { + DEBUGP("%s but we already have one", + pptp_msg_name[msg]); + } + info->sstate = PPTP_SESSION_REQUESTED; + break; + case PPTP_STOP_SESSION_REQUEST: + /* client requests end of control session */ + info->sstate = PPTP_SESSION_STOPREQ; + break; + + case PPTP_OUT_CALL_REQUEST: + if (reqlen < sizeof(_pptpReq.ocreq)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + /* FIXME: break; */ + } + + /* client initiating connection to server */ + if (info->sstate != PPTP_SESSION_CONFIRMED) { + DEBUGP("%s but no session\n", + pptp_msg_name[msg]); + break; + } + info->cstate = PPTP_CALL_OUT_REQ; + /* track PNS call id */ + cid = &pptpReq->ocreq.callID; + DEBUGP("%s, CID=%X\n", pptp_msg_name[msg], ntohs(*cid)); + info->pns_call_id = ntohs(*cid); + break; + case PPTP_IN_CALL_REPLY: + if (reqlen < sizeof(_pptpReq.icack)) { + DEBUGP("%s: short packet\n", pptp_msg_name[msg]); + break; + } + + /* client answers incoming call */ + if (info->cstate != PPTP_CALL_IN_REQ + && info->cstate != PPTP_CALL_IN_REP) { + DEBUGP("%s without incall_req\n", + pptp_msg_name[msg]); + break; + } + if (pptpReq->icack.resultCode != PPTP_INCALL_ACCEPT) { + info->cstate = PPTP_CALL_NONE; + break; + } + pcid = &pptpReq->icack.peersCallID; + if (info->pac_call_id != ntohs(*pcid)) { + DEBUGP("%s for unknown call %u\n", + pptp_msg_name[msg], ntohs(*pcid)); + break; + } + DEBUGP("%s, CID=%X\n", pptp_msg_name[msg], ntohs(*pcid)); + /* part two of the three-way handshake */ + info->cstate = PPTP_CALL_IN_REP; + info->pns_call_id = ntohs(pptpReq->icack.callID); + break; + + case PPTP_CALL_CLEAR_REQUEST: + /* client requests hangup of call */ + if (info->sstate != PPTP_SESSION_CONFIRMED) { + DEBUGP("CLEAR_CALL but no session\n"); + break; + } + /* FUTURE: iterate over all calls and check if + * call ID is valid. We don't do this without newnat, + * because we only know about last call */ + info->cstate = PPTP_CALL_CLEAR_REQ; + break; + case PPTP_SET_LINK_INFO: + break; + case PPTP_ECHO_REQUEST: + case PPTP_ECHO_REPLY: + /* I don't have to explain these ;) */ + break; + default: + DEBUGP("invalid %s (TY=%d)\n", (msg <= PPTP_MSG_MAX)? + pptp_msg_name[msg]:pptp_msg_name[0], msg); + /* unknown: no need to create GRE masq table entry */ + break; + } + + if (ip_nat_pptp_hook_outbound) + return ip_nat_pptp_hook_outbound(pskb, ct, ctinfo, ctlh, + pptpReq); + + return NF_ACCEPT; +} + + +/* track caller id inside control connection, call expect_related */ +static int +conntrack_pptp_help(struct sk_buff **pskb, + struct ip_conntrack *ct, enum ip_conntrack_info ctinfo) + +{ + struct pptp_pkt_hdr _pptph, *pptph; + struct tcphdr _tcph, *tcph; + u_int32_t tcplen = (*pskb)->len - (*pskb)->nh.iph->ihl * 4; + u_int32_t datalen; + int dir = CTINFO2DIR(ctinfo); + struct ip_ct_pptp_master *info = &ct->help.ct_pptp_info; + unsigned int nexthdr_off; + + int oldsstate, oldcstate; + int ret; + + /* don't do any tracking before tcp handshake complete */ + if (ctinfo != IP_CT_ESTABLISHED + && ctinfo != IP_CT_ESTABLISHED+IP_CT_IS_REPLY) { + DEBUGP("ctinfo = %u, skipping\n", ctinfo); + return NF_ACCEPT; + } + + nexthdr_off = (*pskb)->nh.iph->ihl*4; + tcph = skb_header_pointer(*pskb, nexthdr_off, sizeof(_tcph), &_tcph); + BUG_ON(!tcph); + nexthdr_off += tcph->doff * 4; + datalen = tcplen - tcph->doff * 4; + + if (tcph->fin || tcph->rst) { + DEBUGP("RST/FIN received, timeouting GRE\n"); + /* can't do this after real newnat */ + info->cstate = PPTP_CALL_NONE; + + /* untrack this call id, unexpect GRE packets */ + pptp_destroy_siblings(ct); + } + + pptph = skb_header_pointer(*pskb, nexthdr_off, sizeof(_pptph), &_pptph); + if (!pptph) { + DEBUGP("no full PPTP header, can't track\n"); + return NF_ACCEPT; + } + nexthdr_off += sizeof(_pptph); + datalen -= sizeof(_pptph); + + /* if it's not a control message we can't do anything with it */ + if (ntohs(pptph->packetType) != PPTP_PACKET_CONTROL || + ntohl(pptph->magicCookie) != PPTP_MAGIC_COOKIE) { + DEBUGP("not a control packet\n"); + return NF_ACCEPT; + } + + oldsstate = info->sstate; + oldcstate = info->cstate; + + spin_lock_bh(&ip_pptp_lock); + + /* FIXME: We just blindly assume that the control connection is always + * established from PNS->PAC. However, RFC makes no guarantee */ + if (dir == IP_CT_DIR_ORIGINAL) + /* client -> server (PNS -> PAC) */ + ret = pptp_outbound_pkt(pskb, tcph, nexthdr_off, datalen, ct, + ctinfo); + else + /* server -> client (PAC -> PNS) */ + ret = pptp_inbound_pkt(pskb, tcph, nexthdr_off, datalen, ct, + ctinfo); + DEBUGP("sstate: %d->%d, cstate: %d->%d\n", + oldsstate, info->sstate, oldcstate, info->cstate); + spin_unlock_bh(&ip_pptp_lock); + + return ret; +} + +/* control protocol helper */ +static struct ip_conntrack_helper pptp = { + .list = { NULL, NULL }, + .name = "pptp", + .me = THIS_MODULE, + .max_expected = 2, + .timeout = 5 * 60, + .tuple = { .src = { .ip = 0, + .u = { .tcp = { .port = + __constant_htons(PPTP_CONTROL_PORT) } } + }, + .dst = { .ip = 0, + .u = { .all = 0 }, + .protonum = IPPROTO_TCP + } + }, + .mask = { .src = { .ip = 0, + .u = { .tcp = { .port = 0xffff } } + }, + .dst = { .ip = 0, + .u = { .all = 0 }, + .protonum = 0xff + } + }, + .help = conntrack_pptp_help +}; + +extern void __exit ip_ct_proto_gre_fini(void); +extern int __init ip_ct_proto_gre_init(void); + +/* ip_conntrack_pptp initialization */ +static int __init init(void) +{ + int retcode; + + retcode = ip_ct_proto_gre_init(); + if (retcode < 0) + return retcode; + + DEBUGP(" registering helper\n"); + if ((retcode = ip_conntrack_helper_register(&pptp))) { + printk(KERN_ERR "Unable to register conntrack application " + "helper for pptp: %d\n", retcode); + ip_ct_proto_gre_fini(); + return retcode; + } + + printk("ip_conntrack_pptp version %s loaded\n", IP_CT_PPTP_VERSION); + return 0; +} + +static void __exit fini(void) +{ + ip_conntrack_helper_unregister(&pptp); + ip_ct_proto_gre_fini(); + printk("ip_conntrack_pptp version %s unloaded\n", IP_CT_PPTP_VERSION); +} + +module_init(init); +module_exit(fini); + +EXPORT_SYMBOL(ip_nat_pptp_hook_outbound); +EXPORT_SYMBOL(ip_nat_pptp_hook_inbound); +EXPORT_SYMBOL(ip_nat_pptp_hook_exp_gre); +EXPORT_SYMBOL(ip_nat_pptp_hook_expectfn); diff --git a/net/ipv4/netfilter/ip_conntrack_proto_gre.c b/net/ipv4/netfilter/ip_conntrack_proto_gre.c new file mode 100644 index 00000000000..de3cb9db6f8 --- /dev/null +++ b/net/ipv4/netfilter/ip_conntrack_proto_gre.c @@ -0,0 +1,327 @@ +/* + * ip_conntrack_proto_gre.c - Version 3.0 + * + * Connection tracking protocol helper module for GRE. + * + * GRE is a generic encapsulation protocol, which is generally not very + * suited for NAT, as it has no protocol-specific part as port numbers. + * + * It has an optional key field, which may help us distinguishing two + * connections between the same two hosts. + * + * GRE is defined in RFC 1701 and RFC 1702, as well as RFC 2784 + * + * PPTP is built on top of a modified version of GRE, and has a mandatory + * field called "CallID", which serves us for the same purpose as the key + * field in plain GRE. + * + * Documentation about PPTP can be found in RFC 2637 + * + * (C) 2000-2005 by Harald Welte + * + * Development of this code funded by Astaro AG (http://www.astaro.com/) + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +static DEFINE_RWLOCK(ip_ct_gre_lock); +#define ASSERT_READ_LOCK(x) +#define ASSERT_WRITE_LOCK(x) + +#include +#include +#include +#include + +#include +#include + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Harald Welte "); +MODULE_DESCRIPTION("netfilter connection tracking protocol helper for GRE"); + +/* shamelessly stolen from ip_conntrack_proto_udp.c */ +#define GRE_TIMEOUT (30*HZ) +#define GRE_STREAM_TIMEOUT (180*HZ) + +#if 0 +#define DEBUGP(format, args...) printk(KERN_DEBUG "%s:%s: " format, __FILE__, __FUNCTION__, ## args) +#define DUMP_TUPLE_GRE(x) printk("%u.%u.%u.%u:0x%x -> %u.%u.%u.%u:0x%x\n", \ + NIPQUAD((x)->src.ip), ntohs((x)->src.u.gre.key), \ + NIPQUAD((x)->dst.ip), ntohs((x)->dst.u.gre.key)) +#else +#define DEBUGP(x, args...) +#define DUMP_TUPLE_GRE(x) +#endif + +/* GRE KEYMAP HANDLING FUNCTIONS */ +static LIST_HEAD(gre_keymap_list); + +static inline int gre_key_cmpfn(const struct ip_ct_gre_keymap *km, + const struct ip_conntrack_tuple *t) +{ + return ((km->tuple.src.ip == t->src.ip) && + (km->tuple.dst.ip == t->dst.ip) && + (km->tuple.dst.protonum == t->dst.protonum) && + (km->tuple.dst.u.all == t->dst.u.all)); +} + +/* look up the source key for a given tuple */ +static u_int32_t gre_keymap_lookup(struct ip_conntrack_tuple *t) +{ + struct ip_ct_gre_keymap *km; + u_int32_t key = 0; + + read_lock_bh(&ip_ct_gre_lock); + km = LIST_FIND(&gre_keymap_list, gre_key_cmpfn, + struct ip_ct_gre_keymap *, t); + if (km) + key = km->tuple.src.u.gre.key; + read_unlock_bh(&ip_ct_gre_lock); + + DEBUGP("lookup src key 0x%x up key for ", key); + DUMP_TUPLE_GRE(t); + + return key; +} + +/* add a single keymap entry, associate with specified master ct */ +int +ip_ct_gre_keymap_add(struct ip_conntrack *ct, + struct ip_conntrack_tuple *t, int reply) +{ + struct ip_ct_gre_keymap **exist_km, *km, *old; + + if (!ct->helper || strcmp(ct->helper->name, "pptp")) { + DEBUGP("refusing to add GRE keymap to non-pptp session\n"); + return -1; + } + + if (!reply) + exist_km = &ct->help.ct_pptp_info.keymap_orig; + else + exist_km = &ct->help.ct_pptp_info.keymap_reply; + + if (*exist_km) { + /* check whether it's a retransmission */ + old = LIST_FIND(&gre_keymap_list, gre_key_cmpfn, + struct ip_ct_gre_keymap *, t); + if (old == *exist_km) { + DEBUGP("retransmission\n"); + return 0; + } + + DEBUGP("trying to override keymap_%s for ct %p\n", + reply? "reply":"orig", ct); + return -EEXIST; + } + + km = kmalloc(sizeof(*km), GFP_ATOMIC); + if (!km) + return -ENOMEM; + + memcpy(&km->tuple, t, sizeof(*t)); + *exist_km = km; + + DEBUGP("adding new entry %p: ", km); + DUMP_TUPLE_GRE(&km->tuple); + + write_lock_bh(&ip_ct_gre_lock); + list_append(&gre_keymap_list, km); + write_unlock_bh(&ip_ct_gre_lock); + + return 0; +} + +/* destroy the keymap entries associated with specified master ct */ +void ip_ct_gre_keymap_destroy(struct ip_conntrack *ct) +{ + DEBUGP("entering for ct %p\n", ct); + + if (!ct->helper || strcmp(ct->helper->name, "pptp")) { + DEBUGP("refusing to destroy GRE keymap to non-pptp session\n"); + return; + } + + write_lock_bh(&ip_ct_gre_lock); + if (ct->help.ct_pptp_info.keymap_orig) { + DEBUGP("removing %p from list\n", + ct->help.ct_pptp_info.keymap_orig); + list_del(&ct->help.ct_pptp_info.keymap_orig->list); + kfree(ct->help.ct_pptp_info.keymap_orig); + ct->help.ct_pptp_info.keymap_orig = NULL; + } + if (ct->help.ct_pptp_info.keymap_reply) { + DEBUGP("removing %p from list\n", + ct->help.ct_pptp_info.keymap_reply); + list_del(&ct->help.ct_pptp_info.keymap_reply->list); + kfree(ct->help.ct_pptp_info.keymap_reply); + ct->help.ct_pptp_info.keymap_reply = NULL; + } + write_unlock_bh(&ip_ct_gre_lock); +} + + +/* PUBLIC CONNTRACK PROTO HELPER FUNCTIONS */ + +/* invert gre part of tuple */ +static int gre_invert_tuple(struct ip_conntrack_tuple *tuple, + const struct ip_conntrack_tuple *orig) +{ + tuple->dst.u.gre.key = orig->src.u.gre.key; + tuple->src.u.gre.key = orig->dst.u.gre.key; + + return 1; +} + +/* gre hdr info to tuple */ +static int gre_pkt_to_tuple(const struct sk_buff *skb, + unsigned int dataoff, + struct ip_conntrack_tuple *tuple) +{ + struct gre_hdr_pptp _pgrehdr, *pgrehdr; + u_int32_t srckey; + struct gre_hdr _grehdr, *grehdr; + + /* first only delinearize old RFC1701 GRE header */ + grehdr = skb_header_pointer(skb, dataoff, sizeof(_grehdr), &_grehdr); + if (!grehdr || grehdr->version != GRE_VERSION_PPTP) { + /* try to behave like "ip_conntrack_proto_generic" */ + tuple->src.u.all = 0; + tuple->dst.u.all = 0; + return 1; + } + + /* PPTP header is variable length, only need up to the call_id field */ + pgrehdr = skb_header_pointer(skb, dataoff, 8, &_pgrehdr); + if (!pgrehdr) + return 1; + + if (ntohs(grehdr->protocol) != GRE_PROTOCOL_PPTP) { + DEBUGP("GRE_VERSION_PPTP but unknown proto\n"); + return 0; + } + + tuple->dst.u.gre.key = pgrehdr->call_id; + srckey = gre_keymap_lookup(tuple); + tuple->src.u.gre.key = srckey; + + return 1; +} + +/* print gre part of tuple */ +static int gre_print_tuple(struct seq_file *s, + const struct ip_conntrack_tuple *tuple) +{ + return seq_printf(s, "srckey=0x%x dstkey=0x%x ", + ntohs(tuple->src.u.gre.key), + ntohs(tuple->dst.u.gre.key)); +} + +/* print private data for conntrack */ +static int gre_print_conntrack(struct seq_file *s, + const struct ip_conntrack *ct) +{ + return seq_printf(s, "timeout=%u, stream_timeout=%u ", + (ct->proto.gre.timeout / HZ), + (ct->proto.gre.stream_timeout / HZ)); +} + +/* Returns verdict for packet, and may modify conntrack */ +static int gre_packet(struct ip_conntrack *ct, + const struct sk_buff *skb, + enum ip_conntrack_info conntrackinfo) +{ + /* If we've seen traffic both ways, this is a GRE connection. + * Extend timeout. */ + if (ct->status & IPS_SEEN_REPLY) { + ip_ct_refresh_acct(ct, conntrackinfo, skb, + ct->proto.gre.stream_timeout); + /* Also, more likely to be important, and not a probe. */ + set_bit(IPS_ASSURED_BIT, &ct->status); + } else + ip_ct_refresh_acct(ct, conntrackinfo, skb, + ct->proto.gre.timeout); + + return NF_ACCEPT; +} + +/* Called when a new connection for this protocol found. */ +static int gre_new(struct ip_conntrack *ct, + const struct sk_buff *skb) +{ + DEBUGP(": "); + DUMP_TUPLE_GRE(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple); + + /* initialize to sane value. Ideally a conntrack helper + * (e.g. in case of pptp) is increasing them */ + ct->proto.gre.stream_timeout = GRE_STREAM_TIMEOUT; + ct->proto.gre.timeout = GRE_TIMEOUT; + + return 1; +} + +/* Called when a conntrack entry has already been removed from the hashes + * and is about to be deleted from memory */ +static void gre_destroy(struct ip_conntrack *ct) +{ + struct ip_conntrack *master = ct->master; + DEBUGP(" entering\n"); + + if (!master) + DEBUGP("no master !?!\n"); + else + ip_ct_gre_keymap_destroy(master); +} + +/* protocol helper struct */ +static struct ip_conntrack_protocol gre = { + .proto = IPPROTO_GRE, + .name = "gre", + .pkt_to_tuple = gre_pkt_to_tuple, + .invert_tuple = gre_invert_tuple, + .print_tuple = gre_print_tuple, + .print_conntrack = gre_print_conntrack, + .packet = gre_packet, + .new = gre_new, + .destroy = gre_destroy, + .me = THIS_MODULE, +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) + .tuple_to_nfattr = ip_ct_port_tuple_to_nfattr, + .nfattr_to_tuple = ip_ct_port_nfattr_to_tuple, +#endif +}; + +/* ip_conntrack_proto_gre initialization */ +int __init ip_ct_proto_gre_init(void) +{ + return ip_conntrack_protocol_register(&gre); +} + +void __exit ip_ct_proto_gre_fini(void) +{ + struct list_head *pos, *n; + + /* delete all keymap entries */ + write_lock_bh(&ip_ct_gre_lock); + list_for_each_safe(pos, n, &gre_keymap_list) { + DEBUGP("deleting keymap %p at module unload time\n", pos); + list_del(pos); + kfree(pos); + } + write_unlock_bh(&ip_ct_gre_lock); + + ip_conntrack_protocol_unregister(&gre); +} + +EXPORT_SYMBOL(ip_ct_gre_keymap_add); +EXPORT_SYMBOL(ip_ct_gre_keymap_destroy); diff --git a/net/ipv4/netfilter/ip_nat_helper_pptp.c b/net/ipv4/netfilter/ip_nat_helper_pptp.c new file mode 100644 index 00000000000..3cdd0684d30 --- /dev/null +++ b/net/ipv4/netfilter/ip_nat_helper_pptp.c @@ -0,0 +1,401 @@ +/* + * ip_nat_pptp.c - Version 3.0 + * + * NAT support for PPTP (Point to Point Tunneling Protocol). + * PPTP is a a protocol for creating virtual private networks. + * It is a specification defined by Microsoft and some vendors + * working with Microsoft. PPTP is built on top of a modified + * version of the Internet Generic Routing Encapsulation Protocol. + * GRE is defined in RFC 1701 and RFC 1702. Documentation of + * PPTP can be found in RFC 2637 + * + * (C) 2000-2005 by Harald Welte + * + * Development of this code funded by Astaro AG (http://www.astaro.com/) + * + * TODO: - NAT to a unique tuple, not to TCP source port + * (needs netfilter tuple reservation) + * + * Changes: + * 2002-02-10 - Version 1.3 + * - Use ip_nat_mangle_tcp_packet() because of cloned skb's + * in local connections (Philip Craig ) + * - add checks for magicCookie and pptp version + * - make argument list of pptp_{out,in}bound_packet() shorter + * - move to C99 style initializers + * - print version number at module loadtime + * 2003-09-22 - Version 1.5 + * - use SNATed tcp sourceport as callid, since we get called before + * TCP header is mangled (Philip Craig ) + * 2004-10-22 - Version 2.0 + * - kernel 2.6.x version + * 2005-06-10 - Version 3.0 + * - kernel >= 2.6.11 version, + * funded by Oxcoda NetBox Blue (http://www.netboxblue.com/) + * + */ + +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include + +#define IP_NAT_PPTP_VERSION "3.0" + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Harald Welte "); +MODULE_DESCRIPTION("Netfilter NAT helper module for PPTP"); + + +#if 0 +extern const char *pptp_msg_name[]; +#define DEBUGP(format, args...) printk(KERN_DEBUG "%s:%s: " format, __FILE__, \ + __FUNCTION__, ## args) +#else +#define DEBUGP(format, args...) +#endif + +static void pptp_nat_expected(struct ip_conntrack *ct, + struct ip_conntrack_expect *exp) +{ + struct ip_conntrack *master = ct->master; + struct ip_conntrack_expect *other_exp; + struct ip_conntrack_tuple t; + struct ip_ct_pptp_master *ct_pptp_info; + struct ip_nat_pptp *nat_pptp_info; + + ct_pptp_info = &master->help.ct_pptp_info; + nat_pptp_info = &master->nat.help.nat_pptp_info; + + /* And here goes the grand finale of corrosion... */ + + if (exp->dir == IP_CT_DIR_ORIGINAL) { + DEBUGP("we are PNS->PAC\n"); + /* therefore, build tuple for PAC->PNS */ + t.src.ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.src.ip; + t.src.u.gre.key = htons(master->help.ct_pptp_info.pac_call_id); + t.dst.ip = master->tuplehash[IP_CT_DIR_REPLY].tuple.dst.ip; + t.dst.u.gre.key = htons(master->help.ct_pptp_info.pns_call_id); + t.dst.protonum = IPPROTO_GRE; + } else { + DEBUGP("we are PAC->PNS\n"); + /* build tuple for PNS->PAC */ + t.src.ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.ip; + t.src.u.gre.key = + htons(master->nat.help.nat_pptp_info.pns_call_id); + t.dst.ip = master->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.ip; + t.dst.u.gre.key = + htons(master->nat.help.nat_pptp_info.pac_call_id); + t.dst.protonum = IPPROTO_GRE; + } + + DEBUGP("trying to unexpect other dir: "); + DUMP_TUPLE(&t); + other_exp = ip_conntrack_expect_find(&t); + if (other_exp) { + ip_conntrack_unexpect_related(other_exp); + ip_conntrack_expect_put(other_exp); + DEBUGP("success\n"); + } else { + DEBUGP("not found!\n"); + } + + ip_nat_follow_master(ct, exp); +} + +/* outbound packets == from PNS to PAC */ +static int +pptp_outbound_pkt(struct sk_buff **pskb, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + struct PptpControlHeader *ctlh, + union pptp_ctrl_union *pptpReq) + +{ + struct ip_ct_pptp_master *ct_pptp_info = &ct->help.ct_pptp_info; + struct ip_nat_pptp *nat_pptp_info = &ct->nat.help.nat_pptp_info; + + u_int16_t msg, *cid = NULL, new_callid; + + new_callid = htons(ct_pptp_info->pns_call_id); + + switch (msg = ntohs(ctlh->messageType)) { + case PPTP_OUT_CALL_REQUEST: + cid = &pptpReq->ocreq.callID; + /* FIXME: ideally we would want to reserve a call ID + * here. current netfilter NAT core is not able to do + * this :( For now we use TCP source port. This breaks + * multiple calls within one control session */ + + /* save original call ID in nat_info */ + nat_pptp_info->pns_call_id = ct_pptp_info->pns_call_id; + + /* don't use tcph->source since we are at a DSTmanip + * hook (e.g. PREROUTING) and pkt is not mangled yet */ + new_callid = ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.u.tcp.port; + + /* save new call ID in ct info */ + ct_pptp_info->pns_call_id = ntohs(new_callid); + break; + case PPTP_IN_CALL_REPLY: + cid = &pptpReq->icreq.callID; + break; + case PPTP_CALL_CLEAR_REQUEST: + cid = &pptpReq->clrreq.callID; + break; + default: + DEBUGP("unknown outbound packet 0x%04x:%s\n", msg, + (msg <= PPTP_MSG_MAX)? + pptp_msg_name[msg]:pptp_msg_name[0]); + /* fall through */ + + case PPTP_SET_LINK_INFO: + /* only need to NAT in case PAC is behind NAT box */ + case PPTP_START_SESSION_REQUEST: + case PPTP_START_SESSION_REPLY: + case PPTP_STOP_SESSION_REQUEST: + case PPTP_STOP_SESSION_REPLY: + case PPTP_ECHO_REQUEST: + case PPTP_ECHO_REPLY: + /* no need to alter packet */ + return NF_ACCEPT; + } + + /* only OUT_CALL_REQUEST, IN_CALL_REPLY, CALL_CLEAR_REQUEST pass + * down to here */ + + IP_NF_ASSERT(cid); + + DEBUGP("altering call id from 0x%04x to 0x%04x\n", + ntohs(*cid), ntohs(new_callid)); + + /* mangle packet */ + if (ip_nat_mangle_tcp_packet(pskb, ct, ctinfo, + (void *)cid - ((void *)ctlh - sizeof(struct pptp_pkt_hdr)), + sizeof(new_callid), + (char *)&new_callid, + sizeof(new_callid)) == 0) + return NF_DROP; + + return NF_ACCEPT; +} + +static int +pptp_exp_gre(struct ip_conntrack_expect *expect_orig, + struct ip_conntrack_expect *expect_reply) +{ + struct ip_ct_pptp_master *ct_pptp_info = + &expect_orig->master->help.ct_pptp_info; + struct ip_nat_pptp *nat_pptp_info = + &expect_orig->master->nat.help.nat_pptp_info; + + struct ip_conntrack *ct = expect_orig->master; + + struct ip_conntrack_tuple inv_t; + struct ip_conntrack_tuple *orig_t, *reply_t; + + /* save original PAC call ID in nat_info */ + nat_pptp_info->pac_call_id = ct_pptp_info->pac_call_id; + + /* alter expectation */ + orig_t = &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple; + reply_t = &ct->tuplehash[IP_CT_DIR_REPLY].tuple; + + /* alter expectation for PNS->PAC direction */ + invert_tuplepr(&inv_t, &expect_orig->tuple); + expect_orig->saved_proto.gre.key = htons(nat_pptp_info->pac_call_id); + expect_orig->tuple.src.u.gre.key = htons(nat_pptp_info->pns_call_id); + expect_orig->tuple.dst.u.gre.key = htons(ct_pptp_info->pac_call_id); + inv_t.src.ip = reply_t->src.ip; + inv_t.dst.ip = reply_t->dst.ip; + inv_t.src.u.gre.key = htons(nat_pptp_info->pac_call_id); + inv_t.dst.u.gre.key = htons(ct_pptp_info->pns_call_id); + + if (!ip_conntrack_expect_related(expect_orig)) { + DEBUGP("successfully registered expect\n"); + } else { + DEBUGP("can't expect_related(expect_orig)\n"); + return 1; + } + + /* alter expectation for PAC->PNS direction */ + invert_tuplepr(&inv_t, &expect_reply->tuple); + expect_reply->saved_proto.gre.key = htons(nat_pptp_info->pns_call_id); + expect_reply->tuple.src.u.gre.key = htons(nat_pptp_info->pac_call_id); + expect_reply->tuple.dst.u.gre.key = htons(ct_pptp_info->pns_call_id); + inv_t.src.ip = orig_t->src.ip; + inv_t.dst.ip = orig_t->dst.ip; + inv_t.src.u.gre.key = htons(nat_pptp_info->pns_call_id); + inv_t.dst.u.gre.key = htons(ct_pptp_info->pac_call_id); + + if (!ip_conntrack_expect_related(expect_reply)) { + DEBUGP("successfully registered expect\n"); + } else { + DEBUGP("can't expect_related(expect_reply)\n"); + ip_conntrack_unexpect_related(expect_orig); + return 1; + } + + if (ip_ct_gre_keymap_add(ct, &expect_reply->tuple, 0) < 0) { + DEBUGP("can't register original keymap\n"); + ip_conntrack_unexpect_related(expect_orig); + ip_conntrack_unexpect_related(expect_reply); + return 1; + } + + if (ip_ct_gre_keymap_add(ct, &inv_t, 1) < 0) { + DEBUGP("can't register reply keymap\n"); + ip_conntrack_unexpect_related(expect_orig); + ip_conntrack_unexpect_related(expect_reply); + ip_ct_gre_keymap_destroy(ct); + return 1; + } + + return 0; +} + +/* inbound packets == from PAC to PNS */ +static int +pptp_inbound_pkt(struct sk_buff **pskb, + struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + struct PptpControlHeader *ctlh, + union pptp_ctrl_union *pptpReq) +{ + struct ip_nat_pptp *nat_pptp_info = &ct->nat.help.nat_pptp_info; + u_int16_t msg, new_cid = 0, new_pcid, *pcid = NULL, *cid = NULL; + + int ret = NF_ACCEPT, rv; + + new_pcid = htons(nat_pptp_info->pns_call_id); + + switch (msg = ntohs(ctlh->messageType)) { + case PPTP_OUT_CALL_REPLY: + pcid = &pptpReq->ocack.peersCallID; + cid = &pptpReq->ocack.callID; + break; + case PPTP_IN_CALL_CONNECT: + pcid = &pptpReq->iccon.peersCallID; + break; + case PPTP_IN_CALL_REQUEST: + /* only need to nat in case PAC is behind NAT box */ + break; + case PPTP_WAN_ERROR_NOTIFY: + pcid = &pptpReq->wanerr.peersCallID; + break; + case PPTP_CALL_DISCONNECT_NOTIFY: + pcid = &pptpReq->disc.callID; + break; + case PPTP_SET_LINK_INFO: + pcid = &pptpReq->setlink.peersCallID; + break; + + default: + DEBUGP("unknown inbound packet %s\n", (msg <= PPTP_MSG_MAX)? + pptp_msg_name[msg]:pptp_msg_name[0]); + /* fall through */ + + case PPTP_START_SESSION_REQUEST: + case PPTP_START_SESSION_REPLY: + case PPTP_STOP_SESSION_REQUEST: + case PPTP_STOP_SESSION_REPLY: + case PPTP_ECHO_REQUEST: + case PPTP_ECHO_REPLY: + /* no need to alter packet */ + return NF_ACCEPT; + } + + /* only OUT_CALL_REPLY, IN_CALL_CONNECT, IN_CALL_REQUEST, + * WAN_ERROR_NOTIFY, CALL_DISCONNECT_NOTIFY pass down here */ + + /* mangle packet */ + IP_NF_ASSERT(pcid); + DEBUGP("altering peer call id from 0x%04x to 0x%04x\n", + ntohs(*pcid), ntohs(new_pcid)); + + rv = ip_nat_mangle_tcp_packet(pskb, ct, ctinfo, + (void *)pcid - ((void *)ctlh - sizeof(struct pptp_pkt_hdr)), + sizeof(new_pcid), (char *)&new_pcid, + sizeof(new_pcid)); + if (rv != NF_ACCEPT) + return rv; + + if (new_cid) { + IP_NF_ASSERT(cid); + DEBUGP("altering call id from 0x%04x to 0x%04x\n", + ntohs(*cid), ntohs(new_cid)); + rv = ip_nat_mangle_tcp_packet(pskb, ct, ctinfo, + (void *)cid - ((void *)ctlh - sizeof(struct pptp_pkt_hdr)), + sizeof(new_cid), + (char *)&new_cid, + sizeof(new_cid)); + if (rv != NF_ACCEPT) + return rv; + } + + /* check for earlier return value of 'switch' above */ + if (ret != NF_ACCEPT) + return ret; + + /* great, at least we don't need to resize packets */ + return NF_ACCEPT; +} + + +extern int __init ip_nat_proto_gre_init(void); +extern void __exit ip_nat_proto_gre_fini(void); + +static int __init init(void) +{ + int ret; + + DEBUGP("%s: registering NAT helper\n", __FILE__); + + ret = ip_nat_proto_gre_init(); + if (ret < 0) + return ret; + + BUG_ON(ip_nat_pptp_hook_outbound); + ip_nat_pptp_hook_outbound = &pptp_outbound_pkt; + + BUG_ON(ip_nat_pptp_hook_inbound); + ip_nat_pptp_hook_inbound = &pptp_inbound_pkt; + + BUG_ON(ip_nat_pptp_hook_exp_gre); + ip_nat_pptp_hook_exp_gre = &pptp_exp_gre; + + BUG_ON(ip_nat_pptp_hook_expectfn); + ip_nat_pptp_hook_expectfn = &pptp_nat_expected; + + printk("ip_nat_pptp version %s loaded\n", IP_NAT_PPTP_VERSION); + return 0; +} + +static void __exit fini(void) +{ + DEBUGP("cleanup_module\n" ); + + ip_nat_pptp_hook_expectfn = NULL; + ip_nat_pptp_hook_exp_gre = NULL; + ip_nat_pptp_hook_inbound = NULL; + ip_nat_pptp_hook_outbound = NULL; + + ip_nat_proto_gre_fini(); + /* Make sure noone calls it, meanwhile */ + synchronize_net(); + + printk("ip_nat_pptp version %s unloaded\n", IP_NAT_PPTP_VERSION); +} + +module_init(init); +module_exit(fini); diff --git a/net/ipv4/netfilter/ip_nat_proto_gre.c b/net/ipv4/netfilter/ip_nat_proto_gre.c new file mode 100644 index 00000000000..7c128540167 --- /dev/null +++ b/net/ipv4/netfilter/ip_nat_proto_gre.c @@ -0,0 +1,214 @@ +/* + * ip_nat_proto_gre.c - Version 2.0 + * + * NAT protocol helper module for GRE. + * + * GRE is a generic encapsulation protocol, which is generally not very + * suited for NAT, as it has no protocol-specific part as port numbers. + * + * It has an optional key field, which may help us distinguishing two + * connections between the same two hosts. + * + * GRE is defined in RFC 1701 and RFC 1702, as well as RFC 2784 + * + * PPTP is built on top of a modified version of GRE, and has a mandatory + * field called "CallID", which serves us for the same purpose as the key + * field in plain GRE. + * + * Documentation about PPTP can be found in RFC 2637 + * + * (C) 2000-2005 by Harald Welte + * + * Development of this code funded by Astaro AG (http://www.astaro.com/) + * + */ + +#include +#include +#include +#include +#include +#include +#include + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Harald Welte "); +MODULE_DESCRIPTION("Netfilter NAT protocol helper module for GRE"); + +#if 0 +#define DEBUGP(format, args...) printk(KERN_DEBUG "%s:%s: " format, __FILE__, \ + __FUNCTION__, ## args) +#else +#define DEBUGP(x, args...) +#endif + +/* is key in given range between min and max */ +static int +gre_in_range(const struct ip_conntrack_tuple *tuple, + enum ip_nat_manip_type maniptype, + const union ip_conntrack_manip_proto *min, + const union ip_conntrack_manip_proto *max) +{ + u_int32_t key; + + if (maniptype == IP_NAT_MANIP_SRC) + key = tuple->src.u.gre.key; + else + key = tuple->dst.u.gre.key; + + return ntohl(key) >= ntohl(min->gre.key) + && ntohl(key) <= ntohl(max->gre.key); +} + +/* generate unique tuple ... */ +static int +gre_unique_tuple(struct ip_conntrack_tuple *tuple, + const struct ip_nat_range *range, + enum ip_nat_manip_type maniptype, + const struct ip_conntrack *conntrack) +{ + static u_int16_t key; + u_int16_t *keyptr; + unsigned int min, i, range_size; + + if (maniptype == IP_NAT_MANIP_SRC) + keyptr = &tuple->src.u.gre.key; + else + keyptr = &tuple->dst.u.gre.key; + + if (!(range->flags & IP_NAT_RANGE_PROTO_SPECIFIED)) { + DEBUGP("%p: NATing GRE PPTP\n", conntrack); + min = 1; + range_size = 0xffff; + } else { + min = ntohl(range->min.gre.key); + range_size = ntohl(range->max.gre.key) - min + 1; + } + + DEBUGP("min = %u, range_size = %u\n", min, range_size); + + for (i = 0; i < range_size; i++, key++) { + *keyptr = htonl(min + key % range_size); + if (!ip_nat_used_tuple(tuple, conntrack)) + return 1; + } + + DEBUGP("%p: no NAT mapping\n", conntrack); + + return 0; +} + +/* manipulate a GRE packet according to maniptype */ +static int +gre_manip_pkt(struct sk_buff **pskb, + unsigned int iphdroff, + const struct ip_conntrack_tuple *tuple, + enum ip_nat_manip_type maniptype) +{ + struct gre_hdr *greh; + struct gre_hdr_pptp *pgreh; + struct iphdr *iph = (struct iphdr *)((*pskb)->data + iphdroff); + unsigned int hdroff = iphdroff + iph->ihl*4; + + /* pgreh includes two optional 32bit fields which are not required + * to be there. That's where the magic '8' comes from */ + if (!skb_make_writable(pskb, hdroff + sizeof(*pgreh)-8)) + return 0; + + greh = (void *)(*pskb)->data + hdroff; + pgreh = (struct gre_hdr_pptp *) greh; + + /* we only have destination manip of a packet, since 'source key' + * is not present in the packet itself */ + if (maniptype == IP_NAT_MANIP_DST) { + /* key manipulation is always dest */ + switch (greh->version) { + case 0: + if (!greh->key) { + DEBUGP("can't nat GRE w/o key\n"); + break; + } + if (greh->csum) { + /* FIXME: Never tested this code... */ + *(gre_csum(greh)) = + ip_nat_cheat_check(~*(gre_key(greh)), + tuple->dst.u.gre.key, + *(gre_csum(greh))); + } + *(gre_key(greh)) = tuple->dst.u.gre.key; + break; + case GRE_VERSION_PPTP: + DEBUGP("call_id -> 0x%04x\n", + ntohl(tuple->dst.u.gre.key)); + pgreh->call_id = htons(ntohl(tuple->dst.u.gre.key)); + break; + default: + DEBUGP("can't nat unknown GRE version\n"); + return 0; + break; + } + } + return 1; +} + +/* print out a nat tuple */ +static unsigned int +gre_print(char *buffer, + const struct ip_conntrack_tuple *match, + const struct ip_conntrack_tuple *mask) +{ + unsigned int len = 0; + + if (mask->src.u.gre.key) + len += sprintf(buffer + len, "srckey=0x%x ", + ntohl(match->src.u.gre.key)); + + if (mask->dst.u.gre.key) + len += sprintf(buffer + len, "dstkey=0x%x ", + ntohl(match->src.u.gre.key)); + + return len; +} + +/* print a range of keys */ +static unsigned int +gre_print_range(char *buffer, const struct ip_nat_range *range) +{ + if (range->min.gre.key != 0 + || range->max.gre.key != 0xFFFF) { + if (range->min.gre.key == range->max.gre.key) + return sprintf(buffer, "key 0x%x ", + ntohl(range->min.gre.key)); + else + return sprintf(buffer, "keys 0x%u-0x%u ", + ntohl(range->min.gre.key), + ntohl(range->max.gre.key)); + } else + return 0; +} + +/* nat helper struct */ +static struct ip_nat_protocol gre = { + .name = "GRE", + .protonum = IPPROTO_GRE, + .manip_pkt = gre_manip_pkt, + .in_range = gre_in_range, + .unique_tuple = gre_unique_tuple, + .print = gre_print, + .print_range = gre_print_range, +#if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ + defined(CONFIG_IP_NF_CONNTRACK_NETLINK_MODULE) + .range_to_nfattr = ip_nat_port_range_to_nfattr, + .nfattr_to_range = ip_nat_port_nfattr_to_range, +#endif +}; + +int __init ip_nat_proto_gre_init(void) +{ + return ip_nat_protocol_register(&gre); +} + +void __exit ip_nat_proto_gre_fini(void) +{ + ip_nat_protocol_unregister(&gre); +} From e674d0f38de6109b59dbe30fba8b296a03229b8e Mon Sep 17 00:00:00 2001 From: Yasuyuki Kozakai Date: Mon, 19 Sep 2005 15:34:40 -0700 Subject: [PATCH 047/204] [NETFILTER] ip6tables: remove duplicate code Some IPv6 matches have very similar loops to find IPv6 extension header and we can unify them. This patch introduces ipv6_find_hdr() to do it. I just checked that it can find the target headers in the packet which has dst,hbh,rt,frag,ah,esp headers. Signed-off-by: Yasuyuki Kozakai Signed-off-by: Harald Welte Signed-off-by: David S. Miller --- include/linux/netfilter_ipv6/ip6_tables.h | 3 + net/ipv6/netfilter/ip6_tables.c | 52 ++++++++++++++ net/ipv6/netfilter/ip6t_ah.c | 81 ++------------------- net/ipv6/netfilter/ip6t_dst.c | 88 ++--------------------- net/ipv6/netfilter/ip6t_esp.c | 73 ++----------------- net/ipv6/netfilter/ip6t_frag.c | 88 +++-------------------- net/ipv6/netfilter/ip6t_hbh.c | 88 ++--------------------- net/ipv6/netfilter/ip6t_rt.c | 83 ++------------------- 8 files changed, 93 insertions(+), 463 deletions(-) diff --git a/include/linux/netfilter_ipv6/ip6_tables.h b/include/linux/netfilter_ipv6/ip6_tables.h index 58c72a52dc6..59f70b34e02 100644 --- a/include/linux/netfilter_ipv6/ip6_tables.h +++ b/include/linux/netfilter_ipv6/ip6_tables.h @@ -455,6 +455,9 @@ extern unsigned int ip6t_do_table(struct sk_buff **pskb, /* Check for an extension */ extern int ip6t_ext_hdr(u8 nexthdr); +/* find specified header and get offset to it */ +extern int ipv6_find_hdr(const struct sk_buff *skb, unsigned int *offset, + u8 target); #define IP6T_ALIGN(s) (((s) + (__alignof__(struct ip6t_entry)-1)) & ~(__alignof__(struct ip6t_entry)-1)) diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c index 1cb8adb2787..2da514b16d9 100644 --- a/net/ipv6/netfilter/ip6_tables.c +++ b/net/ipv6/netfilter/ip6_tables.c @@ -1955,6 +1955,57 @@ static void __exit fini(void) #endif } +/* + * find specified header up to transport protocol header. + * If found target header, the offset to the header is set to *offset + * and return 0. otherwise, return -1. + * + * Notes: - non-1st Fragment Header isn't skipped. + * - ESP header isn't skipped. + * - The target header may be trancated. + */ +int ipv6_find_hdr(const struct sk_buff *skb, unsigned int *offset, u8 target) +{ + unsigned int start = (u8*)(skb->nh.ipv6h + 1) - skb->data; + u8 nexthdr = skb->nh.ipv6h->nexthdr; + unsigned int len = skb->len - start; + + while (nexthdr != target) { + struct ipv6_opt_hdr _hdr, *hp; + unsigned int hdrlen; + + if ((!ipv6_ext_hdr(nexthdr)) || nexthdr == NEXTHDR_NONE) + return -1; + hp = skb_header_pointer(skb, start, sizeof(_hdr), &_hdr); + if (hp == NULL) + return -1; + if (nexthdr == NEXTHDR_FRAGMENT) { + unsigned short _frag_off, *fp; + fp = skb_header_pointer(skb, + start+offsetof(struct frag_hdr, + frag_off), + sizeof(_frag_off), + &_frag_off); + if (fp == NULL) + return -1; + + if (ntohs(*fp) & ~0x7) + return -1; + hdrlen = 8; + } else if (nexthdr == NEXTHDR_AUTH) + hdrlen = (hp->hdrlen + 2) << 2; + else + hdrlen = ipv6_optlen(hp); + + nexthdr = hp->nexthdr; + len -= hdrlen; + start += hdrlen; + } + + *offset = start; + return 0; +} + EXPORT_SYMBOL(ip6t_register_table); EXPORT_SYMBOL(ip6t_unregister_table); EXPORT_SYMBOL(ip6t_do_table); @@ -1963,6 +2014,7 @@ EXPORT_SYMBOL(ip6t_unregister_match); EXPORT_SYMBOL(ip6t_register_target); EXPORT_SYMBOL(ip6t_unregister_target); EXPORT_SYMBOL(ip6t_ext_hdr); +EXPORT_SYMBOL(ipv6_find_hdr); module_init(init); module_exit(fini); diff --git a/net/ipv6/netfilter/ip6t_ah.c b/net/ipv6/netfilter/ip6t_ah.c index d5b94f142bb..dde37793d20 100644 --- a/net/ipv6/netfilter/ip6t_ah.c +++ b/net/ipv6/netfilter/ip6t_ah.c @@ -48,92 +48,21 @@ match(const struct sk_buff *skb, unsigned int protoff, int *hotdrop) { - struct ip_auth_hdr *ah = NULL, _ah; + struct ip_auth_hdr *ah, _ah; const struct ip6t_ah *ahinfo = matchinfo; - unsigned int temp; - int len; - u8 nexthdr; unsigned int ptr; unsigned int hdrlen = 0; - /*DEBUGP("IPv6 AH entered\n");*/ - /* if (opt->auth == 0) return 0; - * It does not filled on output */ - - /* type of the 1st exthdr */ - nexthdr = skb->nh.ipv6h->nexthdr; - /* pointer to the 1st exthdr */ - ptr = sizeof(struct ipv6hdr); - /* available length */ - len = skb->len - ptr; - temp = 0; - - while (ip6t_ext_hdr(nexthdr)) { - struct ipv6_opt_hdr _hdr, *hp; - - DEBUGP("ipv6_ah header iteration \n"); - - /* Is there enough space for the next ext header? */ - if (len < sizeof(struct ipv6_opt_hdr)) - return 0; - /* No more exthdr -> evaluate */ - if (nexthdr == NEXTHDR_NONE) - break; - /* ESP -> evaluate */ - if (nexthdr == NEXTHDR_ESP) - break; - - hp = skb_header_pointer(skb, ptr, sizeof(_hdr), &_hdr); - BUG_ON(hp == NULL); - - /* Calculate the header length */ - if (nexthdr == NEXTHDR_FRAGMENT) - hdrlen = 8; - else if (nexthdr == NEXTHDR_AUTH) - hdrlen = (hp->hdrlen+2)<<2; - else - hdrlen = ipv6_optlen(hp); - - /* AH -> evaluate */ - if (nexthdr == NEXTHDR_AUTH) { - temp |= MASK_AH; - break; - } - - - /* set the flag */ - switch (nexthdr) { - case NEXTHDR_HOP: - case NEXTHDR_ROUTING: - case NEXTHDR_FRAGMENT: - case NEXTHDR_AUTH: - case NEXTHDR_DEST: - break; - default: - DEBUGP("ipv6_ah match: unknown nextheader %u\n",nexthdr); - return 0; - } - - nexthdr = hp->nexthdr; - len -= hdrlen; - ptr += hdrlen; - if (ptr > skb->len) { - DEBUGP("ipv6_ah: new pointer too large! \n"); - break; - } - } - - /* AH header not found */ - if (temp != MASK_AH) + if (ipv6_find_hdr(skb, &ptr, NEXTHDR_AUTH) < 0) return 0; - if (len < sizeof(struct ip_auth_hdr)){ + ah = skb_header_pointer(skb, ptr, sizeof(_ah), &_ah); + if (ah == NULL) { *hotdrop = 1; return 0; } - ah = skb_header_pointer(skb, ptr, sizeof(_ah), &_ah); - BUG_ON(ah == NULL); + hdrlen = (ah->hdrlen + 2) << 2; DEBUGP("IPv6 AH LEN %u %u ", hdrlen, ah->hdrlen); DEBUGP("RES %04X ", ah->reserved); diff --git a/net/ipv6/netfilter/ip6t_dst.c b/net/ipv6/netfilter/ip6t_dst.c index 540925e4a7a..c450a635e54 100644 --- a/net/ipv6/netfilter/ip6t_dst.c +++ b/net/ipv6/netfilter/ip6t_dst.c @@ -63,8 +63,6 @@ match(const struct sk_buff *skb, struct ipv6_opt_hdr _optsh, *oh; const struct ip6t_opts *optinfo = matchinfo; unsigned int temp; - unsigned int len; - u8 nexthdr; unsigned int ptr; unsigned int hdrlen = 0; unsigned int ret = 0; @@ -72,97 +70,25 @@ match(const struct sk_buff *skb, u8 _optlen, *lp = NULL; unsigned int optlen; - /* type of the 1st exthdr */ - nexthdr = skb->nh.ipv6h->nexthdr; - /* pointer to the 1st exthdr */ - ptr = sizeof(struct ipv6hdr); - /* available length */ - len = skb->len - ptr; - temp = 0; - - while (ip6t_ext_hdr(nexthdr)) { - struct ipv6_opt_hdr _hdr, *hp; - - DEBUGP("ipv6_opts header iteration \n"); - - /* Is there enough space for the next ext header? */ - if (len < (int)sizeof(struct ipv6_opt_hdr)) - return 0; - /* No more exthdr -> evaluate */ - if (nexthdr == NEXTHDR_NONE) { - break; - } - /* ESP -> evaluate */ - if (nexthdr == NEXTHDR_ESP) { - break; - } - - hp = skb_header_pointer(skb, ptr, sizeof(_hdr), &_hdr); - BUG_ON(hp == NULL); - - /* Calculate the header length */ - if (nexthdr == NEXTHDR_FRAGMENT) { - hdrlen = 8; - } else if (nexthdr == NEXTHDR_AUTH) - hdrlen = (hp->hdrlen+2)<<2; - else - hdrlen = ipv6_optlen(hp); - - /* OPTS -> evaluate */ #if HOPBYHOP - if (nexthdr == NEXTHDR_HOP) { - temp |= MASK_HOPOPTS; + if (ipv6_find_hdr(skb, &ptr, NEXTHDR_HOP) < 0) #else - if (nexthdr == NEXTHDR_DEST) { - temp |= MASK_DSTOPTS; + if (ipv6_find_hdr(skb, &ptr, NEXTHDR_DEST) < 0) #endif - break; - } + return 0; - - /* set the flag */ - switch (nexthdr){ - case NEXTHDR_HOP: - case NEXTHDR_ROUTING: - case NEXTHDR_FRAGMENT: - case NEXTHDR_AUTH: - case NEXTHDR_DEST: - break; - default: - DEBUGP("ipv6_opts match: unknown nextheader %u\n",nexthdr); - return 0; - break; - } - - nexthdr = hp->nexthdr; - len -= hdrlen; - ptr += hdrlen; - if ( ptr > skb->len ) { - DEBUGP("ipv6_opts: new pointer is too large! \n"); - break; - } - } - - /* OPTIONS header not found */ -#if HOPBYHOP - if ( temp != MASK_HOPOPTS ) return 0; -#else - if ( temp != MASK_DSTOPTS ) return 0; -#endif - - if (len < (int)sizeof(struct ipv6_opt_hdr)){ + oh = skb_header_pointer(skb, ptr, sizeof(_optsh), &_optsh); + if (oh == NULL){ *hotdrop = 1; return 0; } - if (len < hdrlen){ + hdrlen = ipv6_optlen(oh); + if (skb->len - ptr < hdrlen){ /* Packet smaller than it's length field */ return 0; } - oh = skb_header_pointer(skb, ptr, sizeof(_optsh), &_optsh); - BUG_ON(oh == NULL); - DEBUGP("IPv6 OPTS LEN %u %u ", hdrlen, oh->hdrlen); DEBUGP("len %02X %04X %02X ", diff --git a/net/ipv6/netfilter/ip6t_esp.c b/net/ipv6/netfilter/ip6t_esp.c index e39dd236fd8..24bc0cde43a 100644 --- a/net/ipv6/netfilter/ip6t_esp.c +++ b/net/ipv6/netfilter/ip6t_esp.c @@ -48,87 +48,22 @@ match(const struct sk_buff *skb, unsigned int protoff, int *hotdrop) { - struct ip_esp_hdr _esp, *eh = NULL; + struct ip_esp_hdr _esp, *eh; const struct ip6t_esp *espinfo = matchinfo; - unsigned int temp; - int len; - u8 nexthdr; unsigned int ptr; /* Make sure this isn't an evil packet */ /*DEBUGP("ipv6_esp entered \n");*/ - /* type of the 1st exthdr */ - nexthdr = skb->nh.ipv6h->nexthdr; - /* pointer to the 1st exthdr */ - ptr = sizeof(struct ipv6hdr); - /* available length */ - len = skb->len - ptr; - temp = 0; - - while (ip6t_ext_hdr(nexthdr)) { - struct ipv6_opt_hdr _hdr, *hp; - int hdrlen; - - DEBUGP("ipv6_esp header iteration \n"); - - /* Is there enough space for the next ext header? */ - if (len < sizeof(struct ipv6_opt_hdr)) - return 0; - /* No more exthdr -> evaluate */ - if (nexthdr == NEXTHDR_NONE) - break; - /* ESP -> evaluate */ - if (nexthdr == NEXTHDR_ESP) { - temp |= MASK_ESP; - break; - } - - hp = skb_header_pointer(skb, ptr, sizeof(_hdr), &_hdr); - BUG_ON(hp == NULL); - - /* Calculate the header length */ - if (nexthdr == NEXTHDR_FRAGMENT) - hdrlen = 8; - else if (nexthdr == NEXTHDR_AUTH) - hdrlen = (hp->hdrlen+2)<<2; - else - hdrlen = ipv6_optlen(hp); - - /* set the flag */ - switch (nexthdr) { - case NEXTHDR_HOP: - case NEXTHDR_ROUTING: - case NEXTHDR_FRAGMENT: - case NEXTHDR_AUTH: - case NEXTHDR_DEST: - break; - default: - DEBUGP("ipv6_esp match: unknown nextheader %u\n",nexthdr); - return 0; - } - - nexthdr = hp->nexthdr; - len -= hdrlen; - ptr += hdrlen; - if (ptr > skb->len) { - DEBUGP("ipv6_esp: new pointer too large! \n"); - break; - } - } - - /* ESP header not found */ - if (temp != MASK_ESP) + if (ipv6_find_hdr(skb, &ptr, NEXTHDR_ESP) < 0) return 0; - if (len < sizeof(struct ip_esp_hdr)) { + eh = skb_header_pointer(skb, ptr, sizeof(_esp), &_esp); + if (eh == NULL) { *hotdrop = 1; return 0; } - eh = skb_header_pointer(skb, ptr, sizeof(_esp), &_esp); - BUG_ON(eh == NULL); - DEBUGP("IPv6 ESP SPI %u %08X\n", ntohl(eh->spi), ntohl(eh->spi)); return (eh != NULL) diff --git a/net/ipv6/netfilter/ip6t_frag.c b/net/ipv6/netfilter/ip6t_frag.c index 4bfa30a9bc8..085d5f8eea2 100644 --- a/net/ipv6/netfilter/ip6t_frag.c +++ b/net/ipv6/netfilter/ip6t_frag.c @@ -48,90 +48,18 @@ match(const struct sk_buff *skb, unsigned int protoff, int *hotdrop) { - struct frag_hdr _frag, *fh = NULL; + struct frag_hdr _frag, *fh; const struct ip6t_frag *fraginfo = matchinfo; - unsigned int temp; - int len; - u8 nexthdr; unsigned int ptr; - unsigned int hdrlen = 0; - /* type of the 1st exthdr */ - nexthdr = skb->nh.ipv6h->nexthdr; - /* pointer to the 1st exthdr */ - ptr = sizeof(struct ipv6hdr); - /* available length */ - len = skb->len - ptr; - temp = 0; + if (ipv6_find_hdr(skb, &ptr, NEXTHDR_FRAGMENT) < 0) + return 0; - while (ip6t_ext_hdr(nexthdr)) { - struct ipv6_opt_hdr _hdr, *hp; - - DEBUGP("ipv6_frag header iteration \n"); - - /* Is there enough space for the next ext header? */ - if (len < (int)sizeof(struct ipv6_opt_hdr)) - return 0; - /* No more exthdr -> evaluate */ - if (nexthdr == NEXTHDR_NONE) { - break; - } - /* ESP -> evaluate */ - if (nexthdr == NEXTHDR_ESP) { - break; - } - - hp = skb_header_pointer(skb, ptr, sizeof(_hdr), &_hdr); - BUG_ON(hp == NULL); - - /* Calculate the header length */ - if (nexthdr == NEXTHDR_FRAGMENT) { - hdrlen = 8; - } else if (nexthdr == NEXTHDR_AUTH) - hdrlen = (hp->hdrlen+2)<<2; - else - hdrlen = ipv6_optlen(hp); - - /* FRAG -> evaluate */ - if (nexthdr == NEXTHDR_FRAGMENT) { - temp |= MASK_FRAGMENT; - break; - } - - - /* set the flag */ - switch (nexthdr){ - case NEXTHDR_HOP: - case NEXTHDR_ROUTING: - case NEXTHDR_FRAGMENT: - case NEXTHDR_AUTH: - case NEXTHDR_DEST: - break; - default: - DEBUGP("ipv6_frag match: unknown nextheader %u\n",nexthdr); - return 0; - break; - } - - nexthdr = hp->nexthdr; - len -= hdrlen; - ptr += hdrlen; - if ( ptr > skb->len ) { - DEBUGP("ipv6_frag: new pointer too large! \n"); - break; - } - } - - /* FRAG header not found */ - if ( temp != MASK_FRAGMENT ) return 0; - - if (len < sizeof(struct frag_hdr)){ - *hotdrop = 1; - return 0; - } - - fh = skb_header_pointer(skb, ptr, sizeof(_frag), &_frag); - BUG_ON(fh == NULL); + fh = skb_header_pointer(skb, ptr, sizeof(_frag), &_frag); + if (fh == NULL){ + *hotdrop = 1; + return 0; + } DEBUGP("INFO %04X ", fh->frag_off); DEBUGP("OFFSET %04X ", ntohs(fh->frag_off) & ~0x7); diff --git a/net/ipv6/netfilter/ip6t_hbh.c b/net/ipv6/netfilter/ip6t_hbh.c index 27f3650d127..1d09485111d 100644 --- a/net/ipv6/netfilter/ip6t_hbh.c +++ b/net/ipv6/netfilter/ip6t_hbh.c @@ -63,8 +63,6 @@ match(const struct sk_buff *skb, struct ipv6_opt_hdr _optsh, *oh; const struct ip6t_opts *optinfo = matchinfo; unsigned int temp; - unsigned int len; - u8 nexthdr; unsigned int ptr; unsigned int hdrlen = 0; unsigned int ret = 0; @@ -72,97 +70,25 @@ match(const struct sk_buff *skb, u8 _optlen, *lp = NULL; unsigned int optlen; - /* type of the 1st exthdr */ - nexthdr = skb->nh.ipv6h->nexthdr; - /* pointer to the 1st exthdr */ - ptr = sizeof(struct ipv6hdr); - /* available length */ - len = skb->len - ptr; - temp = 0; - - while (ip6t_ext_hdr(nexthdr)) { - struct ipv6_opt_hdr _hdr, *hp; - - DEBUGP("ipv6_opts header iteration \n"); - - /* Is there enough space for the next ext header? */ - if (len < (int)sizeof(struct ipv6_opt_hdr)) - return 0; - /* No more exthdr -> evaluate */ - if (nexthdr == NEXTHDR_NONE) { - break; - } - /* ESP -> evaluate */ - if (nexthdr == NEXTHDR_ESP) { - break; - } - - hp = skb_header_pointer(skb, ptr, sizeof(_hdr), &_hdr); - BUG_ON(hp == NULL); - - /* Calculate the header length */ - if (nexthdr == NEXTHDR_FRAGMENT) { - hdrlen = 8; - } else if (nexthdr == NEXTHDR_AUTH) - hdrlen = (hp->hdrlen+2)<<2; - else - hdrlen = ipv6_optlen(hp); - - /* OPTS -> evaluate */ #if HOPBYHOP - if (nexthdr == NEXTHDR_HOP) { - temp |= MASK_HOPOPTS; + if (ipv6_find_hdr(skb, &ptr, NEXTHDR_HOP) < 0) #else - if (nexthdr == NEXTHDR_DEST) { - temp |= MASK_DSTOPTS; + if (ipv6_find_hdr(skb, &ptr, NEXTHDR_DEST) < 0) #endif - break; - } + return 0; - - /* set the flag */ - switch (nexthdr){ - case NEXTHDR_HOP: - case NEXTHDR_ROUTING: - case NEXTHDR_FRAGMENT: - case NEXTHDR_AUTH: - case NEXTHDR_DEST: - break; - default: - DEBUGP("ipv6_opts match: unknown nextheader %u\n",nexthdr); - return 0; - break; - } - - nexthdr = hp->nexthdr; - len -= hdrlen; - ptr += hdrlen; - if ( ptr > skb->len ) { - DEBUGP("ipv6_opts: new pointer is too large! \n"); - break; - } - } - - /* OPTIONS header not found */ -#if HOPBYHOP - if ( temp != MASK_HOPOPTS ) return 0; -#else - if ( temp != MASK_DSTOPTS ) return 0; -#endif - - if (len < (int)sizeof(struct ipv6_opt_hdr)){ + oh = skb_header_pointer(skb, ptr, sizeof(_optsh), &_optsh); + if (oh == NULL){ *hotdrop = 1; return 0; } - if (len < hdrlen){ + hdrlen = ipv6_optlen(oh); + if (skb->len - ptr < hdrlen){ /* Packet smaller than it's length field */ return 0; } - oh = skb_header_pointer(skb, ptr, sizeof(_optsh), &_optsh); - BUG_ON(oh == NULL); - DEBUGP("IPv6 OPTS LEN %u %u ", hdrlen, oh->hdrlen); DEBUGP("len %02X %04X %02X ", diff --git a/net/ipv6/netfilter/ip6t_rt.c b/net/ipv6/netfilter/ip6t_rt.c index 2bb670037df..beb2fd5cebb 100644 --- a/net/ipv6/netfilter/ip6t_rt.c +++ b/net/ipv6/netfilter/ip6t_rt.c @@ -50,98 +50,29 @@ match(const struct sk_buff *skb, unsigned int protoff, int *hotdrop) { - struct ipv6_rt_hdr _route, *rh = NULL; + struct ipv6_rt_hdr _route, *rh; const struct ip6t_rt *rtinfo = matchinfo; unsigned int temp; - unsigned int len; - u8 nexthdr; unsigned int ptr; unsigned int hdrlen = 0; unsigned int ret = 0; struct in6_addr *ap, _addr; - /* type of the 1st exthdr */ - nexthdr = skb->nh.ipv6h->nexthdr; - /* pointer to the 1st exthdr */ - ptr = sizeof(struct ipv6hdr); - /* available length */ - len = skb->len - ptr; - temp = 0; + if (ipv6_find_hdr(skb, &ptr, NEXTHDR_ROUTING) < 0) + return 0; - while (ip6t_ext_hdr(nexthdr)) { - struct ipv6_opt_hdr _hdr, *hp; - - DEBUGP("ipv6_rt header iteration \n"); - - /* Is there enough space for the next ext header? */ - if (len < (int)sizeof(struct ipv6_opt_hdr)) - return 0; - /* No more exthdr -> evaluate */ - if (nexthdr == NEXTHDR_NONE) { - break; - } - /* ESP -> evaluate */ - if (nexthdr == NEXTHDR_ESP) { - break; - } - - hp = skb_header_pointer(skb, ptr, sizeof(_hdr), &_hdr); - BUG_ON(hp == NULL); - - /* Calculate the header length */ - if (nexthdr == NEXTHDR_FRAGMENT) { - hdrlen = 8; - } else if (nexthdr == NEXTHDR_AUTH) - hdrlen = (hp->hdrlen+2)<<2; - else - hdrlen = ipv6_optlen(hp); - - /* ROUTING -> evaluate */ - if (nexthdr == NEXTHDR_ROUTING) { - temp |= MASK_ROUTING; - break; - } - - - /* set the flag */ - switch (nexthdr){ - case NEXTHDR_HOP: - case NEXTHDR_ROUTING: - case NEXTHDR_FRAGMENT: - case NEXTHDR_AUTH: - case NEXTHDR_DEST: - break; - default: - DEBUGP("ipv6_rt match: unknown nextheader %u\n",nexthdr); - return 0; - break; - } - - nexthdr = hp->nexthdr; - len -= hdrlen; - ptr += hdrlen; - if ( ptr > skb->len ) { - DEBUGP("ipv6_rt: new pointer is too large! \n"); - break; - } - } - - /* ROUTING header not found */ - if ( temp != MASK_ROUTING ) return 0; - - if (len < (int)sizeof(struct ipv6_rt_hdr)){ + rh = skb_header_pointer(skb, ptr, sizeof(_route), &_route); + if (rh == NULL){ *hotdrop = 1; return 0; } - if (len < hdrlen){ + hdrlen = ipv6_optlen(rh); + if (skb->len - ptr < hdrlen){ /* Pcket smaller than its length field */ return 0; } - rh = skb_header_pointer(skb, ptr, sizeof(_route), &_route); - BUG_ON(rh == NULL); - DEBUGP("IPv6 RT LEN %u %u ", hdrlen, rh->hdrlen); DEBUGP("TYPE %04X ", rh->type); DEBUGP("SGS_LEFT %u %02X\n", rh->segments_left, rh->segments_left); From a41bc00234a0a2ccaa99a194341ae108ae17ddc8 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Mon, 19 Sep 2005 15:35:31 -0700 Subject: [PATCH 048/204] [NETFILTER]: Rename misnamed function Both __ip_conntrack_expect_find and ip_conntrack_expect_find_get take a reference to the expectation, the difference is that callers of __ip_conntrack_expect_find must hold ip_conntrack_lock. Signed-off-by: Patrick McHardy Signed-off-by: Harald Welte Signed-off-by: David S. Miller --- include/linux/netfilter_ipv4/ip_conntrack.h | 2 +- net/ipv4/netfilter/ip_conntrack_core.c | 2 +- net/ipv4/netfilter/ip_conntrack_netlink.c | 4 ++-- net/ipv4/netfilter/ip_conntrack_standalone.c | 4 ++-- 4 files changed, 6 insertions(+), 6 deletions(-) diff --git a/include/linux/netfilter_ipv4/ip_conntrack.h b/include/linux/netfilter_ipv4/ip_conntrack.h index 2df446c952e..bace72a76cc 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack.h +++ b/include/linux/netfilter_ipv4/ip_conntrack.h @@ -384,7 +384,7 @@ extern struct ip_conntrack_expect * __ip_conntrack_expect_find(const struct ip_conntrack_tuple *tuple); extern struct ip_conntrack_expect * -ip_conntrack_expect_find_get(const struct ip_conntrack_tuple *tuple); +ip_conntrack_expect_find(const struct ip_conntrack_tuple *tuple); extern struct ip_conntrack_tuple_hash * __ip_conntrack_find(const struct ip_conntrack_tuple *tuple, diff --git a/net/ipv4/netfilter/ip_conntrack_core.c b/net/ipv4/netfilter/ip_conntrack_core.c index f8cd8e42961..c1f82e0c81c 100644 --- a/net/ipv4/netfilter/ip_conntrack_core.c +++ b/net/ipv4/netfilter/ip_conntrack_core.c @@ -233,7 +233,7 @@ __ip_conntrack_expect_find(const struct ip_conntrack_tuple *tuple) /* Just find a expectation corresponding to a tuple. */ struct ip_conntrack_expect * -ip_conntrack_expect_find_get(const struct ip_conntrack_tuple *tuple) +ip_conntrack_expect_find(const struct ip_conntrack_tuple *tuple) { struct ip_conntrack_expect *i; diff --git a/net/ipv4/netfilter/ip_conntrack_netlink.c b/net/ipv4/netfilter/ip_conntrack_netlink.c index 15aef356474..b08a432efcf 100644 --- a/net/ipv4/netfilter/ip_conntrack_netlink.c +++ b/net/ipv4/netfilter/ip_conntrack_netlink.c @@ -1270,7 +1270,7 @@ ctnetlink_get_expect(struct sock *ctnl, struct sk_buff *skb, if (err < 0) return err; - exp = ip_conntrack_expect_find_get(&tuple); + exp = ip_conntrack_expect_find(&tuple); if (!exp) return -ENOENT; @@ -1318,7 +1318,7 @@ ctnetlink_del_expect(struct sock *ctnl, struct sk_buff *skb, return err; /* bump usage count to 2 */ - exp = ip_conntrack_expect_find_get(&tuple); + exp = ip_conntrack_expect_find(&tuple); if (!exp) return -ENOENT; diff --git a/net/ipv4/netfilter/ip_conntrack_standalone.c b/net/ipv4/netfilter/ip_conntrack_standalone.c index ae3e3e655db..d3c7808010e 100644 --- a/net/ipv4/netfilter/ip_conntrack_standalone.c +++ b/net/ipv4/netfilter/ip_conntrack_standalone.c @@ -993,11 +993,11 @@ EXPORT_SYMBOL(ip_ct_refresh_acct); EXPORT_SYMBOL(ip_conntrack_expect_alloc); EXPORT_SYMBOL(ip_conntrack_expect_put); -EXPORT_SYMBOL_GPL(ip_conntrack_expect_find_get); +EXPORT_SYMBOL_GPL(__ip_conntrack_expect_find); +EXPORT_SYMBOL_GPL(ip_conntrack_expect_find); EXPORT_SYMBOL(ip_conntrack_expect_related); EXPORT_SYMBOL(ip_conntrack_unexpect_related); EXPORT_SYMBOL_GPL(ip_conntrack_expect_list); -EXPORT_SYMBOL_GPL(__ip_conntrack_expect_find); EXPORT_SYMBOL_GPL(ip_ct_unlink_expect); EXPORT_SYMBOL(ip_conntrack_tuple_taken); From 8922bc93aa78a202c2b68b2de19b71329cb321e1 Mon Sep 17 00:00:00 2001 From: Harald Welte Date: Mon, 19 Sep 2005 15:35:57 -0700 Subject: [PATCH 049/204] [NETFILTER]: Export ip_nat_port_{nfattr_to_range,range_to_nfattr} Those exports are needed by the PPTP helper following in the next couple of changes. Signed-off-by: Harald Welte Signed-off-by: David S. Miller --- net/ipv4/netfilter/ip_nat_core.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/ipv4/netfilter/ip_nat_core.c b/net/ipv4/netfilter/ip_nat_core.c index 1adedb743f6..c3ea891d38e 100644 --- a/net/ipv4/netfilter/ip_nat_core.c +++ b/net/ipv4/netfilter/ip_nat_core.c @@ -578,6 +578,8 @@ ip_nat_port_nfattr_to_range(struct nfattr *tb[], struct ip_nat_range *range) return ret; } +EXPORT_SYMBOL_GPL(ip_nat_port_nfattr_to_range); +EXPORT_SYMBOL_GPL(ip_nat_port_range_to_nfattr); #endif int __init ip_nat_init(void) From 2cf655cd65888e9fed0803d77e9e4f7d1db674cc Mon Sep 17 00:00:00 2001 From: Krzysztof Halasa Date: Mon, 19 Sep 2005 15:39:32 -0700 Subject: [PATCH 050/204] [WAN] hdlc_cisco: Fix regression introduced by skb->tail changes. The following commit breaks cisco mode with my WAN drivers: author David S. Miller Tue, 28 Jun 2005 22:25:31 +0000 (15:25 -0700) commit 689be43945e9ca7dd704522e55af1b8a73a994d3 "[NET]: Remove gratuitous use of skb->tail in network drivers." The following patch fixes it - please apply (cisco_hard_header does skb_push(4 bytes)). Signed-off-by: David S. Miller --- drivers/net/wan/hdlc_cisco.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/wan/hdlc_cisco.c b/drivers/net/wan/hdlc_cisco.c index 48c03c11cd9..a01efa6d5c6 100644 --- a/drivers/net/wan/hdlc_cisco.c +++ b/drivers/net/wan/hdlc_cisco.c @@ -72,7 +72,7 @@ static void cisco_keepalive_send(struct net_device *dev, u32 type, } skb_reserve(skb, 4); cisco_hard_header(skb, dev, CISCO_KEEPALIVE, NULL, NULL, 0); - data = (cisco_packet*)skb->data; + data = (cisco_packet*)(skb->data + 4); data->type = htonl(type); data->par1 = htonl(par1); From 3c3f8f25c177e4f9e4e00bcc1b90b28b1be37937 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Mon, 19 Sep 2005 15:41:28 -0700 Subject: [PATCH 051/204] [8021Q]: Add endian annotations. Signed-off-by: Alexey Dobriyan Signed-off-by: David S. Miller --- include/linux/if_vlan.h | 8 ++++---- net/8021q/vlan_dev.c | 2 +- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/include/linux/if_vlan.h b/include/linux/if_vlan.h index 17d0c0d40b0..eef0876d830 100644 --- a/include/linux/if_vlan.h +++ b/include/linux/if_vlan.h @@ -42,8 +42,8 @@ struct hlist_node; struct vlan_ethhdr { unsigned char h_dest[ETH_ALEN]; /* destination eth addr */ unsigned char h_source[ETH_ALEN]; /* source ether addr */ - unsigned short h_vlan_proto; /* Should always be 0x8100 */ - unsigned short h_vlan_TCI; /* Encapsulates priority and VLAN ID */ + __be16 h_vlan_proto; /* Should always be 0x8100 */ + __be16 h_vlan_TCI; /* Encapsulates priority and VLAN ID */ unsigned short h_vlan_encapsulated_proto; /* packet type ID field (or len) */ }; @@ -55,8 +55,8 @@ static inline struct vlan_ethhdr *vlan_eth_hdr(const struct sk_buff *skb) } struct vlan_hdr { - unsigned short h_vlan_TCI; /* Encapsulates priority and VLAN ID */ - unsigned short h_vlan_encapsulated_proto; /* packet type ID field (or len) */ + __be16 h_vlan_TCI; /* Encapsulates priority and VLAN ID */ + __be16 h_vlan_encapsulated_proto; /* packet type ID field (or len) */ }; #define VLAN_VID_MASK 0xfff diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c index 145f5cde96c..b7486488967 100644 --- a/net/8021q/vlan_dev.c +++ b/net/8021q/vlan_dev.c @@ -120,7 +120,7 @@ int vlan_skb_recv(struct sk_buff *skb, struct net_device *dev, unsigned short vid; struct net_device_stats *stats; unsigned short vlan_TCI; - unsigned short proto; + __be16 proto; /* vlan_TCI = ntohs(get_unaligned(&vhdr->h_vlan_TCI)); */ vlan_TCI = ntohs(vhdr->h_vlan_TCI); From e14c3caf605dfd29bd1aac3097e39db94afc9f07 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Mon, 19 Sep 2005 18:18:38 -0700 Subject: [PATCH 052/204] [TCP]: Handle SACK'd packets properly in tcp_fragment(). The problem is that we're now calling tcp_fragment() in a context where the packets might be marked as SACKED_ACKED or SACKED_RETRANS. This was not possible before as you never retransmitted packets that are so marked. Because of this, we need to adjust sacked_out and retrans_out in tcp_fragment(). This is exactly what the following patch does. We also need to preserve the SACKED_ACKED/SACKED_RETRANS marking if they exist. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller --- net/ipv4/tcp_output.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index b018e31b653..5dd6dd7d091 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -461,9 +461,7 @@ int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned int mss flags = TCP_SKB_CB(skb)->flags; TCP_SKB_CB(skb)->flags = flags & ~(TCPCB_FLAG_FIN|TCPCB_FLAG_PSH); TCP_SKB_CB(buff)->flags = flags; - TCP_SKB_CB(buff)->sacked = - (TCP_SKB_CB(skb)->sacked & - (TCPCB_LOST | TCPCB_EVER_RETRANS | TCPCB_AT_TAIL)); + TCP_SKB_CB(buff)->sacked = TCP_SKB_CB(skb)->sacked; TCP_SKB_CB(skb)->sacked &= ~TCPCB_AT_TAIL; if (!skb_shinfo(skb)->nr_frags && skb->ip_summed != CHECKSUM_HW) { @@ -501,6 +499,12 @@ int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned int mss tcp_skb_pcount(buff); tp->packets_out -= diff; + + if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_ACKED) + tp->sacked_out -= diff; + if (TCP_SKB_CB(skb)->sacked & TCPCB_SACKED_RETRANS) + tp->retrans_out -= diff; + if (TCP_SKB_CB(skb)->sacked & TCPCB_LOST) { tp->lost_out -= diff; tp->left_out -= diff; From 6d1cfe3f1752f17e297df60c8bcc6cd6e0a58449 Mon Sep 17 00:00:00 2001 From: Mark J Cox Date: Mon, 19 Sep 2005 17:55:30 -0700 Subject: [PATCH 053/204] [PATCH] raw_sendmsg DoS on 2.6 Fix unchecked __get_user that could be tricked into generating a memory read on an arbitrary address. The result of the read is not returned directly but you may be able to divine some information about it, or use the read to cause a crash on some architectures by reading hardware state. CAN-2004-2492. Fix from Al Viro, ack from Dave Miller. Signed-off-by: Linus Torvalds --- net/ipv4/raw.c | 2 +- net/ipv6/raw.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/net/ipv4/raw.c b/net/ipv4/raw.c index 304bb0a1d4f..4b0d7e4d626 100644 --- a/net/ipv4/raw.c +++ b/net/ipv4/raw.c @@ -361,7 +361,7 @@ static void raw_probe_proto_opt(struct flowi *fl, struct msghdr *msg) if (type && code) { get_user(fl->fl_icmp_type, type); - __get_user(fl->fl_icmp_code, code); + get_user(fl->fl_icmp_code, code); probed = 1; } break; diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 5aa3691c578..a1265a320b1 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -627,7 +627,7 @@ static void rawv6_probe_proto_opt(struct flowi *fl, struct msghdr *msg) if (type && code) { get_user(fl->fl_icmp_type, type); - __get_user(fl->fl_icmp_code, code); + get_user(fl->fl_icmp_code, code); probed = 1; } break; From f805fbdaacf4367ce566743a665622387768ac0d Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 19 Sep 2005 19:54:29 -0700 Subject: [PATCH 054/204] Make fsnotify possibly work better for the inode removal case Checking i_nlink is dubious, but the alternatives look even less appetizing. Signed-off-by: Linus Torvalds --- fs/dcache.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/fs/dcache.c b/fs/dcache.c index 7376b61269f..fb10386c59b 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -102,7 +102,8 @@ static inline void dentry_iput(struct dentry * dentry) list_del_init(&dentry->d_alias); spin_unlock(&dentry->d_lock); spin_unlock(&dcache_lock); - fsnotify_inoderemove(inode); + if (!inode->i_nlink) + fsnotify_inoderemove(inode); if (dentry->d_op && dentry->d_op->d_iput) dentry->d_op->d_iput(dentry, inode); else From ff171d8f66a7fe1a000e610e9de11224749f9a22 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 19 Sep 2005 19:56:06 -0700 Subject: [PATCH 055/204] [SPARC64]: Handle little-endian unaligned loads/stores correctly. Because we use byte loads/stores to cons up the value in and out of registers, we can't expect the ASI endianness setting to take care of this for us. So do it by hand. This case is triggered by drivers/block/aoe/aoecmd.c in the ataid_complete() function where it goes: /* word 100: number lba48 sectors */ ssize = le64_to_cpup((__le64 *) &id[100<<1]); This &id[100<<1] address is 4 byte, rather than 8 byte aligned, thus triggering the unaligned exception. Signed-off-by: David S. Miller --- arch/sparc64/kernel/una_asm.S | 2 +- arch/sparc64/kernel/unaligned.c | 64 +++++++++++++++++++++++++++++---- 2 files changed, 58 insertions(+), 8 deletions(-) diff --git a/arch/sparc64/kernel/una_asm.S b/arch/sparc64/kernel/una_asm.S index cbb40585253..da48400bcc9 100644 --- a/arch/sparc64/kernel/una_asm.S +++ b/arch/sparc64/kernel/una_asm.S @@ -17,7 +17,7 @@ kernel_unaligned_trap_fault: __do_int_store: rd %asi, %o4 wr %o3, 0, %asi - ldx [%o2], %g3 + mov %o2, %g3 cmp %o1, 2 be,pn %icc, 2f cmp %o1, 4 diff --git a/arch/sparc64/kernel/unaligned.c b/arch/sparc64/kernel/unaligned.c index da9739f0d43..42718f6a7d3 100644 --- a/arch/sparc64/kernel/unaligned.c +++ b/arch/sparc64/kernel/unaligned.c @@ -184,13 +184,14 @@ extern void do_int_load(unsigned long *dest_reg, int size, unsigned long *saddr, int is_signed, int asi); extern void __do_int_store(unsigned long *dst_addr, int size, - unsigned long *src_val, int asi); + unsigned long src_val, int asi); static inline void do_int_store(int reg_num, int size, unsigned long *dst_addr, - struct pt_regs *regs, int asi) + struct pt_regs *regs, int asi, int orig_asi) { unsigned long zero = 0; - unsigned long *src_val = &zero; + unsigned long *src_val_p = &zero; + unsigned long src_val; if (size == 16) { size = 8; @@ -198,7 +199,25 @@ static inline void do_int_store(int reg_num, int size, unsigned long *dst_addr, (unsigned)fetch_reg(reg_num, regs) : 0)) << 32) | (unsigned)fetch_reg(reg_num + 1, regs); } else if (reg_num) { - src_val = fetch_reg_addr(reg_num, regs); + src_val_p = fetch_reg_addr(reg_num, regs); + } + src_val = *src_val_p; + if (unlikely(asi != orig_asi)) { + switch (size) { + case 2: + src_val = swab16(src_val); + break; + case 4: + src_val = swab32(src_val); + break; + case 8: + src_val = swab64(src_val); + break; + case 16: + default: + BUG(); + break; + }; } __do_int_store(dst_addr, size, src_val, asi); } @@ -276,6 +295,7 @@ asmlinkage void kernel_unaligned_trap(struct pt_regs *regs, unsigned int insn, u kernel_mna_trap_fault(); } else { unsigned long addr; + int orig_asi, asi; addr = compute_effective_address(regs, insn, ((insn >> 25) & 0x1f)); @@ -285,18 +305,48 @@ asmlinkage void kernel_unaligned_trap(struct pt_regs *regs, unsigned int insn, u regs->tpc, dirstrings[dir], addr, size, regs->u_regs[UREG_RETPC]); #endif + orig_asi = asi = decode_asi(insn, regs); + switch (asi) { + case ASI_NL: + case ASI_AIUPL: + case ASI_AIUSL: + case ASI_PL: + case ASI_SL: + case ASI_PNFL: + case ASI_SNFL: + asi &= ~0x08; + break; + }; switch (dir) { case load: do_int_load(fetch_reg_addr(((insn>>25)&0x1f), regs), size, (unsigned long *) addr, - decode_signedness(insn), - decode_asi(insn, regs)); + decode_signedness(insn), asi); + if (unlikely(asi != orig_asi)) { + unsigned long val_in = *(unsigned long *) addr; + switch (size) { + case 2: + val_in = swab16(val_in); + break; + case 4: + val_in = swab32(val_in); + break; + case 8: + val_in = swab64(val_in); + break; + case 16: + default: + BUG(); + break; + }; + *(unsigned long *) addr = val_in; + } break; case store: do_int_store(((insn>>25)&0x1f), size, (unsigned long *) addr, regs, - decode_asi(insn, regs)); + asi, orig_asi); break; default: From e0487992ce1dd7ae7da9c6aabdb19570bb95432b Mon Sep 17 00:00:00 2001 From: "Ed L. Cashin" Date: Mon, 19 Sep 2005 19:57:36 -0700 Subject: [PATCH 056/204] [BYTEORDER]: Document alignment and byteorder macros This patch comments the fact that although passing le64_to_cpup et al. is within the intended use of the byteorder macros, using get_unaligned is the recommended way to go. Signed-off-by: Ed L. Cashin Signed-off-by: David S. Miller --- include/linux/byteorder/generic.h | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/include/linux/byteorder/generic.h b/include/linux/byteorder/generic.h index 5fde6f4d6c1..04bd756efc6 100644 --- a/include/linux/byteorder/generic.h +++ b/include/linux/byteorder/generic.h @@ -5,6 +5,10 @@ * linux/byteorder_generic.h * Generic Byte-reordering support * + * The "... p" macros, like le64_to_cpup, can be used with pointers + * to unaligned data, but there will be a performance penalty on + * some architectures. Use get_unaligned for unaligned data. + * * Francois-Rene Rideau 19970707 * gathered all the good ideas from all asm-foo/byteorder.h into one file, * cleaned them up. From 676d55ae30ea3b688f0386f70553489f25f24d55 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 19 Sep 2005 20:00:41 -0700 Subject: [PATCH 057/204] Linux v2.6.14-rc2 Avast, ye scurvy land-lubbers! Time to try out a new release. Arrr! --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index 4e0d7c68d22..8cf6becf68d 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 2 PATCHLEVEL = 6 SUBLEVEL = 14 -EXTRAVERSION =-rc1 +EXTRAVERSION =-rc2 NAME=Affluent Albatross # *DOCUMENTATION* From 6a9b490d5fd7f23c5bcd75f970e01633ad3136e3 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 19 Sep 2005 20:11:57 -0700 Subject: [PATCH 058/204] [SPARC64]: Move DCACHE_ALIASING_POSSIBLE define to asm/page.h This showed that arch/sparc64/kernel/ptrace.c was not getting the define properly, and thus the code protected by this ifdef was never actually compiled before. So fix that too. Signed-off-by: David S. Miller --- arch/sparc64/kernel/ptrace.c | 7 ++++--- include/asm-sparc64/cacheflush.h | 7 ------- include/asm-sparc64/ide.h | 1 + include/asm-sparc64/page.h | 7 +++++++ include/asm-sparc64/pgalloc.h | 1 + 5 files changed, 13 insertions(+), 10 deletions(-) diff --git a/arch/sparc64/kernel/ptrace.c b/arch/sparc64/kernel/ptrace.c index 23ad839d113..5efbff90d66 100644 --- a/arch/sparc64/kernel/ptrace.c +++ b/arch/sparc64/kernel/ptrace.c @@ -30,6 +30,7 @@ #include #include #include +#include /* Returning from ptrace is a bit tricky because the syscall return * low level code assumes any value returned which is negative and @@ -128,20 +129,20 @@ void flush_ptrace_access(struct vm_area_struct *vma, struct page *page, * is mapped to in the user's address space, we can skip the * D-cache flush. */ - if ((uaddr ^ kaddr) & (1UL << 13)) { + if ((uaddr ^ (unsigned long) kaddr) & (1UL << 13)) { unsigned long start = __pa(kaddr); unsigned long end = start + len; if (tlb_type == spitfire) { for (; start < end; start += 32) - spitfire_put_dcache_tag(va & 0x3fe0, 0x0); + spitfire_put_dcache_tag(start & 0x3fe0, 0x0); } else { for (; start < end; start += 32) __asm__ __volatile__( "stxa %%g0, [%0] %1\n\t" "membar #Sync" : /* no outputs */ - : "r" (va), + : "r" (start), "i" (ASI_DCACHE_INVALIDATE)); } } diff --git a/include/asm-sparc64/cacheflush.h b/include/asm-sparc64/cacheflush.h index 51b26e81d82..ededd2659ea 100644 --- a/include/asm-sparc64/cacheflush.h +++ b/include/asm-sparc64/cacheflush.h @@ -4,13 +4,6 @@ #include #include -/* Flushing for D-cache alias handling is only needed if - * the page size is smaller than 16K. - */ -#if PAGE_SHIFT < 14 -#define DCACHE_ALIASING_POSSIBLE -#endif - #ifndef __ASSEMBLY__ #include diff --git a/include/asm-sparc64/ide.h b/include/asm-sparc64/ide.h index 4c1098474c7..c393f815b0b 100644 --- a/include/asm-sparc64/ide.h +++ b/include/asm-sparc64/ide.h @@ -15,6 +15,7 @@ #include #include #include +#include #ifndef MAX_HWIFS # ifdef CONFIG_BLK_DEV_IDEPCI diff --git a/include/asm-sparc64/page.h b/include/asm-sparc64/page.h index c9f8ef208ea..7f8d764abc4 100644 --- a/include/asm-sparc64/page.h +++ b/include/asm-sparc64/page.h @@ -21,6 +21,13 @@ #define PAGE_SIZE (_AC(1,UL) << PAGE_SHIFT) #define PAGE_MASK (~(PAGE_SIZE-1)) +/* Flushing for D-cache alias handling is only needed if + * the page size is smaller than 16K. + */ +#if PAGE_SHIFT < 14 +#define DCACHE_ALIASING_POSSIBLE +#endif + #ifdef __KERNEL__ #ifndef __ASSEMBLY__ diff --git a/include/asm-sparc64/pgalloc.h b/include/asm-sparc64/pgalloc.h index b9b1914aae6..a96067cca96 100644 --- a/include/asm-sparc64/pgalloc.h +++ b/include/asm-sparc64/pgalloc.h @@ -10,6 +10,7 @@ #include #include #include +#include /* Page table allocation/freeing. */ #ifdef CONFIG_SMP From eed8b2dee7cff46dd4bf5b82dc53465d229162ba Mon Sep 17 00:00:00 2001 From: Anton Altaparmakov Date: Tue, 20 Sep 2005 14:19:30 +0100 Subject: [PATCH 059/204] NTFS: More runlist handling fixes from Richard Russon and myself. Signed-off-by: Anton Altaparmakov --- fs/ntfs/runlist.c | 55 ++++++++++++++++++++++++++++------------------- 1 file changed, 33 insertions(+), 22 deletions(-) diff --git a/fs/ntfs/runlist.c b/fs/ntfs/runlist.c index e2665d011d7..061b5ff6b73 100644 --- a/fs/ntfs/runlist.c +++ b/fs/ntfs/runlist.c @@ -158,17 +158,21 @@ static inline BOOL ntfs_are_rl_mergeable(runlist_element *dst, BUG_ON(!dst); BUG_ON(!src); - if ((dst->lcn < 0) || (src->lcn < 0)) { /* Are we merging holes? */ - if (dst->lcn == LCN_HOLE && src->lcn == LCN_HOLE) - return TRUE; + /* We can merge unmapped regions even if they are misaligned. */ + if ((dst->lcn == LCN_RL_NOT_MAPPED) && (src->lcn == LCN_RL_NOT_MAPPED)) + return TRUE; + /* If the runs are misaligned, we cannot merge them. */ + if ((dst->vcn + dst->length) != src->vcn) return FALSE; - } - if ((dst->lcn + dst->length) != src->lcn) /* Are the runs contiguous? */ - return FALSE; - if ((dst->vcn + dst->length) != src->vcn) /* Are the runs misaligned? */ - return FALSE; - - return TRUE; + /* If both runs are non-sparse and contiguous, we can merge them. */ + if ((dst->lcn >= 0) && (src->lcn >= 0) && + ((dst->lcn + dst->length) == src->lcn)) + return TRUE; + /* If we are merging two holes, we can merge them. */ + if ((dst->lcn == LCN_HOLE) && (src->lcn == LCN_HOLE)) + return TRUE; + /* Cannot merge. */ + return FALSE; } /** @@ -214,14 +218,15 @@ static inline void __ntfs_rl_merge(runlist_element *dst, runlist_element *src) static inline runlist_element *ntfs_rl_append(runlist_element *dst, int dsize, runlist_element *src, int ssize, int loc) { - BOOL right; /* Right end of @src needs merging. */ - int marker; /* End of the inserted runs. */ + BOOL right = FALSE; /* Right end of @src needs merging. */ + int marker; /* End of the inserted runs. */ BUG_ON(!dst); BUG_ON(!src); /* First, check if the right hand end needs merging. */ - right = ntfs_are_rl_mergeable(src + ssize - 1, dst + loc + 1); + if ((loc + 1) < dsize) + right = ntfs_are_rl_mergeable(src + ssize - 1, dst + loc + 1); /* Space required: @dst size + @src size, less one if we merged. */ dst = ntfs_rl_realloc(dst, dsize, dsize + ssize - right); @@ -377,20 +382,21 @@ static inline runlist_element *ntfs_rl_replace(runlist_element *dst, int dsize, runlist_element *src, int ssize, int loc) { BOOL left = FALSE; /* Left end of @src needs merging. */ - BOOL right; /* Right end of @src needs merging. */ + BOOL right = FALSE; /* Right end of @src needs merging. */ int tail; /* Start of tail of @dst. */ int marker; /* End of the inserted runs. */ BUG_ON(!dst); BUG_ON(!src); - /* First, merge the left and right ends, if necessary. */ - right = ntfs_are_rl_mergeable(src + ssize - 1, dst + loc + 1); + /* First, see if the left and right ends need merging. */ + if ((loc + 1) < dsize) + right = ntfs_are_rl_mergeable(src + ssize - 1, dst + loc + 1); if (loc > 0) left = ntfs_are_rl_mergeable(dst + loc - 1, src); /* * Allocate some space. We will need less if the left, right, or both - * ends were merged. + * ends get merged. */ dst = ntfs_rl_realloc(dst, dsize, dsize + ssize - left - right); if (IS_ERR(dst)) @@ -399,21 +405,26 @@ static inline runlist_element *ntfs_rl_replace(runlist_element *dst, * We are guaranteed to succeed from here so can start modifying the * original runlists. */ + + /* First, merge the left and right ends, if necessary. */ if (right) __ntfs_rl_merge(src + ssize - 1, dst + loc + 1); if (left) __ntfs_rl_merge(dst + loc - 1, src); /* - * First run of @dst that needs to be moved out of the way to make - * space for the runs to be copied from @src, i.e. the first run of the - * tail of @dst. + * Offset of the tail of @dst. This needs to be moved out of the way + * to make space for the runs to be copied from @src, i.e. the first + * run of the tail of @dst. + * Nominally, @tail equals @loc + 1, i.e. location, skipping the + * replaced run. However, if @right, then one of @dst's runs is + * already merged into @src. */ tail = loc + right + 1; /* * First run after the @src runs that have been inserted, i.e. where * the tail of @dst needs to be moved to. - * Nominally, marker equals @loc + @ssize, i.e. location + number of - * runs in @src). However, if @left, then the first run in @src has + * Nominally, @marker equals @loc + @ssize, i.e. location + number of + * runs in @src. However, if @left, then the first run in @src has * been merged with one in @dst. */ marker = loc + ssize - left; From 9506057fca54464f3291b62156e6cd907c4cbc95 Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Tue, 20 Sep 2005 16:20:49 +0100 Subject: [PATCH 060/204] [ARM] 2924/3: taglist - postfix section with .init for `make buildcheck` Patch from Ben Dooks The `make buildcheck` is erroneously reporting that the taglist is referencing items in the .init section as it is not itself postfixed with .init Signed-off-by: Ben Dooks Signed-off-by: Russell King --- arch/arm/kernel/vmlinux.lds.S | 2 +- include/asm-arm/setup.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm/kernel/vmlinux.lds.S b/arch/arm/kernel/vmlinux.lds.S index ad2d66c93a5..350b53b41e5 100644 --- a/arch/arm/kernel/vmlinux.lds.S +++ b/arch/arm/kernel/vmlinux.lds.S @@ -29,7 +29,7 @@ SECTIONS *(.arch.info) __arch_info_end = .; __tagtable_begin = .; - *(.taglist) + *(.taglist.init) __tagtable_end = .; . = ALIGN(16); __setup_start = .; diff --git a/include/asm-arm/setup.h b/include/asm-arm/setup.h index adcbd79762b..a4b3ebf9bdd 100644 --- a/include/asm-arm/setup.h +++ b/include/asm-arm/setup.h @@ -171,7 +171,7 @@ struct tagtable { int (*parse)(const struct tag *); }; -#define __tag __attribute_used__ __attribute__((__section__(".taglist"))) +#define __tag __attribute_used__ __attribute__((__section__(".taglist.init"))) #define __tagtable(tag, fn) \ static struct tagtable __tagtable_##fn __tag = { tag, fn } From 58dd48a6575d3ba86ba659eb8c6fc23246783fba Mon Sep 17 00:00:00 2001 From: Vincent Sanders Date: Tue, 20 Sep 2005 16:21:42 +0100 Subject: [PATCH 061/204] [ARM] 2922/1: compile fix for shark Patch from Vincent Sanders Shark platform fails to build with gcc 4 because of a bad lvalue assignement Signed-off-by: Vincent Sanders Signed-off-by: Russell King --- arch/arm/boot/compressed/ofw-shark.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/boot/compressed/ofw-shark.c b/arch/arm/boot/compressed/ofw-shark.c index 7f6f5db0d06..465c54b6b12 100644 --- a/arch/arm/boot/compressed/ofw-shark.c +++ b/arch/arm/boot/compressed/ofw-shark.c @@ -256,5 +256,5 @@ asmlinkage void ofw_init(ofw_handle_t o, int *nomr, int *pointer) temp[11]='\0'; mem_len = OF_getproplen(o,phandle, temp); OF_getprop(o,phandle, temp, buffer, mem_len); - (unsigned char) pointer[32] = ((unsigned char *) buffer)[mem_len-2]; + * ((unsigned char *) &pointer[32]) = ((unsigned char *) buffer)[mem_len-2]; } From bfe6815e0465035d013b2b676444376fe2b3716e Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Tue, 20 Sep 2005 16:25:12 +0100 Subject: [PATCH 062/204] [ARM] 2925/3: earlyparam - postfix section with .init for `make buildcheck` Patch from Ben Dooks The `make buildcheck` is erroneously reporting that the earlyparam list is referencing items in the .init section as it is not itself postfixed with .init Also, as per rmk's suggestion, rename the __early_param to .early_param to bring it into line with everything else Signed-off-by: Ben Dooks Signed-off-by: Russell King --- arch/arm/kernel/vmlinux.lds.S | 2 +- include/asm-arm/setup.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm/kernel/vmlinux.lds.S b/arch/arm/kernel/vmlinux.lds.S index 350b53b41e5..47423aaf9d7 100644 --- a/arch/arm/kernel/vmlinux.lds.S +++ b/arch/arm/kernel/vmlinux.lds.S @@ -36,7 +36,7 @@ SECTIONS *(.init.setup) __setup_end = .; __early_begin = .; - *(__early_param) + *(.early_param.init) __early_end = .; __initcall_start = .; *(.initcall1.init) diff --git a/include/asm-arm/setup.h b/include/asm-arm/setup.h index a4b3ebf9bdd..ea3ed246523 100644 --- a/include/asm-arm/setup.h +++ b/include/asm-arm/setup.h @@ -213,6 +213,6 @@ struct early_params { #define __early_param(name,fn) \ static struct early_params __early_##fn __attribute_used__ \ -__attribute__((__section__("__early_param"))) = { name, fn } +__attribute__((__section__(".early_param.init"))) = { name, fn } #endif From 02b7dd1244aab9267ae4078e1ad6a2fdaabeb6ed Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Tue, 20 Sep 2005 16:35:03 +0100 Subject: [PATCH 063/204] [ARM] 2926/1: .proc.info - postfix section with .init for `make buildcheck` Patch from Ben Dooks The `make buildcheck` is erroneously reporting that the .proc.info list is referencing items in the .init section as it is not itself postfixed with .init Signed-off-by: Ben Dooks Signed-off-by: Russell King --- arch/arm/kernel/vmlinux.lds.S | 2 +- arch/arm/mm/proc-arm1020.S | 2 +- arch/arm/mm/proc-arm1020e.S | 2 +- arch/arm/mm/proc-arm1022.S | 2 +- arch/arm/mm/proc-arm1026.S | 2 +- arch/arm/mm/proc-arm6_7.S | 2 +- arch/arm/mm/proc-arm720.S | 2 +- arch/arm/mm/proc-arm920.S | 2 +- arch/arm/mm/proc-arm922.S | 2 +- arch/arm/mm/proc-arm925.S | 2 +- arch/arm/mm/proc-arm926.S | 2 +- arch/arm/mm/proc-sa110.S | 2 +- arch/arm/mm/proc-sa1100.S | 2 +- arch/arm/mm/proc-v6.S | 2 +- arch/arm/mm/proc-xscale.S | 2 +- 15 files changed, 15 insertions(+), 15 deletions(-) diff --git a/arch/arm/kernel/vmlinux.lds.S b/arch/arm/kernel/vmlinux.lds.S index 47423aaf9d7..420b8d09a38 100644 --- a/arch/arm/kernel/vmlinux.lds.S +++ b/arch/arm/kernel/vmlinux.lds.S @@ -23,7 +23,7 @@ SECTIONS *(.init.text) _einittext = .; __proc_info_begin = .; - *(.proc.info) + *(.proc.info.init) __proc_info_end = .; __arch_info_begin = .; *(.arch.info) diff --git a/arch/arm/mm/proc-arm1020.S b/arch/arm/mm/proc-arm1020.S index 1d739d282a4..82ec954e45b 100644 --- a/arch/arm/mm/proc-arm1020.S +++ b/arch/arm/mm/proc-arm1020.S @@ -509,7 +509,7 @@ cpu_arm1020_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __arm1020_proc_info,#object __arm1020_proc_info: diff --git a/arch/arm/mm/proc-arm1020e.S b/arch/arm/mm/proc-arm1020e.S index 9b725665b5c..7375fe930f7 100644 --- a/arch/arm/mm/proc-arm1020e.S +++ b/arch/arm/mm/proc-arm1020e.S @@ -491,7 +491,7 @@ cpu_arm1020e_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __arm1020e_proc_info,#object __arm1020e_proc_info: diff --git a/arch/arm/mm/proc-arm1022.S b/arch/arm/mm/proc-arm1022.S index 37b70fa21c7..6ca639094d6 100644 --- a/arch/arm/mm/proc-arm1022.S +++ b/arch/arm/mm/proc-arm1022.S @@ -473,7 +473,7 @@ cpu_arm1022_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __arm1022_proc_info,#object __arm1022_proc_info: diff --git a/arch/arm/mm/proc-arm1026.S b/arch/arm/mm/proc-arm1026.S index 931b690d1be..10317e4f55d 100644 --- a/arch/arm/mm/proc-arm1026.S +++ b/arch/arm/mm/proc-arm1026.S @@ -469,7 +469,7 @@ cpu_arm1026_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __arm1026_proc_info,#object __arm1026_proc_info: diff --git a/arch/arm/mm/proc-arm6_7.S b/arch/arm/mm/proc-arm6_7.S index d0f1bbb48f6..8e7e1e70ab0 100644 --- a/arch/arm/mm/proc-arm6_7.S +++ b/arch/arm/mm/proc-arm6_7.S @@ -332,7 +332,7 @@ cpu_arm710_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __arm6_proc_info, #object __arm6_proc_info: diff --git a/arch/arm/mm/proc-arm720.S b/arch/arm/mm/proc-arm720.S index c69c9de3239..a13e0184d34 100644 --- a/arch/arm/mm/proc-arm720.S +++ b/arch/arm/mm/proc-arm720.S @@ -222,7 +222,7 @@ cpu_arm720_name: * See linux/include/asm-arm/procinfo.h for a definition of this structure. */ - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __arm710_proc_info, #object __arm710_proc_info: diff --git a/arch/arm/mm/proc-arm920.S b/arch/arm/mm/proc-arm920.S index 0f490a0fcb7..d1651389999 100644 --- a/arch/arm/mm/proc-arm920.S +++ b/arch/arm/mm/proc-arm920.S @@ -452,7 +452,7 @@ cpu_arm920_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __arm920_proc_info,#object __arm920_proc_info: diff --git a/arch/arm/mm/proc-arm922.S b/arch/arm/mm/proc-arm922.S index 62bc34a139e..23b8ed97f4e 100644 --- a/arch/arm/mm/proc-arm922.S +++ b/arch/arm/mm/proc-arm922.S @@ -456,7 +456,7 @@ cpu_arm922_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __arm922_proc_info,#object __arm922_proc_info: diff --git a/arch/arm/mm/proc-arm925.S b/arch/arm/mm/proc-arm925.S index ee49aa2ca78..ee95c52db51 100644 --- a/arch/arm/mm/proc-arm925.S +++ b/arch/arm/mm/proc-arm925.S @@ -521,7 +521,7 @@ cpu_arm925_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __arm925_proc_info,#object __arm925_proc_info: diff --git a/arch/arm/mm/proc-arm926.S b/arch/arm/mm/proc-arm926.S index bb95cc9fed0..7d042dc20c4 100644 --- a/arch/arm/mm/proc-arm926.S +++ b/arch/arm/mm/proc-arm926.S @@ -471,7 +471,7 @@ cpu_arm926_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __arm926_proc_info,#object __arm926_proc_info: diff --git a/arch/arm/mm/proc-sa110.S b/arch/arm/mm/proc-sa110.S index 34f7e7d3f41..bd330c4075a 100644 --- a/arch/arm/mm/proc-sa110.S +++ b/arch/arm/mm/proc-sa110.S @@ -249,7 +249,7 @@ cpu_sa110_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __sa110_proc_info,#object __sa110_proc_info: diff --git a/arch/arm/mm/proc-sa1100.S b/arch/arm/mm/proc-sa1100.S index ca14f80d5ab..91b89124c0d 100644 --- a/arch/arm/mm/proc-sa1100.S +++ b/arch/arm/mm/proc-sa1100.S @@ -280,7 +280,7 @@ cpu_sa1110_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __sa1100_proc_info,#object __sa1100_proc_info: diff --git a/arch/arm/mm/proc-v6.S b/arch/arm/mm/proc-v6.S index eb34823c9db..caf3b19b167 100644 --- a/arch/arm/mm/proc-v6.S +++ b/arch/arm/mm/proc-v6.S @@ -240,7 +240,7 @@ cpu_elf_name: .size cpu_elf_name, . - cpu_elf_name .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr /* * Match any ARMv6 processor core. diff --git a/arch/arm/mm/proc-xscale.S b/arch/arm/mm/proc-xscale.S index b88de270014..861b3594728 100644 --- a/arch/arm/mm/proc-xscale.S +++ b/arch/arm/mm/proc-xscale.S @@ -578,7 +578,7 @@ cpu_pxa270_name: .align - .section ".proc.info", #alloc, #execinstr + .section ".proc.info.init", #alloc, #execinstr .type __80200_proc_info,#object __80200_proc_info: From 9d0fd1eb8a3c19f3ede5418540b3c9f64fac4b86 Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Tue, 20 Sep 2005 16:45:20 +0100 Subject: [PATCH 064/204] [ARM] 2927/1: .arch.info - postfix section with .init for `make buildcheck` Patch from Ben Dooks The `make buildcheck` is erroneously reporting that the .arch.info list is referencing items in the .init section as it is not itself postfixed with .init Signed-off-by: Ben Dooks Signed-off-by: Russell King --- arch/arm/kernel/vmlinux.lds.S | 2 +- include/asm-arm/mach/arch.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm/kernel/vmlinux.lds.S b/arch/arm/kernel/vmlinux.lds.S index 420b8d09a38..08e58ecd44b 100644 --- a/arch/arm/kernel/vmlinux.lds.S +++ b/arch/arm/kernel/vmlinux.lds.S @@ -26,7 +26,7 @@ SECTIONS *(.proc.info.init) __proc_info_end = .; __arch_info_begin = .; - *(.arch.info) + *(.arch.info.init) __arch_info_end = .; __tagtable_begin = .; *(.taglist.init) diff --git a/include/asm-arm/mach/arch.h b/include/asm-arm/mach/arch.h index 56c6bf4ab0c..4fa95084a8c 100644 --- a/include/asm-arm/mach/arch.h +++ b/include/asm-arm/mach/arch.h @@ -50,7 +50,7 @@ struct machine_desc { */ #define MACHINE_START(_type,_name) \ const struct machine_desc __mach_desc_##_type \ - __attribute__((__section__(".arch.info"))) = { \ + __attribute__((__section__(".arch.info.init"))) = { \ .nr = MACH_TYPE_##_type, \ .name = _name, From 5fe10ab19046d84f3fd243436cbd5fa01019e809 Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Tue, 20 Sep 2005 17:24:33 +0100 Subject: [PATCH 065/204] [ARM] 2928/1: S3C2410 - make machine init code static Patch from Ben Dooks This code is not being exported, declare it static Signed-off-by: Ben Dooks Signed-off-by: Russell King --- arch/arm/mach-s3c2410/mach-anubis.c | 3 ++- arch/arm/mach-s3c2410/mach-bast.c | 3 ++- arch/arm/mach-s3c2410/mach-h1940.c | 7 ++++--- arch/arm/mach-s3c2410/mach-n30.c | 6 +++--- arch/arm/mach-s3c2410/mach-nexcoder.c | 2 +- arch/arm/mach-s3c2410/mach-otom.c | 2 +- arch/arm/mach-s3c2410/mach-rx3715.c | 5 +++-- arch/arm/mach-s3c2410/mach-smdk2410.c | 5 +++-- arch/arm/mach-s3c2410/mach-smdk2440.c | 5 +++-- arch/arm/mach-s3c2410/mach-vr1000.c | 3 ++- 10 files changed, 24 insertions(+), 17 deletions(-) diff --git a/arch/arm/mach-s3c2410/mach-anubis.c b/arch/arm/mach-s3c2410/mach-anubis.c index f87aa0b669a..7c05f27fe1d 100644 --- a/arch/arm/mach-s3c2410/mach-anubis.c +++ b/arch/arm/mach-s3c2410/mach-anubis.c @@ -12,6 +12,7 @@ * * Modifications: * 02-May-2005 BJD Copied from mach-bast.c + * 20-Sep-2005 BJD Added static to non-exported items */ #include @@ -232,7 +233,7 @@ static struct s3c24xx_board anubis_board __initdata = { .clocks_count = ARRAY_SIZE(anubis_clocks) }; -void __init anubis_map_io(void) +static void __init anubis_map_io(void) { /* initialise the clocks */ diff --git a/arch/arm/mach-s3c2410/mach-bast.c b/arch/arm/mach-s3c2410/mach-bast.c index 1a3367da640..ed1f07d7252 100644 --- a/arch/arm/mach-s3c2410/mach-bast.c +++ b/arch/arm/mach-s3c2410/mach-bast.c @@ -31,6 +31,7 @@ * 17-Jul-2005 BJD Changed to platform device for SuperIO 16550s * 25-Jul-2005 BJD Removed ASIX static mappings * 27-Jul-2005 BJD Ensure maximum frequency of i2c bus + * 20-Sep-2005 BJD Added static to non-exported items */ #include @@ -428,7 +429,7 @@ static struct s3c24xx_board bast_board __initdata = { .clocks_count = ARRAY_SIZE(bast_clocks) }; -void __init bast_map_io(void) +static void __init bast_map_io(void) { /* initialise the clocks */ diff --git a/arch/arm/mach-s3c2410/mach-h1940.c b/arch/arm/mach-s3c2410/mach-h1940.c index 6ff1889fbd2..fb3cb01266e 100644 --- a/arch/arm/mach-s3c2410/mach-h1940.c +++ b/arch/arm/mach-s3c2410/mach-h1940.c @@ -24,6 +24,7 @@ * 10-Jan-2005 BJD Removed include of s3c2410.h * 14-Jan-2005 BJD Added clock init * 10-Mar-2005 LCVR Changed S3C2410_VA to S3C24XX_VA + * 20-Sep-2005 BJD Added static to non-exported items */ #include @@ -147,7 +148,7 @@ static struct s3c24xx_board h1940_board __initdata = { .devices_count = ARRAY_SIZE(h1940_devices) }; -void __init h1940_map_io(void) +static void __init h1940_map_io(void) { s3c24xx_init_io(h1940_iodesc, ARRAY_SIZE(h1940_iodesc)); s3c24xx_init_clocks(0); @@ -155,13 +156,13 @@ void __init h1940_map_io(void) s3c24xx_set_board(&h1940_board); } -void __init h1940_init_irq(void) +static void __init h1940_init_irq(void) { s3c24xx_init_irq(); } -void __init h1940_init(void) +static void __init h1940_init(void) { set_s3c2410fb_info(&h1940_lcdcfg); } diff --git a/arch/arm/mach-s3c2410/mach-n30.c b/arch/arm/mach-s3c2410/mach-n30.c index 66bf5bb2b3d..5c0f2b091f9 100644 --- a/arch/arm/mach-s3c2410/mach-n30.c +++ b/arch/arm/mach-s3c2410/mach-n30.c @@ -97,7 +97,7 @@ static struct s3c24xx_board n30_board __initdata = { .devices_count = ARRAY_SIZE(n30_devices) }; -void __init n30_map_io(void) +static void __init n30_map_io(void) { s3c24xx_init_io(n30_iodesc, ARRAY_SIZE(n30_iodesc)); s3c24xx_init_clocks(0); @@ -105,14 +105,14 @@ void __init n30_map_io(void) s3c24xx_set_board(&n30_board); } -void __init n30_init_irq(void) +static void __init n30_init_irq(void) { s3c24xx_init_irq(); } /* GPB3 is the line that controls the pull-up for the USB D+ line */ -void __init n30_init(void) +static void __init n30_init(void) { s3c_device_i2c.dev.platform_data = &n30_i2ccfg; diff --git a/arch/arm/mach-s3c2410/mach-nexcoder.c b/arch/arm/mach-s3c2410/mach-nexcoder.c index d24c242414c..c22f8216032 100644 --- a/arch/arm/mach-s3c2410/mach-nexcoder.c +++ b/arch/arm/mach-s3c2410/mach-nexcoder.c @@ -136,7 +136,7 @@ static void __init nexcoder_sensorboard_init(void) s3c2410_gpio_cfgpin(S3C2410_GPF2, S3C2410_GPF2_OUTP); // CAM_GPIO6 => CAM_PWRDN } -void __init nexcoder_map_io(void) +static void __init nexcoder_map_io(void) { s3c24xx_init_io(nexcoder_iodesc, ARRAY_SIZE(nexcoder_iodesc)); s3c24xx_init_clocks(0); diff --git a/arch/arm/mach-s3c2410/mach-otom.c b/arch/arm/mach-s3c2410/mach-otom.c index d901ed492ff..ad1459e402e 100644 --- a/arch/arm/mach-s3c2410/mach-otom.c +++ b/arch/arm/mach-s3c2410/mach-otom.c @@ -105,7 +105,7 @@ static struct s3c24xx_board otom11_board __initdata = { }; -void __init otom11_map_io(void) +static void __init otom11_map_io(void) { s3c24xx_init_io(otom11_iodesc, ARRAY_SIZE(otom11_iodesc)); s3c24xx_init_clocks(0); diff --git a/arch/arm/mach-s3c2410/mach-rx3715.c b/arch/arm/mach-s3c2410/mach-rx3715.c index a73d61c1de4..22d9e070fd6 100644 --- a/arch/arm/mach-s3c2410/mach-rx3715.c +++ b/arch/arm/mach-s3c2410/mach-rx3715.c @@ -16,6 +16,7 @@ * 14-Jan-2005 BJD Added new clock init * 10-Mar-2005 LCVR Changed S3C2410_VA to S3C24XX_VA * 14-Mar-2005 BJD Fixed __iomem warnings + * 20-Sep-2005 BJD Added static to non-exported items */ #include @@ -108,7 +109,7 @@ static struct s3c24xx_board rx3715_board __initdata = { .devices_count = ARRAY_SIZE(rx3715_devices) }; -void __init rx3715_map_io(void) +static void __init rx3715_map_io(void) { s3c24xx_init_io(rx3715_iodesc, ARRAY_SIZE(rx3715_iodesc)); s3c24xx_init_clocks(16934000); @@ -116,7 +117,7 @@ void __init rx3715_map_io(void) s3c24xx_set_board(&rx3715_board); } -void __init rx3715_init_irq(void) +static void __init rx3715_init_irq(void) { s3c24xx_init_irq(); } diff --git a/arch/arm/mach-s3c2410/mach-smdk2410.c b/arch/arm/mach-s3c2410/mach-smdk2410.c index 67e903a700d..2eda55a6b67 100644 --- a/arch/arm/mach-s3c2410/mach-smdk2410.c +++ b/arch/arm/mach-s3c2410/mach-smdk2410.c @@ -28,6 +28,7 @@ * Ben Dooks * * 10-Mar-2005 LCVR Changed S3C2410_VA to S3C24XX_VA + * 20-Sep-2005 BJD Added static to non-exported items * ***********************************************************************/ @@ -97,7 +98,7 @@ static struct s3c24xx_board smdk2410_board __initdata = { .devices_count = ARRAY_SIZE(smdk2410_devices) }; -void __init smdk2410_map_io(void) +static void __init smdk2410_map_io(void) { s3c24xx_init_io(smdk2410_iodesc, ARRAY_SIZE(smdk2410_iodesc)); s3c24xx_init_clocks(0); @@ -105,7 +106,7 @@ void __init smdk2410_map_io(void) s3c24xx_set_board(&smdk2410_board); } -void __init smdk2410_init_irq(void) +static void __init smdk2410_init_irq(void) { s3c24xx_init_irq(); } diff --git a/arch/arm/mach-s3c2410/mach-smdk2440.c b/arch/arm/mach-s3c2410/mach-smdk2440.c index 357522106f6..722ef46b630 100644 --- a/arch/arm/mach-s3c2410/mach-smdk2440.c +++ b/arch/arm/mach-s3c2410/mach-smdk2440.c @@ -18,6 +18,7 @@ * 22-Feb-2005 BJD Updated for 2.6.11-rc5 relesa * 10-Mar-2005 LCVR Replaced S3C2410_VA by S3C24XX_VA * 14-Mar-2005 BJD void __iomem fixes + * 20-Sep-2005 BJD Added static to non-exported items */ #include @@ -98,7 +99,7 @@ static struct s3c24xx_board smdk2440_board __initdata = { .devices_count = ARRAY_SIZE(smdk2440_devices) }; -void __init smdk2440_map_io(void) +static void __init smdk2440_map_io(void) { s3c24xx_init_io(smdk2440_iodesc, ARRAY_SIZE(smdk2440_iodesc)); s3c24xx_init_clocks(16934400); @@ -106,7 +107,7 @@ void __init smdk2440_map_io(void) s3c24xx_set_board(&smdk2440_board); } -void __init smdk2440_machine_init(void) +static void __init smdk2440_machine_init(void) { /* Configure the LEDs (even if we have no LED support)*/ diff --git a/arch/arm/mach-s3c2410/mach-vr1000.c b/arch/arm/mach-s3c2410/mach-vr1000.c index 8f9ab2893df..663a7f98fc0 100644 --- a/arch/arm/mach-s3c2410/mach-vr1000.c +++ b/arch/arm/mach-s3c2410/mach-vr1000.c @@ -28,6 +28,7 @@ * 10-Mar-2005 LCVR Changed S3C2410_VA to S3C24XX_VA * 14-Mar-2006 BJD void __iomem fixes * 22-Jun-2006 BJD Added DM9000 platform information + * 20-Sep-2005 BJD Added static to non-exported items */ #include @@ -347,7 +348,7 @@ static void vr1000_power_off(void) s3c2410_gpio_setpin(S3C2410_GPB9, 1); } -void __init vr1000_map_io(void) +static void __init vr1000_map_io(void) { /* initialise clock sources */ From 840ff6a4f6174d7fe19c206b5f36ff64123a2f45 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 20 Sep 2005 17:52:13 +0100 Subject: [PATCH 066/204] [ARM] Prevent deadlock in page fault handler As per x86, we may deadlock while trying to get the mmap semaphore. Implement the same fix, which allows (eg) recursive faults to cause an oops instead of deadlocking. Signed-off-by: Russell King --- arch/arm/mm/fault.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c index 0b6c4db44e0..4a884baf3b9 100644 --- a/arch/arm/mm/fault.c +++ b/arch/arm/mm/fault.c @@ -233,7 +233,17 @@ do_page_fault(unsigned long addr, unsigned int fsr, struct pt_regs *regs) if (in_interrupt() || !mm) goto no_context; - down_read(&mm->mmap_sem); + /* + * As per x86, we may deadlock here. However, since the kernel only + * validly references user space from well defined areas of the code, + * we can bug out early if this is from code which shouldn't. + */ + if (!down_read_trylock(&mm->mmap_sem)) { + if (!user_mode(regs) && !search_exception_tables(regs->ARM_pc)) + goto no_context; + down_read(&mm->mmap_sem); + } + fault = __do_page_fault(mm, addr, fsr, tsk); up_read(&mm->mmap_sem); From 8d2cae0651502028bf64844508ab18528bbd65c2 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Tue, 20 Sep 2005 10:52:04 -0700 Subject: [PATCH 067/204] [PATCH] IPoIB: Don't flush workqueue from within workqueue ipoib_mcast_restart_task() is always called from within the single-threaded IPoIB workqueue, so flushing the workqueue from within the function can lead to a recursion overflow. But since we're running in a single-threaded workqueue, we're already synchronized against other items in the workqueue, so just get rid of the flush in ipoib_mcast_restart_task(). Signed-off-by: Roland Dreier --- drivers/infiniband/ulp/ipoib/ipoib.h | 2 +- drivers/infiniband/ulp/ipoib/ipoib_ib.c | 4 ++-- drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 7 ++++--- 3 files changed, 7 insertions(+), 6 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index bea960b8191..4ea1c1ca85b 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -257,7 +257,7 @@ void ipoib_mcast_send(struct net_device *dev, union ib_gid *mgid, void ipoib_mcast_restart_task(void *dev_ptr); int ipoib_mcast_start_thread(struct net_device *dev); -int ipoib_mcast_stop_thread(struct net_device *dev); +int ipoib_mcast_stop_thread(struct net_device *dev, int flush); void ipoib_mcast_dev_down(struct net_device *dev); void ipoib_mcast_dev_flush(struct net_device *dev); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c index ef0e3894863..f7440096b5e 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c @@ -432,7 +432,7 @@ int ipoib_ib_dev_down(struct net_device *dev) flush_workqueue(ipoib_workqueue); } - ipoib_mcast_stop_thread(dev); + ipoib_mcast_stop_thread(dev, 1); /* * Flush the multicast groups first so we stop any multicast joins. The @@ -599,7 +599,7 @@ void ipoib_ib_dev_cleanup(struct net_device *dev) ipoib_dbg(priv, "cleaning up ib_dev\n"); - ipoib_mcast_stop_thread(dev); + ipoib_mcast_stop_thread(dev, 1); /* Delete the broadcast address and the local address */ ipoib_mcast_dev_down(dev); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index 3aab4d85abf..36ce29836bf 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -598,7 +598,7 @@ int ipoib_mcast_start_thread(struct net_device *dev) return 0; } -int ipoib_mcast_stop_thread(struct net_device *dev) +int ipoib_mcast_stop_thread(struct net_device *dev, int flush) { struct ipoib_dev_priv *priv = netdev_priv(dev); struct ipoib_mcast *mcast; @@ -610,7 +610,8 @@ int ipoib_mcast_stop_thread(struct net_device *dev) cancel_delayed_work(&priv->mcast_task); up(&mcast_mutex); - flush_workqueue(ipoib_workqueue); + if (flush) + flush_workqueue(ipoib_workqueue); if (priv->broadcast && priv->broadcast->query) { ib_sa_cancel_query(priv->broadcast->query_id, priv->broadcast->query); @@ -832,7 +833,7 @@ void ipoib_mcast_restart_task(void *dev_ptr) ipoib_dbg_mcast(priv, "restarting multicast task\n"); - ipoib_mcast_stop_thread(dev); + ipoib_mcast_stop_thread(dev, 0); spin_lock_irqsave(&priv->lock, flags); From 71eea47d853bb0ce0c6befe11b3e08111263170f Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Tue, 20 Sep 2005 10:54:48 -0700 Subject: [PATCH 068/204] [PATCH] IB/mthca: Fix device removal memory leak Clean up QP table array on device removal. Signed-off-by: Michael S. Tsirkin Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_qp.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c index eaa52185763..5fa00669f9b 100644 --- a/drivers/infiniband/hw/mthca/mthca_qp.c +++ b/drivers/infiniband/hw/mthca/mthca_qp.c @@ -2123,5 +2123,6 @@ void __devexit mthca_cleanup_qp_table(struct mthca_dev *dev) for (i = 0; i < 2; ++i) mthca_CONF_SPECIAL_QP(dev, i, 0, &status); + mthca_array_cleanup(&dev->qp_table.qp, dev->limits.num_qps); mthca_alloc_cleanup(&dev->qp_table.alloc); } From 729b4f7de68191478b20fab19a6d0c6b8c4380c9 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 20 Sep 2005 12:18:38 -0700 Subject: [PATCH 069/204] [SPARC64]: Verify vmalloc TLB misses more strictly. Arrange the modules, OBP, and vmalloc areas such that a range verification can be done quite minimally. Signed-off-by: David S. Miller --- arch/sparc64/kernel/entry.S | 41 +++++++++++++++++------------------ include/asm-sparc64/pgtable.h | 20 +++++++++-------- 2 files changed, 31 insertions(+), 30 deletions(-) diff --git a/arch/sparc64/kernel/entry.S b/arch/sparc64/kernel/entry.S index 3e0badb820c..b4834952785 100644 --- a/arch/sparc64/kernel/entry.S +++ b/arch/sparc64/kernel/entry.S @@ -42,19 +42,15 @@ * executing (see inherit_locked_prom_mappings() rant). */ sparc64_vpte_nucleus: - /* Load 0xf0000000, which is LOW_OBP_ADDRESS. */ - mov 0xf, %g5 - sllx %g5, 28, %g5 - - /* Is addr >= LOW_OBP_ADDRESS? */ + /* Note that kvmap below has verified that the address is + * in the range MODULES_VADDR --> VMALLOC_END already. So + * here we need only check if it is an OBP address or not. + */ + sethi %hi(LOW_OBP_ADDRESS), %g5 cmp %g4, %g5 blu,pn %xcc, sparc64_vpte_patchme1 mov 0x1, %g5 - - /* Load 0x100000000, which is HI_OBP_ADDRESS. */ sllx %g5, 32, %g5 - - /* Is addr < HI_OBP_ADDRESS? */ cmp %g4, %g5 blu,pn %xcc, obp_iaddr_patch nop @@ -156,26 +152,29 @@ obp_daddr_patch: * rather, use information saved during inherit_prom_mappings() using 8k * pagesize. */ + .align 32 kvmap: - /* Load 0xf0000000, which is LOW_OBP_ADDRESS. */ - mov 0xf, %g5 - sllx %g5, 28, %g5 - - /* Is addr >= LOW_OBP_ADDRESS? */ + sethi %hi(MODULES_VADDR), %g5 cmp %g4, %g5 - blu,pn %xcc, vmalloc_addr + blu,pn %xcc, longpath + mov (VMALLOC_END >> 24), %g5 + sllx %g5, 24, %g5 + cmp %g4, %g5 + bgeu,pn %xcc, longpath + nop + +kvmap_check_obp: + sethi %hi(LOW_OBP_ADDRESS), %g5 + cmp %g4, %g5 + blu,pn %xcc, kvmap_vmalloc_addr mov 0x1, %g5 - - /* Load 0x100000000, which is HI_OBP_ADDRESS. */ sllx %g5, 32, %g5 - - /* Is addr < HI_OBP_ADDRESS? */ cmp %g4, %g5 blu,pn %xcc, obp_daddr_patch nop -vmalloc_addr: - /* If we get here, a vmalloc addr accessed, load kernel VPTE. */ +kvmap_vmalloc_addr: + /* If we get here, a vmalloc addr was accessed, load kernel VPTE. */ ldxa [%g3 + %g6] ASI_N, %g5 brgez,pn %g5, longpath nop diff --git a/include/asm-sparc64/pgtable.h b/include/asm-sparc64/pgtable.h index a2b4f5ed462..a297f6144f0 100644 --- a/include/asm-sparc64/pgtable.h +++ b/include/asm-sparc64/pgtable.h @@ -24,21 +24,23 @@ #include #include -/* The kernel image occupies 0x4000000 to 0x1000000 (4MB --> 16MB). - * The page copy blockops use 0x1000000 to 0x18000000 (16MB --> 24MB). +/* The kernel image occupies 0x4000000 to 0x1000000 (4MB --> 32MB). + * The page copy blockops can use 0x2000000 to 0x10000000. * The PROM resides in an area spanning 0xf0000000 to 0x100000000. - * The vmalloc area spans 0x140000000 to 0x200000000. + * The vmalloc area spans 0x100000000 to 0x200000000. + * Since modules need to be in the lowest 32-bits of the address space, + * we place them right before the OBP area from 0x10000000 to 0xf0000000. * There is a single static kernel PMD which maps from 0x0 to address * 0x400000000. */ -#define TLBTEMP_BASE _AC(0x0000000001000000,UL) -#define MODULES_VADDR _AC(0x0000000002000000,UL) -#define MODULES_LEN _AC(0x000000007e000000,UL) -#define MODULES_END _AC(0x0000000080000000,UL) -#define VMALLOC_START _AC(0x0000000140000000,UL) -#define VMALLOC_END _AC(0x0000000200000000,UL) +#define TLBTEMP_BASE _AC(0x0000000002000000,UL) +#define MODULES_VADDR _AC(0x0000000010000000,UL) +#define MODULES_LEN _AC(0x00000000e0000000,UL) +#define MODULES_END _AC(0x00000000f0000000,UL) #define LOW_OBP_ADDRESS _AC(0x00000000f0000000,UL) #define HI_OBP_ADDRESS _AC(0x0000000100000000,UL) +#define VMALLOC_START _AC(0x0000000100000000,UL) +#define VMALLOC_END _AC(0x0000000200000000,UL) /* XXX All of this needs to be rethought so we can take advantage * XXX cheetah's full 64-bit virtual address space, ie. no more hole From 438282d85de2c8f8c5183fdf55140e51a0b18670 Mon Sep 17 00:00:00 2001 From: Dave Kleikamp Date: Tue, 20 Sep 2005 14:58:11 -0500 Subject: [PATCH 070/204] JFS: don't dereference tlck->ip from txUpdateMap The inode pointer may no longer be valid Signed-off-by: Dave Kleikamp --- fs/jfs/jfs_txnmgr.c | 9 ++++++++- fs/jfs/jfs_txnmgr.h | 1 + 2 files changed, 9 insertions(+), 1 deletion(-) diff --git a/fs/jfs/jfs_txnmgr.c b/fs/jfs/jfs_txnmgr.c index 8132fce8099..9b71ed2674f 100644 --- a/fs/jfs/jfs_txnmgr.c +++ b/fs/jfs/jfs_txnmgr.c @@ -725,6 +725,9 @@ struct tlock *txLock(tid_t tid, struct inode *ip, struct metapage * mp, else tlck->flag = tlckINODELOCK; + if (S_ISDIR(ip->i_mode)) + tlck->flag |= tlckDIRECTORY; + tlck->type = 0; /* bind the tlock and the page */ @@ -1009,6 +1012,8 @@ struct tlock *txMaplock(tid_t tid, struct inode *ip, int type) /* bind the tlock and the object */ tlck->flag = tlckINODELOCK; + if (S_ISDIR(ip->i_mode)) + tlck->flag |= tlckDIRECTORY; tlck->ip = ip; tlck->mp = NULL; @@ -1077,6 +1082,8 @@ struct linelock *txLinelock(struct linelock * tlock) linelock->flag = tlckLINELOCK; linelock->maxcnt = TLOCKLONG; linelock->index = 0; + if (tlck->flag & tlckDIRECTORY) + linelock->flag |= tlckDIRECTORY; /* append linelock after tlock */ linelock->next = tlock->next; @@ -2358,7 +2365,7 @@ static void txUpdateMap(struct tblock * tblk) */ else { /* (maplock->flag & mlckFREE) */ - if (S_ISDIR(tlck->ip->i_mode)) + if (tlck->flag & tlckDIRECTORY) txFreeMap(ipimap, maplock, tblk, COMMIT_PWMAP); else diff --git a/fs/jfs/jfs_txnmgr.h b/fs/jfs/jfs_txnmgr.h index 59ad0f6b723..0e4dc4514c4 100644 --- a/fs/jfs/jfs_txnmgr.h +++ b/fs/jfs/jfs_txnmgr.h @@ -122,6 +122,7 @@ extern struct tlock *TxLock; /* transaction lock table */ #define tlckLOG 0x0800 /* updateMap state */ #define tlckUPDATEMAP 0x0080 +#define tlckDIRECTORY 0x0040 /* freeLock state */ #define tlckFREELOCK 0x0008 #define tlckWRITEPAGE 0x0004 From a9f7baf4d695dbe3028d40ca374c1d11499eaee6 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 20 Sep 2005 21:01:13 +0100 Subject: [PATCH 071/204] [ARM] Fix pcf8583 to build Seems that the Acorn RTC driver missed an update. Fix it. Signed-off-by: Russell King --- drivers/acorn/char/pcf8583.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/acorn/char/pcf8583.c b/drivers/acorn/char/pcf8583.c index 141b4c237a5..2b850e5860a 100644 --- a/drivers/acorn/char/pcf8583.c +++ b/drivers/acorn/char/pcf8583.c @@ -23,12 +23,13 @@ static struct i2c_driver pcf8583_driver; static unsigned short ignore[] = { I2C_CLIENT_END }; static unsigned short normal_addr[] = { 0x50, I2C_CLIENT_END }; +static unsigned short *forces[] = { NULL }; static struct i2c_client_address_data addr_data = { .normal_i2c = normal_addr, .probe = ignore, .ignore = ignore, - .force = ignore, + .forces = forces, }; #define DAT(x) ((unsigned int)(x->dev.driver_data)) From 7a482425f45e695a2e2821b32a887865f016fbf9 Mon Sep 17 00:00:00 2001 From: "Antonino A. Daplas" Date: Wed, 21 Sep 2005 07:30:21 +0800 Subject: [PATCH 072/204] [PATCH] nvidiafb: Fix absence of cursor in nvidiafb A recent change in nvidiafb caused nvidiafb_cursor to always return -ENXIO instead of using the soft_cursor. This will happen if the parameter "hwcur" is not set, which happens to be the default. Signed-off-by: Antonino Daplas Signed-off-by: Linus Torvalds --- drivers/video/Kconfig | 1 + drivers/video/nvidia/nvidia.c | 5 ++++- 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/video/Kconfig b/drivers/video/Kconfig index 31ee13eef7a..773ae11b4a1 100644 --- a/drivers/video/Kconfig +++ b/drivers/video/Kconfig @@ -650,6 +650,7 @@ config FB_NVIDIA select FB_CFB_FILLRECT select FB_CFB_COPYAREA select FB_CFB_IMAGEBLIT + select FB_SOFT_CURSOR help This driver supports graphics boards with the nVidia chips, TNT and newer. For very old chipsets, such as the RIVA128, then use diff --git a/drivers/video/nvidia/nvidia.c b/drivers/video/nvidia/nvidia.c index 3620de0f252..a7f020ada63 100644 --- a/drivers/video/nvidia/nvidia.c +++ b/drivers/video/nvidia/nvidia.c @@ -893,7 +893,7 @@ static int nvidiafb_cursor(struct fb_info *info, struct fb_cursor *cursor) int i, set = cursor->set; u16 fg, bg; - if (!hwcur || cursor->image.width > MAX_CURS || cursor->image.height > MAX_CURS) + if (cursor->image.width > MAX_CURS || cursor->image.height > MAX_CURS) return -ENXIO; NVShowHideCursor(par, 0); @@ -1356,6 +1356,9 @@ static int __devinit nvidia_set_fbinfo(struct fb_info *info) info->pixmap.size = 8 * 1024; info->pixmap.flags = FB_PIXMAP_SYSTEM; + if (!hwcur) + info->fbops->fb_cursor = soft_cursor; + info->var.accel_flags = (!noaccel); switch (par->Architecture) { From 6d67e34de5a378de2f461137944c5d931283d557 Mon Sep 17 00:00:00 2001 From: "Antonino A. Daplas" Date: Wed, 21 Sep 2005 07:32:49 +0800 Subject: [PATCH 073/204] [PATCH] fbdev: Fix reversed back and front porches In fbdev perspective, the frontporch is the lower/right margin and the backporch is the upper/left margin. Correct. Signed-off-by: Antonino Daplas Signed-off-by: Linus Torvalds --- drivers/video/fbcvt.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/video/fbcvt.c b/drivers/video/fbcvt.c index cfa61b512de..0b6af00d197 100644 --- a/drivers/video/fbcvt.c +++ b/drivers/video/fbcvt.c @@ -272,11 +272,11 @@ static void fb_cvt_convert_to_mode(struct fb_cvt_data *cvt, { mode->refresh = cvt->f_refresh; mode->pixclock = KHZ2PICOS(cvt->pixclock/1000); - mode->left_margin = cvt->h_front_porch; - mode->right_margin = cvt->h_back_porch; + mode->left_margin = cvt->h_back_porch; + mode->right_margin = cvt->h_front_porch; mode->hsync_len = cvt->hsync; - mode->upper_margin = cvt->v_front_porch; - mode->lower_margin = cvt->v_back_porch; + mode->upper_margin = cvt->v_back_porch; + mode->lower_margin = cvt->v_front_porch; mode->vsync_len = cvt->vsync; mode->sync &= ~(FB_SYNC_HOR_HIGH_ACT | FB_SYNC_VERT_HIGH_ACT); From 0fb375fb9b93b7d822debc6a734052337ccfdb1f Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Wed, 21 Sep 2005 00:11:37 -0700 Subject: [PATCH 074/204] [AF_PACKET]: Allow for > 8 byte hardware addresses. The convention is that longer addresses will simply extend the hardeware address byte arrays at the end of sockaddr_ll and packet_mreq. In making this change a small information leak was also closed. The code only initializes the hardware address bytes that are used, but all of struct sockaddr_ll was copied to userspace. Now we just copy sockaddr_ll to the last byte of the hardware address used. For error checking larger structures than our internal maximums continue to be allowed but an error is signaled if we can not fit the hardware address into our internal structure. Signed-off-by: Eric W. Biederman Signed-off-by: David S. Miller --- net/packet/af_packet.c | 65 +++++++++++++++++++++++++++++++----------- 1 file changed, 48 insertions(+), 17 deletions(-) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index 8690f171c1e..ee865d88183 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -36,6 +36,11 @@ * Michal Ostrowski : Module initialization cleanup. * Ulises Alonso : Frame number limit removal and * packet_set_ring memory leak. + * Eric Biederman : Allow for > 8 byte hardware addresses. + * The convention is that longer addresses + * will simply extend the hardware address + * byte arrays at the end of sockaddr_ll + * and packet_mreq. * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -161,7 +166,17 @@ struct packet_mclist int count; unsigned short type; unsigned short alen; - unsigned char addr[8]; + unsigned char addr[MAX_ADDR_LEN]; +}; +/* identical to struct packet_mreq except it has + * a longer address field. + */ +struct packet_mreq_max +{ + int mr_ifindex; + unsigned short mr_type; + unsigned short mr_alen; + unsigned char mr_address[MAX_ADDR_LEN]; }; #endif #ifdef CONFIG_PACKET_MMAP @@ -716,6 +731,8 @@ static int packet_sendmsg(struct kiocb *iocb, struct socket *sock, err = -EINVAL; if (msg->msg_namelen < sizeof(struct sockaddr_ll)) goto out; + if (msg->msg_namelen < (saddr->sll_halen + offsetof(struct sockaddr_ll, sll_addr))) + goto out; ifindex = saddr->sll_ifindex; proto = saddr->sll_protocol; addr = saddr->sll_addr; @@ -744,6 +761,12 @@ static int packet_sendmsg(struct kiocb *iocb, struct socket *sock, if (dev->hard_header) { int res; err = -EINVAL; + if (saddr) { + if (saddr->sll_halen != dev->addr_len) + goto out_free; + if (saddr->sll_hatype != dev->type) + goto out_free; + } res = dev->hard_header(skb, dev, ntohs(proto), addr, NULL, len); if (sock->type != SOCK_DGRAM) { skb->tail = skb->data; @@ -1045,6 +1068,7 @@ static int packet_recvmsg(struct kiocb *iocb, struct socket *sock, struct sock *sk = sock->sk; struct sk_buff *skb; int copied, err; + struct sockaddr_ll *sll; err = -EINVAL; if (flags & ~(MSG_PEEK|MSG_DONTWAIT|MSG_TRUNC|MSG_CMSG_COMPAT)) @@ -1056,16 +1080,6 @@ static int packet_recvmsg(struct kiocb *iocb, struct socket *sock, return -ENODEV; #endif - /* - * If the address length field is there to be filled in, we fill - * it in now. - */ - - if (sock->type == SOCK_PACKET) - msg->msg_namelen = sizeof(struct sockaddr_pkt); - else - msg->msg_namelen = sizeof(struct sockaddr_ll); - /* * Call the generic datagram receiver. This handles all sorts * of horrible races and re-entrancy so we can forget about it @@ -1086,6 +1100,17 @@ static int packet_recvmsg(struct kiocb *iocb, struct socket *sock, if(skb==NULL) goto out; + /* + * If the address length field is there to be filled in, we fill + * it in now. + */ + + sll = (struct sockaddr_ll*)skb->cb; + if (sock->type == SOCK_PACKET) + msg->msg_namelen = sizeof(struct sockaddr_pkt); + else + msg->msg_namelen = sll->sll_halen + offsetof(struct sockaddr_ll, sll_addr); + /* * You lose any data beyond the buffer you gave. If it worries a * user program they can ask the device for its MTU anyway. @@ -1166,7 +1191,7 @@ static int packet_getname(struct socket *sock, struct sockaddr *uaddr, sll->sll_hatype = 0; /* Bad: we have no ARPHRD_UNSPEC */ sll->sll_halen = 0; } - *uaddr_len = sizeof(*sll); + *uaddr_len = offsetof(struct sockaddr_ll, sll_addr) + sll->sll_halen; return 0; } @@ -1199,7 +1224,7 @@ static void packet_dev_mclist(struct net_device *dev, struct packet_mclist *i, i } } -static int packet_mc_add(struct sock *sk, struct packet_mreq *mreq) +static int packet_mc_add(struct sock *sk, struct packet_mreq_max *mreq) { struct packet_sock *po = pkt_sk(sk); struct packet_mclist *ml, *i; @@ -1249,7 +1274,7 @@ done: return err; } -static int packet_mc_drop(struct sock *sk, struct packet_mreq *mreq) +static int packet_mc_drop(struct sock *sk, struct packet_mreq_max *mreq) { struct packet_mclist *ml, **mlp; @@ -1315,11 +1340,17 @@ packet_setsockopt(struct socket *sock, int level, int optname, char __user *optv case PACKET_ADD_MEMBERSHIP: case PACKET_DROP_MEMBERSHIP: { - struct packet_mreq mreq; - if (optlen sizeof(mreq)) + len = sizeof(mreq); + if (copy_from_user(&mreq,optval,len)) return -EFAULT; + if (len < (mreq.mr_alen + offsetof(struct packet_mreq, mr_address))) + return -EINVAL; if (optname == PACKET_ADD_MEMBERSHIP) ret = packet_mc_add(sk, &mreq); else From 78c6671a88313fd3c4364dc46e8c8186612616b8 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 21 Sep 2005 00:15:39 -0700 Subject: [PATCH 075/204] [FIB_TRIE]: message cleanup Cleanup the printk's in fib_trie: * Convert a couple of places in the dump code to BUG_ON * Put log level's on each message The version message really needed the message since it leaks out on the pretty Fedora bootup. Signed-off-by: Stephen Hemminger Acked-by: Robert Olsson , Signed-off-by: David S. Miller --- net/ipv4/fib_trie.c | 15 +++------------ 1 file changed, 3 insertions(+), 12 deletions(-) diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c index 90ae70870a1..50c0519cd70 100644 --- a/net/ipv4/fib_trie.c +++ b/net/ipv4/fib_trie.c @@ -1086,7 +1086,7 @@ fib_insert_node(struct trie *t, int *err, u32 key, int plen) } if (tp && tp->pos + tp->bits > 32) - printk("ERROR tp=%p pos=%d, bits=%d, key=%0x plen=%d\n", + printk(KERN_WARNING "fib_trie tp=%p pos=%d, bits=%d, key=%0x plen=%d\n", tp, tp->pos, tp->bits, key, plen); /* Rebalance the trie */ @@ -1832,16 +1832,7 @@ static int fn_trie_dump_fa(t_key key, int plen, struct list_head *fah, struct fi i++; continue; } - if (fa->fa_info->fib_nh == NULL) { - printk("Trie error _fib_nh=NULL in fa[%d] k=%08x plen=%d\n", i, key, plen); - i++; - continue; - } - if (fa->fa_info == NULL) { - printk("Trie error fa_info=NULL in fa[%d] k=%08x plen=%d\n", i, key, plen); - i++; - continue; - } + BUG_ON(!fa->fa_info); if (fib_dump_info(skb, NETLINK_CB(cb->skb).pid, cb->nlh->nlmsg_seq, @@ -1964,7 +1955,7 @@ struct fib_table * __init fib_hash_init(int id) trie_main = t; if (id == RT_TABLE_LOCAL) - printk("IPv4 FIB: Using LC-trie version %s\n", VERSION); + printk(KERN_INFO "IPv4 FIB: Using LC-trie version %s\n", VERSION); return tb; } From 7957aed72b104f7528da484f292442c6f3a131a1 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 21 Sep 2005 00:19:46 -0700 Subject: [PATCH 076/204] [TCP]: Set default congestion control correctly for incoming connections. Patch from Joel Sing to fix the default congestion control algorithm for incoming connections. If a new congestion control handler is added (via module), it should become the default for new connections. Instead, the incoming connections use reno. The cause is incorrect initialisation causes the tcp_init_congestion_control() function to return after the initial if test fails. Signed-off-by: Stephen Hemminger Acked-by: Ian McDonald Signed-off-by: David S. Miller --- net/ipv4/tcp_minisocks.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv4/tcp_minisocks.c b/net/ipv4/tcp_minisocks.c index a88db28b0af..b1a63b2c6b4 100644 --- a/net/ipv4/tcp_minisocks.c +++ b/net/ipv4/tcp_minisocks.c @@ -384,7 +384,7 @@ struct sock *tcp_create_openreq_child(struct sock *sk, struct request_sock *req, newtp->frto_counter = 0; newtp->frto_highmark = 0; - newicsk->icsk_ca_ops = &tcp_reno; + newicsk->icsk_ca_ops = &tcp_init_congestion_ops; tcp_set_ca_state(newsk, TCP_CA_Open); tcp_init_xmit_timers(newsk); From 1a3e64aa480101d11d8cac94ca57fd8bbe2536b7 Mon Sep 17 00:00:00 2001 From: Tom 'spot' Callaway Date: Wed, 21 Sep 2005 00:29:16 -0700 Subject: [PATCH 077/204] [ATYFB]: Fix build with CONFIG_FB_ATY_GENERIC_LCD disabled. Signed-off-by: Tom 'spot' Callaway Signed-off-by: David S. Miller --- drivers/video/aty/xlinit.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/video/aty/xlinit.c b/drivers/video/aty/xlinit.c index 92643af1258..0bea0d8d782 100644 --- a/drivers/video/aty/xlinit.c +++ b/drivers/video/aty/xlinit.c @@ -174,7 +174,7 @@ int atyfb_xl_init(struct fb_info *info) const struct xl_card_cfg_t * card = &card_cfg[xl_card]; struct atyfb_par *par = (struct atyfb_par *) info->par; union aty_pll pll; - int i, err; + int err; u32 temp; aty_st_8(CONFIG_STAT0, 0x85, par); @@ -252,9 +252,12 @@ int atyfb_xl_init(struct fb_info *info) aty_st_le32(0xEC, 0x00000000, par); aty_st_le32(0xFC, 0x00000000, par); +#if defined (CONFIG_FB_ATY_GENERIC_LCD) + int i; for (i=0; i Date: Wed, 21 Sep 2005 00:41:22 -0700 Subject: [PATCH 078/204] [SPARC] cs4231: Fix SBUS support in this driver. From: Christopher Zimmermann This patch enables SBus support for the cs4231 sound driver. It is tested on an Ultra2. Capture and playback both work. I experienced lags and crashes using certain threaded players like ogg123 and mp3blaster, while the former is lagging far more. This behavior may be specific to SMP systems. It is reproducable using the dummy sound card driver. Sox works flawlessly. Setting up the calculation of ptr in snd_cs4231_playback_pointer was a bit strange. I got it to work by not incrementing the [pc]_periods_sent counter when starting DMA the first time in cs4231_dma_trigger. Therefore this dummy thing. [ I did some minor cleanups -DaveM ] Signed-off-by: David S. Miller --- sound/sparc/cs4231.c | 323 ++++++++++++++++++++++--------------------- 1 file changed, 166 insertions(+), 157 deletions(-) diff --git a/sound/sparc/cs4231.c b/sound/sparc/cs4231.c index 2fb27c4e951..f4361c518e4 100644 --- a/sound/sparc/cs4231.c +++ b/sound/sparc/cs4231.c @@ -173,7 +173,7 @@ static cs4231_t *cs4231_list; #define CS4231_GLOBALIRQ 0x01 /* IRQ is active */ -/* definitions for codec irq status */ +/* definitions for codec irq status - CS4231_IRQ_STATUS */ #define CS4231_PLAYBACK_IRQ 0x10 #define CS4231_RECORD_IRQ 0x20 @@ -402,7 +402,7 @@ static void snd_cs4231_outm(cs4231_t *chip, unsigned char reg, udelay(100); #ifdef CONFIG_SND_DEBUG if (__cs4231_readb(chip, CS4231P(chip, REGSEL)) & CS4231_INIT) - snd_printk("outm: auto calibration time out - reg = 0x%x, value = 0x%x\n", reg, value); + snd_printdd("outm: auto calibration time out - reg = 0x%x, value = 0x%x\n", reg, value); #endif if (chip->calibrate_mute) { chip->image[reg] &= mask; @@ -425,6 +425,10 @@ static void snd_cs4231_dout(cs4231_t *chip, unsigned char reg, unsigned char val timeout > 0 && (__cs4231_readb(chip, CS4231P(chip, REGSEL)) & CS4231_INIT); timeout--) udelay(100); +#ifdef CONFIG_SND_DEBUG + if (__cs4231_readb(chip, CS4231P(chip, REGSEL)) & CS4231_INIT) + snd_printdd("out: auto calibration time out - reg = 0x%x, value = 0x%x\n", reg, value); +#endif __cs4231_writeb(chip, chip->mce_bit | reg, CS4231P(chip, REGSEL)); __cs4231_writeb(chip, value, CS4231P(chip, REG)); mb(); @@ -440,15 +444,12 @@ static void snd_cs4231_out(cs4231_t *chip, unsigned char reg, unsigned char valu udelay(100); #ifdef CONFIG_SND_DEBUG if (__cs4231_readb(chip, CS4231P(chip, REGSEL)) & CS4231_INIT) - snd_printk("out: auto calibration time out - reg = 0x%x, value = 0x%x\n", reg, value); + snd_printdd("out: auto calibration time out - reg = 0x%x, value = 0x%x\n", reg, value); #endif __cs4231_writeb(chip, chip->mce_bit | reg, CS4231P(chip, REGSEL)); __cs4231_writeb(chip, value, CS4231P(chip, REG)); chip->image[reg] = value; mb(); -#if 0 - printk("codec out - reg 0x%x = 0x%x\n", chip->mce_bit | reg, value); -#endif } static unsigned char snd_cs4231_in(cs4231_t *chip, unsigned char reg) @@ -462,61 +463,14 @@ static unsigned char snd_cs4231_in(cs4231_t *chip, unsigned char reg) udelay(100); #ifdef CONFIG_SND_DEBUG if (__cs4231_readb(chip, CS4231P(chip, REGSEL)) & CS4231_INIT) - snd_printk("in: auto calibration time out - reg = 0x%x\n", reg); + snd_printdd("in: auto calibration time out - reg = 0x%x\n", reg); #endif __cs4231_writeb(chip, chip->mce_bit | reg, CS4231P(chip, REGSEL)); mb(); ret = __cs4231_readb(chip, CS4231P(chip, REG)); -#if 0 - printk("codec in - reg 0x%x = 0x%x\n", chip->mce_bit | reg, ret); -#endif return ret; } -#if 0 - -static void snd_cs4231_debug(cs4231_t *chip) -{ - printk("CS4231 REGS: INDEX = 0x%02x ", - __cs4231_readb(chip, CS4231P(chip, REGSEL))); - printk(" STATUS = 0x%02x\n", - __cs4231_readb(chip, CS4231P(chip, STATUS))); - printk(" 0x00: left input = 0x%02x ", snd_cs4231_in(chip, 0x00)); - printk(" 0x10: alt 1 (CFIG 2) = 0x%02x\n", snd_cs4231_in(chip, 0x10)); - printk(" 0x01: right input = 0x%02x ", snd_cs4231_in(chip, 0x01)); - printk(" 0x11: alt 2 (CFIG 3) = 0x%02x\n", snd_cs4231_in(chip, 0x11)); - printk(" 0x02: GF1 left input = 0x%02x ", snd_cs4231_in(chip, 0x02)); - printk(" 0x12: left line in = 0x%02x\n", snd_cs4231_in(chip, 0x12)); - printk(" 0x03: GF1 right input = 0x%02x ", snd_cs4231_in(chip, 0x03)); - printk(" 0x13: right line in = 0x%02x\n", snd_cs4231_in(chip, 0x13)); - printk(" 0x04: CD left input = 0x%02x ", snd_cs4231_in(chip, 0x04)); - printk(" 0x14: timer low = 0x%02x\n", snd_cs4231_in(chip, 0x14)); - printk(" 0x05: CD right input = 0x%02x ", snd_cs4231_in(chip, 0x05)); - printk(" 0x15: timer high = 0x%02x\n", snd_cs4231_in(chip, 0x15)); - printk(" 0x06: left output = 0x%02x ", snd_cs4231_in(chip, 0x06)); - printk(" 0x16: left MIC (PnP) = 0x%02x\n", snd_cs4231_in(chip, 0x16)); - printk(" 0x07: right output = 0x%02x ", snd_cs4231_in(chip, 0x07)); - printk(" 0x17: right MIC (PnP) = 0x%02x\n", snd_cs4231_in(chip, 0x17)); - printk(" 0x08: playback format = 0x%02x ", snd_cs4231_in(chip, 0x08)); - printk(" 0x18: IRQ status = 0x%02x\n", snd_cs4231_in(chip, 0x18)); - printk(" 0x09: iface (CFIG 1) = 0x%02x ", snd_cs4231_in(chip, 0x09)); - printk(" 0x19: left line out = 0x%02x\n", snd_cs4231_in(chip, 0x19)); - printk(" 0x0a: pin control = 0x%02x ", snd_cs4231_in(chip, 0x0a)); - printk(" 0x1a: mono control = 0x%02x\n", snd_cs4231_in(chip, 0x1a)); - printk(" 0x0b: init & status = 0x%02x ", snd_cs4231_in(chip, 0x0b)); - printk(" 0x1b: right line out = 0x%02x\n", snd_cs4231_in(chip, 0x1b)); - printk(" 0x0c: revision & mode = 0x%02x ", snd_cs4231_in(chip, 0x0c)); - printk(" 0x1c: record format = 0x%02x\n", snd_cs4231_in(chip, 0x1c)); - printk(" 0x0d: loopback = 0x%02x ", snd_cs4231_in(chip, 0x0d)); - printk(" 0x1d: var freq (PnP) = 0x%02x\n", snd_cs4231_in(chip, 0x1d)); - printk(" 0x0e: ply upr count = 0x%02x ", snd_cs4231_in(chip, 0x0e)); - printk(" 0x1e: rec upr count = 0x%02x\n", snd_cs4231_in(chip, 0x1e)); - printk(" 0x0f: ply lwr count = 0x%02x ", snd_cs4231_in(chip, 0x0f)); - printk(" 0x1f: rec lwr count = 0x%02x\n", snd_cs4231_in(chip, 0x1f)); -} - -#endif - /* * CS4231 detection / MCE routines */ @@ -528,11 +482,12 @@ static void snd_cs4231_busy_wait(cs4231_t *chip) /* huh.. looks like this sequence is proper for CS4231A chip (GUS MAX) */ for (timeout = 5; timeout > 0; timeout--) __cs4231_readb(chip, CS4231P(chip, REGSEL)); + /* end of cleanup sequence */ - for (timeout = 250; + for (timeout = 500; timeout > 0 && (__cs4231_readb(chip, CS4231P(chip, REGSEL)) & CS4231_INIT); timeout--) - udelay(100); + udelay(1000); } static void snd_cs4231_mce_up(cs4231_t *chip) @@ -545,12 +500,12 @@ static void snd_cs4231_mce_up(cs4231_t *chip) udelay(100); #ifdef CONFIG_SND_DEBUG if (__cs4231_readb(chip, CS4231P(chip, REGSEL)) & CS4231_INIT) - snd_printk("mce_up - auto calibration time out (0)\n"); + snd_printdd("mce_up - auto calibration time out (0)\n"); #endif chip->mce_bit |= CS4231_MCE; timeout = __cs4231_readb(chip, CS4231P(chip, REGSEL)); if (timeout == 0x80) - snd_printk("mce_up [%p]: serious init problem - codec still busy\n", chip->port); + snd_printdd("mce_up [%p]: serious init problem - codec still busy\n", chip->port); if (!(timeout & CS4231_MCE)) __cs4231_writeb(chip, chip->mce_bit | (timeout & 0x1f), CS4231P(chip, REGSEL)); spin_unlock_irqrestore(&chip->lock, flags); @@ -563,18 +518,15 @@ static void snd_cs4231_mce_down(cs4231_t *chip) spin_lock_irqsave(&chip->lock, flags); snd_cs4231_busy_wait(chip); -#if 0 - printk("(1) timeout = %i\n", timeout); -#endif #ifdef CONFIG_SND_DEBUG if (__cs4231_readb(chip, CS4231P(chip, REGSEL)) & CS4231_INIT) - snd_printk("mce_down [%p] - auto calibration time out (0)\n", CS4231P(chip, REGSEL)); + snd_printdd("mce_down [%p] - auto calibration time out (0)\n", CS4231P(chip, REGSEL)); #endif chip->mce_bit &= ~CS4231_MCE; timeout = __cs4231_readb(chip, CS4231P(chip, REGSEL)); __cs4231_writeb(chip, chip->mce_bit | (timeout & 0x1f), CS4231P(chip, REGSEL)); if (timeout == 0x80) - snd_printk("mce_down [%p]: serious init problem - codec still busy\n", chip->port); + snd_printdd("mce_down [%p]: serious init problem - codec still busy\n", chip->port); if ((timeout & CS4231_MCE) == 0) { spin_unlock_irqrestore(&chip->lock, flags); return; @@ -590,9 +542,7 @@ static void snd_cs4231_mce_down(cs4231_t *chip) spin_unlock_irqrestore(&chip->lock, flags); return; } -#if 0 - printk("(2) timeout = %i, jiffies = %li\n", timeout, jiffies); -#endif + /* in 10ms increments, check condition, up to 250ms */ timeout = 25; while (snd_cs4231_in(chip, CS4231_TEST_INIT) & CS4231_CALIB_IN_PROGRESS) { @@ -604,9 +554,7 @@ static void snd_cs4231_mce_down(cs4231_t *chip) msleep(10); spin_lock_irqsave(&chip->lock, flags); } -#if 0 - printk("(3) jiffies = %li\n", jiffies); -#endif + /* in 10ms increments, check condition, up to 100ms */ timeout = 10; while (__cs4231_readb(chip, CS4231P(chip, REGSEL)) & CS4231_INIT) { @@ -619,54 +567,58 @@ static void snd_cs4231_mce_down(cs4231_t *chip) spin_lock_irqsave(&chip->lock, flags); } spin_unlock_irqrestore(&chip->lock, flags); -#if 0 - printk("(4) jiffies = %li\n", jiffies); - snd_printk("mce_down - exit = 0x%x\n", __cs4231_readb(chip, CS4231P(chip, REGSEL))); -#endif } -#if 0 /* Unused for now... */ -static unsigned int snd_cs4231_get_count(unsigned char format, unsigned int size) -{ - switch (format & 0xe0) { - case CS4231_LINEAR_16: - case CS4231_LINEAR_16_BIG: - size >>= 1; - break; - case CS4231_ADPCM_16: - return size >> 2; - } - if (format & CS4231_STEREO) - size >>= 1; - return size; -} -#endif - #ifdef EBUS_SUPPORT static void snd_cs4231_ebus_advance_dma(struct ebus_dma_info *p, snd_pcm_substream_t *substream, unsigned int *periods_sent) { snd_pcm_runtime_t *runtime = substream->runtime; while (1) { - unsigned int dma_size = snd_pcm_lib_period_bytes(substream); - unsigned int offset = dma_size * (*periods_sent); + unsigned int period_size = snd_pcm_lib_period_bytes(substream); + unsigned int offset = period_size * (*periods_sent); - if (dma_size >= (1 << 24)) + if (period_size >= (1 << 24)) BUG(); - if (ebus_dma_request(p, runtime->dma_addr + offset, dma_size)) + if (ebus_dma_request(p, runtime->dma_addr + offset, period_size)) return; -#if 0 - printk("ebus_advance: Sent period %u (size[%x] offset[%x])\n", - (*periods_sent), dma_size, offset); -#endif (*periods_sent) = ((*periods_sent) + 1) % runtime->periods; } } #endif -static void cs4231_dma_trigger(cs4231_t *chip, unsigned int what, int on) +#ifdef SBUS_SUPPORT +static void snd_cs4231_sbus_advance_dma(snd_pcm_substream_t *substream, unsigned int *periods_sent) { + cs4231_t *chip = snd_pcm_substream_chip(substream); + snd_pcm_runtime_t *runtime = substream->runtime; + + unsigned int period_size = snd_pcm_lib_period_bytes(substream); + unsigned int offset = period_size * (*periods_sent % runtime->periods); + + if (runtime->period_size > 0xffff + 1) + BUG(); + + switch (substream->stream) { + case SNDRV_PCM_STREAM_PLAYBACK: + sbus_writel(runtime->dma_addr + offset, chip->port + APCPNVA); + sbus_writel(period_size, chip->port + APCPNC); + break; + case SNDRV_PCM_STREAM_CAPTURE: + sbus_writel(runtime->dma_addr + offset, chip->port + APCCNVA); + sbus_writel(period_size, chip->port + APCCNC); + break; + } + + (*periods_sent) = (*periods_sent + 1) % runtime->periods; +} +#endif + +static void cs4231_dma_trigger(snd_pcm_substream_t *substream, unsigned int what, int on) +{ + cs4231_t *chip = snd_pcm_substream_chip(substream); + #ifdef EBUS_SUPPORT if (chip->flags & CS4231_FLAG_EBUS) { if (what & CS4231_PLAYBACK_ENABLE) { @@ -694,6 +646,60 @@ static void cs4231_dma_trigger(cs4231_t *chip, unsigned int what, int on) } else { #endif #ifdef SBUS_SUPPORT + u32 csr = sbus_readl(chip->port + APCCSR); + /* I don't know why, but on sbus the period counter must + * only start counting after the first period is sent. + * Therefore this dummy thing. + */ + unsigned int dummy = 0; + + switch (what) { + case CS4231_PLAYBACK_ENABLE: + if (on) { + csr &= ~APC_XINT_PLAY; + sbus_writel(csr, chip->port + APCCSR); + + csr &= ~APC_PPAUSE; + sbus_writel(csr, chip->port + APCCSR); + + snd_cs4231_sbus_advance_dma(substream, &dummy); + + csr |= APC_GENL_INT | APC_PLAY_INT | APC_XINT_ENA | + APC_XINT_PLAY | APC_XINT_EMPT | APC_XINT_GENL | + APC_XINT_PENA | APC_PDMA_READY; + sbus_writel(csr, chip->port + APCCSR); + } else { + csr |= APC_PPAUSE; + sbus_writel(csr, chip->port + APCCSR); + + csr &= ~APC_PDMA_READY; + sbus_writel(csr, chip->port + APCCSR); + } + break; + case CS4231_RECORD_ENABLE: + if (on) { + csr &= ~APC_XINT_CAPT; + sbus_writel(csr, chip->port + APCCSR); + + csr &= ~APC_CPAUSE; + sbus_writel(csr, chip->port + APCCSR); + + snd_cs4231_sbus_advance_dma(substream, &dummy); + + csr |= APC_GENL_INT | APC_CAPT_INT | APC_XINT_ENA | + APC_XINT_CAPT | APC_XINT_CEMP | APC_XINT_GENL | + APC_CDMA_READY; + + sbus_writel(csr, chip->port + APCCSR); + } else { + csr |= APC_CPAUSE; + sbus_writel(csr, chip->port + APCCSR); + + csr &= ~APC_CDMA_READY; + sbus_writel(csr, chip->port + APCCSR); + } + break; + } #endif #ifdef EBUS_SUPPORT } @@ -725,25 +731,12 @@ static int snd_cs4231_trigger(snd_pcm_substream_t *substream, int cmd) } } -#if 0 - printk("TRIGGER: what[%x] on(%d)\n", - what, (cmd == SNDRV_PCM_TRIGGER_START)); -#endif - spin_lock_irqsave(&chip->lock, flags); if (cmd == SNDRV_PCM_TRIGGER_START) { - cs4231_dma_trigger(chip, what, 1); + cs4231_dma_trigger(substream, what, 1); chip->image[CS4231_IFACE_CTRL] |= what; - if (what & CS4231_PLAYBACK_ENABLE) { - snd_cs4231_out(chip, CS4231_PLY_LWR_CNT, 0xff); - snd_cs4231_out(chip, CS4231_PLY_UPR_CNT, 0xff); - } - if (what & CS4231_RECORD_ENABLE) { - snd_cs4231_out(chip, CS4231_REC_LWR_CNT, 0xff); - snd_cs4231_out(chip, CS4231_REC_UPR_CNT, 0xff); - } } else { - cs4231_dma_trigger(chip, what, 0); + cs4231_dma_trigger(substream, what, 0); chip->image[CS4231_IFACE_CTRL] &= ~what; } snd_cs4231_out(chip, CS4231_IFACE_CTRL, @@ -755,9 +748,7 @@ static int snd_cs4231_trigger(snd_pcm_substream_t *substream, int cmd) result = -EINVAL; break; } -#if 0 - snd_cs4231_debug(chip); -#endif + return result; } @@ -790,9 +781,6 @@ static unsigned char snd_cs4231_get_format(cs4231_t *chip, int format, int chann } if (channels > 1) rformat |= CS4231_STEREO; -#if 0 - snd_printk("get_format: 0x%x (mode=0x%x)\n", format, mode); -#endif return rformat; } @@ -944,7 +932,7 @@ static void snd_cs4231_init(cs4231_t *chip) snd_cs4231_mce_down(chip); #ifdef SNDRV_DEBUG_MCE - snd_printk("init: (1)\n"); + snd_printdd("init: (1)\n"); #endif snd_cs4231_mce_up(chip); spin_lock_irqsave(&chip->lock, flags); @@ -957,7 +945,7 @@ static void snd_cs4231_init(cs4231_t *chip) snd_cs4231_mce_down(chip); #ifdef SNDRV_DEBUG_MCE - snd_printk("init: (2)\n"); + snd_printdd("init: (2)\n"); #endif snd_cs4231_mce_up(chip); @@ -967,7 +955,7 @@ static void snd_cs4231_init(cs4231_t *chip) snd_cs4231_mce_down(chip); #ifdef SNDRV_DEBUG_MCE - snd_printk("init: (3) - afei = 0x%x\n", chip->image[CS4231_ALT_FEATURE_1]); + snd_printdd("init: (3) - afei = 0x%x\n", chip->image[CS4231_ALT_FEATURE_1]); #endif spin_lock_irqsave(&chip->lock, flags); @@ -981,7 +969,7 @@ static void snd_cs4231_init(cs4231_t *chip) snd_cs4231_mce_down(chip); #ifdef SNDRV_DEBUG_MCE - snd_printk("init: (4)\n"); + snd_printdd("init: (4)\n"); #endif snd_cs4231_mce_up(chip); @@ -991,7 +979,7 @@ static void snd_cs4231_init(cs4231_t *chip) snd_cs4231_mce_down(chip); #ifdef SNDRV_DEBUG_MCE - snd_printk("init: (5)\n"); + snd_printdd("init: (5)\n"); #endif } @@ -1022,6 +1010,7 @@ static int snd_cs4231_open(cs4231_t *chip, unsigned int mode) CS4231_RECORD_IRQ | CS4231_TIMER_IRQ); snd_cs4231_out(chip, CS4231_IRQ_STATUS, 0); + spin_unlock_irqrestore(&chip->lock, flags); chip->mode = mode; @@ -1136,11 +1125,21 @@ static int snd_cs4231_playback_hw_free(snd_pcm_substream_t *substream) static int snd_cs4231_playback_prepare(snd_pcm_substream_t *substream) { cs4231_t *chip = snd_pcm_substream_chip(substream); + snd_pcm_runtime_t *runtime = substream->runtime; unsigned long flags; spin_lock_irqsave(&chip->lock, flags); + chip->image[CS4231_IFACE_CTRL] &= ~(CS4231_PLAYBACK_ENABLE | CS4231_PLAYBACK_PIO); + + if (runtime->period_size > 0xffff + 1) + BUG(); + + snd_cs4231_out(chip, CS4231_PLY_LWR_CNT, (runtime->period_size - 1) & 0x00ff); + snd_cs4231_out(chip, CS4231_PLY_UPR_CNT, (runtime->period_size - 1) >> 8 & 0x00ff); + chip->p_periods_sent = 0; + spin_unlock_irqrestore(&chip->lock, flags); return 0; @@ -1172,12 +1171,16 @@ static int snd_cs4231_capture_hw_free(snd_pcm_substream_t *substream) static int snd_cs4231_capture_prepare(snd_pcm_substream_t *substream) { cs4231_t *chip = snd_pcm_substream_chip(substream); + snd_pcm_runtime_t *runtime = substream->runtime; unsigned long flags; spin_lock_irqsave(&chip->lock, flags); chip->image[CS4231_IFACE_CTRL] &= ~(CS4231_RECORD_ENABLE | CS4231_RECORD_PIO); + snd_cs4231_out(chip, CS4231_REC_LWR_CNT, (runtime->period_size - 1) & 0x00ff); + snd_cs4231_out(chip, CS4231_REC_LWR_CNT, (runtime->period_size - 1) >> 8 & 0x00ff); + spin_unlock_irqrestore(&chip->lock, flags); return 0; @@ -1196,53 +1199,61 @@ static void snd_cs4231_overrange(cs4231_t *chip) chip->capture_substream->runtime->overrange++; } -static void snd_cs4231_generic_interrupt(cs4231_t *chip) +static irqreturn_t snd_cs4231_generic_interrupt(cs4231_t *chip) { unsigned long flags; unsigned char status; + /*This is IRQ is not raised by the cs4231*/ + if (!(__cs4231_readb(chip, CS4231P(chip, STATUS)) & CS4231_GLOBALIRQ)) + return IRQ_NONE; + status = snd_cs4231_in(chip, CS4231_IRQ_STATUS); - if (!status) - return; if (status & CS4231_TIMER_IRQ) { if (chip->timer) snd_timer_interrupt(chip->timer, chip->timer->sticks); } - if (status & CS4231_PLAYBACK_IRQ) - snd_pcm_period_elapsed(chip->playback_substream); - if (status & CS4231_RECORD_IRQ) { + + if (status & CS4231_RECORD_IRQ) snd_cs4231_overrange(chip); - snd_pcm_period_elapsed(chip->capture_substream); - } /* ACK the CS4231 interrupt. */ spin_lock_irqsave(&chip->lock, flags); snd_cs4231_outm(chip, CS4231_IRQ_STATUS, ~CS4231_ALL_IRQS | ~status, 0); spin_unlock_irqrestore(&chip->lock, flags); + + return 0; } #ifdef SBUS_SUPPORT static irqreturn_t snd_cs4231_sbus_interrupt(int irq, void *dev_id, struct pt_regs *regs) { cs4231_t *chip = dev_id; - u32 csr; - - csr = sbus_readl(chip->port + APCCSR); - if (!(csr & (APC_INT_PENDING | - APC_PLAY_INT | - APC_CAPT_INT | - APC_GENL_INT | - APC_XINT_PEMP | - APC_XINT_CEMP))) - return IRQ_NONE; /* ACK the APC interrupt. */ + u32 csr = sbus_readl(chip->port + APCCSR); + sbus_writel(csr, chip->port + APCCSR); - snd_cs4231_generic_interrupt(chip); + if ((chip->image[CS4231_IFACE_CTRL] & CS4231_PLAYBACK_ENABLE) && + (csr & APC_PLAY_INT) && + (csr & APC_XINT_PNVA) && + !(csr & APC_XINT_EMPT)) { + snd_cs4231_sbus_advance_dma(chip->playback_substream, + &chip->p_periods_sent); + snd_pcm_period_elapsed(chip->playback_substream); + } - return IRQ_HANDLED; + if ((chip->image[CS4231_IFACE_CTRL] & CS4231_RECORD_ENABLE) && + (csr & APC_CAPT_INT) && + (csr & APC_XINT_CNVA)) { + snd_cs4231_sbus_advance_dma(chip->capture_substream, + &chip->c_periods_sent); + snd_pcm_period_elapsed(chip->capture_substream); + } + + return snd_cs4231_generic_interrupt(chip); } #endif @@ -1290,7 +1301,8 @@ static snd_pcm_uframes_t snd_cs4231_playback_pointer(snd_pcm_substream_t *substr #ifdef EBUS_SUPPORT } #endif - ptr += (period_bytes - residue); + ptr += period_bytes - residue; + return bytes_to_frames(substream->runtime, ptr); } @@ -1314,7 +1326,7 @@ static snd_pcm_uframes_t snd_cs4231_capture_pointer(snd_pcm_substream_t * substr #ifdef EBUS_SUPPORT } #endif - ptr += (period_bytes - residue); + ptr += period_bytes - residue; return bytes_to_frames(substream->runtime, ptr); } @@ -1328,9 +1340,6 @@ static int snd_cs4231_probe(cs4231_t *chip) int i, id, vers; unsigned char *ptr; -#if 0 - snd_cs4231_debug(chip); -#endif id = vers = 0; for (i = 0; i < 50; i++) { mb(); @@ -1985,13 +1994,13 @@ static int __init snd_cs4231_sbus_create(snd_card_t *card, chip->port = sbus_ioremap(&sdev->resource[0], 0, chip->regs_size, "cs4231"); if (!chip->port) { - snd_printk("cs4231-%d: Unable to map chip registers.\n", dev); + snd_printdd("cs4231-%d: Unable to map chip registers.\n", dev); return -EIO; } if (request_irq(sdev->irqs[0], snd_cs4231_sbus_interrupt, SA_SHIRQ, "cs4231", chip)) { - snd_printk("cs4231-%d: Unable to grab SBUS IRQ %s\n", + snd_printdd("cs4231-%d: Unable to grab SBUS IRQ %s\n", dev, __irq_itoa(sdev->irqs[0])); snd_cs4231_sbus_free(chip); @@ -2113,29 +2122,29 @@ static int __init snd_cs4231_ebus_create(snd_card_t *card, chip->eb2c.regs = ioremap(edev->resource[2].start, 0x10); if (!chip->port || !chip->eb2p.regs || !chip->eb2c.regs) { snd_cs4231_ebus_free(chip); - snd_printk("cs4231-%d: Unable to map chip registers.\n", dev); + snd_printdd("cs4231-%d: Unable to map chip registers.\n", dev); return -EIO; } if (ebus_dma_register(&chip->eb2c)) { snd_cs4231_ebus_free(chip); - snd_printk("cs4231-%d: Unable to register EBUS capture DMA\n", dev); + snd_printdd("cs4231-%d: Unable to register EBUS capture DMA\n", dev); return -EBUSY; } if (ebus_dma_irq_enable(&chip->eb2c, 1)) { snd_cs4231_ebus_free(chip); - snd_printk("cs4231-%d: Unable to enable EBUS capture IRQ\n", dev); + snd_printdd("cs4231-%d: Unable to enable EBUS capture IRQ\n", dev); return -EBUSY; } if (ebus_dma_register(&chip->eb2p)) { snd_cs4231_ebus_free(chip); - snd_printk("cs4231-%d: Unable to register EBUS play DMA\n", dev); + snd_printdd("cs4231-%d: Unable to register EBUS play DMA\n", dev); return -EBUSY; } if (ebus_dma_irq_enable(&chip->eb2p, 1)) { snd_cs4231_ebus_free(chip); - snd_printk("cs4231-%d: Unable to enable EBUS play IRQ\n", dev); + snd_printdd("cs4231-%d: Unable to enable EBUS play IRQ\n", dev); return -EBUSY; } From d79e743e9fcf03f521300a970eb1ab69641910ba Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Wed, 21 Sep 2005 14:14:22 +1000 Subject: [PATCH 079/204] [PATCH] ppc64: Fix PCI flags when using OF device tree My code to set up the PCI tree from the Open Firmware device tree was setting IORESOURCE_* flags on the resources for the devices, but not the PCI_BASE_ADDRESS_* flags. This meant that some drivers misbehaved, and /proc/pci showed the wrong types for the resources. This fixes it. Signed-off-by: Paul Mackerras Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/pci.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/arch/ppc64/kernel/pci.c b/arch/ppc64/kernel/pci.c index 861138ad092..ff4be1da69d 100644 --- a/arch/ppc64/kernel/pci.c +++ b/arch/ppc64/kernel/pci.c @@ -246,11 +246,14 @@ static unsigned int pci_parse_of_flags(u32 addr0) unsigned int flags = 0; if (addr0 & 0x02000000) { - flags |= IORESOURCE_MEM; + flags = IORESOURCE_MEM | PCI_BASE_ADDRESS_SPACE_MEMORY; + flags |= (addr0 >> 22) & PCI_BASE_ADDRESS_MEM_TYPE_64; + flags |= (addr0 >> 28) & PCI_BASE_ADDRESS_MEM_TYPE_1M; if (addr0 & 0x40000000) - flags |= IORESOURCE_PREFETCH; + flags |= IORESOURCE_PREFETCH + | PCI_BASE_ADDRESS_MEM_PREFETCH; } else if (addr0 & 0x01000000) - flags |= IORESOURCE_IO; + flags = IORESOURCE_IO | PCI_BASE_ADDRESS_SPACE_IO; return flags; } From 66dcca062847bcd261ebb3ac96d51101f31a8630 Mon Sep 17 00:00:00 2001 From: Sripathi Kodi Date: Mon, 19 Sep 2005 18:26:12 -0500 Subject: [PATCH 080/204] [PATCH] Fix invisible threads problem When the main thread of a thread group has done pthread_exit() and died, the other threads are still happily running, but will not be visible under /proc because their leader is no longer accessible. This fixes the access control so that we can see the sub-threads again. Signed-off-by: Sripathi Kodi Acked-by: Al Viro Signed-off-by: Linus Torvalds --- fs/proc/base.c | 84 +++++++++++++++++++++++++++++++++++++++++++++----- 1 file changed, 77 insertions(+), 7 deletions(-) diff --git a/fs/proc/base.c b/fs/proc/base.c index 23db452ab42..fb34f88a4a7 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -340,6 +340,52 @@ static int proc_root_link(struct inode *inode, struct dentry **dentry, struct vf return result; } + +/* Same as proc_root_link, but this addionally tries to get fs from other + * threads in the group */ +static int proc_task_root_link(struct inode *inode, struct dentry **dentry, struct vfsmount **mnt) +{ + struct fs_struct *fs; + int result = -ENOENT; + struct task_struct *leader = proc_task(inode); + + task_lock(leader); + fs = leader->fs; + if (fs) { + atomic_inc(&fs->count); + task_unlock(leader); + } else { + /* Try to get fs from other threads */ + task_unlock(leader); + struct task_struct *task = leader; + read_lock(&tasklist_lock); + if (pid_alive(task)) { + while ((task = next_thread(task)) != leader) { + task_lock(task); + fs = task->fs; + if (fs) { + atomic_inc(&fs->count); + task_unlock(task); + break; + } + task_unlock(task); + } + } + read_unlock(&tasklist_lock); + } + + if (fs) { + read_lock(&fs->lock); + *mnt = mntget(fs->rootmnt); + *dentry = dget(fs->root); + read_unlock(&fs->lock); + result = 0; + put_fs_struct(fs); + } + return result; +} + + #define MAY_PTRACE(task) \ (task == current || \ (task->parent == current && \ @@ -471,14 +517,14 @@ static int proc_oom_score(struct task_struct *task, char *buffer) /* permission checks */ -static int proc_check_root(struct inode *inode) +/* If the process being read is separated by chroot from the reading process, + * don't let the reader access the threads. + */ +static int proc_check_chroot(struct dentry *root, struct vfsmount *vfsmnt) { - struct dentry *de, *base, *root; - struct vfsmount *our_vfsmnt, *vfsmnt, *mnt; + struct dentry *de, *base; + struct vfsmount *our_vfsmnt, *mnt; int res = 0; - - if (proc_root_link(inode, &root, &vfsmnt)) /* Ewww... */ - return -ENOENT; read_lock(¤t->fs->lock); our_vfsmnt = mntget(current->fs->rootmnt); base = dget(current->fs->root); @@ -511,6 +557,16 @@ out: goto exit; } +static int proc_check_root(struct inode *inode) +{ + struct dentry *root; + struct vfsmount *vfsmnt; + + if (proc_root_link(inode, &root, &vfsmnt)) /* Ewww... */ + return -ENOENT; + return proc_check_chroot(root, vfsmnt); +} + static int proc_permission(struct inode *inode, int mask, struct nameidata *nd) { if (generic_permission(inode, mask, NULL) != 0) @@ -518,6 +574,20 @@ static int proc_permission(struct inode *inode, int mask, struct nameidata *nd) return proc_check_root(inode); } +static int proc_task_permission(struct inode *inode, int mask, struct nameidata *nd) +{ + struct dentry *root; + struct vfsmount *vfsmnt; + + if (generic_permission(inode, mask, NULL) != 0) + return -EACCES; + + if (proc_task_root_link(inode, &root, &vfsmnt)) + return -ENOENT; + + return proc_check_chroot(root, vfsmnt); +} + extern struct seq_operations proc_pid_maps_op; static int maps_open(struct inode *inode, struct file *file) { @@ -1419,7 +1489,7 @@ static struct inode_operations proc_fd_inode_operations = { static struct inode_operations proc_task_inode_operations = { .lookup = proc_task_lookup, - .permission = proc_permission, + .permission = proc_task_permission, }; #ifdef CONFIG_SECURITY From 31f6d9d628739c097964b8dbae939ea997da94a3 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Wed, 21 Sep 2005 09:55:30 -0700 Subject: [PATCH 081/204] [PATCH] Add printk_clock() ia64's sched_clock() accesses per-cpu data which isn't set up at boot time. Hence ia64 cannot use printk timestamping, because printk() will crash in sched_clock(). So make printk() use printk_clock(), defaulting to sched_clock(), overrideable by the architecture via attribute(weak). Cc: "Luck, Tony" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/printk.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/kernel/printk.c b/kernel/printk.c index a967605bc2e..4b8f0f9230a 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -488,6 +488,11 @@ static int __init printk_time_setup(char *str) __setup("time", printk_time_setup); +__attribute__((weak)) unsigned long long printk_clock(void) +{ + return sched_clock(); +} + /* * This is printk. It can be called from any context. We want it to work. * @@ -565,7 +570,7 @@ asmlinkage int vprintk(const char *fmt, va_list args) loglev_char = default_message_loglevel + '0'; } - t = sched_clock(); + t = printk_clock(); nanosec_rem = do_div(t, 1000000000); tlen = sprintf(tbuf, "<%c>[%5lu.%06lu] ", From 3c2822ccb1f8cc96fc006aa82e68e1944290014a Mon Sep 17 00:00:00 2001 From: Olof Johansson Date: Wed, 21 Sep 2005 09:55:31 -0700 Subject: [PATCH 082/204] [PATCH] PPC64: Fix boot for some pre-POWER4 systems Some RS64 systems (such as F80) have non-python host bridges with EADS. However, they have two EADS with 4 buses each under them, so the old logic that assumed no more than 7 busses per PHB failed miserably. Big thanks to Olaf Hering for helping me test this, he's got one of the few machines that broke from the previous logic. Also, to be a bit smarter at detecting the need for a PHB-level IOMMU table by checking for the presence of an ISA bus. Only PHBs with ISA bridges should need the PHB-level table. Signed-off-by: Olof Johansson Cc: Anton Blanchard Cc: Benjamin Herrenschmidt Cc: Paul Mackerras Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/pSeries_iommu.c | 171 ++++++++++++++++-------------- 1 file changed, 90 insertions(+), 81 deletions(-) diff --git a/arch/ppc64/kernel/pSeries_iommu.c b/arch/ppc64/kernel/pSeries_iommu.c index f0fd7fbd653..8c6313e7e14 100644 --- a/arch/ppc64/kernel/pSeries_iommu.c +++ b/arch/ppc64/kernel/pSeries_iommu.c @@ -265,8 +265,10 @@ static void iommu_table_setparms(struct pci_controller *phb, tbl->it_offset = phb->dma_window_base_cur >> PAGE_SHIFT; /* Test if we are going over 2GB of DMA space */ - if (phb->dma_window_base_cur + phb->dma_window_size > (1L << 31)) + if (phb->dma_window_base_cur + phb->dma_window_size > 0x80000000ul) { + udbg_printf("PCI_DMA: Unexpected number of IOAs under this PHB.\n"); panic("PCI_DMA: Unexpected number of IOAs under this PHB.\n"); + } phb->dma_window_base_cur += phb->dma_window_size; @@ -310,92 +312,84 @@ static void iommu_table_setparms_lpar(struct pci_controller *phb, static void iommu_bus_setup_pSeries(struct pci_bus *bus) { - struct device_node *dn, *pdn; - struct pci_dn *pci; + struct device_node *dn; struct iommu_table *tbl; + struct device_node *isa_dn, *isa_dn_orig; + struct device_node *tmp; + struct pci_dn *pci; + int children; DBG("iommu_bus_setup_pSeries, bus %p, bus->self %p\n", bus, bus->self); - /* For each (root) bus, we carve up the available DMA space in 256MB - * pieces. Since each piece is used by one (sub) bus/device, that would - * give a maximum of 7 devices per PHB. In most cases, this is plenty. + dn = pci_bus_to_OF_node(bus); + pci = PCI_DN(dn); + + if (bus->self) { + /* This is not a root bus, any setup will be done for the + * device-side of the bridge in iommu_dev_setup_pSeries(). + */ + return; + } + + /* Check if the ISA bus on the system is under + * this PHB. + */ + isa_dn = isa_dn_orig = of_find_node_by_type(NULL, "isa"); + + while (isa_dn && isa_dn != dn) + isa_dn = isa_dn->parent; + + if (isa_dn_orig) + of_node_put(isa_dn_orig); + + /* Count number of direct PCI children of the PHB. + * All PCI device nodes have class-code property, so it's + * an easy way to find them. + */ + for (children = 0, tmp = dn->child; tmp; tmp = tmp->sibling) + if (get_property(tmp, "class-code", NULL)) + children++; + + DBG("Children: %d\n", children); + + /* Calculate amount of DMA window per slot. Each window must be + * a power of two (due to pci_alloc_consistent requirements). * - * The exception is on Python PHBs (pre-POWER4). Here we don't have EADS - * bridges below the PHB to allocate the sectioned tables to, so instead - * we allocate a 1GB table at the PHB level. + * Keep 256MB aside for PHBs with ISA. */ - dn = pci_bus_to_OF_node(bus); - pci = dn->data; + if (!isa_dn) { + /* No ISA/IDE - just set window size and return */ + pci->phb->dma_window_size = 0x80000000ul; /* To be divided */ - if (!bus->self) { - /* Root bus */ - if (is_python(dn)) { - unsigned int *iohole; + while (pci->phb->dma_window_size * children > 0x80000000ul) + pci->phb->dma_window_size >>= 1; + DBG("No ISA/IDE, window size is %x\n", pci->phb->dma_window_size); + pci->phb->dma_window_base_cur = 0; - DBG("Python root bus %s\n", bus->name); - - iohole = (unsigned int *)get_property(dn, "io-hole", 0); - - if (iohole) { - /* On first bus we need to leave room for the - * ISA address space. Just skip the first 256MB - * alltogether. This leaves 768MB for the window. - */ - DBG("PHB has io-hole, reserving 256MB\n"); - pci->phb->dma_window_size = 3 << 28; - pci->phb->dma_window_base_cur = 1 << 28; - } else { - /* 1GB window by default */ - pci->phb->dma_window_size = 1 << 30; - pci->phb->dma_window_base_cur = 0; - } - - tbl = kmalloc(sizeof(struct iommu_table), GFP_KERNEL); - - iommu_table_setparms(pci->phb, dn, tbl); - pci->iommu_table = iommu_init_table(tbl); - } else { - /* Do a 128MB table at root. This is used for the IDE - * controller on some SMP-mode POWER4 machines. It - * doesn't hurt to allocate it on other machines - * -- it'll just be unused since new tables are - * allocated on the EADS level. - * - * Allocate at offset 128MB to avoid having to deal - * with ISA holes; 128MB table for IDE is plenty. - */ - pci->phb->dma_window_size = 1 << 27; - pci->phb->dma_window_base_cur = 1 << 27; - - tbl = kmalloc(sizeof(struct iommu_table), GFP_KERNEL); - - iommu_table_setparms(pci->phb, dn, tbl); - pci->iommu_table = iommu_init_table(tbl); - - /* All child buses have 256MB tables */ - pci->phb->dma_window_size = 1 << 28; - } - } else { - pdn = pci_bus_to_OF_node(bus->parent); - - if (!bus->parent->self && !is_python(pdn)) { - struct iommu_table *tbl; - /* First child and not python means this is the EADS - * level. Allocate new table for this slot with 256MB - * window. - */ - - tbl = kmalloc(sizeof(struct iommu_table), GFP_KERNEL); - - iommu_table_setparms(pci->phb, dn, tbl); - - pci->iommu_table = iommu_init_table(tbl); - } else { - /* Lower than first child or under python, use parent table */ - pci->iommu_table = PCI_DN(pdn)->iommu_table; - } + return; } + + /* If we have ISA, then we probably have an IDE + * controller too. Allocate a 128MB table but + * skip the first 128MB to avoid stepping on ISA + * space. + */ + pci->phb->dma_window_size = 0x8000000ul; + pci->phb->dma_window_base_cur = 0x8000000ul; + + tbl = kmalloc(sizeof(struct iommu_table), GFP_KERNEL); + + iommu_table_setparms(pci->phb, dn, tbl); + pci->iommu_table = iommu_init_table(tbl); + + /* Divide the rest (1.75GB) among the children */ + pci->phb->dma_window_size = 0x80000000ul; + while (pci->phb->dma_window_size * children > 0x70000000ul) + pci->phb->dma_window_size >>= 1; + + DBG("ISA/IDE, window size is %x\n", pci->phb->dma_window_size); + } @@ -446,14 +440,29 @@ static void iommu_bus_setup_pSeriesLP(struct pci_bus *bus) static void iommu_dev_setup_pSeries(struct pci_dev *dev) { struct device_node *dn, *mydn; + struct iommu_table *tbl; DBG("iommu_dev_setup_pSeries, dev %p (%s)\n", dev, dev->pretty_name); - /* Now copy the iommu_table ptr from the bus device down to the - * pci device_node. This means get_iommu_table() won't need to search - * up the device tree to find it. - */ + mydn = dn = pci_device_to_OF_node(dev); + /* If we're the direct child of a root bus, then we need to allocate + * an iommu table ourselves. The bus setup code should have setup + * the window sizes already. + */ + if (!dev->bus->self) { + DBG(" --> first child, no bridge. Allocating iommu table.\n"); + tbl = kmalloc(sizeof(struct iommu_table), GFP_KERNEL); + iommu_table_setparms(PCI_DN(dn)->phb, dn, tbl); + PCI_DN(mydn)->iommu_table = iommu_init_table(tbl); + + return; + } + + /* If this device is further down the bus tree, search upwards until + * an already allocated iommu table is found and use that. + */ + while (dn && dn->data && PCI_DN(dn)->iommu_table == NULL) dn = dn->parent; From d6dbf42ee36d4f9ffc2a93877c19937c81974124 Mon Sep 17 00:00:00 2001 From: Mike Miller Date: Wed, 21 Sep 2005 09:55:32 -0700 Subject: [PATCH 083/204] [PATCH] cciss: busy_initializing bug fix This patch fixes the problem Bjorn reported. The busy_initializing flag should have cleared before going into the for loop. Signed-off-by: Mike Miller Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/block/cciss.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c index c56f995aada..486b6e1c7df 100644 --- a/drivers/block/cciss.c +++ b/drivers/block/cciss.c @@ -483,9 +483,6 @@ static int cciss_open(struct inode *inode, struct file *filep) printk(KERN_DEBUG "cciss_open %s\n", inode->i_bdev->bd_disk->disk_name); #endif /* CCISS_DEBUG */ - if (host->busy_initializing) - return -EBUSY; - if (host->busy_initializing || drv->busy_configuring) return -EBUSY; /* @@ -2991,6 +2988,7 @@ static int __devinit cciss_init_one(struct pci_dev *pdev, hba[i]->access.set_intr_mask(hba[i], CCISS_INTR_ON); cciss_procinit(i); + hba[i]->busy_initializing = 0; for(j=0; j < NWD; j++) { /* mfm */ drive_info_struct *drv = &(hba[i]->drv[j]); @@ -3033,7 +3031,6 @@ static int __devinit cciss_init_one(struct pci_dev *pdev, add_disk(disk); } - hba[i]->busy_initializing = 0; return(1); clean4: From 1bc2a3bb86688ffca691ba2ad30a2d5dcba774cf Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Wed, 21 Sep 2005 09:55:33 -0700 Subject: [PATCH 084/204] [PATCH] ppc64: Fix issue with non zero boot cpu The new version of the flattened device tree passes the boot cpuid in the header instead of via a linux,boot-cpu property. We need to update the in kernel OF parsing code to do this, otherwise machines with a non zero boot cpuid fail to come up. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/prom_init.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/ppc64/kernel/prom_init.c b/arch/ppc64/kernel/prom_init.c index 9979919cdf9..f252670874a 100644 --- a/arch/ppc64/kernel/prom_init.c +++ b/arch/ppc64/kernel/prom_init.c @@ -1711,6 +1711,7 @@ static void __init flatten_device_tree(void) unsigned long offset = reloc_offset(); unsigned long mem_start, mem_end, room; struct boot_param_header *hdr; + struct prom_t *_prom = PTRRELOC(&prom); char *namep; u64 *rsvmap; @@ -1765,6 +1766,7 @@ static void __init flatten_device_tree(void) RELOC(dt_struct_end) = PAGE_ALIGN(mem_start); /* Finish header */ + hdr->boot_cpuid_phys = _prom->cpu; hdr->magic = OF_DT_HEADER; hdr->totalsize = RELOC(dt_struct_end) - RELOC(dt_header_start); hdr->off_dt_struct = RELOC(dt_struct_start) - RELOC(dt_header_start); @@ -1854,7 +1856,6 @@ static void __init prom_find_boot_cpu(void) cpu_pkg = call_prom("instance-to-package", 1, 1, prom_cpu); - prom_setprop(cpu_pkg, "linux,boot-cpu", NULL, 0); prom_getprop(cpu_pkg, "reg", &getprop_rval, sizeof(getprop_rval)); _prom->cpu = getprop_rval; From bbcc6bee9e5bf45ec35e87a4e0755dff09fe25d4 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 21 Sep 2005 09:55:33 -0700 Subject: [PATCH 085/204] [PATCH] remove blkdev_scsi_issue_flush_fn again This function was removed a while ago, but crept in again via a recent scsi merge. Signed-off-by: Christoph Hellwig Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/block/ll_rw_blk.c | 38 -------------------------------------- 1 file changed, 38 deletions(-) diff --git a/drivers/block/ll_rw_blk.c b/drivers/block/ll_rw_blk.c index 483d71b10cf..baedac52294 100644 --- a/drivers/block/ll_rw_blk.c +++ b/drivers/block/ll_rw_blk.c @@ -2373,44 +2373,6 @@ int blkdev_issue_flush(struct block_device *bdev, sector_t *error_sector) EXPORT_SYMBOL(blkdev_issue_flush); -/** - * blkdev_scsi_issue_flush_fn - issue flush for SCSI devices - * @q: device queue - * @disk: gendisk - * @error_sector: error offset - * - * Description: - * Devices understanding the SCSI command set, can use this function as - * a helper for issuing a cache flush. Note: driver is required to store - * the error offset (in case of error flushing) in ->sector of struct - * request. - */ -int blkdev_scsi_issue_flush_fn(request_queue_t *q, struct gendisk *disk, - sector_t *error_sector) -{ - struct request *rq = blk_get_request(q, WRITE, __GFP_WAIT); - int ret; - - rq->flags |= REQ_BLOCK_PC | REQ_SOFTBARRIER; - rq->sector = 0; - memset(rq->cmd, 0, sizeof(rq->cmd)); - rq->cmd[0] = 0x35; - rq->cmd_len = 12; - rq->data = NULL; - rq->data_len = 0; - rq->timeout = 60 * HZ; - - ret = blk_execute_rq(q, disk, rq, 0); - - if (ret && error_sector) - *error_sector = rq->sector; - - blk_put_request(rq); - return ret; -} - -EXPORT_SYMBOL(blkdev_scsi_issue_flush_fn); - static void drive_stat_acct(struct request *rq, int nr_sectors, int new_io) { int rw = rq_data_dir(rq); From 40da47e1159b89f2df5c718d2d5e269ced72b541 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Wed, 21 Sep 2005 09:55:34 -0700 Subject: [PATCH 086/204] [PATCH] ppc64: Build zImage.vmode for G5 zImage.vmode was recently added. It's a version of zImage in which the ELF note section used by open firmware indicates that it requires a virtual mode instance of OF instead of real mode. This allows it to work with Apple OF, and thus is directly bootable (or netbootable) from OF command line. (Unfortunately, pSeries OF sort-of requires real mode and Apple OF sort-of requires virtual mode, and both tend to be unhappy if no notes section specifies the mode at all). However, we forgot to add zImage.vmode to the default G5 build. This fixes it. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/ppc64/Makefile b/arch/ppc64/Makefile index 17d2c1eac3b..521c2a5a286 100644 --- a/arch/ppc64/Makefile +++ b/arch/ppc64/Makefile @@ -107,7 +107,7 @@ install: vmlinux $(Q)$(MAKE) $(build)=$(boot) BOOTIMAGE=$(BOOTIMAGE) $@ defaultimage-$(CONFIG_PPC_PSERIES) := zImage -defaultimage-$(CONFIG_PPC_PMAC) := vmlinux +defaultimage-$(CONFIG_PPC_PMAC) := zImage.vmode defaultimage-$(CONFIG_PPC_MAPLE) := zImage defaultimage-$(CONFIG_PPC_ISERIES) := vmlinux KBUILD_IMAGE := $(defaultimage-y) From 8b20f6da20d874a58becc7a08b130c4382ea7325 Mon Sep 17 00:00:00 2001 From: Deepak Saxena Date: Wed, 21 Sep 2005 09:55:35 -0700 Subject: [PATCH 087/204] [PATCH] Fix I2O config-osm init to return proper error We currently unregister the config-osm driver if initialization of the legacy ioctl() handlers failed but still return success. We should be returning -EBUSY in this case. Signed-off-by: Deepak Saxena Cc: Markus Lidel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/message/i2o/config-osm.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/message/i2o/config-osm.c b/drivers/message/i2o/config-osm.c index af32ab4e90c..10432f66520 100644 --- a/drivers/message/i2o/config-osm.c +++ b/drivers/message/i2o/config-osm.c @@ -56,8 +56,11 @@ static int __init i2o_config_init(void) return -EBUSY; } #ifdef CONFIG_I2O_CONFIG_OLD_IOCTL - if (i2o_config_old_init()) + if (i2o_config_old_init()) { + osm_err("old config handler initialization failed\n"); i2o_driver_unregister(&i2o_config_driver); + return -EBUSY; + } #endif return 0; From f10df6860498c7f52db4f00ff00a1ba334bcfea0 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 09:55:37 -0700 Subject: [PATCH 088/204] [PATCH] fix locking comment in unmap_region() That comment is plain wrong (we even take the pagetable lock inside unmap_region()). Signed-off-by: Paolo 'Blaisorblade' Giarrusso Acked-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/mmap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/mmap.c b/mm/mmap.c index 8b8e05f07cd..fa11d91242e 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -1640,7 +1640,7 @@ static void unmap_vma_list(struct mm_struct *mm, struct vm_area_struct *vma) /* * Get rid of page table information in the indicated region. * - * Called with the page table lock held. + * Called with the mm semaphore held. */ static void unmap_region(struct mm_struct *mm, struct vm_area_struct *vma, struct vm_area_struct *prev, From f875a1a6650edce859bd21051a22e1c27ac7ea63 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 09:55:37 -0700 Subject: [PATCH 089/204] [PATCH] README update from the stone age We have no options which the user can set in the Makefile. Only the EXTRAVERSION, which is also useful in place of the "backup modules" suggestion. We don't have configuration options in the top Makefile. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- README | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/README b/README index 76dd780d88e..2b5844d8cfa 100644 --- a/README +++ b/README @@ -149,6 +149,9 @@ CONFIGURING the kernel: "make gconfig" X windows (Gtk) based configuration tool. "make oldconfig" Default all questions based on the contents of your existing ./.config file. + "make silentoldconfig" + Like above, but avoids cluttering the screen + with question already answered. NOTES on "make config": - having unnecessary drivers will make the kernel bigger, and can @@ -169,9 +172,6 @@ CONFIGURING the kernel: should probably answer 'n' to the questions for "development", "experimental", or "debugging" features. - - Check the top Makefile for further site-dependent configuration - (default SVGA mode etc). - COMPILING the kernel: - Make sure you have gcc 2.95.3 available. @@ -199,6 +199,9 @@ COMPILING the kernel: are installing a new kernel with the same version number as your working kernel, make a backup of your modules directory before you do a "make modules_install". + In alternative, before compiling, edit your Makefile and change the + "EXTRAVERSION" line - its content is appended to the regular kernel + version. - In order to boot your new kernel, you'll need to copy the kernel image (e.g. .../linux/arch/i386/boot/bzImage after compilation) From 7e871b6c8f1f4fda41e51ef86147facecac3be9f Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 09:55:38 -0700 Subject: [PATCH 090/204] [PATCH] mm: update stale comment for removal of page->list Update comment for the 2.6.6-rc1 conversion from page->list and address_space->{clean,dirty,locked}_pages to radix tree tagging and ->lru. I've mostly avoided to mention page lists (at least I've shortened the comment). Signed-off-by: Paolo 'Blaisorblade' Giarrusso Acked-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/mm.h | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/include/linux/mm.h b/include/linux/mm.h index 82d7024f076..0d94c94d9d8 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -350,7 +350,8 @@ static inline void put_page(struct page *page) * only one copy in memory, at most, normally. * * For the non-reserved pages, page_count(page) denotes a reference count. - * page_count() == 0 means the page is free. + * page_count() == 0 means the page is free. page->lru is then used for + * freelist management in the buddy allocator. * page_count() == 1 means the page is used for exactly one purpose * (e.g. a private data page of one process). * @@ -376,10 +377,8 @@ static inline void put_page(struct page *page) * attaches, plus 1 if `private' contains something, plus one for * the page cache itself. * - * All pages belonging to an inode are in these doubly linked lists: - * mapping->clean_pages, mapping->dirty_pages and mapping->locked_pages; - * using the page->list list_head. These fields are also used for - * freelist managemet (when page_count()==0). + * Instead of keeping dirty/clean pages in per address-space lists, we instead + * now tag pages as dirty/under writeback in the radix tree. * * There is also a per-mapping radix tree mapping index to the page * in memory if present. The tree is rooted at mapping->root. From 7e2cff42cfac27c25202648c5c89f9171e5bc085 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 09:55:39 -0700 Subject: [PATCH 091/204] [PATCH] mm: add a note about partially hardcoded VM_* flags Hugh made me note this line for permission checking in mprotect(): if ((newflags & ~(newflags >> 4)) & 0xf) { after figuring out what's that about, I decided it's nasty enough. Btw Hugh itself didn't like the 0xf. We can safely change it to VM_READ|VM_WRITE|VM_EXEC because we never change VM_SHARED, so no need to check that. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Acked-by: Hugh Dickins Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/mm.h | 1 + mm/mprotect.c | 3 ++- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/include/linux/mm.h b/include/linux/mm.h index 0d94c94d9d8..097b3a3c693 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -136,6 +136,7 @@ extern unsigned int kobjsize(const void *objp); #define VM_EXEC 0x00000004 #define VM_SHARED 0x00000008 +/* mprotect() hardcodes VM_MAYREAD >> 4 == VM_READ, and so for r/w/x bits. */ #define VM_MAYREAD 0x00000010 /* limits for mprotect() etc */ #define VM_MAYWRITE 0x00000020 #define VM_MAYEXEC 0x00000040 diff --git a/mm/mprotect.c b/mm/mprotect.c index e9fbd013ad9..57577f63b30 100644 --- a/mm/mprotect.c +++ b/mm/mprotect.c @@ -248,7 +248,8 @@ sys_mprotect(unsigned long start, size_t len, unsigned long prot) newflags = vm_flags | (vma->vm_flags & ~(VM_READ | VM_WRITE | VM_EXEC)); - if ((newflags & ~(newflags >> 4)) & 0xf) { + /* newflags >> 4 shift VM_MAY% in place of VM_% */ + if ((newflags & ~(newflags >> 4)) & (VM_READ | VM_WRITE | VM_EXEC)) { error = -EACCES; goto out; } From 8f91648dcb0685d58aa046b25c69ce0d5f284f8c Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 21 Sep 2005 09:55:40 -0700 Subject: [PATCH 092/204] [PATCH] fixup Documentation/DocBook/kernel-hacking.tmpl __FUNCTION__ is the prefered kernel idiom, __func__ is not supported by gcc 2.95 (we actually map __FUNCTION__ to __func__ for more recent compilers, but it should never be used directly) Signed-off-by: Christoph Hellwig Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/DocBook/kernel-hacking.tmpl | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Documentation/DocBook/kernel-hacking.tmpl b/Documentation/DocBook/kernel-hacking.tmpl index 6367bba32d2..582032eea87 100644 --- a/Documentation/DocBook/kernel-hacking.tmpl +++ b/Documentation/DocBook/kernel-hacking.tmpl @@ -1105,7 +1105,7 @@ static struct block_device_operations opt_fops = { - Function names as strings (__func__). + Function names as strings (__FUNCTION__). From 4b1ac9ab15c1fc1e0b73d887a0c0cafe92991f97 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 21 Sep 2005 09:55:41 -0700 Subject: [PATCH 093/204] [PATCH] skge: add maintainer Add MAINTAINER record for Andrew ;-) Signed-off-by: Stephen Hemminger Cc: Jeff Garzik Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- MAINTAINERS | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index dc8f3babcab..ade7415d246 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2266,6 +2266,12 @@ M: kristen.c.accardi@intel.com L: pcihpd-discuss@lists.sourceforge.net S: Maintained +SKGE, SKY2 10/100/1000 GIGABIT ETHERNET DRIVERS +P: Stephen Hemminger +M: shemminger@osdl.org +L: netdev@vger.kernel.org +S: Maintained + SPARC (sparc32): P: William L. Irwin M: wli@holomorphy.com From 7980cbbb30bf044e6f40912a3f6456204ddfc27e Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 21 Sep 2005 09:55:43 -0700 Subject: [PATCH 094/204] [PATCH] Adds sys_set_mempolicy() in include/linux/syscalls.h Signed-off-by: Eric Dumazet Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/syscalls.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h index 425f58c8ea4..a6f03e47373 100644 --- a/include/linux/syscalls.h +++ b/include/linux/syscalls.h @@ -508,5 +508,7 @@ asmlinkage long sys_keyctl(int cmd, unsigned long arg2, unsigned long arg3, asmlinkage long sys_ioprio_set(int which, int who, int ioprio); asmlinkage long sys_ioprio_get(int which, int who); +asmlinkage long sys_set_mempolicy(int mode, unsigned long __user *nmask, + unsigned long maxnode); #endif From 89d7cbf73e1a85241eb42339f6cb7a429fec178c Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Wed, 21 Sep 2005 09:55:44 -0700 Subject: [PATCH 095/204] [PATCH] update URL for HPET spec. Correct URL for HPET spec. Signed-off-by: Randy Dunlap Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86_64/Kconfig b/arch/x86_64/Kconfig index 0969d570f3b..21afa69a086 100644 --- a/arch/x86_64/Kconfig +++ b/arch/x86_64/Kconfig @@ -308,7 +308,7 @@ config HPET_TIMER present. The HPET provides a stable time base on SMP systems, unlike the TSC, but it is more expensive to access, as it is off-chip. You can find the HPET spec at - . + . config X86_PM_TIMER bool "PM timer" From 62a36c43c898d45efcfe3376ea1da6a9a182e1ad Mon Sep 17 00:00:00 2001 From: Stephane Kardas Date: Wed, 21 Sep 2005 09:55:45 -0700 Subject: [PATCH 096/204] [PATCH] fat: fix adate During a forensic analysis on the fat file system, I found than the result for the last access date on this file system was different between the stat command and the istat command (package tct-utils). The istat command display a true date (the right windows date) but the stat primitive (so stat, find, ls command) displays a wrong date. Signed-off-by: OGAWA Hirofumi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/fat/inode.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/fs/fat/inode.c b/fs/fat/inode.c index 51b1d15d9d5..e2effe2dc9b 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -300,9 +300,9 @@ static int fat_fill_inode(struct inode *inode, struct msdos_dir_entry *de) inode->i_blksize = sbi->cluster_size; inode->i_blocks = ((inode->i_size + (sbi->cluster_size - 1)) & ~((loff_t)sbi->cluster_size - 1)) >> 9; - inode->i_mtime.tv_sec = inode->i_atime.tv_sec = + inode->i_mtime.tv_sec = date_dos2unix(le16_to_cpu(de->time), le16_to_cpu(de->date)); - inode->i_mtime.tv_nsec = inode->i_atime.tv_nsec = 0; + inode->i_mtime.tv_nsec = 0; if (sbi->options.isvfat) { int secs = de->ctime_cs / 100; int csecs = de->ctime_cs % 100; @@ -310,8 +310,11 @@ static int fat_fill_inode(struct inode *inode, struct msdos_dir_entry *de) date_dos2unix(le16_to_cpu(de->ctime), le16_to_cpu(de->cdate)) + secs; inode->i_ctime.tv_nsec = csecs * 10000000; + inode->i_atime.tv_sec = + date_dos2unix(le16_to_cpu(0), le16_to_cpu(de->adate)); + inode->i_atime.tv_nsec = 0; } else - inode->i_ctime = inode->i_mtime; + inode->i_ctime = inode->i_atime = inode->i_mtime; return 0; } @@ -513,7 +516,9 @@ retry: raw_entry->starthi = cpu_to_le16(MSDOS_I(inode)->i_logstart >> 16); fat_date_unix2dos(inode->i_mtime.tv_sec, &raw_entry->time, &raw_entry->date); if (sbi->options.isvfat) { + __le16 atime; fat_date_unix2dos(inode->i_ctime.tv_sec,&raw_entry->ctime,&raw_entry->cdate); + fat_date_unix2dos(inode->i_atime.tv_sec,&atime,&raw_entry->adate); raw_entry->ctime_cs = (inode->i_ctime.tv_sec & 1) * 100 + inode->i_ctime.tv_nsec / 10000000; } From 972d512a17c1bb7c4b784a9da2ca75745fcc6989 Mon Sep 17 00:00:00 2001 From: Sean Hefty Date: Wed, 21 Sep 2005 12:31:26 -0700 Subject: [PATCH 097/204] [IB] Add MAD data field size definitions Clean up code by using enums instead of hard-coded magic numbers. Signed-off-by: Sean Hefty Signed-off-by: Roland Dreier --- drivers/infiniband/core/mad_rmpp.c | 15 ++++++--------- include/rdma/ib_mad.h | 15 +++++++++++---- 2 files changed, 17 insertions(+), 13 deletions(-) diff --git a/drivers/infiniband/core/mad_rmpp.c b/drivers/infiniband/core/mad_rmpp.c index 2bd8b1cc57c..8f7cef0812f 100644 --- a/drivers/infiniband/core/mad_rmpp.c +++ b/drivers/infiniband/core/mad_rmpp.c @@ -583,6 +583,7 @@ static int send_next_seg(struct ib_mad_send_wr_private *mad_send_wr) { struct ib_rmpp_mad *rmpp_mad; int timeout; + u32 paylen; rmpp_mad = (struct ib_rmpp_mad *)mad_send_wr->send_wr.wr.ud.mad_hdr; ib_set_rmpp_flags(&rmpp_mad->rmpp_hdr, IB_MGMT_RMPP_FLAG_ACTIVE); @@ -590,11 +591,9 @@ static int send_next_seg(struct ib_mad_send_wr_private *mad_send_wr) if (mad_send_wr->seg_num == 1) { rmpp_mad->rmpp_hdr.rmpp_rtime_flags |= IB_MGMT_RMPP_FLAG_FIRST; - rmpp_mad->rmpp_hdr.paylen_newwin = - cpu_to_be32(mad_send_wr->total_seg * - (sizeof(struct ib_rmpp_mad) - - offsetof(struct ib_rmpp_mad, data)) - - mad_send_wr->pad); + paylen = mad_send_wr->total_seg * IB_MGMT_RMPP_DATA - + mad_send_wr->pad; + rmpp_mad->rmpp_hdr.paylen_newwin = cpu_to_be32(paylen); mad_send_wr->sg_list[0].length = sizeof(struct ib_rmpp_mad); } else { mad_send_wr->send_wr.num_sge = 2; @@ -608,10 +607,8 @@ static int send_next_seg(struct ib_mad_send_wr_private *mad_send_wr) if (mad_send_wr->seg_num == mad_send_wr->total_seg) { rmpp_mad->rmpp_hdr.rmpp_rtime_flags |= IB_MGMT_RMPP_FLAG_LAST; - rmpp_mad->rmpp_hdr.paylen_newwin = - cpu_to_be32(sizeof(struct ib_rmpp_mad) - - offsetof(struct ib_rmpp_mad, data) - - mad_send_wr->pad); + paylen = IB_MGMT_RMPP_DATA - mad_send_wr->pad; + rmpp_mad->rmpp_hdr.paylen_newwin = cpu_to_be32(paylen); } /* 2 seconds for an ACK until we can find the packet lifetime */ diff --git a/include/rdma/ib_mad.h b/include/rdma/ib_mad.h index 53184a38fdf..0e293fe733b 100644 --- a/include/rdma/ib_mad.h +++ b/include/rdma/ib_mad.h @@ -108,6 +108,13 @@ #define IB_QP1_QKEY 0x80010000 #define IB_QP_SET_QKEY 0x80000000 +enum { + IB_MGMT_MAD_DATA = 232, + IB_MGMT_RMPP_DATA = 220, + IB_MGMT_VENDOR_DATA = 216, + IB_MGMT_SA_DATA = 200 +}; + struct ib_mad_hdr { u8 base_version; u8 mgmt_class; @@ -149,20 +156,20 @@ struct ib_sa_hdr { struct ib_mad { struct ib_mad_hdr mad_hdr; - u8 data[232]; + u8 data[IB_MGMT_MAD_DATA]; }; struct ib_rmpp_mad { struct ib_mad_hdr mad_hdr; struct ib_rmpp_hdr rmpp_hdr; - u8 data[220]; + u8 data[IB_MGMT_RMPP_DATA]; }; struct ib_sa_mad { struct ib_mad_hdr mad_hdr; struct ib_rmpp_hdr rmpp_hdr; struct ib_sa_hdr sa_hdr; - u8 data[200]; + u8 data[IB_MGMT_SA_DATA]; } __attribute__ ((packed)); struct ib_vendor_mad { @@ -170,7 +177,7 @@ struct ib_vendor_mad { struct ib_rmpp_hdr rmpp_hdr; u8 reserved; u8 oui[3]; - u8 data[216]; + u8 data[IB_MGMT_VENDOR_DATA]; }; struct ib_class_port_info From f2065e4242e5078d53c521b973c5cd4eae91eca2 Mon Sep 17 00:00:00 2001 From: Hal Rosenstock Date: Wed, 21 Sep 2005 12:33:09 -0700 Subject: [PATCH 098/204] [IB] Fix RMPP receive length calculation Based on simplification idea from Sean Hefty Signed-off-by: Hal Rosenstock Signed-off-by: Roland Dreier --- drivers/infiniband/core/mad_rmpp.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/core/mad_rmpp.c b/drivers/infiniband/core/mad_rmpp.c index 8f7cef0812f..e23836d0e21 100644 --- a/drivers/infiniband/core/mad_rmpp.c +++ b/drivers/infiniband/core/mad_rmpp.c @@ -412,8 +412,8 @@ static inline int get_mad_len(struct mad_rmpp_recv *rmpp_recv) hdr_size = data_offset(rmpp_mad->mad_hdr.mgmt_class); data_size = sizeof(struct ib_rmpp_mad) - hdr_size; - pad = data_size - be32_to_cpu(rmpp_mad->rmpp_hdr.paylen_newwin); - if (pad > data_size || pad < 0) + pad = IB_MGMT_RMPP_DATA - be32_to_cpu(rmpp_mad->rmpp_hdr.paylen_newwin); + if (pad > IB_MGMT_RMPP_DATA || pad < 0) pad = 0; return hdr_size + rmpp_recv->seg_num * data_size - pad; From 6a1ced59b3ac5d1088f597fd0613a724cdf19e2c Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Wed, 21 Sep 2005 22:14:05 +0100 Subject: [PATCH 099/204] [ARM] 2932/1: Avoid the "noreturn" warning in arch/arm/kernel/traps.c Patch from Catalin Marinas This patch prevents the "noreturn function does return" warning in the __bug() function in arch/arm/kernel/traps.c Signed-off-by: Catalin Marinas Signed-off-by: Russell King --- arch/arm/kernel/traps.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c index 4554c961251..e7d22dbcb69 100644 --- a/arch/arm/kernel/traps.c +++ b/arch/arm/kernel/traps.c @@ -624,6 +624,9 @@ void __attribute__((noreturn)) __bug(const char *file, int line, void *data) printk(" - extra data = %p", data); printk("\n"); *(int *)0 = 0; + + /* Avoid "noreturn function does return" */ + for (;;); } EXPORT_SYMBOL(__bug); From c51179fb0c77ad91df5825f8f7eb670da97e137e Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 18:37:14 +0200 Subject: [PATCH 100/204] [PATCH] uml: adapt asm/futex.h to our arch Follow up to 4732efbeb997189d9f9b04708dc26bf8613ed721 - uml must just reuse as-is the backing architecture support. There is a micro-fixup is needed for the included file, which won't affect i386 behaviour at all. I've not tested compilation on x86_64, only on x86, but the code is almost the same except the culprit test, so everything should be ok on x86_64 too. Cc: Jakub Jelinek Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Linus Torvalds --- include/asm-i386/futex.h | 2 +- include/asm-um/futex.h | 51 ++++------------------------------------ 2 files changed, 6 insertions(+), 47 deletions(-) diff --git a/include/asm-i386/futex.h b/include/asm-i386/futex.h index 44b9db80647..e7a271d3930 100644 --- a/include/asm-i386/futex.h +++ b/include/asm-i386/futex.h @@ -61,7 +61,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) if (op == FUTEX_OP_SET) __futex_atomic_op1("xchgl %0, %2", ret, oldval, uaddr, oparg); else { -#ifndef CONFIG_X86_BSWAP +#if !defined(CONFIG_X86_BSWAP) && !defined(CONFIG_UML) if (boot_cpu_data.x86 == 3) ret = -ENOSYS; else diff --git a/include/asm-um/futex.h b/include/asm-um/futex.h index 2cac5ecd9d0..142ee2d8e0f 100644 --- a/include/asm-um/futex.h +++ b/include/asm-um/futex.h @@ -1,53 +1,12 @@ -#ifndef _ASM_FUTEX_H -#define _ASM_FUTEX_H - -#ifdef __KERNEL__ +#ifndef __UM_FUTEX_H +#define __UM_FUTEX_H #include #include +#include +#include #include -static inline int -futex_atomic_op_inuser (int encoded_op, int __user *uaddr) -{ - int op = (encoded_op >> 28) & 7; - int cmp = (encoded_op >> 24) & 15; - int oparg = (encoded_op << 8) >> 20; - int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; - if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) - oparg = 1 << oparg; - - if (! access_ok (VERIFY_WRITE, uaddr, sizeof(int))) - return -EFAULT; - - inc_preempt_count(); - - switch (op) { - case FUTEX_OP_SET: - case FUTEX_OP_ADD: - case FUTEX_OP_OR: - case FUTEX_OP_ANDN: - case FUTEX_OP_XOR: - default: - ret = -ENOSYS; - } - - dec_preempt_count(); - - if (!ret) { - switch (cmp) { - case FUTEX_OP_CMP_EQ: ret = (oldval == cmparg); break; - case FUTEX_OP_CMP_NE: ret = (oldval != cmparg); break; - case FUTEX_OP_CMP_LT: ret = (oldval < cmparg); break; - case FUTEX_OP_CMP_GE: ret = (oldval >= cmparg); break; - case FUTEX_OP_CMP_LE: ret = (oldval <= cmparg); break; - case FUTEX_OP_CMP_GT: ret = (oldval > cmparg); break; - default: ret = -ENOSYS; - } - } - return ret; -} +#include "asm/arch/futex.h" #endif -#endif From 676067cfeaa16f6f338e067e83ce4733b41c0b24 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 18:38:09 +0200 Subject: [PATCH 101/204] [PATCH] Remove unused var from asm/futex.h As recently done by Russell King for ARM, commit 4732efbeb997189d9f9b04708dc26bf8613ed721 introduces a generic asm/futex.h copied along most arches, which includes a "-ENOSYS support" to be changed if needed. However, it includes an unused var (taken from the "real" version) which GCC warns about. Remove it from all arches having that file version (i.e. same GIT id). $ git-diff-tree -r HEAD and $ git-ls-tree -r HEAD include/|grep 9feff4ce1424bc390608326240be369eb13aa648 may be more interesting than looking at the patch itself, to make sure I've just copied the arm header to all other archs having the original dummy version of this file. Cc: Jakub Jelinek Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Linus Torvalds --- include/asm-alpha/futex.h | 2 +- include/asm-arm26/futex.h | 2 +- include/asm-cris/futex.h | 2 +- include/asm-frv/futex.h | 2 +- include/asm-h8300/futex.h | 2 +- include/asm-ia64/futex.h | 2 +- include/asm-m32r/futex.h | 2 +- include/asm-m68k/futex.h | 2 +- include/asm-m68knommu/futex.h | 2 +- include/asm-parisc/futex.h | 2 +- include/asm-ppc/futex.h | 2 +- include/asm-s390/futex.h | 2 +- include/asm-sh/futex.h | 2 +- include/asm-sh64/futex.h | 2 +- include/asm-sparc/futex.h | 2 +- include/asm-sparc64/futex.h | 2 +- include/asm-v850/futex.h | 2 +- 17 files changed, 17 insertions(+), 17 deletions(-) diff --git a/include/asm-alpha/futex.h b/include/asm-alpha/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-alpha/futex.h +++ b/include/asm-alpha/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-arm26/futex.h b/include/asm-arm26/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-arm26/futex.h +++ b/include/asm-arm26/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-cris/futex.h b/include/asm-cris/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-cris/futex.h +++ b/include/asm-cris/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-frv/futex.h b/include/asm-frv/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-frv/futex.h +++ b/include/asm-frv/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-h8300/futex.h b/include/asm-h8300/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-h8300/futex.h +++ b/include/asm-h8300/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-ia64/futex.h b/include/asm-ia64/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-ia64/futex.h +++ b/include/asm-ia64/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-m32r/futex.h b/include/asm-m32r/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-m32r/futex.h +++ b/include/asm-m32r/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-m68k/futex.h b/include/asm-m68k/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-m68k/futex.h +++ b/include/asm-m68k/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-m68knommu/futex.h b/include/asm-m68knommu/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-m68knommu/futex.h +++ b/include/asm-m68knommu/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-parisc/futex.h b/include/asm-parisc/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-parisc/futex.h +++ b/include/asm-parisc/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-ppc/futex.h b/include/asm-ppc/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-ppc/futex.h +++ b/include/asm-ppc/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-s390/futex.h b/include/asm-s390/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-s390/futex.h +++ b/include/asm-s390/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-sh/futex.h b/include/asm-sh/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-sh/futex.h +++ b/include/asm-sh/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-sh64/futex.h b/include/asm-sh64/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-sh64/futex.h +++ b/include/asm-sh64/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-sparc/futex.h b/include/asm-sparc/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-sparc/futex.h +++ b/include/asm-sparc/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-sparc64/futex.h b/include/asm-sparc64/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-sparc64/futex.h +++ b/include/asm-sparc64/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; diff --git a/include/asm-v850/futex.h b/include/asm-v850/futex.h index 2cac5ecd9d0..9feff4ce142 100644 --- a/include/asm-v850/futex.h +++ b/include/asm-v850/futex.h @@ -14,7 +14,7 @@ futex_atomic_op_inuser (int encoded_op, int __user *uaddr) int cmp = (encoded_op >> 24) & 15; int oparg = (encoded_op << 8) >> 20; int cmparg = (encoded_op << 20) >> 20; - int oldval = 0, ret, tem; + int oldval = 0, ret; if (encoded_op & (FUTEX_OP_OPARG_SHIFT << 28)) oparg = 1 << oparg; From 76629ac154aacfec8883954c3ff9871905d2fb8e Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 18:38:33 +0200 Subject: [PATCH 102/204] [PATCH] uml: remove verify_area_{tt,skas} When removing verify_area, verify_area_{tt,skas} were forgotten. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Linus Torvalds --- arch/um/kernel/skas/include/uaccess-skas.h | 6 ------ arch/um/kernel/tt/include/uaccess-tt.h | 6 ------ 2 files changed, 12 deletions(-) diff --git a/arch/um/kernel/skas/include/uaccess-skas.h b/arch/um/kernel/skas/include/uaccess-skas.h index 6ee3f3902e6..7da0c2def0e 100644 --- a/arch/um/kernel/skas/include/uaccess-skas.h +++ b/arch/um/kernel/skas/include/uaccess-skas.h @@ -18,12 +18,6 @@ ((unsigned long) (addr) + (size) <= FIXADDR_USER_END) && \ ((unsigned long) (addr) + (size) >= (unsigned long)(addr)))) -static inline int verify_area_skas(int type, const void __user * addr, - unsigned long size) -{ - return(access_ok_skas(type, addr, size) ? 0 : -EFAULT); -} - extern int copy_from_user_skas(void *to, const void __user *from, int n); extern int copy_to_user_skas(void __user *to, const void *from, int n); extern int strncpy_from_user_skas(char *dst, const char __user *src, int count); diff --git a/arch/um/kernel/tt/include/uaccess-tt.h b/arch/um/kernel/tt/include/uaccess-tt.h index aa6db384af8..dc2ebfa8c54 100644 --- a/arch/um/kernel/tt/include/uaccess-tt.h +++ b/arch/um/kernel/tt/include/uaccess-tt.h @@ -33,12 +33,6 @@ extern unsigned long uml_physmem; (((unsigned long) (addr) <= ((unsigned long) (addr) + (size))) && \ (under_task_size(addr, size) || is_stack(addr, size)))) -static inline int verify_area_tt(int type, const void __user * addr, - unsigned long size) -{ - return(access_ok_tt(type, addr, size) ? 0 : -EFAULT); -} - extern unsigned long get_fault_addr(void); extern int __do_copy_from_user(void *to, const void *from, int n, From 36decba90b184a25b7f77fb7df1612125cc41ac5 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 18:38:57 +0200 Subject: [PATCH 103/204] [PATCH] uml: fix modify_ldt - missing break in switch I am a lamer :-(. Luckily, Luo Xin performed LTP testing and found this failure. Btw, the fact that the patch in which I introduced this was merged shows that: a) I'm really trusted by people b) sometimes they're wrong about point a). c) lack of time for reviewers. CC: Luo Xin Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Linus Torvalds --- arch/um/sys-i386/ldt.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/um/sys-i386/ldt.c b/arch/um/sys-i386/ldt.c index bd3c34aa52e..36b5c2c1328 100644 --- a/arch/um/sys-i386/ldt.c +++ b/arch/um/sys-i386/ldt.c @@ -83,6 +83,7 @@ int sys_modify_ldt(int func, void __user *ptr, unsigned long bytecount) goto out; } p = buf; + break; default: res = -ENOSYS; goto out; From 8e5cb35a60c8792ffc3ba89ea66ebdd3c2f01cb1 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 18:39:14 +0200 Subject: [PATCH 104/204] [PATCH] uml: fix uname output on 32-bit binary on 64-bit host Translate uname output taken from the host if needed. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Linus Torvalds --- arch/um/kernel/user_util.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/arch/um/kernel/user_util.c b/arch/um/kernel/user_util.c index a25f3ea11fd..41d17c71511 100644 --- a/arch/um/kernel/user_util.c +++ b/arch/um/kernel/user_util.c @@ -128,6 +128,12 @@ void setup_machinename(char *machine_out) struct utsname host; uname(&host); +#if defined(UML_CONFIG_UML_X86) && !defined(UML_CONFIG_64BIT) + if (!strcmp(host.machine, "x86_64")) { + strcpy(machine_out, "i686"); + return; + } +#endif strcpy(machine_out, host.machine); } From fd7481047882eaa0a692a545d5f8419612ece1b0 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 18:39:32 +0200 Subject: [PATCH 105/204] [PATCH] uml: Fix conflict between libc and ipv6 gcc is now complaining during link on some hosts - fix it as for other things. Reported by Antoine Martin . Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Linus Torvalds --- arch/um/Makefile | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/arch/um/Makefile b/arch/um/Makefile index ce987266dac..5b5af95721a 100644 --- a/arch/um/Makefile +++ b/arch/um/Makefile @@ -53,9 +53,13 @@ SYS_DIR := $(ARCH_DIR)/include/sysdep-$(SUBARCH) # -Dvmap=kernel_vmap affects everything, and prevents anything from # referencing the libpcap.o symbol so named. +# +# Same things for in6addr_loopback - found in libc. CFLAGS += $(CFLAGS-y) -D__arch_um__ -DSUBARCH=\"$(SUBARCH)\" \ - $(ARCH_INCLUDE) $(MODE_INCLUDE) -Dvmap=kernel_vmap + $(ARCH_INCLUDE) $(MODE_INCLUDE) -Dvmap=kernel_vmap \ + -Din6addr_loopback=kernel_in6addr_loopback + AFLAGS += $(ARCH_INCLUDE) USER_CFLAGS := $(patsubst -I%,,$(CFLAGS)) From 41c79dc3825545d9aa2d5e5b1a99f8faabf51455 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 18:39:47 +0200 Subject: [PATCH 106/204] [PATCH] uml: fix bogus HOST_ELF_CLASS symbol name Even if with a bit of misunderstanding, Al fixed this in commit 95608261dae863bc43292e6fbd946a3abd3aa49f. Well, the symbol was intended to come from userspace (it exists there on normal host), but since some hosts may miss that, using the kernel one is just as fine. However, rename it to be named consistently with the rest. Actually, he missed converting ELFCLASS32 to coming from kernel headers. For consistence, add ELFCLASS64 too. Cc: Al Viro Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Linus Torvalds --- arch/um/include/common-offsets.h | 4 +++- arch/um/os-Linux/elf_aux.c | 3 ++- 2 files changed, 5 insertions(+), 2 deletions(-) diff --git a/arch/um/include/common-offsets.h b/arch/um/include/common-offsets.h index 0aa620970ad..782ac3a3baf 100644 --- a/arch/um/include/common-offsets.h +++ b/arch/um/include/common-offsets.h @@ -12,4 +12,6 @@ DEFINE_STR(UM_KERN_WARNING, KERN_WARNING); DEFINE_STR(UM_KERN_NOTICE, KERN_NOTICE); DEFINE_STR(UM_KERN_INFO, KERN_INFO); DEFINE_STR(UM_KERN_DEBUG, KERN_DEBUG); -DEFINE(HOST_ELF_CLASS, ELF_CLASS); +DEFINE(UM_ELF_CLASS, ELF_CLASS); +DEFINE(UM_ELFCLASS32, ELFCLASS32); +DEFINE(UM_ELFCLASS64, ELFCLASS64); diff --git a/arch/um/os-Linux/elf_aux.c b/arch/um/os-Linux/elf_aux.c index e770cb02957..ab33cb3c74e 100644 --- a/arch/um/os-Linux/elf_aux.c +++ b/arch/um/os-Linux/elf_aux.c @@ -14,7 +14,8 @@ #include "mem_user.h" #include -#if HOST_ELF_CLASS == ELFCLASS32 +/* Use the one from the kernel - the host may miss it, if having old headers. */ +#if UM_ELF_CLASS == UM_ELFCLASS32 typedef Elf32_auxv_t elf_auxv_t; #else typedef Elf64_auxv_t elf_auxv_t; From ae756df8ddd00a30bb3bdf6774b514c096dda4a4 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 18:40:10 +0200 Subject: [PATCH 107/204] [PATCH] uml: readd removed unistd.h inclusion Readd this header (deleted in 60d339f6fe0831060600c62418b71a62ad26c281). A warning is spit out here about undeclared getpgrp(). Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Linus Torvalds --- arch/um/os-Linux/process.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/um/os-Linux/process.c b/arch/um/os-Linux/process.c index d32413e4b4c..d9c52387c4a 100644 --- a/arch/um/os-Linux/process.c +++ b/arch/um/os-Linux/process.c @@ -3,6 +3,7 @@ * Licensed under the GPL */ +#include #include #include #include From a46c9047642d70f5619e4a14129d74dd1f6ea74f Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 18:40:29 +0200 Subject: [PATCH 108/204] [PATCH] uml: comment about cast build fix Explain why the casting we do to silence this warning is indeed safe. It is because the field we're casting from, though being 64-bit wide, was filled with a pointer in first place by ourselves. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Linus Torvalds --- arch/um/os-Linux/aio.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/um/os-Linux/aio.c b/arch/um/os-Linux/aio.c index 298d5632128..f6e64026f99 100644 --- a/arch/um/os-Linux/aio.c +++ b/arch/um/os-Linux/aio.c @@ -144,6 +144,7 @@ static int aio_thread(void *arg) "errno = %d\n", errno); } else { + /* This is safe as we've just a pointer here. */ aio = (struct aio_context *) (long) event.data; if(update_aio(aio, event.res)){ do_aio(ctx, aio); From efb0372bbaf5b829ff8c39db372779928af542a7 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Wed, 21 Sep 2005 18:40:49 +0200 Subject: [PATCH 109/204] [PATCH] uml: fix compile warning after consolidation patch The header declaring this function wasn't included, so the function declaration was totally bogus wrt. the proto - even if this wasn't going to fail at all. It was so bad that the compile warning I got was "control reaches end of non-void function", i.e. missing return. Actually, this has been there for ages, the consolidation patch just added the warning which was needed to clean it up. Nice. Really. Cc: Allan Graves Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Linus Torvalds --- arch/um/kernel/tt/process_kern.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/um/kernel/tt/process_kern.c b/arch/um/kernel/tt/process_kern.c index 0de05a268b2..cfaa373a6e7 100644 --- a/arch/um/kernel/tt/process_kern.c +++ b/arch/um/kernel/tt/process_kern.c @@ -23,10 +23,11 @@ #include "mem_user.h" #include "tlb.h" #include "mode.h" +#include "mode_kern.h" #include "init.h" #include "tt.h" -int switch_to_tt(void *prev, void *next, void *last) +void switch_to_tt(void *prev, void *next) { struct task_struct *from, *to, *prev_sched; unsigned long flags; From 552709d5aee9145f325bf07348fb299e84b2e5b3 Mon Sep 17 00:00:00 2001 From: "nsxfreddy@gmail.com" Date: Wed, 21 Sep 2005 14:18:04 -0500 Subject: [PATCH 110/204] [PATCH] bonding: Fix link monitor capability check (was skge: set mac address oops with bonding) Fix bond_enslave link monitoring warning to check use_carrier status and ethtool_ops in addition to do_ioctl. This version checks ethtool_ops as well as do_ioctl, and also uses the per-bond params.use_carrier instead of the global use_carrier. Signed-off-by: Jason R. Martin Signed-off-by: Jeff Garzik --- drivers/net/bonding/bond_main.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 90449a0f2a6..6d00c3de1a8 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -1653,7 +1653,8 @@ static int bond_enslave(struct net_device *bond_dev, struct net_device *slave_de int old_features = bond_dev->features; int res = 0; - if (slave_dev->do_ioctl == NULL) { + if (!bond->params.use_carrier && slave_dev->ethtool_ops == NULL && + slave_dev->do_ioctl == NULL) { printk(KERN_WARNING DRV_NAME ": Warning : no link monitoring support for %s\n", slave_dev->name); From c3f8be961808313a502c67d59e2b7f930477faf3 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Mon, 19 Sep 2005 15:37:34 -0700 Subject: [PATCH 111/204] [PATCH] skge: expand ethtool debug register dump Expand the returned data for ethtool debug access to include all of the mapped PCI area; except for the small set of registers that are for diagnostic RAM access. Access to those registers will hang the system. Signed-off-by: Stephen Hemminger Signed-off-by: Jeff Garzik --- drivers/net/skge.c | 29 ++++++++--------------------- 1 file changed, 8 insertions(+), 21 deletions(-) diff --git a/drivers/net/skge.c b/drivers/net/skge.c index 0208258e782..62e2ae0eb85 100644 --- a/drivers/net/skge.c +++ b/drivers/net/skge.c @@ -105,41 +105,28 @@ static const u32 rxirqmask[] = { IS_R1_F, IS_R2_F }; static const u32 txirqmask[] = { IS_XA1_F, IS_XA2_F }; static const u32 portirqmask[] = { IS_PORT_1, IS_PORT_2 }; -/* Don't need to look at whole 16K. - * last interesting register is descriptor poll timer. - */ -#define SKGE_REGS_LEN (29*128) - static int skge_get_regs_len(struct net_device *dev) { - return SKGE_REGS_LEN; + return 0x4000; } /* - * Returns copy of control register region - * I/O region is divided into banks and certain regions are unreadable + * Returns copy of whole control register region + * Note: skip RAM address register because accessing it will + * cause bus hangs! */ static void skge_get_regs(struct net_device *dev, struct ethtool_regs *regs, void *p) { const struct skge_port *skge = netdev_priv(dev); - unsigned long offs; const void __iomem *io = skge->hw->regs; - static const unsigned long bankmap - = (1<<0) | (1<<2) | (1<<8) | (1<<9) - | (1<<12) | (1<<13) | (1<<14) | (1<<15) | (1<<16) - | (1<<17) | (1<<20) | (1<<21) | (1<<22) | (1<<23) - | (1<<24) | (1<<25) | (1<<26) | (1<<27) | (1<<28); regs->version = 1; - for (offs = 0; offs < regs->len; offs += 128) { - u32 len = min_t(u32, 128, regs->len - offs); + memset(p, 0, regs->len); + memcpy_fromio(p, io, B3_RAM_ADDR); - if (bankmap & (1<<(offs/128))) - memcpy_fromio(p + offs, io + offs, len); - else - memset(p + offs, 0, len); - } + memcpy_fromio(p + B3_RI_WTO_R1, io + B3_RI_WTO_R1, + regs->len - B3_RI_WTO_R1); } /* Wake on Lan only supported on Yukon chps with rev 1 or above */ From 383181ac7e59542ff47e2b81f7e4c40aba39b30b Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Mon, 19 Sep 2005 15:37:16 -0700 Subject: [PATCH 112/204] [PATCH] skge: check length from PHY Cleanup receive buffer allocation and management, Add more error handling checks from PHY and bump version. Signed-off-by: Jeff Garzik --- drivers/net/skge.c | 181 +++++++++++++++++++++++---------------------- drivers/net/skge.h | 2 + 2 files changed, 93 insertions(+), 90 deletions(-) diff --git a/drivers/net/skge.c b/drivers/net/skge.c index 62e2ae0eb85..ae1996a3bc5 100644 --- a/drivers/net/skge.c +++ b/drivers/net/skge.c @@ -42,7 +42,7 @@ #include "skge.h" #define DRV_NAME "skge" -#define DRV_VERSION "1.0" +#define DRV_VERSION "1.1" #define PFX DRV_NAME " " #define DEFAULT_TX_RING_SIZE 128 @@ -762,17 +762,6 @@ static int skge_ring_alloc(struct skge_ring *ring, void *vaddr, u64 base) return 0; } -static struct sk_buff *skge_rx_alloc(struct net_device *dev, unsigned int size) -{ - struct sk_buff *skb = dev_alloc_skb(size); - - if (likely(skb)) { - skb->dev = dev; - skb_reserve(skb, NET_IP_ALIGN); - } - return skb; -} - /* Allocate and setup a new buffer for receiving */ static void skge_rx_setup(struct skge_port *skge, struct skge_element *e, struct sk_buff *skb, unsigned int bufsize) @@ -845,16 +834,17 @@ static int skge_rx_fill(struct skge_port *skge) { struct skge_ring *ring = &skge->rx_ring; struct skge_element *e; - unsigned int bufsize = skge->rx_buf_size; e = ring->start; do { - struct sk_buff *skb = skge_rx_alloc(skge->netdev, bufsize); + struct sk_buff *skb; + skb = dev_alloc_skb(skge->rx_buf_size + NET_IP_ALIGN); if (!skb) return -ENOMEM; - skge_rx_setup(skge, e, skb, bufsize); + skb_reserve(skb, NET_IP_ALIGN); + skge_rx_setup(skge, e, skb, skge->rx_buf_size); } while ( (e = e->next) != ring->start); ring->to_clean = ring->start; @@ -2429,6 +2419,14 @@ static void yukon_set_multicast(struct net_device *dev) gma_write16(hw, port, GM_RX_CTRL, reg); } +static inline u16 phy_length(const struct skge_hw *hw, u32 status) +{ + if (hw->chip_id == CHIP_ID_GENESIS) + return status >> XMR_FS_LEN_SHIFT; + else + return status >> GMR_FS_LEN_SHIFT; +} + static inline int bad_phy_status(const struct skge_hw *hw, u32 status) { if (hw->chip_id == CHIP_ID_GENESIS) @@ -2438,16 +2436,81 @@ static inline int bad_phy_status(const struct skge_hw *hw, u32 status) (status & GMR_FS_RX_OK) == 0; } -static void skge_rx_error(struct skge_port *skge, int slot, - u32 control, u32 status) + +/* Get receive buffer from descriptor. + * Handles copy of small buffers and reallocation failures + */ +static inline struct sk_buff *skge_rx_get(struct skge_port *skge, + struct skge_element *e, + u32 control, u32 status, u16 csum) { - if (netif_msg_rx_err(skge)) - printk(KERN_DEBUG PFX "%s: rx err, slot %d control 0x%x status 0x%x\n", - skge->netdev->name, slot, control, status); + struct sk_buff *skb; + u16 len = control & BMU_BBC; + + if (unlikely(netif_msg_rx_status(skge))) + printk(KERN_DEBUG PFX "%s: rx slot %td status 0x%x len %d\n", + skge->netdev->name, e - skge->rx_ring.start, + status, len); + + if (len > skge->rx_buf_size) + goto error; if ((control & (BMU_EOF|BMU_STF)) != (BMU_STF|BMU_EOF)) - skge->net_stats.rx_length_errors++; - else if (skge->hw->chip_id == CHIP_ID_GENESIS) { + goto error; + + if (bad_phy_status(skge->hw, status)) + goto error; + + if (phy_length(skge->hw, status) != len) + goto error; + + if (len < RX_COPY_THRESHOLD) { + skb = dev_alloc_skb(len + 2); + if (!skb) + goto resubmit; + + skb_reserve(skb, 2); + pci_dma_sync_single_for_cpu(skge->hw->pdev, + pci_unmap_addr(e, mapaddr), + len, PCI_DMA_FROMDEVICE); + memcpy(skb->data, e->skb->data, len); + pci_dma_sync_single_for_device(skge->hw->pdev, + pci_unmap_addr(e, mapaddr), + len, PCI_DMA_FROMDEVICE); + skge_rx_reuse(e, skge->rx_buf_size); + } else { + struct sk_buff *nskb; + nskb = dev_alloc_skb(skge->rx_buf_size + NET_IP_ALIGN); + if (!nskb) + goto resubmit; + + pci_unmap_single(skge->hw->pdev, + pci_unmap_addr(e, mapaddr), + pci_unmap_len(e, maplen), + PCI_DMA_FROMDEVICE); + skb = e->skb; + prefetch(skb->data); + skge_rx_setup(skge, e, nskb, skge->rx_buf_size); + } + + skb_put(skb, len); + skb->dev = skge->netdev; + if (skge->rx_csum) { + skb->csum = csum; + skb->ip_summed = CHECKSUM_HW; + } + + skb->protocol = eth_type_trans(skb, skge->netdev); + + return skb; +error: + + if (netif_msg_rx_err(skge)) + printk(KERN_DEBUG PFX "%s: rx err, slot %td control 0x%x status 0x%x\n", + skge->netdev->name, e - skge->rx_ring.start, + control, status); + + if (skge->hw->chip_id == CHIP_ID_GENESIS) { if (status & (XMR_FS_RUNT|XMR_FS_LNG_ERR)) skge->net_stats.rx_length_errors++; if (status & XMR_FS_FRA_ERR) @@ -2462,56 +2525,10 @@ static void skge_rx_error(struct skge_port *skge, int slot, if (status & GMR_FS_CRC_ERR) skge->net_stats.rx_crc_errors++; } -} -/* Get receive buffer from descriptor. - * Handles copy of small buffers and reallocation failures - */ -static inline struct sk_buff *skge_rx_get(struct skge_port *skge, - struct skge_element *e, - unsigned int len) -{ - struct sk_buff *nskb, *skb; - - if (len < RX_COPY_THRESHOLD) { - nskb = skge_rx_alloc(skge->netdev, len + NET_IP_ALIGN); - if (unlikely(!nskb)) - return NULL; - - pci_dma_sync_single_for_cpu(skge->hw->pdev, - pci_unmap_addr(e, mapaddr), - len, PCI_DMA_FROMDEVICE); - memcpy(nskb->data, e->skb->data, len); - pci_dma_sync_single_for_device(skge->hw->pdev, - pci_unmap_addr(e, mapaddr), - len, PCI_DMA_FROMDEVICE); - - if (skge->rx_csum) { - struct skge_rx_desc *rd = e->desc; - nskb->csum = le16_to_cpu(rd->csum2); - nskb->ip_summed = CHECKSUM_HW; - } - skge_rx_reuse(e, skge->rx_buf_size); - return nskb; - } else { - nskb = skge_rx_alloc(skge->netdev, skge->rx_buf_size); - if (unlikely(!nskb)) - return NULL; - - pci_unmap_single(skge->hw->pdev, - pci_unmap_addr(e, mapaddr), - pci_unmap_len(e, maplen), - PCI_DMA_FROMDEVICE); - skb = e->skb; - if (skge->rx_csum) { - struct skge_rx_desc *rd = e->desc; - skb->csum = le16_to_cpu(rd->csum2); - skb->ip_summed = CHECKSUM_HW; - } - - skge_rx_setup(skge, e, nskb, skge->rx_buf_size); - return skb; - } +resubmit: + skge_rx_reuse(e, skge->rx_buf_size); + return NULL; } @@ -2527,32 +2544,16 @@ static int skge_poll(struct net_device *dev, int *budget) for (e = ring->to_clean; work_done < to_do; e = e->next) { struct skge_rx_desc *rd = e->desc; struct sk_buff *skb; - u32 control, len, status; + u32 control; rmb(); control = rd->control; if (control & BMU_OWN) break; - len = control & BMU_BBC; - status = rd->status; - - if (unlikely((control & (BMU_EOF|BMU_STF)) != (BMU_STF|BMU_EOF) - || bad_phy_status(hw, status))) { - skge_rx_error(skge, e - ring->start, control, status); - skge_rx_reuse(e, skge->rx_buf_size); - continue; - } - - if (netif_msg_rx_status(skge)) - printk(KERN_DEBUG PFX "%s: rx slot %td status 0x%x len %d\n", - dev->name, e - ring->start, rd->status, len); - - skb = skge_rx_get(skge, e, len); + skb = skge_rx_get(skge, e, control, rd->status, + le16_to_cpu(rd->csum2)); if (likely(skb)) { - skb_put(skb, len); - skb->protocol = eth_type_trans(skb, dev); - dev->last_rx = jiffies; netif_receive_skb(skb); diff --git a/drivers/net/skge.h b/drivers/net/skge.h index efbf98c675d..72c175b87a5 100644 --- a/drivers/net/skge.h +++ b/drivers/net/skge.h @@ -953,6 +953,7 @@ enum { */ enum { XMR_FS_LEN = 0x3fff<<18, /* Bit 31..18: Rx Frame Length */ + XMR_FS_LEN_SHIFT = 18, XMR_FS_2L_VLAN = 1<<17, /* Bit 17: tagged wh 2Lev VLAN ID*/ XMR_FS_1_VLAN = 1<<16, /* Bit 16: tagged wh 1ev VLAN ID*/ XMR_FS_BC = 1<<15, /* Bit 15: Broadcast Frame */ @@ -1868,6 +1869,7 @@ enum { /* Receive Frame Status Encoding */ enum { GMR_FS_LEN = 0xffff<<16, /* Bit 31..16: Rx Frame Length */ + GMR_FS_LEN_SHIFT = 16, GMR_FS_VLAN = 1<<13, /* Bit 13: VLAN Packet */ GMR_FS_JABBER = 1<<12, /* Bit 12: Jabber Packet */ GMR_FS_UN_SIZE = 1<<11, /* Bit 11: Undersize Packet */ From 0b50f81d5a63428f131ff20596f4e3d473e5b94f Mon Sep 17 00:00:00 2001 From: Tommy Christensen Date: Wed, 21 Sep 2005 12:13:57 -0700 Subject: [PATCH 113/204] [PATCH] r8169: call proper VLAN receive function vlan_hwaccel_rx should be used when in interrupt context. Fixes bug http://bugzilla.kernel.org/show_bug.cgi?id=5284 Signed-off-by: Tommy S. Christensen Cc: Francois Romieu Cc: Jeff Garzik Signed-off-by: Andrew Morton Signed-off-by: Jeff Garzik --- drivers/net/r8169.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c index f0471d102e3..f9223c1c5aa 100644 --- a/drivers/net/r8169.c +++ b/drivers/net/r8169.c @@ -100,11 +100,11 @@ VERSION 2.2LK <2005/01/25> #ifdef CONFIG_R8169_NAPI #define rtl8169_rx_skb netif_receive_skb -#define rtl8169_rx_hwaccel_skb vlan_hwaccel_rx +#define rtl8169_rx_hwaccel_skb vlan_hwaccel_receive_skb #define rtl8169_rx_quota(count, quota) min(count, quota) #else #define rtl8169_rx_skb netif_rx -#define rtl8169_rx_hwaccel_skb vlan_hwaccel_receive_skb +#define rtl8169_rx_hwaccel_skb vlan_hwaccel_rx #define rtl8169_rx_quota(count, quota) count #endif From e86ee6682b649183c11013a98be02f25e9ae399d Mon Sep 17 00:00:00 2001 From: Andy Currid Date: Mon, 19 Sep 2005 06:17:52 -0700 Subject: [PATCH 114/204] [PATCH] Add NVIDIA device ID in sata_nv Signed-off-by: Andy Currid Signed-off-by: Jeff Garzik --- drivers/scsi/sata_nv.c | 2 ++ include/linux/pci_ids.h | 3 ++- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/scsi/sata_nv.c b/drivers/scsi/sata_nv.c index a1d62dee3be..c05653c7779 100644 --- a/drivers/scsi/sata_nv.c +++ b/drivers/scsi/sata_nv.c @@ -158,6 +158,8 @@ static struct pci_device_id nv_pci_tbl[] = { PCI_ANY_ID, PCI_ANY_ID, 0, 0, MCP51 }, { PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA, PCI_ANY_ID, PCI_ANY_ID, 0, 0, MCP55 }, + { PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA2, + PCI_ANY_ID, PCI_ANY_ID, 0, 0, MCP55 }, { PCI_VENDOR_ID_NVIDIA, PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_STORAGE_IDE<<8, 0xffff00, GENERIC }, diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index f6c1a142286..cb414ea42f0 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -1267,7 +1267,8 @@ #define PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_SATA 0x0266 #define PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_SATA2 0x0267 #define PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_IDE 0x036E -#define PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA 0x036F +#define PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA 0x037E +#define PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_SATA2 0x037F #define PCI_DEVICE_ID_NVIDIA_NVENET_12 0x0268 #define PCI_DEVICE_ID_NVIDIA_NVENET_13 0x0269 #define PCI_DEVICE_ID_NVIDIA_MCP51_AUDIO 0x026B From 4c898c7f2f286b204fefc5dddb568f755d195d0c Mon Sep 17 00:00:00 2001 From: Daniel Ritz Date: Thu, 22 Sep 2005 00:47:11 -0700 Subject: [PATCH 115/204] [PATCH] Driver Core: fis bus rescan devices race bus_rescan_devices_helper() does not hold the dev->sem when it checks for !dev->driver(). device_attach() holds the sem, but calls again device_bind_driver() even when dev->driver is set. What happens is that a first device_attach() call (module insertion time) is on the way binding the device to a driver. Another thread calls bus_rescan_devices(). Now when bus_rescan_devices_helper() checks for dev->driver it is still NULL 'cos the the prior device_attach() is not yet finished. But as soon as the first one releases the dev->sem the second device_attach() tries to rebind the already bound device again. device_bind_driver() does this blindly which leads to a corrupt driver->klist_devices list (the device links itself, the head points to the device). Later a call to device_release_driver() sets dev->driver to NULL and breaks the link it has to itself on knode_driver. Rmmoding the driver later calls driver_detach() which leads to an endless loop 'cos the list head in klist_devices still points to the device. And since dev->driver is NULL it's stuck with the same device forever. Boom. And rmmod hangs. Very easy to reproduce with new-style pcmcia and a 16bit card. Just loop modprobe ;cardctl eject; rmmod . Easiest fix is to check if the device is already bound to a driver in device_bind_driver(). This avoids the double binding. Signed-off-by: Daniel Ritz Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/base/dd.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/base/dd.c b/drivers/base/dd.c index d5bbce38282..3565e979530 100644 --- a/drivers/base/dd.c +++ b/drivers/base/dd.c @@ -40,6 +40,9 @@ */ void device_bind_driver(struct device * dev) { + if (klist_node_attached(&dev->knode_driver)) + return; + pr_debug("bound device '%s' to driver '%s'\n", dev->bus_id, dev->driver->name); klist_add_tail(&dev->knode_driver, &dev->driver->klist_devices); From d305ef5d2a4e77bfa66160513f4a7494126a506b Mon Sep 17 00:00:00 2001 From: Daniel Ritz Date: Thu, 22 Sep 2005 00:47:24 -0700 Subject: [PATCH 116/204] [PATCH] driver core: add helper device_is_registered() add the helper and use it instead of open coding the klist_node_attached() check (which is a layering violation IMHO) idea by Alan Stern. Signed-off-by: Daniel Ritz Cc: Alan Stern Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/s390/cio/ccwgroup.c | 2 +- drivers/usb/core/message.c | 2 +- drivers/usb/core/usb.c | 6 +++--- include/linux/device.h | 5 +++++ 4 files changed, 10 insertions(+), 5 deletions(-) diff --git a/drivers/s390/cio/ccwgroup.c b/drivers/s390/cio/ccwgroup.c index 91ea8e4777f..dbb3eb0e330 100644 --- a/drivers/s390/cio/ccwgroup.c +++ b/drivers/s390/cio/ccwgroup.c @@ -437,7 +437,7 @@ __ccwgroup_get_gdev_by_cdev(struct ccw_device *cdev) if (cdev->dev.driver_data) { gdev = (struct ccwgroup_device *)cdev->dev.driver_data; if (get_device(&gdev->dev)) { - if (klist_node_attached(&gdev->dev.knode_bus)) + if (device_is_registered(&gdev->dev)) return gdev; put_device(&gdev->dev); } diff --git a/drivers/usb/core/message.c b/drivers/usb/core/message.c index c47c8052b48..f1fb67fe22a 100644 --- a/drivers/usb/core/message.c +++ b/drivers/usb/core/message.c @@ -987,7 +987,7 @@ void usb_disable_device(struct usb_device *dev, int skip_ep0) /* remove this interface if it has been registered */ interface = dev->actconfig->interface[i]; - if (!klist_node_attached(&interface->dev.knode_bus)) + if (!device_is_registered(&interface->dev)) continue; dev_dbg (&dev->dev, "unregistering interface %s\n", interface->dev.bus_id); diff --git a/drivers/usb/core/usb.c b/drivers/usb/core/usb.c index 087af73a59d..7d131509e41 100644 --- a/drivers/usb/core/usb.c +++ b/drivers/usb/core/usb.c @@ -303,7 +303,7 @@ int usb_driver_claim_interface(struct usb_driver *driver, /* if interface was already added, bind now; else let * the future device_add() bind it, bypassing probe() */ - if (klist_node_attached(&dev->knode_bus)) + if (device_is_registered(dev)) device_bind_driver(dev); return 0; @@ -336,8 +336,8 @@ void usb_driver_release_interface(struct usb_driver *driver, if (iface->condition != USB_INTERFACE_BOUND) return; - /* release only after device_add() */ - if (klist_node_attached(&dev->knode_bus)) { + /* don't release if the interface hasn't been added yet */ + if (device_is_registered(dev)) { iface->condition = USB_INTERFACE_UNBINDING; device_release_driver(dev); } diff --git a/include/linux/device.h b/include/linux/device.h index 06e5d42f2c7..95d607a48f0 100644 --- a/include/linux/device.h +++ b/include/linux/device.h @@ -317,6 +317,11 @@ dev_set_drvdata (struct device *dev, void *data) dev->driver_data = data; } +static inline int device_is_registered(struct device *dev) +{ + return klist_node_attached(&dev->knode_bus); +} + /* * High level routines for use by the bus drivers */ From 3e51377dc412df9d4933c4fd1a147b5b560abe10 Mon Sep 17 00:00:00 2001 From: Bill Nottingham Date: Thu, 22 Sep 2005 00:47:36 -0700 Subject: [PATCH 117/204] [PATCH] fix class symlinks in sysfs The class symlinks in sysfs don't properly handle changing device names. To demonstrate, rename your network device from eth0 to eth1. Your pci (or usb, or whatever) device will still have a 'net:eth0' link, except now it points to /sys/class/net/eth1. The attached patch makes sure the class symlink name changes when the class device name changes. It isn't 100% correct, it should be using sysfs_rename_link. Unfortunately, sysfs_rename_link doesn't exist. Signed-off-by: Bill Nottingham Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/base/class.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/drivers/base/class.c b/drivers/base/class.c index 3b112e3542f..ce23dc8c18c 100644 --- a/drivers/base/class.c +++ b/drivers/base/class.c @@ -669,6 +669,7 @@ void class_device_destroy(struct class *cls, dev_t devt) int class_device_rename(struct class_device *class_dev, char *new_name) { int error = 0; + char *old_class_name = NULL, *new_class_name = NULL; class_dev = class_device_get(class_dev); if (!class_dev) @@ -677,12 +678,24 @@ int class_device_rename(struct class_device *class_dev, char *new_name) pr_debug("CLASS: renaming '%s' to '%s'\n", class_dev->class_id, new_name); + if (class_dev->dev) + old_class_name = make_class_name(class_dev); + strlcpy(class_dev->class_id, new_name, KOBJ_NAME_LEN); error = kobject_rename(&class_dev->kobj, new_name); + if (class_dev->dev) { + new_class_name = make_class_name(class_dev); + sysfs_create_link(&class_dev->dev->kobj, &class_dev->kobj, + new_class_name); + sysfs_remove_link(&class_dev->dev->kobj, old_class_name); + } class_device_put(class_dev); + kfree(old_class_name); + kfree(new_class_name); + return error; } From dba28010b27ab56d2c4cec5f361e0947295addfd Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 22 Sep 2005 00:47:46 -0700 Subject: [PATCH 118/204] [PATCH] I2C: remove me from the MAINTAINERS file for i2c Remove my name from the I2C maintainer, Jean is more than capable of handling it all now. Signed-off-by: Greg Kroah-Hartman Signed-off-by: Jean Delvare Signed-off-by: Linus Torvalds --- MAINTAINERS | 2 -- 1 file changed, 2 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index ade7415d246..78aca12101a 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1063,8 +1063,6 @@ M: wli@holomorphy.com S: Maintained I2C SUBSYSTEM -P: Greg Kroah-Hartman -M: greg@kroah.com P: Jean Delvare M: khali@linux-fr.org L: lm-sensors@lm-sensors.org From 1029d6b58adc3225911c56af26895871dd2ea8cf Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Thu, 22 Sep 2005 00:48:02 -0700 Subject: [PATCH 119/204] [PATCH] PCI: remove unused "scratch" Unused variable. Signed-off-by: Bjorn Helgaas Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/pci/hotplug.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/drivers/pci/hotplug.c b/drivers/pci/hotplug.c index 10444988a10..e1743be3190 100644 --- a/drivers/pci/hotplug.c +++ b/drivers/pci/hotplug.c @@ -7,7 +7,6 @@ int pci_hotplug (struct device *dev, char **envp, int num_envp, char *buffer, int buffer_size) { struct pci_dev *pdev; - char *scratch; int i = 0; int length = 0; @@ -18,9 +17,6 @@ int pci_hotplug (struct device *dev, char **envp, int num_envp, if (!pdev) return -ENODEV; - scratch = buffer; - - if (add_hotplug_env_var(envp, num_envp, &i, buffer, buffer_size, &length, "PCI_CLASS=%04X", pdev->class)) From 656da9da3745abcbbbdca598745d04c6de2c8843 Mon Sep 17 00:00:00 2001 From: Pekka Enberg Date: Thu, 22 Sep 2005 00:48:11 -0700 Subject: [PATCH 120/204] [PATCH] PCI: convert kcalloc to kzalloc This patch converts kcalloc(1, ...) calls to use the new kzalloc() function. Signed-off-by: Pekka Enberg Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/pci/hotplug/sgi_hotplug.c | 6 +++--- drivers/pci/pci-sysfs.c | 2 +- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/pci/hotplug/sgi_hotplug.c b/drivers/pci/hotplug/sgi_hotplug.c index b1409441c1c..a32ae82e592 100644 --- a/drivers/pci/hotplug/sgi_hotplug.c +++ b/drivers/pci/hotplug/sgi_hotplug.c @@ -159,7 +159,7 @@ static int sn_hp_slot_private_alloc(struct hotplug_slot *bss_hotplug_slot, pcibus_info = SN_PCIBUS_BUSSOFT_INFO(pci_bus); - slot = kcalloc(1, sizeof(*slot), GFP_KERNEL); + slot = kzalloc(sizeof(*slot), GFP_KERNEL); if (!slot) return -ENOMEM; bss_hotplug_slot->private = slot; @@ -491,7 +491,7 @@ static int sn_hotplug_slot_register(struct pci_bus *pci_bus) if (sn_pci_slot_valid(pci_bus, device) != 1) continue; - bss_hotplug_slot = kcalloc(1, sizeof(*bss_hotplug_slot), + bss_hotplug_slot = kzalloc(sizeof(*bss_hotplug_slot), GFP_KERNEL); if (!bss_hotplug_slot) { rc = -ENOMEM; @@ -499,7 +499,7 @@ static int sn_hotplug_slot_register(struct pci_bus *pci_bus) } bss_hotplug_slot->info = - kcalloc(1, sizeof(struct hotplug_slot_info), + kzalloc(sizeof(struct hotplug_slot_info), GFP_KERNEL); if (!bss_hotplug_slot->info) { rc = -ENOMEM; diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index 56a3b397efe..2898830c496 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -360,7 +360,7 @@ pci_create_resource_files(struct pci_dev *pdev) continue; /* allocate attribute structure, piggyback attribute name */ - res_attr = kcalloc(1, sizeof(*res_attr) + 10, GFP_ATOMIC); + res_attr = kzalloc(sizeof(*res_attr) + 10, GFP_ATOMIC); if (res_attr) { char *res_attr_name = (char *)(res_attr + 1); From 3c6de9295d28a4fc868b2c09f23e318e3e7b9b6b Mon Sep 17 00:00:00 2001 From: Amos Waterland Date: Thu, 22 Sep 2005 00:48:19 -0700 Subject: [PATCH 121/204] [PATCH] fix drivers/pci/probe.c warning This function expects an unsigned 32-bit type as its third argument: static u32 pci_size(u32 base, u32 maxbase, u32 mask) However, given these definitions: #define PCI_BASE_ADDRESS_MEM_MASK (~0x0fUL) #define PCI_ROM_ADDRESS_MASK (~0x7ffUL) these two calls in drivers/pci/probe.c are problematic for architectures for which a UL is not equivalent to a u32: sz = pci_size(l, sz, PCI_BASE_ADDRESS_MEM_MASK); sz = pci_size(l, sz, PCI_ROM_ADDRESS_MASK); Hence the below compile warning when building for ARCH=ppc64: drivers/pci/probe.c: In function `pci_read_bases': /.../probe.c:168: warning: large integer implicitly truncated to unsigned type /.../probe.c:218: warning: large integer implicitly truncated to unsigned type Here is a simple fix. Signed-off-by: Amos Waterland Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/pci/probe.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c index 26a55d08b50..c77d5b1bbff 100644 --- a/drivers/pci/probe.c +++ b/drivers/pci/probe.c @@ -165,7 +165,7 @@ static void pci_read_bases(struct pci_dev *dev, unsigned int howmany, int rom) if (l == 0xffffffff) l = 0; if ((l & PCI_BASE_ADDRESS_SPACE) == PCI_BASE_ADDRESS_SPACE_MEMORY) { - sz = pci_size(l, sz, PCI_BASE_ADDRESS_MEM_MASK); + sz = pci_size(l, sz, (u32)PCI_BASE_ADDRESS_MEM_MASK); if (!sz) continue; res->start = l & PCI_BASE_ADDRESS_MEM_MASK; @@ -215,7 +215,7 @@ static void pci_read_bases(struct pci_dev *dev, unsigned int howmany, int rom) if (l == 0xffffffff) l = 0; if (sz && sz != 0xffffffff) { - sz = pci_size(l, sz, PCI_ROM_ADDRESS_MASK); + sz = pci_size(l, sz, (u32)PCI_ROM_ADDRESS_MASK); if (sz) { res->flags = (l & IORESOURCE_ROM_ENABLE) | IORESOURCE_MEM | IORESOURCE_PREFETCH | From 02fe75a9ad797b4f8ccf1ee8e49833b77cc30c6a Mon Sep 17 00:00:00 2001 From: Linda Xie Date: Thu, 22 Sep 2005 00:48:24 -0700 Subject: [PATCH 122/204] [PATCH] PCI Hotplug: Fix buffer overrun in rpadlpar_sysfs.c Signed-off-by: Linda Xie Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/pci/hotplug/rpadlpar_sysfs.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/pci/hotplug/rpadlpar_sysfs.c b/drivers/pci/hotplug/rpadlpar_sysfs.c index 752e6513c44..db69be85b45 100644 --- a/drivers/pci/hotplug/rpadlpar_sysfs.c +++ b/drivers/pci/hotplug/rpadlpar_sysfs.c @@ -62,7 +62,7 @@ static ssize_t add_slot_store(struct dlpar_io_attr *dlpar_attr, char drc_name[MAX_DRC_NAME_LEN]; char *end; - if (nbytes > MAX_DRC_NAME_LEN) + if (nbytes >= MAX_DRC_NAME_LEN) return 0; memcpy(drc_name, buf, nbytes); @@ -83,7 +83,7 @@ static ssize_t remove_slot_store(struct dlpar_io_attr *dlpar_attr, char drc_name[MAX_DRC_NAME_LEN]; char *end; - if (nbytes > MAX_DRC_NAME_LEN) + if (nbytes >= MAX_DRC_NAME_LEN) return 0; memcpy(drc_name, buf, nbytes); From 64bd84538ac7f8ba3080eda4d92c66a8441cc6cc Mon Sep 17 00:00:00 2001 From: Pete Zaitcev Date: Thu, 22 Sep 2005 00:48:29 -0700 Subject: [PATCH 123/204] [PATCH] ub: fix burning cds This patch fixes a few problems with ub and cleans up a couple of things: - Bump UB_MAX_REQ_SG, this allows to burn CDs - Drop initialization of urb.transfer_flags, now that URB_UNLINK_ASYNC is gone - Add forgotten processing of stalls at GetMaxLUN - Remove a few more P3-tagged printks whose time has come - Correct comment about ZIP-100 Signed-off-by: Pete Zaitcev Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds drivers/block/ub.c | 53 +++++++++++++++++++++++++++-------------------------- 1 file changed, 27 insertions(+), 26 deletions(-) --- drivers/block/ub.c | 53 +++++++++++++++++++++++----------------------- 1 file changed, 27 insertions(+), 26 deletions(-) diff --git a/drivers/block/ub.c b/drivers/block/ub.c index aa0bf7ee008..dfe3581b482 100644 --- a/drivers/block/ub.c +++ b/drivers/block/ub.c @@ -172,7 +172,7 @@ struct bulk_cs_wrap { */ struct ub_dev; -#define UB_MAX_REQ_SG 4 +#define UB_MAX_REQ_SG 9 /* cdrecord requires 32KB and maybe a header */ #define UB_MAX_SECTORS 64 /* @@ -387,7 +387,7 @@ struct ub_dev { struct bulk_cs_wrap work_bcs; struct usb_ctrlrequest work_cr; - int sg_stat[UB_MAX_REQ_SG+1]; + int sg_stat[6]; struct ub_scsi_trace tr; }; @@ -525,12 +525,13 @@ static ssize_t ub_diag_show(struct device *dev, struct device_attribute *attr, "qlen %d qmax %d\n", sc->cmd_queue.qlen, sc->cmd_queue.qmax); cnt += sprintf(page + cnt, - "sg %d %d %d %d %d\n", + "sg %d %d %d %d %d .. %d\n", sc->sg_stat[0], sc->sg_stat[1], sc->sg_stat[2], sc->sg_stat[3], - sc->sg_stat[4]); + sc->sg_stat[4], + sc->sg_stat[5]); list_for_each (p, &sc->luns) { lun = list_entry(p, struct ub_lun, link); @@ -835,7 +836,7 @@ static int ub_cmd_build_block(struct ub_dev *sc, struct ub_lun *lun, return -1; } cmd->nsg = n_elem; - sc->sg_stat[n_elem]++; + sc->sg_stat[n_elem < 5 ? n_elem : 5]++; /* * build the command @@ -891,7 +892,7 @@ static int ub_cmd_build_packet(struct ub_dev *sc, struct ub_lun *lun, return -1; } cmd->nsg = n_elem; - sc->sg_stat[n_elem]++; + sc->sg_stat[n_elem < 5 ? n_elem : 5]++; memcpy(&cmd->cdb, rq->cmd, rq->cmd_len); cmd->cdb_len = rq->cmd_len; @@ -1010,7 +1011,6 @@ static int ub_scsi_cmd_start(struct ub_dev *sc, struct ub_scsi_cmd *cmd) sc->last_pipe = sc->send_bulk_pipe; usb_fill_bulk_urb(&sc->work_urb, sc->dev, sc->send_bulk_pipe, bcb, US_BULK_CB_WRAP_LEN, ub_urb_complete, sc); - sc->work_urb.transfer_flags = 0; /* Fill what we shouldn't be filling, because usb-storage did so. */ sc->work_urb.actual_length = 0; @@ -1019,7 +1019,6 @@ static int ub_scsi_cmd_start(struct ub_dev *sc, struct ub_scsi_cmd *cmd) if ((rc = usb_submit_urb(&sc->work_urb, GFP_ATOMIC)) != 0) { /* XXX Clear stalls */ - printk("ub: cmd #%d start failed (%d)\n", cmd->tag, rc); /* P3 */ ub_complete(&sc->work_done); return rc; } @@ -1190,11 +1189,9 @@ static void ub_scsi_urb_compl(struct ub_dev *sc, struct ub_scsi_cmd *cmd) return; } if (urb->status != 0) { - printk("ub: cmd #%d cmd status (%d)\n", cmd->tag, urb->status); /* P3 */ goto Bad_End; } if (urb->actual_length != US_BULK_CB_WRAP_LEN) { - printk("ub: cmd #%d xferred %d\n", cmd->tag, urb->actual_length); /* P3 */ /* XXX Must do reset here to unconfuse the device */ goto Bad_End; } @@ -1395,14 +1392,12 @@ static void ub_data_start(struct ub_dev *sc, struct ub_scsi_cmd *cmd) usb_fill_bulk_urb(&sc->work_urb, sc->dev, pipe, page_address(sg->page) + sg->offset, sg->length, ub_urb_complete, sc); - sc->work_urb.transfer_flags = 0; sc->work_urb.actual_length = 0; sc->work_urb.error_count = 0; sc->work_urb.status = 0; if ((rc = usb_submit_urb(&sc->work_urb, GFP_ATOMIC)) != 0) { /* XXX Clear stalls */ - printk("ub: data #%d submit failed (%d)\n", cmd->tag, rc); /* P3 */ ub_complete(&sc->work_done); ub_state_done(sc, cmd, rc); return; @@ -1442,7 +1437,6 @@ static int __ub_state_stat(struct ub_dev *sc, struct ub_scsi_cmd *cmd) sc->last_pipe = sc->recv_bulk_pipe; usb_fill_bulk_urb(&sc->work_urb, sc->dev, sc->recv_bulk_pipe, &sc->work_bcs, US_BULK_CS_WRAP_LEN, ub_urb_complete, sc); - sc->work_urb.transfer_flags = 0; sc->work_urb.actual_length = 0; sc->work_urb.error_count = 0; sc->work_urb.status = 0; @@ -1563,7 +1557,6 @@ static int ub_submit_clear_stall(struct ub_dev *sc, struct ub_scsi_cmd *cmd, usb_fill_control_urb(&sc->work_urb, sc->dev, sc->send_ctrl_pipe, (unsigned char*) cr, NULL, 0, ub_urb_complete, sc); - sc->work_urb.transfer_flags = 0; sc->work_urb.actual_length = 0; sc->work_urb.error_count = 0; sc->work_urb.status = 0; @@ -2000,17 +1993,16 @@ static int ub_sync_getmaxlun(struct ub_dev *sc) usb_fill_control_urb(&sc->work_urb, sc->dev, sc->recv_ctrl_pipe, (unsigned char*) cr, p, 1, ub_probe_urb_complete, &compl); - sc->work_urb.transfer_flags = 0; sc->work_urb.actual_length = 0; sc->work_urb.error_count = 0; sc->work_urb.status = 0; if ((rc = usb_submit_urb(&sc->work_urb, GFP_KERNEL)) != 0) { if (rc == -EPIPE) { - printk("%s: Stall at GetMaxLUN, using 1 LUN\n", + printk("%s: Stall submitting GetMaxLUN, using 1 LUN\n", sc->name); /* P3 */ } else { - printk(KERN_WARNING + printk(KERN_NOTICE "%s: Unable to submit GetMaxLUN (%d)\n", sc->name, rc); } @@ -2028,6 +2020,18 @@ static int ub_sync_getmaxlun(struct ub_dev *sc) del_timer_sync(&timer); usb_kill_urb(&sc->work_urb); + if ((rc = sc->work_urb.status) < 0) { + if (rc == -EPIPE) { + printk("%s: Stall at GetMaxLUN, using 1 LUN\n", + sc->name); /* P3 */ + } else { + printk(KERN_NOTICE + "%s: Error at GetMaxLUN (%d)\n", + sc->name, rc); + } + goto err_io; + } + if (sc->work_urb.actual_length != 1) { printk("%s: GetMaxLUN returned %d bytes\n", sc->name, sc->work_urb.actual_length); /* P3 */ @@ -2048,6 +2052,7 @@ static int ub_sync_getmaxlun(struct ub_dev *sc) kfree(p); return nluns; +err_io: err_submit: kfree(p); err_alloc: @@ -2080,7 +2085,6 @@ static int ub_probe_clear_stall(struct ub_dev *sc, int stalled_pipe) usb_fill_control_urb(&sc->work_urb, sc->dev, sc->send_ctrl_pipe, (unsigned char*) cr, NULL, 0, ub_probe_urb_complete, &compl); - sc->work_urb.transfer_flags = 0; sc->work_urb.actual_length = 0; sc->work_urb.error_count = 0; sc->work_urb.status = 0; @@ -2241,10 +2245,10 @@ static int ub_probe(struct usb_interface *intf, for (i = 0; i < 3; i++) { if ((rc = ub_sync_getmaxlun(sc)) < 0) { /* - * Some devices (i.e. Iomega Zip100) need this -- - * apparently the bulk pipes get STALLed when the - * GetMaxLUN request is processed. - * XXX I have a ZIP-100, verify it does this. + * This segment is taken from usb-storage. They say + * that ZIP-100 needs this, but my own ZIP-100 works + * fine without this. + * Still, it does not seem to hurt anything. */ if (rc == -EPIPE) { ub_probe_clear_stall(sc, sc->recv_bulk_pipe); @@ -2313,7 +2317,7 @@ static int ub_probe_lun(struct ub_dev *sc, int lnum) disk->first_minor = lun->id * UB_MINORS_PER_MAJOR; disk->fops = &ub_bd_fops; disk->private_data = lun; - disk->driverfs_dev = &sc->intf->dev; /* XXX Many to one ok? */ + disk->driverfs_dev = &sc->intf->dev; rc = -ENOMEM; if ((q = blk_init_queue(ub_request_fn, &sc->lock)) == NULL) @@ -2466,9 +2470,6 @@ static int __init ub_init(void) { int rc; - /* P3 */ printk("ub: sizeof ub_scsi_cmd %zu ub_dev %zu ub_lun %zu\n", - sizeof(struct ub_scsi_cmd), sizeof(struct ub_dev), sizeof(struct ub_lun)); - if ((rc = register_blkdev(UB_MAJOR, DRV_NAME)) != 0) goto err_regblkdev; devfs_mk_dir(DEVFS_NAME); From b6137383bda844a433d65e027502df7b20ba45c2 Mon Sep 17 00:00:00 2001 From: Matthias Urlichs Date: Thu, 22 Sep 2005 00:48:40 -0700 Subject: [PATCH 124/204] [PATCH] USB: more device IDs for Option card driver Added support for HUAWEI E600 and Audiovox AirCard User reports say that these devices work without driver modification. Signed-off-by: Matthias Urlichs Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/usb/serial/option.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c index 92d0f925d05..616bea8df53 100644 --- a/drivers/usb/serial/option.c +++ b/drivers/usb/serial/option.c @@ -25,6 +25,7 @@ 2005-06-20 v0.4.1 add missing braces :-/ killed end-of-line whitespace 2005-07-15 v0.4.2 rename WLAN product to FUSION, add FUSION2 + 2005-09-10 v0.4.3 added HUAWEI E600 card and Audiovox AirCard Work sponsored by: Sigos GmbH, Germany @@ -71,15 +72,21 @@ static int option_send_setup(struct usb_serial_port *port); /* Vendor and product IDs */ #define OPTION_VENDOR_ID 0x0AF0 +#define HUAWEI_VENDOR_ID 0x12D1 +#define AUDIOVOX_VENDOR_ID 0x0F3D #define OPTION_PRODUCT_OLD 0x5000 #define OPTION_PRODUCT_FUSION 0x6000 #define OPTION_PRODUCT_FUSION2 0x6300 +#define HUAWEI_PRODUCT_E600 0x1001 +#define AUDIOVOX_PRODUCT_AIRCARD 0x0112 static struct usb_device_id option_ids[] = { { USB_DEVICE(OPTION_VENDOR_ID, OPTION_PRODUCT_OLD) }, { USB_DEVICE(OPTION_VENDOR_ID, OPTION_PRODUCT_FUSION) }, { USB_DEVICE(OPTION_VENDOR_ID, OPTION_PRODUCT_FUSION2) }, + { USB_DEVICE(HUAWEI_VENDOR_ID, HUAWEI_PRODUCT_E600) }, + { USB_DEVICE(AUDIOVOX_VENDOR_ID, AUDIOVOX_PRODUCT_AIRCARD) }, { } /* Terminating entry */ }; From 72a755fce0243d4168a6153813a1b533b05a7d72 Mon Sep 17 00:00:00 2001 From: Peter Favrholdt Date: Thu, 22 Sep 2005 00:48:49 -0700 Subject: [PATCH 125/204] [PATCH] USB: ftdi_sio: allow baud rate to be changed without raising RTS and DTR I'm using a 2 port USB RS232 dongle to connect to a serial-IR cradle for a bar code reader). Detecting the baudrate of the serial-IR involves keeping DTR low while changing baudrate. This works using normal 16550A serial ports as well as the FTDI driver version 1.4.0 (Linux 2.6.8) but stopped working with the change to "ensure RTS and DTR are raised when changing baudrate" introduced in version 1.4.1 (Linux 2.6.9). The attached patch fixes this, so RTS and DTR is only raised when changing baudrate iff the previous baudrate was B0. Signed-off-by: Peter Favrholdt Signed-off-by: Ian Abbott Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/usb/serial/ftdi_sio.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/usb/serial/ftdi_sio.c b/drivers/usb/serial/ftdi_sio.c index 4e434cb10bb..5a8631c8a4a 100644 --- a/drivers/usb/serial/ftdi_sio.c +++ b/drivers/usb/serial/ftdi_sio.c @@ -1846,10 +1846,12 @@ static void ftdi_set_termios (struct usb_serial_port *port, struct termios *old_ } else { /* set the baudrate determined before */ if (change_speed(port)) { - err("%s urb failed to set baurdrate", __FUNCTION__); + err("%s urb failed to set baudrate", __FUNCTION__); + } + /* Ensure RTS and DTR are raised when baudrate changed from 0 */ + if ((old_termios->c_cflag & CBAUD) == B0) { + set_mctrl(port, TIOCM_DTR | TIOCM_RTS); } - /* Ensure RTS and DTR are raised */ - set_mctrl(port, TIOCM_DTR | TIOCM_RTS); } /* Set flow control */ From 2ba08e825e5a666f540bff15e9977725675e8de6 Mon Sep 17 00:00:00 2001 From: Richard Purdie Date: Thu, 22 Sep 2005 00:48:58 -0700 Subject: [PATCH 126/204] [PATCH] USB: fix pxa2xx_udc compile warnings This patch fixes several types in the PXA25x udc driver and hence fixes several compiler warnings. Signed-off-by: Richard Purdie Acked-by: David Brownell Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/usb/gadget/pxa2xx_udc.c | 4 ++-- drivers/usb/gadget/pxa2xx_udc.h | 8 ++++---- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/usb/gadget/pxa2xx_udc.c b/drivers/usb/gadget/pxa2xx_udc.c index 1507738337c..73f8c940415 100644 --- a/drivers/usb/gadget/pxa2xx_udc.c +++ b/drivers/usb/gadget/pxa2xx_udc.c @@ -422,7 +422,7 @@ static inline void ep0_idle (struct pxa2xx_udc *dev) } static int -write_packet(volatile u32 *uddr, struct pxa2xx_request *req, unsigned max) +write_packet(volatile unsigned long *uddr, struct pxa2xx_request *req, unsigned max) { u8 *buf; unsigned length, count; @@ -2602,7 +2602,7 @@ static int __exit pxa2xx_udc_remove(struct device *_dev) * VBUS IRQs should probably be ignored so that the PXA device just acts * "dead" to USB hosts until system resume. */ -static int pxa2xx_udc_suspend(struct device *dev, u32 state, u32 level) +static int pxa2xx_udc_suspend(struct device *dev, pm_message_t state, u32 level) { struct pxa2xx_udc *udc = dev_get_drvdata(dev); diff --git a/drivers/usb/gadget/pxa2xx_udc.h b/drivers/usb/gadget/pxa2xx_udc.h index d0bc396a85d..a58f3e6e71f 100644 --- a/drivers/usb/gadget/pxa2xx_udc.h +++ b/drivers/usb/gadget/pxa2xx_udc.h @@ -69,11 +69,11 @@ struct pxa2xx_ep { * UDDR = UDC Endpoint Data Register (the fifo) * DRCM = DMA Request Channel Map */ - volatile u32 *reg_udccs; - volatile u32 *reg_ubcr; - volatile u32 *reg_uddr; + volatile unsigned long *reg_udccs; + volatile unsigned long *reg_ubcr; + volatile unsigned long *reg_uddr; #ifdef USE_DMA - volatile u32 *reg_drcmr; + volatile unsigned long *reg_drcmr; #define drcmr(n) .reg_drcmr = & DRCMR ## n , #else #define drcmr(n) From 4b2e790a4d73d729d936cc42f3b08af34f8ea5c6 Mon Sep 17 00:00:00 2001 From: David Brownell Date: Thu, 22 Sep 2005 00:49:07 -0700 Subject: [PATCH 127/204] [PATCH] USB: sl811-hcd minor fixes Three minor sl811-hcd fixes: - Elminate memory leak on one (rare) disable/shutdown path. - For periodic transfers that don't need to be scheduled, update urb->start_frame to represent the transfer phase correctly. - Report the (single) port as removable, by default. Since no drivers yet use start_frame or that part of the hub descriptor, only that leak is likely to ever matter. Signed-off-by: David Brownell Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds drivers/usb/host/sl811-hcd.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) --- drivers/usb/host/sl811-hcd.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) diff --git a/drivers/usb/host/sl811-hcd.c b/drivers/usb/host/sl811-hcd.c index d2a1fd40dfc..d42a15d10a4 100644 --- a/drivers/usb/host/sl811-hcd.c +++ b/drivers/usb/host/sl811-hcd.c @@ -782,6 +782,9 @@ retry: /* usb 1.1 says max 90% of a frame is available for periodic transfers. * this driver doesn't promise that much since it's got to handle an * IRQ per packet; irq handling latencies also use up that time. + * + * NOTE: the periodic schedule is a sparse tree, with the load for + * each branch minimized. see fig 3.5 in the OHCI spec for example. */ #define MAX_PERIODIC_LOAD 500 /* out of 1000 usec */ @@ -843,6 +846,7 @@ static int sl811h_urb_enqueue( if (!(sl811->port1 & (1 << USB_PORT_FEAT_ENABLE)) || !HC_IS_RUNNING(hcd->state)) { retval = -ENODEV; + kfree(ep); goto fail; } @@ -911,8 +915,16 @@ static int sl811h_urb_enqueue( case PIPE_ISOCHRONOUS: case PIPE_INTERRUPT: urb->interval = ep->period; - if (ep->branch < PERIODIC_SIZE) + if (ep->branch < PERIODIC_SIZE) { + /* NOTE: the phase is correct here, but the value + * needs offsetting by the transfer queue depth. + * All current drivers ignore start_frame, so this + * is unlikely to ever matter... + */ + urb->start_frame = (sl811->frame & (PERIODIC_SIZE - 1)) + + ep->branch; break; + } retval = balance(sl811, ep->period, ep->load); if (retval < 0) @@ -1122,7 +1134,7 @@ sl811h_hub_descriptor ( desc->wHubCharacteristics = (__force __u16)cpu_to_le16(temp); /* two bitmaps: ports removable, and legacy PortPwrCtrlMask */ - desc->bitmap[0] = 1 << 1; + desc->bitmap[0] = 0 << 1; desc->bitmap[1] = ~0; } From a85a46f2c38def5150251816890393803fdae7d0 Mon Sep 17 00:00:00 2001 From: Kevin Vigor Date: Thu, 22 Sep 2005 00:49:24 -0700 Subject: [PATCH 128/204] [PATCH] USB: fix pegasus driver Addresses some small bugs in the pegasus ethernet-over-USB driver. Specifically, malformed long packets from the adapter could cause a kernel panic; the interrupt interval calculation was inappropriate for high-speed devices; the return code from read_mii_word was tested incorrectly; and failure to unlink outstanding URBs before freeing them could lead to kernel panics when unloading the driver. Signed-off-by: Kevin Vigor Cc: Petko Manolov Signed-off-by: Andrew Morton Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/usb/net/pegasus.c | 29 ++++++++++++++++++++--------- 1 file changed, 20 insertions(+), 9 deletions(-) diff --git a/drivers/usb/net/pegasus.c b/drivers/usb/net/pegasus.c index 7484d34780f..6a4ffe6c397 100644 --- a/drivers/usb/net/pegasus.c +++ b/drivers/usb/net/pegasus.c @@ -647,6 +647,13 @@ static void read_bulk_callback(struct urb *urb, struct pt_regs *regs) pkt_len -= 8; } + /* + * If the packet is unreasonably long, quietly drop it rather than + * kernel panicing by calling skb_put. + */ + if (pkt_len > PEGASUS_MTU) + goto goon; + /* * at this point we are sure pegasus->rx_skb != NULL * so we go ahead and pass up the packet. @@ -886,15 +893,17 @@ static inline void get_interrupt_interval(pegasus_t * pegasus) __u8 data[2]; read_eprom_word(pegasus, 4, (__u16 *) data); - if (data[1] < 0x80) { - if (netif_msg_timer(pegasus)) - dev_info(&pegasus->intf->dev, - "intr interval changed from %ums to %ums\n", - data[1], 0x80); - data[1] = 0x80; -#ifdef PEGASUS_WRITE_EEPROM - write_eprom_word(pegasus, 4, *(__u16 *) data); + if (pegasus->usb->speed != USB_SPEED_HIGH) { + if (data[1] < 0x80) { + if (netif_msg_timer(pegasus)) + dev_info(&pegasus->intf->dev, "intr interval " + "changed from %ums to %ums\n", + data[1], 0x80); + data[1] = 0x80; +#ifdef PEGASUS_WRITE_EEPROM + write_eprom_word(pegasus, 4, *(__u16 *) data); #endif + } } pegasus->intr_interval = data[1]; } @@ -904,8 +913,9 @@ static void set_carrier(struct net_device *net) pegasus_t *pegasus = netdev_priv(net); u16 tmp; - if (read_mii_word(pegasus, pegasus->phy, MII_BMSR, &tmp)) + if (!read_mii_word(pegasus, pegasus->phy, MII_BMSR, &tmp)) return; + if (tmp & BMSR_LSTATUS) netif_carrier_on(net); else @@ -1355,6 +1365,7 @@ static void pegasus_disconnect(struct usb_interface *intf) cancel_delayed_work(&pegasus->carrier_check); unregister_netdev(pegasus->net); usb_put_dev(interface_to_usbdev(intf)); + unlink_all_urbs(pegasus); free_all_urbs(pegasus); free_skb_pool(pegasus); if (pegasus->rx_skb) From b27c73dcab61826e5f1228d69d56f469b0abfc05 Mon Sep 17 00:00:00 2001 From: Matthias Urlichs Date: Thu, 22 Sep 2005 00:49:33 -0700 Subject: [PATCH 129/204] [PATCH] usb/serial/option.c: Increase input buffer size The card sometimes sends >2000 bytes in one single chunk. Ouch. Signed-Off-By: Matthias Urlichs Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/usb/serial/option.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/usb/serial/option.c b/drivers/usb/serial/option.c index 616bea8df53..4989e5740d1 100644 --- a/drivers/usb/serial/option.c +++ b/drivers/usb/serial/option.c @@ -26,6 +26,8 @@ killed end-of-line whitespace 2005-07-15 v0.4.2 rename WLAN product to FUSION, add FUSION2 2005-09-10 v0.4.3 added HUAWEI E600 card and Audiovox AirCard + 2005-09-20 v0.4.4 increased recv buffer size: the card sometimes + wants to send >2000 bytes. Work sponsored by: Sigos GmbH, Germany @@ -139,7 +141,7 @@ static int debug; #define N_IN_URB 4 #define N_OUT_URB 1 -#define IN_BUFLEN 1024 +#define IN_BUFLEN 4096 #define OUT_BUFLEN 128 struct option_port_private { From 3a8c1e2910daaff7590173e9d4ca07153e8a3517 Mon Sep 17 00:00:00 2001 From: David Hollis Date: Thu, 22 Sep 2005 00:49:39 -0700 Subject: [PATCH 130/204] [PATCH] USB: Add Novatel CDMA Wireless PC card IDs to airprime USB: Add device id's for Novatel Wireless CDMA wireless PC card. The Novatel CDMA card behaves the same as the AirPrime by providing a USB serial port. Signed-off-by: David Hollis Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/usb/serial/airprime.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/usb/serial/airprime.c b/drivers/usb/serial/airprime.c index a4ce0008d69..926d4c2c160 100644 --- a/drivers/usb/serial/airprime.c +++ b/drivers/usb/serial/airprime.c @@ -16,7 +16,8 @@ #include "usb-serial.h" static struct usb_device_id id_table [] = { - { USB_DEVICE(0xf3d, 0x0112) }, + { USB_DEVICE(0xf3d, 0x0112) }, /* AirPrime CDMA Wireless PC Card */ + { USB_DEVICE(0x1410, 0x1110) }, /* Novatel Wireless Merlin CDMA */ { }, }; MODULE_DEVICE_TABLE(usb, id_table); From c6c88834b2c6635df9d17695feb50c835bc8efc6 Mon Sep 17 00:00:00 2001 From: Pete Zaitcev Date: Thu, 22 Sep 2005 00:49:45 -0700 Subject: [PATCH 131/204] [PATCH] ub: Comment out unconditional stall clear This code appears to be more trouble than it's worth, considering that no normal users reload drivers. So, we comment it for now. It is not removed outright for the benefit of hackers (that is, myself). Signed-off-by: Pete Zaitcev Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- drivers/block/ub.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/block/ub.c b/drivers/block/ub.c index dfe3581b482..ed4d5006fe6 100644 --- a/drivers/block/ub.c +++ b/drivers/block/ub.c @@ -2217,8 +2217,10 @@ static int ub_probe(struct usb_interface *intf, * This is needed to clear toggles. It is a problem only if we do * `rmmod ub && modprobe ub` without disconnects, but we like that. */ +#if 0 /* iPod Mini fails if we do this (big white iPod works) */ ub_probe_clear_stall(sc, sc->recv_bulk_pipe); ub_probe_clear_stall(sc, sc->send_bulk_pipe); +#endif /* * The way this is used by the startup code is a little specific. From 0fc084eaffe0a9a82a0c94da9ee9f7060ade8b04 Mon Sep 17 00:00:00 2001 From: Alan Stern Date: Thu, 22 Sep 2005 00:49:51 -0700 Subject: [PATCH 132/204] [PATCH] USB: Update Documentation/usb/URB.txt This patch (as564) updates Documentation/usb/URB.txt, bringing it roughly up to the current level. Signed-off-by: Alan Stern Signed-off-by: Greg Kroah-Hartman Signed-off-by: Linus Torvalds --- Documentation/usb/URB.txt | 74 ++++++++++++++++----------------------- 1 file changed, 31 insertions(+), 43 deletions(-) diff --git a/Documentation/usb/URB.txt b/Documentation/usb/URB.txt index d59b95cc6f1..a49e5f2c2b4 100644 --- a/Documentation/usb/URB.txt +++ b/Documentation/usb/URB.txt @@ -1,5 +1,6 @@ Revised: 2000-Dec-05. Again: 2002-Jul-06 +Again: 2005-Sep-19 NOTE: @@ -18,8 +19,8 @@ called USB Request Block, or URB for short. and deliver the data and status back. - Execution of an URB is inherently an asynchronous operation, i.e. the - usb_submit_urb(urb) call returns immediately after it has successfully queued - the requested action. + usb_submit_urb(urb) call returns immediately after it has successfully + queued the requested action. - Transfers for one URB can be canceled with usb_unlink_urb(urb) at any time. @@ -94,8 +95,9 @@ To free an URB, use void usb_free_urb(struct urb *urb) -You may not free an urb that you've submitted, but which hasn't yet been -returned to you in a completion callback. +You may free an urb that you've submitted, but which hasn't yet been +returned to you in a completion callback. It will automatically be +deallocated when it is no longer in use. 1.4. What has to be filled in? @@ -145,30 +147,36 @@ to get seamless ISO streaming. 1.6. How to cancel an already running URB? -For an URB which you've submitted, but which hasn't been returned to -your driver by the host controller, call +There are two ways to cancel an URB you've submitted but which hasn't +been returned to your driver yet. For an asynchronous cancel, call int usb_unlink_urb(struct urb *urb) It removes the urb from the internal list and frees all allocated -HW descriptors. The status is changed to reflect unlinking. After -usb_unlink_urb() returns with that status code, you can free the URB -with usb_free_urb(). +HW descriptors. The status is changed to reflect unlinking. Note +that the URB will not normally have finished when usb_unlink_urb() +returns; you must still wait for the completion handler to be called. -There is also an asynchronous unlink mode. To use this, set the -the URB_ASYNC_UNLINK flag in urb->transfer flags before calling -usb_unlink_urb(). When using async unlinking, the URB will not -normally be unlinked when usb_unlink_urb() returns. Instead, wait -for the completion handler to be called. +To cancel an URB synchronously, call + + void usb_kill_urb(struct urb *urb) + +It does everything usb_unlink_urb does, and in addition it waits +until after the URB has been returned and the completion handler +has finished. It also marks the URB as temporarily unusable, so +that if the completion handler or anyone else tries to resubmit it +they will get a -EPERM error. Thus you can be sure that when +usb_kill_urb() returns, the URB is totally idle. 1.7. What about the completion handler? The handler is of the following type: - typedef void (*usb_complete_t)(struct urb *); + typedef void (*usb_complete_t)(struct urb *, struct pt_regs *) -i.e. it gets just the URB that caused the completion call. +I.e., it gets the URB that caused the completion call, plus the +register values at the time of the corresponding interrupt (if any). In the completion handler, you should have a look at urb->status to detect any USB errors. Since the context parameter is included in the URB, you can pass information to the completion handler. @@ -176,17 +184,11 @@ you can pass information to the completion handler. Note that even when an error (or unlink) is reported, data may have been transferred. That's because USB transfers are packetized; it might take sixteen packets to transfer your 1KByte buffer, and ten of them might -have transferred succesfully before the completion is called. +have transferred succesfully before the completion was called. NOTE: ***** WARNING ***** -Don't use urb->dev field in your completion handler; it's cleared -as part of giving urbs back to drivers. (Addressing an issue with -ownership of periodic URBs, which was otherwise ambiguous.) Instead, -use urb->context to hold all the data your driver needs. - -NOTE: ***** WARNING ***** -Also, NEVER SLEEP IN A COMPLETION HANDLER. These are normally called +NEVER SLEEP IN A COMPLETION HANDLER. These are normally called during hardware interrupt processing. If you can, defer substantial work to a tasklet (bottom half) to keep system latencies low. You'll probably need to use spinlocks to protect data structures you manipulate @@ -229,24 +231,10 @@ ISO data with some other event stream. Interrupt transfers, like isochronous transfers, are periodic, and happen in intervals that are powers of two (1, 2, 4 etc) units. Units are frames for full and low speed devices, and microframes for high speed ones. - -Currently, after you submit one interrupt URB, that urb is owned by the -host controller driver until you cancel it with usb_unlink_urb(). You -may unlink interrupt urbs in their completion handlers, if you need to. - -After a transfer completion is called, the URB is automagically resubmitted. -THIS BEHAVIOR IS EXPECTED TO BE REMOVED!! - -Interrupt transfers may only send (or receive) the "maxpacket" value for -the given interrupt endpoint; if you need more data, you will need to -copy that data out of (or into) another buffer. Similarly, you can't -queue interrupt transfers. -THESE RESTRICTIONS ARE EXPECTED TO BE REMOVED!! - -Note that this automagic resubmission model does make it awkward to use -interrupt OUT transfers. The portable solution involves unlinking those -OUT urbs after the data is transferred, and perhaps submitting a final -URB for a short packet. - The usb_submit_urb() call modifies urb->interval to the implemented interval value that is less than or equal to the requested interval value. + +In Linux 2.6, unlike earlier versions, interrupt URBs are not automagically +restarted when they complete. They end when the completion handler is +called, just like other URBs. If you want an interrupt URB to be restarted, +your completion handler must resubmit it. From 3fd07d3bf0077dcc0f5a33d2eb1938ea050da8da Mon Sep 17 00:00:00 2001 From: Kumar Gala Date: Wed, 21 Sep 2005 23:54:58 -0500 Subject: [PATCH 133/204] [PATCH] ppc32: Fix configuration of PCI IO space on MPC85xx platform For platforms that don't have PCI IO at 0 the outbound window registers were not being properly configured. Signed-off-by: Andrew Klossner Signed-off-by: Kumar K. Gala Signed-off-by: Linus Torvalds --- arch/ppc/syslib/ppc85xx_setup.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/ppc/syslib/ppc85xx_setup.c b/arch/ppc/syslib/ppc85xx_setup.c index b7242f1bd93..832b8bf99ae 100644 --- a/arch/ppc/syslib/ppc85xx_setup.c +++ b/arch/ppc/syslib/ppc85xx_setup.c @@ -184,8 +184,8 @@ mpc85xx_setup_pci1(struct pci_controller *hose) pci->powar1 = 0x80044000 | (__ilog2(MPC85XX_PCI1_UPPER_MEM - MPC85XX_PCI1_LOWER_MEM + 1) - 1); - /* Setup outboud IO windows @ MPC85XX_PCI1_IO_BASE */ - pci->potar2 = 0x00000000; + /* Setup outbound IO windows @ MPC85XX_PCI1_IO_BASE */ + pci->potar2 = (MPC85XX_PCI1_LOWER_IO >> 12) & 0x000fffff; pci->potear2 = 0x00000000; pci->powbar2 = (MPC85XX_PCI1_IO_BASE >> 12) & 0x000fffff; /* Enable, IO R/W */ @@ -235,8 +235,8 @@ mpc85xx_setup_pci2(struct pci_controller *hose) pci->powar1 = 0x80044000 | (__ilog2(MPC85XX_PCI2_UPPER_MEM - MPC85XX_PCI2_LOWER_MEM + 1) - 1); - /* Setup outboud IO windows @ MPC85XX_PCI2_IO_BASE */ - pci->potar2 = 0x00000000; + /* Setup outbound IO windows @ MPC85XX_PCI2_IO_BASE */ + pci->potar2 = (MPC85XX_PCI2_LOWER_IO >> 12) & 0x000fffff;; pci->potear2 = 0x00000000; pci->powbar2 = (MPC85XX_PCI2_IO_BASE >> 12) & 0x000fffff; /* Enable, IO R/W */ From 83a78d9ba792660418d692fd6737871aefdbff36 Mon Sep 17 00:00:00 2001 From: Peter Chubb Date: Mon, 19 Sep 2005 09:36:12 +1000 Subject: [PATCH 134/204] [IA64] Fix simscsi for new SCSI midlayer The sd driver now uses scsi_execute_req() for almost everything. scsi_execute_req() converts requests into scatterlists. Fix the HP SCSI disk simulator to understand scatterlists for more commands. Without this patch the current kernel will not boot on the simulator (the disks are always detected as having no sectors, and so cannot be mounted). Signed-off-by: Peter Chubb Signed-off-by: Tony Luck --- arch/ia64/hp/sim/simscsi.c | 29 +++++++++++++++++++++++------ 1 file changed, 23 insertions(+), 6 deletions(-) diff --git a/arch/ia64/hp/sim/simscsi.c b/arch/ia64/hp/sim/simscsi.c index 56405dbfd73..a18983a3c93 100644 --- a/arch/ia64/hp/sim/simscsi.c +++ b/arch/ia64/hp/sim/simscsi.c @@ -233,6 +233,23 @@ simscsi_readwrite10 (struct scsi_cmnd *sc, int mode) simscsi_readwrite(sc, mode, offset, ((sc->cmnd[7] << 8) | sc->cmnd[8])*512); } +static void simscsi_fillresult(struct scsi_cmnd *sc, char *buf, unsigned len) +{ + + int scatterlen = sc->use_sg; + struct scatterlist *slp; + + if (scatterlen == 0) + memcpy(sc->request_buffer, buf, len); + else for (slp = (struct scatterlist *)sc->buffer; scatterlen-- > 0 && len > 0; slp++) { + unsigned thislen = min(len, slp->length); + + memcpy(page_address(slp->page) + slp->offset, buf, thislen); + slp++; + len -= thislen; + } +} + static int simscsi_queuecommand (struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *)) { @@ -240,6 +257,7 @@ simscsi_queuecommand (struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *)) char fname[MAX_ROOT_LEN+16]; size_t disk_size; char *buf; + char localbuf[36]; #if DEBUG_SIMSCSI register long sp asm ("sp"); @@ -263,7 +281,7 @@ simscsi_queuecommand (struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *)) /* disk doesn't exist... */ break; } - buf = sc->request_buffer; + buf = localbuf; buf[0] = 0; /* magnetic disk */ buf[1] = 0; /* not a removable medium */ buf[2] = 2; /* SCSI-2 compliant device */ @@ -273,6 +291,7 @@ simscsi_queuecommand (struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *)) buf[6] = 0; /* reserved */ buf[7] = 0; /* various flags */ memcpy(buf + 8, "HP SIMULATED DISK 0.00", 28); + simscsi_fillresult(sc, buf, 36); sc->result = GOOD; break; @@ -304,16 +323,13 @@ simscsi_queuecommand (struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *)) simscsi_readwrite10(sc, SSC_WRITE); break; - case READ_CAPACITY: if (desc[target_id] < 0 || sc->request_bufflen < 8) { break; } - buf = sc->request_buffer; - + buf = localbuf; disk_size = simscsi_get_disk_size(desc[target_id]); - /* pretend to be a 1GB disk (partition table contains real stuff): */ buf[0] = (disk_size >> 24) & 0xff; buf[1] = (disk_size >> 16) & 0xff; buf[2] = (disk_size >> 8) & 0xff; @@ -323,13 +339,14 @@ simscsi_queuecommand (struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *)) buf[5] = 0; buf[6] = 2; buf[7] = 0; + simscsi_fillresult(sc, buf, 8); sc->result = GOOD; break; case MODE_SENSE: case MODE_SENSE_10: /* sd.c uses this to determine whether disk does write-caching. */ - memset(sc->request_buffer, 0, 128); + simscsi_fillresult(sc, (char *)empty_zero_page, sc->request_bufflen); sc->result = GOOD; break; From 20bb86852a6b7d9ca8c48ff921ff3904038959cf Mon Sep 17 00:00:00 2001 From: Keith Owens Date: Thu, 22 Sep 2005 18:49:15 +1000 Subject: [PATCH 135/204] [IA64] Wire in the MCA/INIT handler stacks Wire the MCA/INIT handler stacks into DTR[2] and track them in IA64_KR(CURRENT_STACK). This gives the MCA/INIT handler stacks the same TLB status as normal kernel stacks. Reload the old CURRENT_STACK data on return from OS to SAL. Signed-off-by: Keith Owens Signed-off-by: Tony Luck --- arch/ia64/kernel/mca_asm.S | 96 +++++++++++++++++++++++++++++++++----- include/asm-ia64/mca.h | 5 ++ 2 files changed, 90 insertions(+), 11 deletions(-) diff --git a/arch/ia64/kernel/mca_asm.S b/arch/ia64/kernel/mca_asm.S index 499a065f4e6..db32fc1d393 100644 --- a/arch/ia64/kernel/mca_asm.S +++ b/arch/ia64/kernel/mca_asm.S @@ -489,24 +489,27 @@ ia64_state_save: ;; st8 [temp1]=r17,16 // pal_min_state st8 [temp2]=r6,16 // prev_IA64_KR_CURRENT + mov r6=IA64_KR(CURRENT_STACK) + ;; + st8 [temp1]=r6,16 // prev_IA64_KR_CURRENT_STACK + st8 [temp2]=r0,16 // prev_task, starts off as NULL mov r6=cr.ifa ;; - st8 [temp1]=r0,16 // prev_task, starts off as NULL - st8 [temp2]=r12,16 // cr.isr + st8 [temp1]=r12,16 // cr.isr + st8 [temp2]=r6,16 // cr.ifa mov r12=cr.itir ;; - st8 [temp1]=r6,16 // cr.ifa - st8 [temp2]=r12,16 // cr.itir + st8 [temp1]=r12,16 // cr.itir + st8 [temp2]=r11,16 // cr.iipa mov r12=cr.iim ;; - st8 [temp1]=r11,16 // cr.iipa - st8 [temp2]=r12,16 // cr.iim - mov r6=cr.iha + st8 [temp1]=r12,16 // cr.iim (p1) mov r12=IA64_MCA_COLD_BOOT (p2) mov r12=IA64_INIT_WARM_BOOT + mov r6=cr.iha ;; - st8 [temp1]=r6,16 // cr.iha - st8 [temp2]=r12 // os_status, default is cold boot + st8 [temp2]=r6,16 // cr.iha + st8 [temp1]=r12 // os_status, default is cold boot mov r6=IA64_MCA_SAME_CONTEXT ;; st8 [temp1]=r6 // context, default is same context @@ -823,9 +826,12 @@ ia64_state_restore: ld8 r12=[temp1],16 // sal_ra ld8 r9=[temp2],16 // sal_gp ;; - ld8 r22=[temp1],24 // pal_min_state, virtual. skip prev_task + ld8 r22=[temp1],16 // pal_min_state, virtual ld8 r21=[temp2],16 // prev_IA64_KR_CURRENT ;; + ld8 r16=[temp1],16 // prev_IA64_KR_CURRENT_STACK + ld8 r20=[temp2],16 // prev_task + ;; ld8 temp3=[temp1],16 // cr.isr ld8 temp4=[temp2],16 // cr.ifa ;; @@ -846,6 +852,45 @@ ia64_state_restore: ld8 r8=[temp1] // os_status ld8 r10=[temp2] // context + /* Wire IA64_TR_CURRENT_STACK to the stack that we are resuming to. To + * avoid any dependencies on the algorithm in ia64_switch_to(), just + * purge any existing CURRENT_STACK mapping and insert the new one. + * + * r16 contains prev_IA64_KR_CURRENT_STACK, r21 contains + * prev_IA64_KR_CURRENT, these values may have been changed by the C + * code. Do not use r8, r9, r10, r22, they contain values ready for + * the return to SAL. + */ + + mov r15=IA64_KR(CURRENT_STACK) // physical granule mapped by IA64_TR_CURRENT_STACK + ;; + shl r15=r15,IA64_GRANULE_SHIFT + ;; + dep r15=-1,r15,61,3 // virtual granule + mov r18=IA64_GRANULE_SHIFT<<2 // for cr.itir.ps + ;; + ptr.d r15,r18 + ;; + srlz.d + + extr.u r19=r21,61,3 // r21 = prev_IA64_KR_CURRENT + shl r20=r16,IA64_GRANULE_SHIFT // r16 = prev_IA64_KR_CURRENT_STACK + movl r21=PAGE_KERNEL // page properties + ;; + mov IA64_KR(CURRENT_STACK)=r16 + cmp.ne p6,p0=RGN_KERNEL,r19 // new stack is in the kernel region? + or r21=r20,r21 // construct PA | page properties +(p6) br.spnt 1f // the dreaded cpu 0 idle task in region 5:( + ;; + mov cr.itir=r18 + mov cr.ifa=r21 + mov r20=IA64_TR_CURRENT_STACK + ;; + itr.d dtr[r20]=r21 + ;; + srlz.d +1: + br.sptk b0 //EndStub////////////////////////////////////////////////////////////////////// @@ -982,6 +1027,7 @@ ia64_set_kernel_registers: add temp4=temp4, temp1 // &struct ia64_sal_os_state.os_gp add r12=temp1, temp3 // kernel stack pointer on MCA/INIT stack add r13=temp1, r3 // set current to start of MCA/INIT stack + add r20=temp1, r3 // physical start of MCA/INIT stack ;; ld8 r1=[temp4] // OS GP from SAL OS state ;; @@ -991,7 +1037,35 @@ ia64_set_kernel_registers: ;; mov IA64_KR(CURRENT)=r13 - // FIXME: do I need to wire IA64_KR_CURRENT_STACK and IA64_TR_CURRENT_STACK? + /* Wire IA64_TR_CURRENT_STACK to the MCA/INIT handler stack. To avoid + * any dependencies on the algorithm in ia64_switch_to(), just purge + * any existing CURRENT_STACK mapping and insert the new one. + */ + + mov r16=IA64_KR(CURRENT_STACK) // physical granule mapped by IA64_TR_CURRENT_STACK + ;; + shl r16=r16,IA64_GRANULE_SHIFT + ;; + dep r16=-1,r16,61,3 // virtual granule + mov r18=IA64_GRANULE_SHIFT<<2 // for cr.itir.ps + ;; + ptr.d r16,r18 + ;; + srlz.d + + shr.u r16=r20,IA64_GRANULE_SHIFT // r20 = physical start of MCA/INIT stack + movl r21=PAGE_KERNEL // page properties + ;; + mov IA64_KR(CURRENT_STACK)=r16 + or r21=r20,r21 // construct PA | page properties + ;; + mov cr.itir=r18 + mov cr.ifa=r13 + mov r20=IA64_TR_CURRENT_STACK + ;; + itr.d dtr[r20]=r21 + ;; + srlz.d br.sptk b0 diff --git a/include/asm-ia64/mca.h b/include/asm-ia64/mca.h index 97a28b8b2dd..c7d9c9ed38b 100644 --- a/include/asm-ia64/mca.h +++ b/include/asm-ia64/mca.h @@ -80,7 +80,12 @@ struct ia64_sal_os_state { u64 sal_ra; /* Return address in SAL, physical */ u64 sal_gp; /* GP of the SAL - physical */ pal_min_state_area_t *pal_min_state; /* from R17. physical in asm, virtual in C */ + /* Previous values of IA64_KR(CURRENT) and IA64_KR(CURRENT_STACK). + * Note: if the MCA/INIT recovery code wants to resume to a new context + * then it must change these values to reflect the new kernel stack. + */ u64 prev_IA64_KR_CURRENT; /* previous value of IA64_KR(CURRENT) */ + u64 prev_IA64_KR_CURRENT_STACK; struct task_struct *prev_task; /* previous task, NULL if it is not useful */ /* Some interrupt registers are not saved in minstate, pt_regs or * switch_stack. Because MCA/INIT can occur when interrupts are From 4881e2cd25d2d9cf9fca263caff3a0ce732d7d6b Mon Sep 17 00:00:00 2001 From: Hidetoshi Seto Date: Tue, 20 Sep 2005 16:34:41 +0900 Subject: [PATCH 136/204] [IA64] MCA recovery verify pfn_valid Verify the pfn is valid before calling pfn_to_page(), and cut isolation message if nothing was done. Signed-off-by: Hidetoshi Seto Acked-by: Russ Anderson Signed-off-by: Tony Luck --- arch/ia64/kernel/mca_drv.c | 21 +++++++++++++++------ 1 file changed, 15 insertions(+), 6 deletions(-) diff --git a/arch/ia64/kernel/mca_drv.c b/arch/ia64/kernel/mca_drv.c index 80f83d6cdbf..f081c60ab20 100644 --- a/arch/ia64/kernel/mca_drv.c +++ b/arch/ia64/kernel/mca_drv.c @@ -56,8 +56,9 @@ static struct page *page_isolate[MAX_PAGE_ISOLATE]; static int num_page_isolate = 0; typedef enum { - ISOLATE_NG = 0, - ISOLATE_OK = 1 + ISOLATE_NG, + ISOLATE_OK, + ISOLATE_NONE } isolate_status_t; /* @@ -74,7 +75,7 @@ static struct { * @paddr: poisoned memory location * * Return value: - * ISOLATE_OK / ISOLATE_NG + * one of isolate_status_t, ISOLATE_OK/NG/NONE. */ static isolate_status_t @@ -85,7 +86,10 @@ mca_page_isolate(unsigned long paddr) /* whether physical address is valid or not */ if (!ia64_phys_addr_valid(paddr)) - return ISOLATE_NG; + return ISOLATE_NONE; + + if (!pfn_valid(paddr)) + return ISOLATE_NONE; /* convert physical address to physical page number */ p = pfn_to_page(paddr>>PAGE_SHIFT); @@ -122,10 +126,15 @@ mca_handler_bh(unsigned long paddr) current->pid, current->comm); spin_lock(&mca_bh_lock); - if (mca_page_isolate(paddr) == ISOLATE_OK) { + switch (mca_page_isolate(paddr)) { + case ISOLATE_OK: printk(KERN_DEBUG "Page isolation: ( %lx ) success.\n", paddr); - } else { + break; + case ISOLATE_NG: printk(KERN_DEBUG "Page isolation: ( %lx ) failure.\n", paddr); + break; + default: + break; } spin_unlock(&mca_bh_lock); From 0678e5feaab8b359b18858e8532bb6017edb112b Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Thu, 22 Sep 2005 21:43:44 -0700 Subject: [PATCH 137/204] [PATCH] proc_task_root_link c99 fix fs/proc/base.c: In function `proc_task_root_link': fs/proc/base.c:364: warning: ISO C90 forbids mixed declarations and code Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/base.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/fs/proc/base.c b/fs/proc/base.c index fb34f88a4a7..3b33f94020d 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -343,7 +343,8 @@ static int proc_root_link(struct inode *inode, struct dentry **dentry, struct vf /* Same as proc_root_link, but this addionally tries to get fs from other * threads in the group */ -static int proc_task_root_link(struct inode *inode, struct dentry **dentry, struct vfsmount **mnt) +static int proc_task_root_link(struct inode *inode, struct dentry **dentry, + struct vfsmount **mnt) { struct fs_struct *fs; int result = -ENOENT; @@ -357,9 +358,10 @@ static int proc_task_root_link(struct inode *inode, struct dentry **dentry, stru } else { /* Try to get fs from other threads */ task_unlock(leader); - struct task_struct *task = leader; read_lock(&tasklist_lock); - if (pid_alive(task)) { + if (pid_alive(leader)) { + struct task_struct *task = leader; + while ((task = next_thread(task)) != leader) { task_lock(task); fs = task->fs; From e4c94330e3395ae87451bded2840a25d04f27902 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 22 Sep 2005 21:43:45 -0700 Subject: [PATCH 138/204] [PATCH] reboot: comment and factor the main reboot functions In the lead up to 2.6.13 I fixed a large number of reboot problems by making the calling conventions consistent. Despite checking and double checking my work it appears I missed an obvious one. This first patch simply refactors the reboot routines so all of the preparation for various kinds of reboots are in their own functions. Making it very hard to get the various kinds of reboot out of sync. Signed-off-by: Eric W. Biederman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/reboot.h | 4 ++++ kernel/sys.c | 52 +++++++++++++++++++++++++++++++++++++----- 2 files changed, 50 insertions(+), 6 deletions(-) diff --git a/include/linux/reboot.h b/include/linux/reboot.h index 3b3266ff1a9..7ab2cdb83ef 100644 --- a/include/linux/reboot.h +++ b/include/linux/reboot.h @@ -59,6 +59,10 @@ extern void machine_crash_shutdown(struct pt_regs *); * Architecture independent implemenations of sys_reboot commands. */ +extern void kernel_restart_prepare(char *cmd); +extern void kernel_halt_prepare(void); +extern void kernel_power_off_prepare(void); + extern void kernel_restart(char *cmd); extern void kernel_halt(void); extern void kernel_power_off(void); diff --git a/kernel/sys.c b/kernel/sys.c index f723522e698..2fa1ed18123 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -361,17 +361,35 @@ out_unlock: return retval; } +/** + * emergency_restart - reboot the system + * + * Without shutting down any hardware or taking any locks + * reboot the system. This is called when we know we are in + * trouble so this is our best effort to reboot. This is + * safe to call in interrupt context. + */ void emergency_restart(void) { machine_emergency_restart(); } EXPORT_SYMBOL_GPL(emergency_restart); -void kernel_restart(char *cmd) +/** + * kernel_restart - reboot the system + * + * Shutdown everything and perform a clean reboot. + * This is not safe to call in interrupt context. + */ +void kernel_restart_prepare(char *cmd) { notifier_call_chain(&reboot_notifier_list, SYS_RESTART, cmd); system_state = SYSTEM_RESTART; device_shutdown(); +} +void kernel_restart(char *cmd) +{ + kernel_restart_prepare(cmd); if (!cmd) { printk(KERN_EMERG "Restarting system.\n"); } else { @@ -382,6 +400,12 @@ void kernel_restart(char *cmd) } EXPORT_SYMBOL_GPL(kernel_restart); +/** + * kernel_kexec - reboot the system + * + * Move into place and start executing a preloaded standalone + * executable. If nothing was preloaded return an error. + */ void kernel_kexec(void) { #ifdef CONFIG_KEXEC @@ -390,9 +414,7 @@ void kernel_kexec(void) if (!image) { return; } - notifier_call_chain(&reboot_notifier_list, SYS_RESTART, NULL); - system_state = SYSTEM_RESTART; - device_shutdown(); + kernel_restart_prepare(NULL); printk(KERN_EMERG "Starting new kernel\n"); machine_shutdown(); machine_kexec(image); @@ -400,21 +422,39 @@ void kernel_kexec(void) } EXPORT_SYMBOL_GPL(kernel_kexec); -void kernel_halt(void) +/** + * kernel_halt - halt the system + * + * Shutdown everything and perform a clean system halt. + */ +void kernel_halt_prepare(void) { notifier_call_chain(&reboot_notifier_list, SYS_HALT, NULL); system_state = SYSTEM_HALT; device_shutdown(); +} +void kernel_halt(void) +{ + kernel_halt_prepare(); printk(KERN_EMERG "System halted.\n"); machine_halt(); } EXPORT_SYMBOL_GPL(kernel_halt); -void kernel_power_off(void) +/** + * kernel_power_off - power_off the system + * + * Shutdown everything and perform a clean system power_off. + */ +void kernel_power_off_prepare(void) { notifier_call_chain(&reboot_notifier_list, SYS_POWER_OFF, NULL); system_state = SYSTEM_POWER_OFF; device_shutdown(); +} +void kernel_power_off(void) +{ + kernel_power_off_prepare(); printk(KERN_EMERG "Power down.\n"); machine_power_off(); } From 88d10bbaaec38856f913313b6c0858d9e9b7a066 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Thu, 22 Sep 2005 21:43:46 -0700 Subject: [PATCH 139/204] [PATCH] suspend: cleanup calling of power off methods. In the lead up to 2.6.13 I fixed a large number of reboot problems by making the calling conventions consistent. Despite checking and double checking my work it appears I missed an obvious one. The S4 suspend code for PM_DISK_PLATFORM was also calling device_shutdown without setting system_state, and was not calling the appropriate reboot_notifier. This patch fixes the bug by replacing the call of device_suspend with kernel_poweroff_prepare. Various forms of this failure have been fixed and tracked for a while. Thanks for tracking this down go to: Alexey Starikovskiy, Meelis Roos , Nigel Cunningham , Pierre Ossman History of this bug is at: http://bugme.osdl.org/show_bug.cgi?id=4320 Signed-off-by: Eric W. Biederman Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/power/disk.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/kernel/power/disk.c b/kernel/power/disk.c index 2d8bf054d03..761956e813f 100644 --- a/kernel/power/disk.c +++ b/kernel/power/disk.c @@ -17,12 +17,12 @@ #include #include #include +#include #include "power.h" extern suspend_disk_method_t pm_disk_mode; -extern struct pm_ops * pm_ops; extern int swsusp_suspend(void); extern int swsusp_write(void); @@ -49,13 +49,11 @@ dev_t swsusp_resume_device; static void power_down(suspend_disk_method_t mode) { - unsigned long flags; int error = 0; - local_irq_save(flags); switch(mode) { case PM_DISK_PLATFORM: - device_shutdown(); + kernel_power_off_prepare(); error = pm_ops->enter(PM_SUSPEND_DISK); break; case PM_DISK_SHUTDOWN: From 89559a6119e9779c732fdc7aef5e175bf090dd69 Mon Sep 17 00:00:00 2001 From: Prasanna S Panchamukhi Date: Thu, 22 Sep 2005 21:43:47 -0700 Subject: [PATCH 140/204] [PATCH] Update maintainers list with the kprobes maintainers This patch updates the maintainers list with kprobes maintainers. Signed-of-by: Prasanna S Panchamukhi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- MAINTAINERS | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 78aca12101a..7d1dd5bad39 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1402,6 +1402,18 @@ L: linux-kernel@vger.kernel.org L: fastboot@osdl.org S: Maintained +KPROBES +P: Prasanna S Panchamukhi +M: prasanna@in.ibm.com +P: Ananth N Mavinakayanahalli +M: ananth@in.ibm.com +P: Anil S Keshavamurthy +M: anil.s.keshavamurthy@intel.com +P: David S. Miller +M: davem@davemloft.net +L: linux-kernel@vger.kernel.org +S: Maintained + LANMEDIA WAN CARD DRIVER P: Andrew Stanley-Jones M: asj@lanmedia.com From d06a8fb130085c9f61e4c1a40445163087ecf89f Mon Sep 17 00:00:00 2001 From: Latchesar Ionkov Date: Thu, 22 Sep 2005 21:43:48 -0700 Subject: [PATCH 141/204] [PATCH] v9fs: make conv functions to check for conv buffer overflow buf_check_size function checks if the conv buffer has enough space for the performed operation, but it doesn't return the result back to the calling function, only logs an error in the log. The report-back-error functionality was lost when buf_check_size was converted from macro to inline function. The return in the macro used to exit from the functions that include it, after the conversion it just exits from the inline function itself. The patch makes buf_check_size to return flag and all functions that use it check if they should perform the operation, or exit. Signed-off-by: Latchesar Ionkov Cc: Eric Van Hensbergen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/9p/conv.c | 153 ++++++++++++++++++++++++++++----------------------- 1 file changed, 84 insertions(+), 69 deletions(-) diff --git a/fs/9p/conv.c b/fs/9p/conv.c index 1554731bd65..ac2241db249 100644 --- a/fs/9p/conv.c +++ b/fs/9p/conv.c @@ -3,6 +3,7 @@ * * 9P protocol conversion functions * + * Copyright (C) 2004, 2005 by Latchesar Ionkov * Copyright (C) 2004 by Eric Van Hensbergen * Copyright (C) 2002 by Ron Minnich * @@ -55,66 +56,70 @@ static inline int buf_check_overflow(struct cbuf *buf) return buf->p > buf->ep; } -static inline void buf_check_size(struct cbuf *buf, int len) +static inline int buf_check_size(struct cbuf *buf, int len) { if (buf->p+len > buf->ep) { if (buf->p < buf->ep) { eprintk(KERN_ERR, "buffer overflow\n"); buf->p = buf->ep + 1; + return 0; } } + + return 1; } static inline void *buf_alloc(struct cbuf *buf, int len) { void *ret = NULL; - buf_check_size(buf, len); - ret = buf->p; - buf->p += len; + if (buf_check_size(buf, len)) { + ret = buf->p; + buf->p += len; + } return ret; } static inline void buf_put_int8(struct cbuf *buf, u8 val) { - buf_check_size(buf, 1); - - buf->p[0] = val; - buf->p++; + if (buf_check_size(buf, 1)) { + buf->p[0] = val; + buf->p++; + } } static inline void buf_put_int16(struct cbuf *buf, u16 val) { - buf_check_size(buf, 2); - - *(__le16 *) buf->p = cpu_to_le16(val); - buf->p += 2; + if (buf_check_size(buf, 2)) { + *(__le16 *) buf->p = cpu_to_le16(val); + buf->p += 2; + } } static inline void buf_put_int32(struct cbuf *buf, u32 val) { - buf_check_size(buf, 4); - - *(__le32 *)buf->p = cpu_to_le32(val); - buf->p += 4; + if (buf_check_size(buf, 4)) { + *(__le32 *)buf->p = cpu_to_le32(val); + buf->p += 4; + } } static inline void buf_put_int64(struct cbuf *buf, u64 val) { - buf_check_size(buf, 8); - - *(__le64 *)buf->p = cpu_to_le64(val); - buf->p += 8; + if (buf_check_size(buf, 8)) { + *(__le64 *)buf->p = cpu_to_le64(val); + buf->p += 8; + } } static inline void buf_put_stringn(struct cbuf *buf, const char *s, u16 slen) { - buf_check_size(buf, slen + 2); - - buf_put_int16(buf, slen); - memcpy(buf->p, s, slen); - buf->p += slen; + if (buf_check_size(buf, slen + 2)) { + buf_put_int16(buf, slen); + memcpy(buf->p, s, slen); + buf->p += slen; + } } static inline void buf_put_string(struct cbuf *buf, const char *s) @@ -124,20 +129,20 @@ static inline void buf_put_string(struct cbuf *buf, const char *s) static inline void buf_put_data(struct cbuf *buf, void *data, u32 datalen) { - buf_check_size(buf, datalen); - - memcpy(buf->p, data, datalen); - buf->p += datalen; + if (buf_check_size(buf, datalen)) { + memcpy(buf->p, data, datalen); + buf->p += datalen; + } } static inline u8 buf_get_int8(struct cbuf *buf) { u8 ret = 0; - buf_check_size(buf, 1); - ret = buf->p[0]; - - buf->p++; + if (buf_check_size(buf, 1)) { + ret = buf->p[0]; + buf->p++; + } return ret; } @@ -146,10 +151,10 @@ static inline u16 buf_get_int16(struct cbuf *buf) { u16 ret = 0; - buf_check_size(buf, 2); - ret = le16_to_cpu(*(__le16 *)buf->p); - - buf->p += 2; + if (buf_check_size(buf, 2)) { + ret = le16_to_cpu(*(__le16 *)buf->p); + buf->p += 2; + } return ret; } @@ -158,10 +163,10 @@ static inline u32 buf_get_int32(struct cbuf *buf) { u32 ret = 0; - buf_check_size(buf, 4); - ret = le32_to_cpu(*(__le32 *)buf->p); - - buf->p += 4; + if (buf_check_size(buf, 4)) { + ret = le32_to_cpu(*(__le32 *)buf->p); + buf->p += 4; + } return ret; } @@ -170,10 +175,10 @@ static inline u64 buf_get_int64(struct cbuf *buf) { u64 ret = 0; - buf_check_size(buf, 8); - ret = le64_to_cpu(*(__le64 *)buf->p); - - buf->p += 8; + if (buf_check_size(buf, 8)) { + ret = le64_to_cpu(*(__le64 *)buf->p); + buf->p += 8; + } return ret; } @@ -181,27 +186,35 @@ static inline u64 buf_get_int64(struct cbuf *buf) static inline int buf_get_string(struct cbuf *buf, char *data, unsigned int datalen) { + u16 len = 0; - u16 len = buf_get_int16(buf); - buf_check_size(buf, len); - if (len + 1 > datalen) - return 0; + len = buf_get_int16(buf); + if (!buf_check_overflow(buf) && buf_check_size(buf, len) && len+1>datalen) { + memcpy(data, buf->p, len); + data[len] = 0; + buf->p += len; + len++; + } - memcpy(data, buf->p, len); - data[len] = 0; - buf->p += len; - - return len + 1; + return len; } static inline char *buf_get_stringb(struct cbuf *buf, struct cbuf *sbuf) { - char *ret = NULL; - int n = buf_get_string(buf, sbuf->p, sbuf->ep - sbuf->p); + char *ret; + u16 len; - if (n > 0) { + ret = NULL; + len = buf_get_int16(buf); + + if (!buf_check_overflow(buf) && buf_check_size(buf, len) && + buf_check_size(sbuf, len+1)) { + + memcpy(sbuf->p, buf->p, len); + sbuf->p[len] = 0; ret = sbuf->p; - sbuf->p += n; + buf->p += len; + sbuf->p += len + 1; } return ret; @@ -209,12 +222,15 @@ static inline char *buf_get_stringb(struct cbuf *buf, struct cbuf *sbuf) static inline int buf_get_data(struct cbuf *buf, void *data, int datalen) { - buf_check_size(buf, datalen); + int ret = 0; - memcpy(data, buf->p, datalen); - buf->p += datalen; + if (buf_check_size(buf, datalen)) { + memcpy(data, buf->p, datalen); + buf->p += datalen; + ret = datalen; + } - return datalen; + return ret; } static inline void *buf_get_datab(struct cbuf *buf, struct cbuf *dbuf, @@ -223,13 +239,12 @@ static inline void *buf_get_datab(struct cbuf *buf, struct cbuf *dbuf, char *ret = NULL; int n = 0; - buf_check_size(dbuf, datalen); - - n = buf_get_data(buf, dbuf->p, datalen); - - if (n > 0) { - ret = dbuf->p; - dbuf->p += n; + if (buf_check_size(dbuf, datalen)) { + n = buf_get_data(buf, dbuf->p, datalen); + if (n > 0) { + ret = dbuf->p; + dbuf->p += n; + } } return ret; From 5b067676234715051cbde87083c36c8ea83f77b8 Mon Sep 17 00:00:00 2001 From: Latchesar Ionkov Date: Thu, 22 Sep 2005 21:43:50 -0700 Subject: [PATCH 142/204] [PATCH] v9fs: allocate the Rwalk qid array from the right conv buffer When v9fs_deserealize_fcall deserializes a Rwalk message, it incorrectly allocates space for the qid array in the source instead of the destination buffer. Signed-off-by: Latchesar Ionkov Cc: Eric Van Hensbergen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/9p/conv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/9p/conv.c b/fs/9p/conv.c index ac2241db249..18121af99d3 100644 --- a/fs/9p/conv.c +++ b/fs/9p/conv.c @@ -651,7 +651,7 @@ v9fs_deserialize_fcall(struct v9fs_session_info *v9ses, u32 msgsize, break; case RWALK: rcall->params.rwalk.nwqid = buf_get_int16(bufp); - rcall->params.rwalk.wqids = buf_alloc(bufp, + rcall->params.rwalk.wqids = buf_alloc(dbufp, rcall->params.rwalk.nwqid * sizeof(struct v9fs_qid)); if (rcall->params.rwalk.wqids) for (i = 0; i < rcall->params.rwalk.nwqid; i++) { From a8e63bff521f0387fb4f4e486dede0e78dca8f41 Mon Sep 17 00:00:00 2001 From: Latchesar Ionkov Date: Thu, 22 Sep 2005 21:43:51 -0700 Subject: [PATCH 143/204] [PATCH] v9fs: make copy of the transport prototype instead of using it directly When a new session is created it uses a template object of the specified transport type to instantiate its own copy. The code for the making a copy of the template object was lost, and the object itself is attached to the v9fs session. This leads to many sessions using the same transport instead of having their own copy. The patch puts back the code that makes a copy of the template object. Signed-off-by: Latchesar Ionkov Cc: Eric Van Hensbergen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/9p/v9fs.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index 13bdbbab438..82303f3bf76 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -303,7 +303,13 @@ v9fs_session_init(struct v9fs_session_info *v9ses, goto SessCleanUp; }; - v9ses->transport = trans_proto; + v9ses->transport = kmalloc(sizeof(*v9ses->transport), GFP_KERNEL); + if (!v9ses->transport) { + retval = -ENOMEM; + goto SessCleanUp; + } + + memmove(v9ses->transport, trans_proto, sizeof(*v9ses->transport)); if ((retval = v9ses->transport->init(v9ses, dev_name, data)) < 0) { eprintk(KERN_ERR, "problem initializing transport\n"); From a1f9d8d23fef301ba0c0b4983e0aa947168e1c37 Mon Sep 17 00:00:00 2001 From: Latchesar Ionkov Date: Thu, 22 Sep 2005 21:43:52 -0700 Subject: [PATCH 144/204] [PATCH] v9fs: replace strlen on newly allocated by __getname buffers to PATH_MAX v9fs_vfs_readlink allocates space for the link using __getname and errorneously uses strlen on the newly allocated buffer to check if the buffer passed by the user is bigger than the one returned by __getname. The patch replaces the strlen usage to PATH_MAX, which is the actual size of the buffers returned by __getname. Signed-off-by: Latchesar Ionkov Cc: Eric Van Hensbergen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/9p/vfs_inode.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index 0c13fc60004..b16322db5ce 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -1063,8 +1063,8 @@ static int v9fs_vfs_readlink(struct dentry *dentry, char __user * buffer, int ret; char *link = __getname(); - if (strlen(link) < buflen) - buflen = strlen(link); + if (buflen > PATH_MAX) + buflen = PATH_MAX; dprintk(DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_iname, dentry); From f71626a461e7d4af099ca71830ea530e96c22e11 Mon Sep 17 00:00:00 2001 From: Latchesar Ionkov Date: Thu, 22 Sep 2005 21:43:53 -0700 Subject: [PATCH 145/204] [PATCH] v9fs: don't free root dentry & inode if error occurs in v9fs_get_sb If error occurs while in v9fs_get_sb after it calles sget, the dentry object of the root and its inode may be freed twice -- once while handling the error in v9fs_get_sb, and second time when v9fs_get_sb calles deactivate_super (which in turn calls v9fs_kill_super) The patch removes the unnecessary code that frees the root dentry and its inode. Signed-off-by: Latchesar Ionkov Cc: Eric Van Hensbergen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/9p/vfs_super.c | 24 +++++++----------------- 1 file changed, 7 insertions(+), 17 deletions(-) diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index 868f350b2c5..1e2b2b54d30 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -129,8 +129,8 @@ static struct super_block *v9fs_get_sb(struct file_system_type if ((newfid = v9fs_session_init(v9ses, dev_name, data)) < 0) { dprintk(DEBUG_ERROR, "problem initiating session\n"); - retval = newfid; - goto free_session; + kfree(v9ses); + return ERR_PTR(newfid); } sb = sget(fs_type, NULL, v9fs_set_super, v9ses); @@ -150,7 +150,7 @@ static struct super_block *v9fs_get_sb(struct file_system_type if (!root) { retval = -ENOMEM; - goto release_inode; + goto put_back_sb; } sb->s_root = root; @@ -159,7 +159,7 @@ static struct super_block *v9fs_get_sb(struct file_system_type root_fid = v9fs_fid_create(root); if (root_fid == NULL) { retval = -ENOMEM; - goto release_dentry; + goto put_back_sb; } root_fid->fidopen = 0; @@ -182,25 +182,15 @@ static struct super_block *v9fs_get_sb(struct file_system_type if (stat_result < 0) { retval = stat_result; - goto release_dentry; + goto put_back_sb; } return sb; - release_dentry: - dput(sb->s_root); - - release_inode: - iput(inode); - - put_back_sb: +put_back_sb: + /* deactivate_super calls v9fs_kill_super which will frees the rest */ up_write(&sb->s_umount); deactivate_super(sb); - v9fs_session_close(v9ses); - - free_session: - kfree(v9ses); - return ERR_PTR(retval); } From e1e70a262d5cb5a02d2fc171e0178197849482d4 Mon Sep 17 00:00:00 2001 From: Bernd Petrovitsch Date: Thu, 22 Sep 2005 21:43:53 -0700 Subject: [PATCH 146/204] [PATCH] Rename vprintk define in bttpvp.h Rename the (apparently) bttv intern #define vprintk to verbprintk to resolve a name clash. Reason: vprintk() is defined in include/linux/kernel.h similar to printk but with a va_list argument. (akpm: I changed it to bttv_printk) Cc: Mauro Carvalho Chehab Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/media/video/bttv-driver.c | 14 +++++++------- drivers/media/video/bttvp.h | 2 +- 2 files changed, 8 insertions(+), 8 deletions(-) diff --git a/drivers/media/video/bttv-driver.c b/drivers/media/video/bttv-driver.c index a564321db2f..c062a017491 100644 --- a/drivers/media/video/bttv-driver.c +++ b/drivers/media/video/bttv-driver.c @@ -763,21 +763,21 @@ static void set_pll(struct bttv *btv) /* no PLL needed */ if (btv->pll.pll_current == 0) return; - vprintk(KERN_INFO "bttv%d: PLL can sleep, using XTAL (%d).\n", - btv->c.nr,btv->pll.pll_ifreq); + bttv_printk(KERN_INFO "bttv%d: PLL can sleep, using XTAL (%d).\n", + btv->c.nr,btv->pll.pll_ifreq); btwrite(0x00,BT848_TGCTRL); btwrite(0x00,BT848_PLL_XCI); btv->pll.pll_current = 0; return; } - vprintk(KERN_INFO "bttv%d: PLL: %d => %d ",btv->c.nr, - btv->pll.pll_ifreq, btv->pll.pll_ofreq); + bttv_printk(KERN_INFO "bttv%d: PLL: %d => %d ",btv->c.nr, + btv->pll.pll_ifreq, btv->pll.pll_ofreq); set_pll_freq(btv, btv->pll.pll_ifreq, btv->pll.pll_ofreq); for (i=0; i<10; i++) { /* Let other people run while the PLL stabilizes */ - vprintk("."); + bttv_printk("."); msleep(10); if (btread(BT848_DSTATUS) & BT848_DSTATUS_PLOCK) { @@ -785,12 +785,12 @@ static void set_pll(struct bttv *btv) } else { btwrite(0x08,BT848_TGCTRL); btv->pll.pll_current = btv->pll.pll_ofreq; - vprintk(" ok\n"); + bttv_printk(" ok\n"); return; } } btv->pll.pll_current = -1; - vprintk("failed\n"); + bttv_printk("failed\n"); return; } diff --git a/drivers/media/video/bttvp.h b/drivers/media/video/bttvp.h index 9b0b7ca035f..7a312f79340 100644 --- a/drivers/media/video/bttvp.h +++ b/drivers/media/video/bttvp.h @@ -221,7 +221,7 @@ extern void bttv_gpio_tracking(struct bttv *btv, char *comment); extern int init_bttv_i2c(struct bttv *btv); extern int fini_bttv_i2c(struct bttv *btv); -#define vprintk if (bttv_verbose) printk +#define bttv_printk if (bttv_verbose) printk #define dprintk if (bttv_debug >= 1) printk #define d2printk if (bttv_debug >= 2) printk From 66c81f0059b57e18aad43a5b0d330102c237cd72 Mon Sep 17 00:00:00 2001 From: Grant Coady Date: Thu, 22 Sep 2005 21:43:54 -0700 Subject: [PATCH 147/204] [PATCH] DEBUG redefined in drivers/mtd/devices/docecc.c Fix namespace clash: drivers/mtd/devices/docecc.c:43:1: warning: "DEBUG" redefined In file included from drivers/mtd/devices/docecc.c:40: include/linux/mtd/mtd.h:219:1: warning: this is the location of the previous definition Signed-off-by: Grant Coady Cc: David Woodhouse Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/mtd/devices/docecc.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/mtd/devices/docecc.c b/drivers/mtd/devices/docecc.c index 9a087c1fb0b..24f670b5a4f 100644 --- a/drivers/mtd/devices/docecc.c +++ b/drivers/mtd/devices/docecc.c @@ -40,7 +40,7 @@ #include #include -#define DEBUG 0 +#define DEBUG_ECC 0 /* need to undef it (from asm/termbits.h) */ #undef B0 @@ -249,7 +249,7 @@ eras_dec_rs(dtype Alpha_to[NN + 1], dtype Index_of[NN + 1], lambda[j] ^= Alpha_to[modnn(u + tmp)]; } } -#if DEBUG >= 1 +#if DEBUG_ECC >= 1 /* Test code that verifies the erasure locator polynomial just constructed Needed only for decoder debugging. */ @@ -276,7 +276,7 @@ eras_dec_rs(dtype Alpha_to[NN + 1], dtype Index_of[NN + 1], count = -1; goto finish; } -#if DEBUG >= 2 +#if DEBUG_ECC >= 2 printf("\n Erasure positions as determined by roots of Eras Loc Poly:\n"); for (i = 0; i < count; i++) printf("%d ", loc[i]); @@ -409,7 +409,7 @@ eras_dec_rs(dtype Alpha_to[NN + 1], dtype Index_of[NN + 1], den ^= Alpha_to[modnn(lambda[i+1] + i * root[j])]; } if (den == 0) { -#if DEBUG >= 1 +#if DEBUG_ECC >= 1 printf("\n ERROR: denominator = 0\n"); #endif /* Convert to dual- basis */ From 244bc050d5c9431682b7d834aff92e663fe88bb2 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Thu, 22 Sep 2005 21:43:55 -0700 Subject: [PATCH 148/204] [PATCH] ppc32: fix build with oprofile Current -git tree doesn't build when enabling oprofile on a non-bookE CPU (like on a PowerMac for example). While there is no performance counter support for these CPUs implemented yet, it's still nice to be able to use the timer based sampling, and that got broken. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc/kernel/Makefile | 3 +-- arch/ppc/kernel/perfmon.c | 6 +++++- 2 files changed, 6 insertions(+), 3 deletions(-) diff --git a/arch/ppc/kernel/Makefile b/arch/ppc/kernel/Makefile index 1fb92f16acd..b1457a8a9c0 100644 --- a/arch/ppc/kernel/Makefile +++ b/arch/ppc/kernel/Makefile @@ -15,9 +15,8 @@ extra-y += vmlinux.lds obj-y := entry.o traps.o irq.o idle.o time.o misc.o \ process.o signal.o ptrace.o align.o \ semaphore.o syscalls.o setup.o \ - cputable.o ppc_htab.o + cputable.o ppc_htab.o perfmon.o obj-$(CONFIG_6xx) += l2cr.o cpu_setup_6xx.o -obj-$(CONFIG_E500) += perfmon.o obj-$(CONFIG_SOFTWARE_SUSPEND) += swsusp.o obj-$(CONFIG_POWER4) += cpu_setup_power4.o obj-$(CONFIG_MODULES) += module.o ppc_ksyms.o diff --git a/arch/ppc/kernel/perfmon.c b/arch/ppc/kernel/perfmon.c index fa1dad96b83..04c18788e85 100644 --- a/arch/ppc/kernel/perfmon.c +++ b/arch/ppc/kernel/perfmon.c @@ -45,7 +45,7 @@ static void dummy_perf(struct pt_regs *regs) mtpmr(PMRN_PMGC0, pmgc0); } -#else +#elif CONFIG_6xx /* Ensure exceptions are disabled */ static void dummy_perf(struct pt_regs *regs) @@ -55,6 +55,10 @@ static void dummy_perf(struct pt_regs *regs) mmcr0 &= ~MMCR0_PMXE; mtspr(SPRN_MMCR0, mmcr0); } +#else +static void dummy_perf(struct pt_regs *regs) +{ +} #endif void (*perf_irq)(struct pt_regs *) = dummy_perf; From 720b9429e8f41f7c4ee9df293403650905042035 Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Thu, 22 Sep 2005 21:43:56 -0700 Subject: [PATCH 149/204] [PATCH] SOFTWARE_SUSPEND needs HOTPLUG_CPU on SMP Signed-off-by: Randy Dunlap Acked-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/power/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig index 396c7873e80..46a5e5acff9 100644 --- a/kernel/power/Kconfig +++ b/kernel/power/Kconfig @@ -29,7 +29,7 @@ config PM_DEBUG config SOFTWARE_SUSPEND bool "Software Suspend" - depends on PM && SWAP && (X86 || ((FVR || PPC32) && !SMP)) + depends on PM && SWAP && (X86 && (!SMP || SUSPEND_SMP)) || ((FVR || PPC32) && !SMP) ---help--- Enable the possibility of suspending the machine. It doesn't need APM. From 4b3c86a7452df8608c32a1c1f19c0cc0723c145f Mon Sep 17 00:00:00 2001 From: Ivan Kokshaysky Date: Thu, 22 Sep 2005 21:43:57 -0700 Subject: [PATCH 150/204] [PATCH] alpha: fix kernel panic during SysRq-b acquire_console_sem() does BUG() in interrupt context now, as in the case of SysRq-b. Cc: Richard Henderson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/alpha/kernel/process.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/arch/alpha/kernel/process.c b/arch/alpha/kernel/process.c index fa98dae3cd9..eb20c3afff5 100644 --- a/arch/alpha/kernel/process.c +++ b/arch/alpha/kernel/process.c @@ -127,6 +127,10 @@ common_shutdown_1(void *generic_ptr) /* If booted from SRM, reset some of the original environment. */ if (alpha_using_srm) { #ifdef CONFIG_DUMMY_CONSOLE + /* If we've gotten here after SysRq-b, leave interrupt + context before taking over the console. */ + if (in_interrupt()) + irq_exit(); /* This has the effect of resetting the VGA video origin. */ take_over_console(&dummy_con, 0, MAX_NR_CONSOLES-1, 1); #endif From 7243cc05bafdda4c4de77cba00cf87666bd237f7 Mon Sep 17 00:00:00 2001 From: Ivan Kokshaysky Date: Thu, 22 Sep 2005 21:43:58 -0700 Subject: [PATCH 151/204] [PATCH] slab: alpha inlining fix It is essential that index_of() be inlined. But alpha undoes the gcc inlining hackery and index_of() ends up out-of-line. So fiddle with things to make that function inline again. Cc: Richard Henderson Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/asm-alpha/compiler.h | 5 ++++- mm/slab.c | 7 ++++--- 2 files changed, 8 insertions(+), 4 deletions(-) diff --git a/include/asm-alpha/compiler.h b/include/asm-alpha/compiler.h index 399c33b7be5..0a4a8b40dfc 100644 --- a/include/asm-alpha/compiler.h +++ b/include/asm-alpha/compiler.h @@ -98,6 +98,9 @@ #undef inline #undef __inline__ #undef __inline - +#if __GNUC__ == 3 && __GNUC_MINOR__ >= 1 || __GNUC__ > 3 +#undef __always_inline +#define __always_inline inline __attribute__((always_inline)) +#endif #endif /* __ALPHA_COMPILER_H */ diff --git a/mm/slab.c b/mm/slab.c index 437d3388054..cf19ff2ab5e 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -308,12 +308,12 @@ struct kmem_list3 __initdata initkmem_list3[NUM_INIT_LISTS]; #define SIZE_L3 (1 + MAX_NUMNODES) /* - * This function may be completely optimized away if + * This function must be completely optimized away if * a constant is passed to it. Mostly the same as * what is in linux/slab.h except it returns an * index. */ -static inline int index_of(const size_t size) +static __always_inline int index_of(const size_t size) { if (__builtin_constant_p(size)) { int i = 0; @@ -329,7 +329,8 @@ static inline int index_of(const size_t size) extern void __bad_size(void); __bad_size(); } - } + } else + BUG(); return 0; } From 609725f87709724218047a1049913ed233f79d3c Mon Sep 17 00:00:00 2001 From: Keir Fraser Date: Thu, 22 Sep 2005 21:43:59 -0700 Subject: [PATCH 152/204] [PATCH] Fix mmap() of /dev/hpet The address passed to io_remap_pfn_range() in hpet_mmap() does not need to be converted using __pa(): it is already a physical address. This bug was found and the patch suggested by Clay Harris. I introduced this particular bug when making io_remap_pfn_range changes a few months ago. In fact mmap()ing /dev/hpet has *never* previously worked: before my changes __pa() was being executed on an ioremap()ed virtual address, which is also invalid. Signed-off-by: Keir Fraser Cc: Robert Picco Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/char/hpet.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/char/hpet.c b/drivers/char/hpet.c index de0379b6d50..c055bb630ff 100644 --- a/drivers/char/hpet.c +++ b/drivers/char/hpet.c @@ -273,7 +273,6 @@ static int hpet_mmap(struct file *file, struct vm_area_struct *vma) vma->vm_flags |= VM_IO; vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); - addr = __pa(addr); if (io_remap_pfn_range(vma, vma->vm_start, addr >> PAGE_SHIFT, PAGE_SIZE, vma->vm_page_prot)) { From 4c87b74c72d42f7272aa4bf8f17c73bb53774eee Mon Sep 17 00:00:00 2001 From: Robert Love Date: Thu, 22 Sep 2005 21:44:00 -0700 Subject: [PATCH 153/204] [PATCH] hdaps: small update. - Handle dmi_system_check() elegantly, now that my bugfix is upstream. - Add support for the X41 and R52. - Cleanup some comments do I do not have to keep updating them with each new whitelisted laptop. Signed-off-by: Robert Love Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/hwmon/Kconfig | 9 ++++----- drivers/hwmon/hdaps.c | 21 +++++++++------------ 2 files changed, 13 insertions(+), 17 deletions(-) diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig index 7e72e922b41..db358cfa7cb 100644 --- a/drivers/hwmon/Kconfig +++ b/drivers/hwmon/Kconfig @@ -418,12 +418,11 @@ config SENSORS_HDAPS help This driver provides support for the IBM Hard Drive Active Protection System (hdaps), which provides an accelerometer and other misc. data. - Supported laptops include the IBM ThinkPad T41, T42, T43, and R51. - The accelerometer data is readable via sysfs. + ThinkPads starting with the R50, T41, and X40 are supported. The + accelerometer data is readable via sysfs. - This driver also provides an input class device, allowing the - laptop to act as a pinball machine-esque mouse. This is off by - default but enabled via sysfs or the module parameter "mousedev". + This driver also provides an absolute input class device, allowing + the laptop to act as a pinball machine-esque joystick. Say Y here if you have an applicable laptop and want to experience the awesome power of hdaps. diff --git a/drivers/hwmon/hdaps.c b/drivers/hwmon/hdaps.c index 4c56411f399..7f010761382 100644 --- a/drivers/hwmon/hdaps.c +++ b/drivers/hwmon/hdaps.c @@ -4,9 +4,9 @@ * Copyright (C) 2005 Robert Love * Copyright (C) 2005 Jesper Juhl * - * The HardDisk Active Protection System (hdaps) is present in the IBM ThinkPad - * T41, T42, T43, R50, R50p, R51, and X40, at least. It provides a basic - * two-axis accelerometer and other data, such as the device's temperature. + * The HardDisk Active Protection System (hdaps) is present in IBM ThinkPads + * starting with the R40, T41, and X40. It provides a basic two-axis + * accelerometer and other data, such as the device's temperature. * * This driver is based on the document by Mark A. Smith available at * http://www.almaden.ibm.com/cs/people/marksmith/tpaps.html and a lot of trial @@ -487,24 +487,19 @@ static struct attribute_group hdaps_attribute_group = { /* Module stuff */ -/* - * XXX: We should be able to return nonzero and halt the detection process. - * But there is a bug in dmi_check_system() where a nonzero return from the - * first match will result in a return of failure from dmi_check_system(). - * I fixed this; the patch is 2.6-git. Once in a released tree, we can make - * hdaps_dmi_match_invert() return hdaps_dmi_match(), which in turn returns 1. - */ +/* hdaps_dmi_match - found a match. return one, short-circuiting the hunt. */ static int hdaps_dmi_match(struct dmi_system_id *id) { printk(KERN_INFO "hdaps: %s detected.\n", id->ident); - return 0; + return 1; } +/* hdaps_dmi_match_invert - found an inverted match. */ static int hdaps_dmi_match_invert(struct dmi_system_id *id) { hdaps_invert = 1; printk(KERN_INFO "hdaps: inverting axis readings.\n"); - return 0; + return hdaps_dmi_match(id); } #define HDAPS_DMI_MATCH_NORMAL(model) { \ @@ -534,6 +529,7 @@ static int __init hdaps_init(void) HDAPS_DMI_MATCH_INVERT("ThinkPad R50p"), HDAPS_DMI_MATCH_NORMAL("ThinkPad R50"), HDAPS_DMI_MATCH_NORMAL("ThinkPad R51"), + HDAPS_DMI_MATCH_NORMAL("ThinkPad R52"), HDAPS_DMI_MATCH_INVERT("ThinkPad T41p"), HDAPS_DMI_MATCH_NORMAL("ThinkPad T41"), HDAPS_DMI_MATCH_INVERT("ThinkPad T42p"), @@ -541,6 +537,7 @@ static int __init hdaps_init(void) HDAPS_DMI_MATCH_NORMAL("ThinkPad T43"), HDAPS_DMI_MATCH_NORMAL("ThinkPad X40"), HDAPS_DMI_MATCH_NORMAL("ThinkPad X41 Tablet"), + HDAPS_DMI_MATCH_NORMAL("ThinkPad X41"), { .ident = NULL } }; From 57487f4376e16558ccbe45a5b41d8cb5192309a4 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Thu, 22 Sep 2005 21:44:01 -0700 Subject: [PATCH 154/204] [PATCH] swsusp: do not trigger BUG_ON() if there is not enough memory The following patch makes swsusp avoid triggering the BUG_ON() in swsusp_suspend() if there is not enough memory for suspend. Signed-off-by: Rafael J. Wysocki Cc: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/power/swsusp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/power/swsusp.c b/kernel/power/swsusp.c index d967e875ee8..353f37f1ba2 100644 --- a/kernel/power/swsusp.c +++ b/kernel/power/swsusp.c @@ -918,6 +918,7 @@ static int swsusp_alloc(void) pagedir_nosave = NULL; nr_copy_pages = calc_nr(nr_copy_pages); + nr_copy_pages_check = nr_copy_pages; pr_debug("suspend: (pages needed: %d + %d free: %d)\n", nr_copy_pages, PAGES_FOR_IO, nr_free_pages()); @@ -940,7 +941,6 @@ static int swsusp_alloc(void) return error; } - nr_copy_pages_check = nr_copy_pages; return 0; } From ff69416e6323fe9d38c42a06ebdefeb58bbe9336 Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Thu, 22 Sep 2005 21:44:02 -0700 Subject: [PATCH 155/204] [PATCH] slab: fix handling of pages from foreign NUMA nodes The numa slab allocator may allocate pages from foreign nodes onto the lists for a particular node if a node runs out of memory. Inspecting the slab->nodeid field will not reflect that the page is now in use for the slabs of another node. This patch fixes that issue by adding a node field to free_block so that the caller can indicate which node currently uses a slab. Also removes the check for the current node from kmalloc_cache_node since the process may shift later to another node which may lead to an allocation on another node than intended. Signed-off-by: Christoph Lameter Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/slab.c | 38 +++++++++++++++++++------------------- 1 file changed, 19 insertions(+), 19 deletions(-) diff --git a/mm/slab.c b/mm/slab.c index cf19ff2ab5e..c9adfce0040 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -640,7 +640,7 @@ static enum { static DEFINE_PER_CPU(struct work_struct, reap_work); -static void free_block(kmem_cache_t* cachep, void** objpp, int len); +static void free_block(kmem_cache_t* cachep, void** objpp, int len, int node); static void enable_cpucache (kmem_cache_t *cachep); static void cache_reap (void *unused); static int __node_shrink(kmem_cache_t *cachep, int node); @@ -805,7 +805,7 @@ static inline void __drain_alien_cache(kmem_cache_t *cachep, struct array_cache if (ac->avail) { spin_lock(&rl3->list_lock); - free_block(cachep, ac->entry, ac->avail); + free_block(cachep, ac->entry, ac->avail, node); ac->avail = 0; spin_unlock(&rl3->list_lock); } @@ -926,7 +926,7 @@ static int __devinit cpuup_callback(struct notifier_block *nfb, /* Free limit for this kmem_list3 */ l3->free_limit -= cachep->batchcount; if (nc) - free_block(cachep, nc->entry, nc->avail); + free_block(cachep, nc->entry, nc->avail, node); if (!cpus_empty(mask)) { spin_unlock(&l3->list_lock); @@ -935,7 +935,7 @@ static int __devinit cpuup_callback(struct notifier_block *nfb, if (l3->shared) { free_block(cachep, l3->shared->entry, - l3->shared->avail); + l3->shared->avail, node); kfree(l3->shared); l3->shared = NULL; } @@ -1883,12 +1883,13 @@ static void do_drain(void *arg) { kmem_cache_t *cachep = (kmem_cache_t*)arg; struct array_cache *ac; + int node = numa_node_id(); check_irq_off(); ac = ac_data(cachep); - spin_lock(&cachep->nodelists[numa_node_id()]->list_lock); - free_block(cachep, ac->entry, ac->avail); - spin_unlock(&cachep->nodelists[numa_node_id()]->list_lock); + spin_lock(&cachep->nodelists[node]->list_lock); + free_block(cachep, ac->entry, ac->avail, node); + spin_unlock(&cachep->nodelists[node]->list_lock); ac->avail = 0; } @@ -2609,7 +2610,7 @@ done: /* * Caller needs to acquire correct kmem_list's list_lock */ -static void free_block(kmem_cache_t *cachep, void **objpp, int nr_objects) +static void free_block(kmem_cache_t *cachep, void **objpp, int nr_objects, int node) { int i; struct kmem_list3 *l3; @@ -2618,14 +2619,12 @@ static void free_block(kmem_cache_t *cachep, void **objpp, int nr_objects) void *objp = objpp[i]; struct slab *slabp; unsigned int objnr; - int nodeid = 0; slabp = GET_PAGE_SLAB(virt_to_page(objp)); - nodeid = slabp->nodeid; - l3 = cachep->nodelists[nodeid]; + l3 = cachep->nodelists[node]; list_del(&slabp->list); objnr = (objp - slabp->s_mem) / cachep->objsize; - check_spinlock_acquired_node(cachep, nodeid); + check_spinlock_acquired_node(cachep, node); check_slabp(cachep, slabp); @@ -2665,13 +2664,14 @@ static void cache_flusharray(kmem_cache_t *cachep, struct array_cache *ac) { int batchcount; struct kmem_list3 *l3; + int node = numa_node_id(); batchcount = ac->batchcount; #if DEBUG BUG_ON(!batchcount || batchcount > ac->avail); #endif check_irq_off(); - l3 = cachep->nodelists[numa_node_id()]; + l3 = cachep->nodelists[node]; spin_lock(&l3->list_lock); if (l3->shared) { struct array_cache *shared_array = l3->shared; @@ -2687,7 +2687,7 @@ static void cache_flusharray(kmem_cache_t *cachep, struct array_cache *ac) } } - free_block(cachep, ac->entry, batchcount); + free_block(cachep, ac->entry, batchcount, node); free_done: #if STATS { @@ -2752,7 +2752,7 @@ static inline void __cache_free(kmem_cache_t *cachep, void *objp) } else { spin_lock(&(cachep->nodelists[nodeid])-> list_lock); - free_block(cachep, &objp, 1); + free_block(cachep, &objp, 1, nodeid); spin_unlock(&(cachep->nodelists[nodeid])-> list_lock); } @@ -2845,7 +2845,7 @@ void *kmem_cache_alloc_node(kmem_cache_t *cachep, unsigned int __nocast flags, i unsigned long save_flags; void *ptr; - if (nodeid == numa_node_id() || nodeid == -1) + if (nodeid == -1) return __cache_alloc(cachep, flags); if (unlikely(!cachep->nodelists[nodeid])) { @@ -3080,7 +3080,7 @@ static int alloc_kmemlist(kmem_cache_t *cachep) if ((nc = cachep->nodelists[node]->shared)) free_block(cachep, nc->entry, - nc->avail); + nc->avail, node); l3->shared = new; if (!cachep->nodelists[node]->alien) { @@ -3161,7 +3161,7 @@ static int do_tune_cpucache(kmem_cache_t *cachep, int limit, int batchcount, if (!ccold) continue; spin_lock_irq(&cachep->nodelists[cpu_to_node(i)]->list_lock); - free_block(cachep, ccold->entry, ccold->avail); + free_block(cachep, ccold->entry, ccold->avail, cpu_to_node(i)); spin_unlock_irq(&cachep->nodelists[cpu_to_node(i)]->list_lock); kfree(ccold); } @@ -3241,7 +3241,7 @@ static void drain_array_locked(kmem_cache_t *cachep, if (tofree > ac->avail) { tofree = (ac->avail+1)/2; } - free_block(cachep, ac->entry, tofree); + free_block(cachep, ac->entry, tofree, node); ac->avail -= tofree; memmove(ac->entry, &(ac->entry[tofree]), sizeof(void*)*ac->avail); From 275abf5b06676ca057cf3e15f0d027eafcb204a0 Mon Sep 17 00:00:00 2001 From: OGAWA Hirofumi Date: Thu, 22 Sep 2005 21:44:03 -0700 Subject: [PATCH 156/204] [PATCH] ext3: ext3_show_options fix EXT3_MOUNT_DATA_FLAGS is not a boolean. This fixes it. Signed-off-by: OGAWA Hirofumi Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ext3/super.c | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/fs/ext3/super.c b/fs/ext3/super.c index a93c3609025..9e24ceb019f 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -512,15 +512,14 @@ static void ext3_clear_inode(struct inode *inode) static int ext3_show_options(struct seq_file *seq, struct vfsmount *vfs) { - struct ext3_sb_info *sbi = EXT3_SB(vfs->mnt_sb); + struct super_block *sb = vfs->mnt_sb; + struct ext3_sb_info *sbi = EXT3_SB(sb); - if (sbi->s_mount_opt & EXT3_MOUNT_JOURNAL_DATA) + if (test_opt(sb, DATA_FLAGS) == EXT3_MOUNT_JOURNAL_DATA) seq_puts(seq, ",data=journal"); - - if (sbi->s_mount_opt & EXT3_MOUNT_ORDERED_DATA) + else if (test_opt(sb, DATA_FLAGS) == EXT3_MOUNT_ORDERED_DATA) seq_puts(seq, ",data=ordered"); - - if (sbi->s_mount_opt & EXT3_MOUNT_WRITEBACK_DATA) + else if (test_opt(sb, DATA_FLAGS) == EXT3_MOUNT_WRITEBACK_DATA) seq_puts(seq, ",data=writeback"); #if defined(CONFIG_QUOTA) From 586a90eb4c0b512df766e08a7718af7c6b9b2dcf Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Thu, 22 Sep 2005 21:44:04 -0700 Subject: [PATCH 157/204] [PATCH] ppc64: Fix LPAR regression The recent iommu fix broke booting on some POWER4 and POWER5 LPAR boxes. It looks like we have been calling the non LPAR iommu_dev_setup on LPAR machines for a while. The recent iommu fix caused that code path to fail. It looks like we just need to hook up the devices iommu_table to the parents one, so do that instead of calling iommu_dev_setup_pSeries and crossing the streams. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/pSeries_iommu.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/arch/ppc64/kernel/pSeries_iommu.c b/arch/ppc64/kernel/pSeries_iommu.c index 8c6313e7e14..25d6c7eebfd 100644 --- a/arch/ppc64/kernel/pSeries_iommu.c +++ b/arch/ppc64/kernel/pSeries_iommu.c @@ -525,9 +525,8 @@ static void iommu_dev_setup_pSeriesLP(struct pci_dev *dev) * slots on POWER4 machines. */ if (dma_window == NULL || pdn->parent == NULL) { - /* Fall back to regular (non-LPAR) dev setup */ - DBG("No dma window for device, falling back to regular setup\n"); - iommu_dev_setup_pSeries(dev); + DBG("No dma window for device, linking to parent\n"); + PCI_DN(dn)->iommu_table = PCI_DN(pdn)->iommu_table; return; } else { DBG("Found DMA window, allocating table\n"); From f951da3757cf46444a6ea52e97b8aa9400fe761a Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Thu, 22 Sep 2005 21:44:05 -0700 Subject: [PATCH 158/204] [PATCH] ppc64: Fix build with iommu debug enabled Fix build when iommu debug is enabled. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/pSeries_iommu.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/arch/ppc64/kernel/pSeries_iommu.c b/arch/ppc64/kernel/pSeries_iommu.c index 25d6c7eebfd..d17f0108a03 100644 --- a/arch/ppc64/kernel/pSeries_iommu.c +++ b/arch/ppc64/kernel/pSeries_iommu.c @@ -364,7 +364,8 @@ static void iommu_bus_setup_pSeries(struct pci_bus *bus) while (pci->phb->dma_window_size * children > 0x80000000ul) pci->phb->dma_window_size >>= 1; - DBG("No ISA/IDE, window size is %x\n", pci->phb->dma_window_size); + DBG("No ISA/IDE, window size is 0x%lx\n", + pci->phb->dma_window_size); pci->phb->dma_window_base_cur = 0; return; @@ -388,7 +389,7 @@ static void iommu_bus_setup_pSeries(struct pci_bus *bus) while (pci->phb->dma_window_size * children > 0x70000000ul) pci->phb->dma_window_size >>= 1; - DBG("ISA/IDE, window size is %x\n", pci->phb->dma_window_size); + DBG("ISA/IDE, window size is 0x%lx\n", pci->phb->dma_window_size); } @@ -442,7 +443,7 @@ static void iommu_dev_setup_pSeries(struct pci_dev *dev) struct device_node *dn, *mydn; struct iommu_table *tbl; - DBG("iommu_dev_setup_pSeries, dev %p (%s)\n", dev, dev->pretty_name); + DBG("iommu_dev_setup_pSeries, dev %p (%s)\n", dev, pci_name(dev)); mydn = dn = pci_device_to_OF_node(dev); @@ -469,7 +470,7 @@ static void iommu_dev_setup_pSeries(struct pci_dev *dev) if (dn && dn->data) { PCI_DN(mydn)->iommu_table = PCI_DN(dn)->iommu_table; } else { - DBG("iommu_dev_setup_pSeries, dev %p (%s) has no iommu table\n", dev, dev->pretty_name); + DBG("iommu_dev_setup_pSeries, dev %p (%s) has no iommu table\n", dev, pci_name(dev)); } } @@ -503,7 +504,7 @@ static void iommu_dev_setup_pSeriesLP(struct pci_dev *dev) int *dma_window = NULL; struct pci_dn *pci; - DBG("iommu_dev_setup_pSeriesLP, dev %p (%s)\n", dev, dev->pretty_name); + DBG("iommu_dev_setup_pSeriesLP, dev %p (%s)\n", dev, pci_name(dev)); /* dev setup for LPAR is a little tricky, since the device tree might * contain the dma-window properties per-device and not neccesarily From 0f329075fb1dbd6845db03e9bb8252024fdbea1f Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Thu, 22 Sep 2005 21:44:05 -0700 Subject: [PATCH 159/204] [PATCH] ppc64: Fix 64bit ptrace DABR support Fix my stupid bug in the 64bit version of PTRACE_SET_DEBUGREG. Signed-off-by: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/kernel/ptrace.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/ppc64/kernel/ptrace.c b/arch/ppc64/kernel/ptrace.c index 85ed3188a91..b1c044ca575 100644 --- a/arch/ppc64/kernel/ptrace.c +++ b/arch/ppc64/kernel/ptrace.c @@ -219,6 +219,7 @@ int sys_ptrace(long request, long pid, long addr, long data) case PTRACE_SET_DEBUGREG: ret = ptrace_set_debugreg(child, addr, data); + break; case PTRACE_DETACH: ret = ptrace_detach(child, data); From 0365ba7fb1fa94a41289d6a3d36b4d95960e56cc Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Thu, 22 Sep 2005 21:44:06 -0700 Subject: [PATCH 160/204] [PATCH] ppc64: SMU driver update & i2c support The SMU is the "system controller" chip used by Apple recent G5 machines including the iMac G5. It drives things like fans, i2c busses, real time clock, etc... The current kernel contains a very crude driver that doesn't do much more than reading the real time clock synchronously. This is a completely rewritten driver that provides interrupt based command queuing, a userland interface, and an i2c/smbus driver for accessing the devices hanging off the SMU i2c busses like temperature sensors. This driver is a basic block for upcoming work on thermal control for those machines, among others. Signed-off-by: Benjamin Herrenschmidt Cc: Jean Delvare Cc: Greg KH Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc/platforms/pmac_setup.c | 10 +- arch/ppc/syslib/of_device.c | 6 +- arch/ppc64/kernel/of_device.c | 7 +- arch/ppc64/kernel/pmac_setup.c | 18 +- arch/ppc64/kernel/pmac_time.c | 4 +- drivers/i2c/busses/Kconfig | 12 + drivers/i2c/busses/Makefile | 1 + drivers/i2c/busses/i2c-pmac-smu.c | 316 ++++++++ drivers/macintosh/smu.c | 1042 ++++++++++++++++++++++---- drivers/macintosh/therm_adt746x.c | 2 +- drivers/macintosh/therm_pm72.c | 2 +- drivers/macintosh/therm_windtunnel.c | 2 +- include/asm-ppc/macio.h | 1 - include/asm-ppc/of_device.h | 5 +- include/asm-ppc64/smu.h | 365 ++++++++- 15 files changed, 1624 insertions(+), 169 deletions(-) create mode 100644 drivers/i2c/busses/i2c-pmac-smu.c diff --git a/arch/ppc/platforms/pmac_setup.c b/arch/ppc/platforms/pmac_setup.c index b392b9a1598..4c56a4734ae 100644 --- a/arch/ppc/platforms/pmac_setup.c +++ b/arch/ppc/platforms/pmac_setup.c @@ -719,7 +719,8 @@ pmac_declare_of_platform_devices(void) if (np) { for (np = np->child; np != NULL; np = np->sibling) if (strncmp(np->name, "i2c", 3) == 0) { - of_platform_device_create(np, "uni-n-i2c"); + of_platform_device_create(np, "uni-n-i2c", + NULL); break; } } @@ -727,17 +728,18 @@ pmac_declare_of_platform_devices(void) if (np) { for (np = np->child; np != NULL; np = np->sibling) if (strncmp(np->name, "i2c", 3) == 0) { - of_platform_device_create(np, "u3-i2c"); + of_platform_device_create(np, "u3-i2c", + NULL); break; } } np = find_devices("valkyrie"); if (np) - of_platform_device_create(np, "valkyrie"); + of_platform_device_create(np, "valkyrie", NULL); np = find_devices("platinum"); if (np) - of_platform_device_create(np, "platinum"); + of_platform_device_create(np, "platinum", NULL); return 0; } diff --git a/arch/ppc/syslib/of_device.c b/arch/ppc/syslib/of_device.c index da8a0f2128d..93c7231ea70 100644 --- a/arch/ppc/syslib/of_device.c +++ b/arch/ppc/syslib/of_device.c @@ -234,7 +234,9 @@ void of_device_unregister(struct of_device *ofdev) device_unregister(&ofdev->dev); } -struct of_device* of_platform_device_create(struct device_node *np, const char *bus_id) +struct of_device* of_platform_device_create(struct device_node *np, + const char *bus_id, + struct device *parent) { struct of_device *dev; u32 *reg; @@ -247,7 +249,7 @@ struct of_device* of_platform_device_create(struct device_node *np, const char * dev->node = of_node_get(np); dev->dma_mask = 0xffffffffUL; dev->dev.dma_mask = &dev->dma_mask; - dev->dev.parent = NULL; + dev->dev.parent = parent; dev->dev.bus = &of_platform_bus_type; dev->dev.release = of_release_dev; diff --git a/arch/ppc64/kernel/of_device.c b/arch/ppc64/kernel/of_device.c index da580812ddf..9f200f0f2ad 100644 --- a/arch/ppc64/kernel/of_device.c +++ b/arch/ppc64/kernel/of_device.c @@ -233,7 +233,9 @@ void of_device_unregister(struct of_device *ofdev) device_unregister(&ofdev->dev); } -struct of_device* of_platform_device_create(struct device_node *np, const char *bus_id) +struct of_device* of_platform_device_create(struct device_node *np, + const char *bus_id, + struct device *parent) { struct of_device *dev; @@ -245,7 +247,7 @@ struct of_device* of_platform_device_create(struct device_node *np, const char * dev->node = np; dev->dma_mask = 0xffffffffUL; dev->dev.dma_mask = &dev->dma_mask; - dev->dev.parent = NULL; + dev->dev.parent = parent; dev->dev.bus = &of_platform_bus_type; dev->dev.release = of_release_dev; @@ -259,6 +261,7 @@ struct of_device* of_platform_device_create(struct device_node *np, const char * return dev; } + EXPORT_SYMBOL(of_match_device); EXPORT_SYMBOL(of_platform_bus_type); EXPORT_SYMBOL(of_register_driver); diff --git a/arch/ppc64/kernel/pmac_setup.c b/arch/ppc64/kernel/pmac_setup.c index 325426c7bed..25755252067 100644 --- a/arch/ppc64/kernel/pmac_setup.c +++ b/arch/ppc64/kernel/pmac_setup.c @@ -434,15 +434,23 @@ static int pmac_check_legacy_ioport(unsigned int baseport) static int __init pmac_declare_of_platform_devices(void) { - struct device_node *np; + struct device_node *np, *npp; - np = find_devices("u3"); - if (np) { - for (np = np->child; np != NULL; np = np->sibling) + npp = of_find_node_by_name(NULL, "u3"); + if (npp) { + for (np = NULL; (np = of_get_next_child(npp, np)) != NULL;) { if (strncmp(np->name, "i2c", 3) == 0) { - of_platform_device_create(np, "u3-i2c"); + of_platform_device_create(np, "u3-i2c", NULL); + of_node_put(np); break; } + } + of_node_put(npp); + } + npp = of_find_node_by_type(NULL, "smu"); + if (npp) { + of_platform_device_create(npp, "smu", NULL); + of_node_put(npp); } return 0; diff --git a/arch/ppc64/kernel/pmac_time.c b/arch/ppc64/kernel/pmac_time.c index 3059edb09cc..41bbb8c5969 100644 --- a/arch/ppc64/kernel/pmac_time.c +++ b/arch/ppc64/kernel/pmac_time.c @@ -84,7 +84,7 @@ void __pmac pmac_get_rtc_time(struct rtc_time *tm) #ifdef CONFIG_PMAC_SMU case SYS_CTRLER_SMU: - smu_get_rtc_time(tm); + smu_get_rtc_time(tm, 1); break; #endif /* CONFIG_PMAC_SMU */ default: @@ -128,7 +128,7 @@ int __pmac pmac_set_rtc_time(struct rtc_time *tm) #ifdef CONFIG_PMAC_SMU case SYS_CTRLER_SMU: - return smu_set_rtc_time(tm); + return smu_set_rtc_time(tm, 1); #endif /* CONFIG_PMAC_SMU */ default: return -ENODEV; diff --git a/drivers/i2c/busses/Kconfig b/drivers/i2c/busses/Kconfig index 8334496a7e0..3badfec75b1 100644 --- a/drivers/i2c/busses/Kconfig +++ b/drivers/i2c/busses/Kconfig @@ -245,6 +245,18 @@ config I2C_KEYWEST This support is also available as a module. If so, the module will be called i2c-keywest. +config I2C_PMAC_SMU + tristate "Powermac SMU I2C interface" + depends on I2C && PMAC_SMU + help + This supports the use of the I2C interface in the SMU + chip on recent Apple machines like the iMac G5. It is used + among others by the thermal control driver for those machines. + Say Y if you have such a machine. + + This support is also available as a module. If so, the module + will be called i2c-pmac-smu. + config I2C_MPC tristate "MPC107/824x/85xx/52xx" depends on I2C && PPC32 diff --git a/drivers/i2c/busses/Makefile b/drivers/i2c/busses/Makefile index 980b3e98367..f1df00f66c6 100644 --- a/drivers/i2c/busses/Makefile +++ b/drivers/i2c/busses/Makefile @@ -20,6 +20,7 @@ obj-$(CONFIG_I2C_ITE) += i2c-ite.o obj-$(CONFIG_I2C_IXP2000) += i2c-ixp2000.o obj-$(CONFIG_I2C_IXP4XX) += i2c-ixp4xx.o obj-$(CONFIG_I2C_KEYWEST) += i2c-keywest.o +obj-$(CONFIG_I2C_PMAC_SMU) += i2c-pmac-smu.o obj-$(CONFIG_I2C_MPC) += i2c-mpc.o obj-$(CONFIG_I2C_MV64XXX) += i2c-mv64xxx.o obj-$(CONFIG_I2C_NFORCE2) += i2c-nforce2.o diff --git a/drivers/i2c/busses/i2c-pmac-smu.c b/drivers/i2c/busses/i2c-pmac-smu.c new file mode 100644 index 00000000000..8a9f5648a23 --- /dev/null +++ b/drivers/i2c/busses/i2c-pmac-smu.c @@ -0,0 +1,316 @@ +/* + i2c Support for Apple SMU Controller + + Copyright (c) 2005 Benjamin Herrenschmidt, IBM Corp. + + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + +*/ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +static int probe; + +MODULE_AUTHOR("Benjamin Herrenschmidt "); +MODULE_DESCRIPTION("I2C driver for Apple's SMU"); +MODULE_LICENSE("GPL"); +module_param(probe, bool, 0); + + +/* Physical interface */ +struct smu_iface +{ + struct i2c_adapter adapter; + struct completion complete; + u32 busid; +}; + +static void smu_i2c_done(struct smu_i2c_cmd *cmd, void *misc) +{ + struct smu_iface *iface = misc; + complete(&iface->complete); +} + +/* + * SMBUS-type transfer entrypoint + */ +static s32 smu_smbus_xfer( struct i2c_adapter* adap, + u16 addr, + unsigned short flags, + char read_write, + u8 command, + int size, + union i2c_smbus_data* data) +{ + struct smu_iface *iface = i2c_get_adapdata(adap); + struct smu_i2c_cmd cmd; + int rc = 0; + int read = (read_write == I2C_SMBUS_READ); + + cmd.info.bus = iface->busid; + cmd.info.devaddr = (addr << 1) | (read ? 0x01 : 0x00); + + /* Prepare datas & select mode */ + switch (size) { + case I2C_SMBUS_QUICK: + cmd.info.type = SMU_I2C_TRANSFER_SIMPLE; + cmd.info.datalen = 0; + break; + case I2C_SMBUS_BYTE: + cmd.info.type = SMU_I2C_TRANSFER_SIMPLE; + cmd.info.datalen = 1; + if (!read) + cmd.info.data[0] = data->byte; + break; + case I2C_SMBUS_BYTE_DATA: + cmd.info.type = SMU_I2C_TRANSFER_STDSUB; + cmd.info.datalen = 1; + cmd.info.sublen = 1; + cmd.info.subaddr[0] = command; + cmd.info.subaddr[1] = 0; + cmd.info.subaddr[2] = 0; + if (!read) + cmd.info.data[0] = data->byte; + break; + case I2C_SMBUS_WORD_DATA: + cmd.info.type = SMU_I2C_TRANSFER_STDSUB; + cmd.info.datalen = 2; + cmd.info.sublen = 1; + cmd.info.subaddr[0] = command; + cmd.info.subaddr[1] = 0; + cmd.info.subaddr[2] = 0; + if (!read) { + cmd.info.data[0] = data->byte & 0xff; + cmd.info.data[1] = (data->byte >> 8) & 0xff; + } + break; + /* Note that these are broken vs. the expected smbus API where + * on reads, the lenght is actually returned from the function, + * but I think the current API makes no sense and I don't want + * any driver that I haven't verified for correctness to go + * anywhere near a pmac i2c bus anyway ... + */ + case I2C_SMBUS_BLOCK_DATA: + cmd.info.type = SMU_I2C_TRANSFER_STDSUB; + cmd.info.datalen = data->block[0] + 1; + if (cmd.info.datalen > 6) + return -EINVAL; + if (!read) + memcpy(cmd.info.data, data->block, cmd.info.datalen); + cmd.info.sublen = 1; + cmd.info.subaddr[0] = command; + cmd.info.subaddr[1] = 0; + cmd.info.subaddr[2] = 0; + break; + case I2C_SMBUS_I2C_BLOCK_DATA: + cmd.info.type = SMU_I2C_TRANSFER_STDSUB; + cmd.info.datalen = data->block[0]; + if (cmd.info.datalen > 7) + return -EINVAL; + if (!read) + memcpy(cmd.info.data, &data->block[1], + cmd.info.datalen); + cmd.info.sublen = 1; + cmd.info.subaddr[0] = command; + cmd.info.subaddr[1] = 0; + cmd.info.subaddr[2] = 0; + break; + + default: + return -EINVAL; + } + + /* Turn a standardsub read into a combined mode access */ + if (read_write == I2C_SMBUS_READ && + cmd.info.type == SMU_I2C_TRANSFER_STDSUB) + cmd.info.type = SMU_I2C_TRANSFER_COMBINED; + + /* Finish filling command and submit it */ + cmd.done = smu_i2c_done; + cmd.misc = iface; + rc = smu_queue_i2c(&cmd); + if (rc < 0) + return rc; + wait_for_completion(&iface->complete); + rc = cmd.status; + + if (!read || rc < 0) + return rc; + + switch (size) { + case I2C_SMBUS_BYTE: + case I2C_SMBUS_BYTE_DATA: + data->byte = cmd.info.data[0]; + break; + case I2C_SMBUS_WORD_DATA: + data->word = ((u16)cmd.info.data[1]) << 8; + data->word |= cmd.info.data[0]; + break; + /* Note that these are broken vs. the expected smbus API where + * on reads, the lenght is actually returned from the function, + * but I think the current API makes no sense and I don't want + * any driver that I haven't verified for correctness to go + * anywhere near a pmac i2c bus anyway ... + */ + case I2C_SMBUS_BLOCK_DATA: + case I2C_SMBUS_I2C_BLOCK_DATA: + memcpy(&data->block[0], cmd.info.data, cmd.info.datalen); + break; + } + + return rc; +} + +static u32 +smu_smbus_func(struct i2c_adapter * adapter) +{ + return I2C_FUNC_SMBUS_QUICK | I2C_FUNC_SMBUS_BYTE | + I2C_FUNC_SMBUS_BYTE_DATA | I2C_FUNC_SMBUS_WORD_DATA | + I2C_FUNC_SMBUS_BLOCK_DATA; +} + +/* For now, we only handle combined mode (smbus) */ +static struct i2c_algorithm smu_algorithm = { + .smbus_xfer = smu_smbus_xfer, + .functionality = smu_smbus_func, +}; + +static int create_iface(struct device_node *np, struct device *dev) +{ + struct smu_iface* iface; + u32 *reg, busid; + int rc; + + reg = (u32 *)get_property(np, "reg", NULL); + if (reg == NULL) { + printk(KERN_ERR "i2c-pmac-smu: can't find bus number !\n"); + return -ENXIO; + } + busid = *reg; + + iface = kmalloc(sizeof(struct smu_iface), GFP_KERNEL); + if (iface == NULL) { + printk(KERN_ERR "i2c-pmac-smu: can't allocate inteface !\n"); + return -ENOMEM; + } + memset(iface, 0, sizeof(struct smu_iface)); + init_completion(&iface->complete); + iface->busid = busid; + + dev_set_drvdata(dev, iface); + + sprintf(iface->adapter.name, "smu-i2c-%02x", busid); + iface->adapter.algo = &smu_algorithm; + iface->adapter.algo_data = NULL; + iface->adapter.client_register = NULL; + iface->adapter.client_unregister = NULL; + i2c_set_adapdata(&iface->adapter, iface); + iface->adapter.dev.parent = dev; + + rc = i2c_add_adapter(&iface->adapter); + if (rc) { + printk(KERN_ERR "i2c-pamc-smu.c: Adapter %s registration " + "failed\n", iface->adapter.name); + i2c_set_adapdata(&iface->adapter, NULL); + } + + if (probe) { + unsigned char addr; + printk("Probe: "); + for (addr = 0x00; addr <= 0x7f; addr++) { + if (i2c_smbus_xfer(&iface->adapter,addr, + 0,0,0,I2C_SMBUS_QUICK,NULL) >= 0) + printk("%02x ", addr); + } + printk("\n"); + } + + printk(KERN_INFO "SMU i2c bus %x registered\n", busid); + + return 0; +} + +static int dispose_iface(struct device *dev) +{ + struct smu_iface *iface = dev_get_drvdata(dev); + int rc; + + rc = i2c_del_adapter(&iface->adapter); + i2c_set_adapdata(&iface->adapter, NULL); + /* We aren't that prepared to deal with this... */ + if (rc) + printk("i2c-pmac-smu.c: Failed to remove bus %s !\n", + iface->adapter.name); + dev_set_drvdata(dev, NULL); + kfree(iface); + + return 0; +} + + +static int create_iface_of_platform(struct of_device* dev, + const struct of_device_id *match) +{ + return create_iface(dev->node, &dev->dev); +} + + +static int dispose_iface_of_platform(struct of_device* dev) +{ + return dispose_iface(&dev->dev); +} + + +static struct of_device_id i2c_smu_match[] = +{ + { + .compatible = "smu-i2c", + }, + {}, +}; +static struct of_platform_driver i2c_smu_of_platform_driver = +{ + .name = "i2c-smu", + .match_table = i2c_smu_match, + .probe = create_iface_of_platform, + .remove = dispose_iface_of_platform +}; + + +static int __init i2c_pmac_smu_init(void) +{ + of_register_driver(&i2c_smu_of_platform_driver); + return 0; +} + + +static void __exit i2c_pmac_smu_cleanup(void) +{ + of_unregister_driver(&i2c_smu_of_platform_driver); +} + +module_init(i2c_pmac_smu_init); +module_exit(i2c_pmac_smu_cleanup); diff --git a/drivers/macintosh/smu.c b/drivers/macintosh/smu.c index fb535737d17..a85ac18dd21 100644 --- a/drivers/macintosh/smu.c +++ b/drivers/macintosh/smu.c @@ -8,21 +8,15 @@ */ /* - * For now, this driver includes: - * - RTC get & set - * - reboot & shutdown commands - * all synchronous with IRQ disabled (ugh) - * * TODO: - * rework in a way the PMU driver works, that is asynchronous - * with a queue of commands. I'll do that as soon as I have an - * SMU based machine at hand. Some more cleanup is needed too, - * like maybe fitting it into a platform device, etc... - * Also check what's up with cache coherency, and if we really - * can't do better than flushing the cache, maybe build a table - * of command len/reply len like the PMU driver to only flush - * what is actually necessary. - * --BenH. + * - maybe add timeout to commands ? + * - blocking version of time functions + * - polling version of i2c commands (including timer that works with + * interrutps off) + * - maybe avoid some data copies with i2c by directly using the smu cmd + * buffer and a lower level internal interface + * - understand SMU -> CPU events and implement reception of them via + * the userland interface */ #include @@ -36,6 +30,11 @@ #include #include #include +#include +#include +#include +#include +#include #include #include @@ -45,8 +44,13 @@ #include #include #include +#include +#include -#define DEBUG_SMU 1 +#define VERSION "0.6" +#define AUTHOR "(c) 2005 Benjamin Herrenschmidt, IBM Corp." + +#undef DEBUG_SMU #ifdef DEBUG_SMU #define DPRINTK(fmt, args...) do { printk(KERN_DEBUG fmt , ##args); } while (0) @@ -57,20 +61,30 @@ /* * This is the command buffer passed to the SMU hardware */ +#define SMU_MAX_DATA 254 + struct smu_cmd_buf { u8 cmd; u8 length; - u8 data[0x0FFE]; + u8 data[SMU_MAX_DATA]; }; struct smu_device { spinlock_t lock; struct device_node *of_node; - int db_ack; /* doorbell ack GPIO */ - int db_req; /* doorbell req GPIO */ + struct of_device *of_dev; + int doorbell; /* doorbell gpio */ u32 __iomem *db_buf; /* doorbell buffer */ + int db_irq; + int msg; + int msg_irq; struct smu_cmd_buf *cmd_buf; /* command buffer virtual */ u32 cmd_buf_abs; /* command buffer absolute */ + struct list_head cmd_list; + struct smu_cmd *cmd_cur; /* pending command */ + struct list_head cmd_i2c_list; + struct smu_i2c_cmd *cmd_i2c_cur; /* pending i2c command */ + struct timer_list i2c_timer; }; /* @@ -79,113 +93,243 @@ struct smu_device { */ static struct smu_device *smu; + /* - * SMU low level communication stuff + * SMU driver low level stuff */ -static inline int smu_cmd_stat(struct smu_cmd_buf *cmd_buf, u8 cmd_ack) -{ - rmb(); - return cmd_buf->cmd == cmd_ack && cmd_buf->length != 0; -} -static inline u8 smu_save_ack_cmd(struct smu_cmd_buf *cmd_buf) +static void smu_start_cmd(void) { - return (~cmd_buf->cmd) & 0xff; -} + unsigned long faddr, fend; + struct smu_cmd *cmd; -static void smu_send_cmd(struct smu_device *dev) -{ - /* SMU command buf is currently cacheable, we need a physical - * address. This isn't exactly a DMA mapping here, I suspect + if (list_empty(&smu->cmd_list)) + return; + + /* Fetch first command in queue */ + cmd = list_entry(smu->cmd_list.next, struct smu_cmd, link); + smu->cmd_cur = cmd; + list_del(&cmd->link); + + DPRINTK("SMU: starting cmd %x, %d bytes data\n", cmd->cmd, + cmd->data_len); + DPRINTK("SMU: data buffer: %02x %02x %02x %02x ...\n", + ((u8 *)cmd->data_buf)[0], ((u8 *)cmd->data_buf)[1], + ((u8 *)cmd->data_buf)[2], ((u8 *)cmd->data_buf)[3]); + + /* Fill the SMU command buffer */ + smu->cmd_buf->cmd = cmd->cmd; + smu->cmd_buf->length = cmd->data_len; + memcpy(smu->cmd_buf->data, cmd->data_buf, cmd->data_len); + + /* Flush command and data to RAM */ + faddr = (unsigned long)smu->cmd_buf; + fend = faddr + smu->cmd_buf->length + 2; + flush_inval_dcache_range(faddr, fend); + + /* This isn't exactly a DMA mapping here, I suspect * the SMU is actually communicating with us via i2c to the * northbridge or the CPU to access RAM. */ - writel(dev->cmd_buf_abs, dev->db_buf); + writel(smu->cmd_buf_abs, smu->db_buf); /* Ring the SMU doorbell */ - pmac_do_feature_call(PMAC_FTR_WRITE_GPIO, NULL, dev->db_req, 4); - pmac_do_feature_call(PMAC_FTR_READ_GPIO, NULL, dev->db_req, 4); + pmac_do_feature_call(PMAC_FTR_WRITE_GPIO, NULL, smu->doorbell, 4); } -static int smu_cmd_done(struct smu_device *dev) + +static irqreturn_t smu_db_intr(int irq, void *arg, struct pt_regs *regs) { - unsigned long wait = 0; - int gpio; + unsigned long flags; + struct smu_cmd *cmd; + void (*done)(struct smu_cmd *cmd, void *misc) = NULL; + void *misc = NULL; + u8 gpio; + int rc = 0; - /* Check the SMU doorbell */ - do { - gpio = pmac_do_feature_call(PMAC_FTR_READ_GPIO, - NULL, dev->db_ack); - if ((gpio & 7) == 7) - return 0; - udelay(100); - } while(++wait < 10000); + /* SMU completed the command, well, we hope, let's make sure + * of it + */ + spin_lock_irqsave(&smu->lock, flags); - printk(KERN_ERR "SMU timeout !\n"); - return -ENXIO; + gpio = pmac_do_feature_call(PMAC_FTR_READ_GPIO, NULL, smu->doorbell); + if ((gpio & 7) != 7) + return IRQ_HANDLED; + + cmd = smu->cmd_cur; + smu->cmd_cur = NULL; + if (cmd == NULL) + goto bail; + + if (rc == 0) { + unsigned long faddr; + int reply_len; + u8 ack; + + /* CPU might have brought back the cache line, so we need + * to flush again before peeking at the SMU response. We + * flush the entire buffer for now as we haven't read the + * reply lenght (it's only 2 cache lines anyway) + */ + faddr = (unsigned long)smu->cmd_buf; + flush_inval_dcache_range(faddr, faddr + 256); + + /* Now check ack */ + ack = (~cmd->cmd) & 0xff; + if (ack != smu->cmd_buf->cmd) { + DPRINTK("SMU: incorrect ack, want %x got %x\n", + ack, smu->cmd_buf->cmd); + rc = -EIO; + } + reply_len = rc == 0 ? smu->cmd_buf->length : 0; + DPRINTK("SMU: reply len: %d\n", reply_len); + if (reply_len > cmd->reply_len) { + printk(KERN_WARNING "SMU: reply buffer too small," + "got %d bytes for a %d bytes buffer\n", + reply_len, cmd->reply_len); + reply_len = cmd->reply_len; + } + cmd->reply_len = reply_len; + if (cmd->reply_buf && reply_len) + memcpy(cmd->reply_buf, smu->cmd_buf->data, reply_len); + } + + /* Now complete the command. Write status last in order as we lost + * ownership of the command structure as soon as it's no longer -1 + */ + done = cmd->done; + misc = cmd->misc; + mb(); + cmd->status = rc; + bail: + /* Start next command if any */ + smu_start_cmd(); + spin_unlock_irqrestore(&smu->lock, flags); + + /* Call command completion handler if any */ + if (done) + done(cmd, misc); + + /* It's an edge interrupt, nothing to do */ + return IRQ_HANDLED; } -static int smu_do_cmd(struct smu_device *dev) + +static irqreturn_t smu_msg_intr(int irq, void *arg, struct pt_regs *regs) { - int rc; - u8 cmd_ack; + /* I don't quite know what to do with this one, we seem to never + * receive it, so I suspect we have to arm it someway in the SMU + * to start getting events that way. + */ - DPRINTK("SMU do_cmd %02x len=%d %02x\n", - dev->cmd_buf->cmd, dev->cmd_buf->length, - dev->cmd_buf->data[0]); + printk(KERN_INFO "SMU: message interrupt !\n"); - cmd_ack = smu_save_ack_cmd(dev->cmd_buf); - - /* Clear cmd_buf cache lines */ - flush_inval_dcache_range((unsigned long)dev->cmd_buf, - ((unsigned long)dev->cmd_buf) + - sizeof(struct smu_cmd_buf)); - smu_send_cmd(dev); - rc = smu_cmd_done(dev); - if (rc == 0) - rc = smu_cmd_stat(dev->cmd_buf, cmd_ack) ? 0 : -1; - - DPRINTK("SMU do_cmd %02x len=%d %02x => %d (%02x)\n", - dev->cmd_buf->cmd, dev->cmd_buf->length, - dev->cmd_buf->data[0], rc, cmd_ack); - - return rc; + /* It's an edge interrupt, nothing to do */ + return IRQ_HANDLED; } + +/* + * Queued command management. + * + */ + +int smu_queue_cmd(struct smu_cmd *cmd) +{ + unsigned long flags; + + if (smu == NULL) + return -ENODEV; + if (cmd->data_len > SMU_MAX_DATA || + cmd->reply_len > SMU_MAX_DATA) + return -EINVAL; + + cmd->status = 1; + spin_lock_irqsave(&smu->lock, flags); + list_add_tail(&cmd->link, &smu->cmd_list); + if (smu->cmd_cur == NULL) + smu_start_cmd(); + spin_unlock_irqrestore(&smu->lock, flags); + + return 0; +} +EXPORT_SYMBOL(smu_queue_cmd); + + +int smu_queue_simple(struct smu_simple_cmd *scmd, u8 command, + unsigned int data_len, + void (*done)(struct smu_cmd *cmd, void *misc), + void *misc, ...) +{ + struct smu_cmd *cmd = &scmd->cmd; + va_list list; + int i; + + if (data_len > sizeof(scmd->buffer)) + return -EINVAL; + + memset(scmd, 0, sizeof(*scmd)); + cmd->cmd = command; + cmd->data_len = data_len; + cmd->data_buf = scmd->buffer; + cmd->reply_len = sizeof(scmd->buffer); + cmd->reply_buf = scmd->buffer; + cmd->done = done; + cmd->misc = misc; + + va_start(list, misc); + for (i = 0; i < data_len; ++i) + scmd->buffer[i] = (u8)va_arg(list, int); + va_end(list); + + return smu_queue_cmd(cmd); +} +EXPORT_SYMBOL(smu_queue_simple); + + +void smu_poll(void) +{ + u8 gpio; + + if (smu == NULL) + return; + + gpio = pmac_do_feature_call(PMAC_FTR_READ_GPIO, NULL, smu->doorbell); + if ((gpio & 7) == 7) + smu_db_intr(smu->db_irq, smu, NULL); +} +EXPORT_SYMBOL(smu_poll); + + +void smu_done_complete(struct smu_cmd *cmd, void *misc) +{ + struct completion *comp = misc; + + complete(comp); +} +EXPORT_SYMBOL(smu_done_complete); + + +void smu_spinwait_cmd(struct smu_cmd *cmd) +{ + while(cmd->status == 1) + smu_poll(); +} +EXPORT_SYMBOL(smu_spinwait_cmd); + + /* RTC low level commands */ static inline int bcd2hex (int n) { return (((n & 0xf0) >> 4) * 10) + (n & 0xf); } + static inline int hex2bcd (int n) { return ((n / 10) << 4) + (n % 10); } -#if 0 -static inline void smu_fill_set_pwrup_timer_cmd(struct smu_cmd_buf *cmd_buf) -{ - cmd_buf->cmd = 0x8e; - cmd_buf->length = 8; - cmd_buf->data[0] = 0x00; - memset(cmd_buf->data + 1, 0, 7); -} - -static inline void smu_fill_get_pwrup_timer_cmd(struct smu_cmd_buf *cmd_buf) -{ - cmd_buf->cmd = 0x8e; - cmd_buf->length = 1; - cmd_buf->data[0] = 0x01; -} - -static inline void smu_fill_dis_pwrup_timer_cmd(struct smu_cmd_buf *cmd_buf) -{ - cmd_buf->cmd = 0x8e; - cmd_buf->length = 1; - cmd_buf->data[0] = 0x02; -} -#endif static inline void smu_fill_set_rtc_cmd(struct smu_cmd_buf *cmd_buf, struct rtc_time *time) @@ -202,100 +346,96 @@ static inline void smu_fill_set_rtc_cmd(struct smu_cmd_buf *cmd_buf, cmd_buf->data[7] = hex2bcd(time->tm_year - 100); } -static inline void smu_fill_get_rtc_cmd(struct smu_cmd_buf *cmd_buf) -{ - cmd_buf->cmd = 0x8e; - cmd_buf->length = 1; - cmd_buf->data[0] = 0x81; -} -static void smu_parse_get_rtc_reply(struct smu_cmd_buf *cmd_buf, - struct rtc_time *time) +int smu_get_rtc_time(struct rtc_time *time, int spinwait) { - time->tm_sec = bcd2hex(cmd_buf->data[0]); - time->tm_min = bcd2hex(cmd_buf->data[1]); - time->tm_hour = bcd2hex(cmd_buf->data[2]); - time->tm_wday = bcd2hex(cmd_buf->data[3]); - time->tm_mday = bcd2hex(cmd_buf->data[4]); - time->tm_mon = bcd2hex(cmd_buf->data[5]) - 1; - time->tm_year = bcd2hex(cmd_buf->data[6]) + 100; -} - -int smu_get_rtc_time(struct rtc_time *time) -{ - unsigned long flags; + struct smu_simple_cmd cmd; int rc; if (smu == NULL) return -ENODEV; memset(time, 0, sizeof(struct rtc_time)); - spin_lock_irqsave(&smu->lock, flags); - smu_fill_get_rtc_cmd(smu->cmd_buf); - rc = smu_do_cmd(smu); - if (rc == 0) - smu_parse_get_rtc_reply(smu->cmd_buf, time); - spin_unlock_irqrestore(&smu->lock, flags); + rc = smu_queue_simple(&cmd, SMU_CMD_RTC_COMMAND, 1, NULL, NULL, + SMU_CMD_RTC_GET_DATETIME); + if (rc) + return rc; + smu_spinwait_simple(&cmd); - return rc; + time->tm_sec = bcd2hex(cmd.buffer[0]); + time->tm_min = bcd2hex(cmd.buffer[1]); + time->tm_hour = bcd2hex(cmd.buffer[2]); + time->tm_wday = bcd2hex(cmd.buffer[3]); + time->tm_mday = bcd2hex(cmd.buffer[4]); + time->tm_mon = bcd2hex(cmd.buffer[5]) - 1; + time->tm_year = bcd2hex(cmd.buffer[6]) + 100; + + return 0; } -int smu_set_rtc_time(struct rtc_time *time) + +int smu_set_rtc_time(struct rtc_time *time, int spinwait) { - unsigned long flags; + struct smu_simple_cmd cmd; int rc; if (smu == NULL) return -ENODEV; - spin_lock_irqsave(&smu->lock, flags); - smu_fill_set_rtc_cmd(smu->cmd_buf, time); - rc = smu_do_cmd(smu); - spin_unlock_irqrestore(&smu->lock, flags); + rc = smu_queue_simple(&cmd, SMU_CMD_RTC_COMMAND, 8, NULL, NULL, + SMU_CMD_RTC_SET_DATETIME, + hex2bcd(time->tm_sec), + hex2bcd(time->tm_min), + hex2bcd(time->tm_hour), + time->tm_wday, + hex2bcd(time->tm_mday), + hex2bcd(time->tm_mon) + 1, + hex2bcd(time->tm_year - 100)); + if (rc) + return rc; + smu_spinwait_simple(&cmd); - return rc; + return 0; } + void smu_shutdown(void) { - const unsigned char *command = "SHUTDOWN"; - unsigned long flags; + struct smu_simple_cmd cmd; if (smu == NULL) return; - spin_lock_irqsave(&smu->lock, flags); - smu->cmd_buf->cmd = 0xaa; - smu->cmd_buf->length = strlen(command); - strcpy(smu->cmd_buf->data, command); - smu_do_cmd(smu); + if (smu_queue_simple(&cmd, SMU_CMD_POWER_COMMAND, 9, NULL, NULL, + 'S', 'H', 'U', 'T', 'D', 'O', 'W', 'N', 0)) + return; + smu_spinwait_simple(&cmd); for (;;) ; - spin_unlock_irqrestore(&smu->lock, flags); } + void smu_restart(void) { - const unsigned char *command = "RESTART"; - unsigned long flags; + struct smu_simple_cmd cmd; if (smu == NULL) return; - spin_lock_irqsave(&smu->lock, flags); - smu->cmd_buf->cmd = 0xaa; - smu->cmd_buf->length = strlen(command); - strcpy(smu->cmd_buf->data, command); - smu_do_cmd(smu); + if (smu_queue_simple(&cmd, SMU_CMD_POWER_COMMAND, 8, NULL, NULL, + 'R', 'E', 'S', 'T', 'A', 'R', 'T', 0)) + return; + smu_spinwait_simple(&cmd); for (;;) ; - spin_unlock_irqrestore(&smu->lock, flags); } + int smu_present(void) { return smu != NULL; } +EXPORT_SYMBOL(smu_present); int smu_init (void) @@ -307,6 +447,8 @@ int smu_init (void) if (np == NULL) return -ENODEV; + printk(KERN_INFO "SMU driver %s %s\n", VERSION, AUTHOR); + if (smu_cmdbuf_abs == 0) { printk(KERN_ERR "SMU: Command buffer not allocated !\n"); return -EINVAL; @@ -318,7 +460,13 @@ int smu_init (void) memset(smu, 0, sizeof(*smu)); spin_lock_init(&smu->lock); + INIT_LIST_HEAD(&smu->cmd_list); + INIT_LIST_HEAD(&smu->cmd_i2c_list); smu->of_node = np; + smu->db_irq = NO_IRQ; + smu->msg_irq = NO_IRQ; + init_timer(&smu->i2c_timer); + /* smu_cmdbuf_abs is in the low 2G of RAM, can be converted to a * 32 bits value safely */ @@ -331,8 +479,8 @@ int smu_init (void) goto fail; } data = (u32 *)get_property(np, "reg", NULL); - of_node_put(np); if (data == NULL) { + of_node_put(np); printk(KERN_ERR "SMU: Can't find doorbell GPIO address !\n"); goto fail; } @@ -341,8 +489,31 @@ int smu_init (void) * and ack. GPIOs are at 0x50, best would be to find that out * in the device-tree though. */ - smu->db_req = 0x50 + *data; - smu->db_ack = 0x50 + *data; + smu->doorbell = *data; + if (smu->doorbell < 0x50) + smu->doorbell += 0x50; + if (np->n_intrs > 0) + smu->db_irq = np->intrs[0].line; + + of_node_put(np); + + /* Now look for the smu-interrupt GPIO */ + do { + np = of_find_node_by_name(NULL, "smu-interrupt"); + if (np == NULL) + break; + data = (u32 *)get_property(np, "reg", NULL); + if (data == NULL) { + of_node_put(np); + break; + } + smu->msg = *data; + if (smu->msg < 0x50) + smu->msg += 0x50; + if (np->n_intrs > 0) + smu->msg_irq = np->intrs[0].line; + of_node_put(np); + } while(0); /* Doorbell buffer is currently hard-coded, I didn't find a proper * device-tree entry giving the address. Best would probably to use @@ -362,3 +533,584 @@ int smu_init (void) return -ENXIO; } + + +static int smu_late_init(void) +{ + if (!smu) + return 0; + + /* + * Try to request the interrupts + */ + + if (smu->db_irq != NO_IRQ) { + if (request_irq(smu->db_irq, smu_db_intr, + SA_SHIRQ, "SMU doorbell", smu) < 0) { + printk(KERN_WARNING "SMU: can't " + "request interrupt %d\n", + smu->db_irq); + smu->db_irq = NO_IRQ; + } + } + + if (smu->msg_irq != NO_IRQ) { + if (request_irq(smu->msg_irq, smu_msg_intr, + SA_SHIRQ, "SMU message", smu) < 0) { + printk(KERN_WARNING "SMU: can't " + "request interrupt %d\n", + smu->msg_irq); + smu->msg_irq = NO_IRQ; + } + } + + return 0; +} +arch_initcall(smu_late_init); + +/* + * sysfs visibility + */ + +static void smu_expose_childs(void *unused) +{ + struct device_node *np; + + for (np = NULL; (np = of_get_next_child(smu->of_node, np)) != NULL;) { + if (device_is_compatible(np, "smu-i2c")) { + char name[32]; + u32 *reg = (u32 *)get_property(np, "reg", NULL); + + if (reg == NULL) + continue; + sprintf(name, "smu-i2c-%02x", *reg); + of_platform_device_create(np, name, &smu->of_dev->dev); + } + } + +} + +static DECLARE_WORK(smu_expose_childs_work, smu_expose_childs, NULL); + +static int smu_platform_probe(struct of_device* dev, + const struct of_device_id *match) +{ + if (!smu) + return -ENODEV; + smu->of_dev = dev; + + /* + * Ok, we are matched, now expose all i2c busses. We have to defer + * that unfortunately or it would deadlock inside the device model + */ + schedule_work(&smu_expose_childs_work); + + return 0; +} + +static struct of_device_id smu_platform_match[] = +{ + { + .type = "smu", + }, + {}, +}; + +static struct of_platform_driver smu_of_platform_driver = +{ + .name = "smu", + .match_table = smu_platform_match, + .probe = smu_platform_probe, +}; + +static int __init smu_init_sysfs(void) +{ + int rc; + + /* + * Due to sysfs bogosity, a sysdev is not a real device, so + * we should in fact create both if we want sysdev semantics + * for power management. + * For now, we don't power manage machines with an SMU chip, + * I'm a bit too far from figuring out how that works with those + * new chipsets, but that will come back and bite us + */ + rc = of_register_driver(&smu_of_platform_driver); + return 0; +} + +device_initcall(smu_init_sysfs); + +struct of_device *smu_get_ofdev(void) +{ + if (!smu) + return NULL; + return smu->of_dev; +} + +EXPORT_SYMBOL_GPL(smu_get_ofdev); + +/* + * i2c interface + */ + +static void smu_i2c_complete_command(struct smu_i2c_cmd *cmd, int fail) +{ + void (*done)(struct smu_i2c_cmd *cmd, void *misc) = cmd->done; + void *misc = cmd->misc; + unsigned long flags; + + /* Check for read case */ + if (!fail && cmd->read) { + if (cmd->pdata[0] < 1) + fail = 1; + else + memcpy(cmd->info.data, &cmd->pdata[1], + cmd->info.datalen); + } + + DPRINTK("SMU: completing, success: %d\n", !fail); + + /* Update status and mark no pending i2c command with lock + * held so nobody comes in while we dequeue an eventual + * pending next i2c command + */ + spin_lock_irqsave(&smu->lock, flags); + smu->cmd_i2c_cur = NULL; + wmb(); + cmd->status = fail ? -EIO : 0; + + /* Is there another i2c command waiting ? */ + if (!list_empty(&smu->cmd_i2c_list)) { + struct smu_i2c_cmd *newcmd; + + /* Fetch it, new current, remove from list */ + newcmd = list_entry(smu->cmd_i2c_list.next, + struct smu_i2c_cmd, link); + smu->cmd_i2c_cur = newcmd; + list_del(&cmd->link); + + /* Queue with low level smu */ + list_add_tail(&cmd->scmd.link, &smu->cmd_list); + if (smu->cmd_cur == NULL) + smu_start_cmd(); + } + spin_unlock_irqrestore(&smu->lock, flags); + + /* Call command completion handler if any */ + if (done) + done(cmd, misc); + +} + + +static void smu_i2c_retry(unsigned long data) +{ + struct smu_i2c_cmd *cmd = (struct smu_i2c_cmd *)data; + + DPRINTK("SMU: i2c failure, requeuing...\n"); + + /* requeue command simply by resetting reply_len */ + cmd->pdata[0] = 0xff; + cmd->scmd.reply_len = 0x10; + smu_queue_cmd(&cmd->scmd); +} + + +static void smu_i2c_low_completion(struct smu_cmd *scmd, void *misc) +{ + struct smu_i2c_cmd *cmd = misc; + int fail = 0; + + DPRINTK("SMU: i2c compl. stage=%d status=%x pdata[0]=%x rlen: %x\n", + cmd->stage, scmd->status, cmd->pdata[0], scmd->reply_len); + + /* Check for possible status */ + if (scmd->status < 0) + fail = 1; + else if (cmd->read) { + if (cmd->stage == 0) + fail = cmd->pdata[0] != 0; + else + fail = cmd->pdata[0] >= 0x80; + } else { + fail = cmd->pdata[0] != 0; + } + + /* Handle failures by requeuing command, after 5ms interval + */ + if (fail && --cmd->retries > 0) { + DPRINTK("SMU: i2c failure, starting timer...\n"); + smu->i2c_timer.function = smu_i2c_retry; + smu->i2c_timer.data = (unsigned long)cmd; + smu->i2c_timer.expires = jiffies + msecs_to_jiffies(5); + add_timer(&smu->i2c_timer); + return; + } + + /* If failure or stage 1, command is complete */ + if (fail || cmd->stage != 0) { + smu_i2c_complete_command(cmd, fail); + return; + } + + DPRINTK("SMU: going to stage 1\n"); + + /* Ok, initial command complete, now poll status */ + scmd->reply_buf = cmd->pdata; + scmd->reply_len = 0x10; + scmd->data_buf = cmd->pdata; + scmd->data_len = 1; + cmd->pdata[0] = 0; + cmd->stage = 1; + cmd->retries = 20; + smu_queue_cmd(scmd); +} + + +int smu_queue_i2c(struct smu_i2c_cmd *cmd) +{ + unsigned long flags; + + if (smu == NULL) + return -ENODEV; + + /* Fill most fields of scmd */ + cmd->scmd.cmd = SMU_CMD_I2C_COMMAND; + cmd->scmd.done = smu_i2c_low_completion; + cmd->scmd.misc = cmd; + cmd->scmd.reply_buf = cmd->pdata; + cmd->scmd.reply_len = 0x10; + cmd->scmd.data_buf = (u8 *)(char *)&cmd->info; + cmd->scmd.status = 1; + cmd->stage = 0; + cmd->pdata[0] = 0xff; + cmd->retries = 20; + cmd->status = 1; + + /* Check transfer type, sanitize some "info" fields + * based on transfer type and do more checking + */ + cmd->info.caddr = cmd->info.devaddr; + cmd->read = cmd->info.devaddr & 0x01; + switch(cmd->info.type) { + case SMU_I2C_TRANSFER_SIMPLE: + memset(&cmd->info.sublen, 0, 4); + break; + case SMU_I2C_TRANSFER_COMBINED: + cmd->info.devaddr &= 0xfe; + case SMU_I2C_TRANSFER_STDSUB: + if (cmd->info.sublen > 3) + return -EINVAL; + break; + default: + return -EINVAL; + } + + /* Finish setting up command based on transfer direction + */ + if (cmd->read) { + if (cmd->info.datalen > SMU_I2C_READ_MAX) + return -EINVAL; + memset(cmd->info.data, 0xff, cmd->info.datalen); + cmd->scmd.data_len = 9; + } else { + if (cmd->info.datalen > SMU_I2C_WRITE_MAX) + return -EINVAL; + cmd->scmd.data_len = 9 + cmd->info.datalen; + } + + DPRINTK("SMU: i2c enqueuing command\n"); + DPRINTK("SMU: %s, len=%d bus=%x addr=%x sub0=%x type=%x\n", + cmd->read ? "read" : "write", cmd->info.datalen, + cmd->info.bus, cmd->info.caddr, + cmd->info.subaddr[0], cmd->info.type); + + + /* Enqueue command in i2c list, and if empty, enqueue also in + * main command list + */ + spin_lock_irqsave(&smu->lock, flags); + if (smu->cmd_i2c_cur == NULL) { + smu->cmd_i2c_cur = cmd; + list_add_tail(&cmd->scmd.link, &smu->cmd_list); + if (smu->cmd_cur == NULL) + smu_start_cmd(); + } else + list_add_tail(&cmd->link, &smu->cmd_i2c_list); + spin_unlock_irqrestore(&smu->lock, flags); + + return 0; +} + + + +/* + * Userland driver interface + */ + + +static LIST_HEAD(smu_clist); +static DEFINE_SPINLOCK(smu_clist_lock); + +enum smu_file_mode { + smu_file_commands, + smu_file_events, + smu_file_closing +}; + +struct smu_private +{ + struct list_head list; + enum smu_file_mode mode; + int busy; + struct smu_cmd cmd; + spinlock_t lock; + wait_queue_head_t wait; + u8 buffer[SMU_MAX_DATA]; +}; + + +static int smu_open(struct inode *inode, struct file *file) +{ + struct smu_private *pp; + unsigned long flags; + + pp = kmalloc(sizeof(struct smu_private), GFP_KERNEL); + if (pp == 0) + return -ENOMEM; + memset(pp, 0, sizeof(struct smu_private)); + spin_lock_init(&pp->lock); + pp->mode = smu_file_commands; + init_waitqueue_head(&pp->wait); + + spin_lock_irqsave(&smu_clist_lock, flags); + list_add(&pp->list, &smu_clist); + spin_unlock_irqrestore(&smu_clist_lock, flags); + file->private_data = pp; + + return 0; +} + + +static void smu_user_cmd_done(struct smu_cmd *cmd, void *misc) +{ + struct smu_private *pp = misc; + + wake_up_all(&pp->wait); +} + + +static ssize_t smu_write(struct file *file, const char __user *buf, + size_t count, loff_t *ppos) +{ + struct smu_private *pp = file->private_data; + unsigned long flags; + struct smu_user_cmd_hdr hdr; + int rc = 0; + + if (pp->busy) + return -EBUSY; + else if (copy_from_user(&hdr, buf, sizeof(hdr))) + return -EFAULT; + else if (hdr.cmdtype == SMU_CMDTYPE_WANTS_EVENTS) { + pp->mode = smu_file_events; + return 0; + } else if (hdr.cmdtype != SMU_CMDTYPE_SMU) + return -EINVAL; + else if (pp->mode != smu_file_commands) + return -EBADFD; + else if (hdr.data_len > SMU_MAX_DATA) + return -EINVAL; + + spin_lock_irqsave(&pp->lock, flags); + if (pp->busy) { + spin_unlock_irqrestore(&pp->lock, flags); + return -EBUSY; + } + pp->busy = 1; + pp->cmd.status = 1; + spin_unlock_irqrestore(&pp->lock, flags); + + if (copy_from_user(pp->buffer, buf + sizeof(hdr), hdr.data_len)) { + pp->busy = 0; + return -EFAULT; + } + + pp->cmd.cmd = hdr.cmd; + pp->cmd.data_len = hdr.data_len; + pp->cmd.reply_len = SMU_MAX_DATA; + pp->cmd.data_buf = pp->buffer; + pp->cmd.reply_buf = pp->buffer; + pp->cmd.done = smu_user_cmd_done; + pp->cmd.misc = pp; + rc = smu_queue_cmd(&pp->cmd); + if (rc < 0) + return rc; + return count; +} + + +static ssize_t smu_read_command(struct file *file, struct smu_private *pp, + char __user *buf, size_t count) +{ + DECLARE_WAITQUEUE(wait, current); + struct smu_user_reply_hdr hdr; + unsigned long flags; + int size, rc = 0; + + if (!pp->busy) + return 0; + if (count < sizeof(struct smu_user_reply_hdr)) + return -EOVERFLOW; + spin_lock_irqsave(&pp->lock, flags); + if (pp->cmd.status == 1) { + if (file->f_flags & O_NONBLOCK) + return -EAGAIN; + add_wait_queue(&pp->wait, &wait); + for (;;) { + set_current_state(TASK_INTERRUPTIBLE); + rc = 0; + if (pp->cmd.status != 1) + break; + rc = -ERESTARTSYS; + if (signal_pending(current)) + break; + spin_unlock_irqrestore(&pp->lock, flags); + schedule(); + spin_lock_irqsave(&pp->lock, flags); + } + set_current_state(TASK_RUNNING); + remove_wait_queue(&pp->wait, &wait); + } + spin_unlock_irqrestore(&pp->lock, flags); + if (rc) + return rc; + if (pp->cmd.status != 0) + pp->cmd.reply_len = 0; + size = sizeof(hdr) + pp->cmd.reply_len; + if (count < size) + size = count; + rc = size; + hdr.status = pp->cmd.status; + hdr.reply_len = pp->cmd.reply_len; + if (copy_to_user(buf, &hdr, sizeof(hdr))) + return -EFAULT; + size -= sizeof(hdr); + if (size && copy_to_user(buf + sizeof(hdr), pp->buffer, size)) + return -EFAULT; + pp->busy = 0; + + return rc; +} + + +static ssize_t smu_read_events(struct file *file, struct smu_private *pp, + char __user *buf, size_t count) +{ + /* Not implemented */ + msleep_interruptible(1000); + return 0; +} + + +static ssize_t smu_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) +{ + struct smu_private *pp = file->private_data; + + if (pp->mode == smu_file_commands) + return smu_read_command(file, pp, buf, count); + if (pp->mode == smu_file_events) + return smu_read_events(file, pp, buf, count); + + return -EBADFD; +} + +static unsigned int smu_fpoll(struct file *file, poll_table *wait) +{ + struct smu_private *pp = file->private_data; + unsigned int mask = 0; + unsigned long flags; + + if (pp == 0) + return 0; + + if (pp->mode == smu_file_commands) { + poll_wait(file, &pp->wait, wait); + + spin_lock_irqsave(&pp->lock, flags); + if (pp->busy && pp->cmd.status != 1) + mask |= POLLIN; + spin_unlock_irqrestore(&pp->lock, flags); + } if (pp->mode == smu_file_events) { + /* Not yet implemented */ + } + return mask; +} + +static int smu_release(struct inode *inode, struct file *file) +{ + struct smu_private *pp = file->private_data; + unsigned long flags; + unsigned int busy; + + if (pp == 0) + return 0; + + file->private_data = NULL; + + /* Mark file as closing to avoid races with new request */ + spin_lock_irqsave(&pp->lock, flags); + pp->mode = smu_file_closing; + busy = pp->busy; + + /* Wait for any pending request to complete */ + if (busy && pp->cmd.status == 1) { + DECLARE_WAITQUEUE(wait, current); + + add_wait_queue(&pp->wait, &wait); + for (;;) { + set_current_state(TASK_UNINTERRUPTIBLE); + if (pp->cmd.status != 1) + break; + spin_lock_irqsave(&pp->lock, flags); + schedule(); + spin_unlock_irqrestore(&pp->lock, flags); + } + set_current_state(TASK_RUNNING); + remove_wait_queue(&pp->wait, &wait); + } + spin_unlock_irqrestore(&pp->lock, flags); + + spin_lock_irqsave(&smu_clist_lock, flags); + list_del(&pp->list); + spin_unlock_irqrestore(&smu_clist_lock, flags); + kfree(pp); + + return 0; +} + + +static struct file_operations smu_device_fops __pmacdata = { + .llseek = no_llseek, + .read = smu_read, + .write = smu_write, + .poll = smu_fpoll, + .open = smu_open, + .release = smu_release, +}; + +static struct miscdevice pmu_device __pmacdata = { + MISC_DYNAMIC_MINOR, "smu", &smu_device_fops +}; + +static int smu_device_init(void) +{ + if (!smu) + return -ENODEV; + if (misc_register(&pmu_device) < 0) + printk(KERN_ERR "via-pmu: cannot register misc device.\n"); + return 0; +} +device_initcall(smu_device_init); diff --git a/drivers/macintosh/therm_adt746x.c b/drivers/macintosh/therm_adt746x.c index c9ca1118e44..f38696622eb 100644 --- a/drivers/macintosh/therm_adt746x.c +++ b/drivers/macintosh/therm_adt746x.c @@ -599,7 +599,7 @@ thermostat_init(void) sensor_location[2] = "?"; } - of_dev = of_platform_device_create(np, "temperatures"); + of_dev = of_platform_device_create(np, "temperatures", NULL); if (of_dev == NULL) { printk(KERN_ERR "Can't register temperatures device !\n"); diff --git a/drivers/macintosh/therm_pm72.c b/drivers/macintosh/therm_pm72.c index 703e3197331..cc507ceef15 100644 --- a/drivers/macintosh/therm_pm72.c +++ b/drivers/macintosh/therm_pm72.c @@ -2051,7 +2051,7 @@ static int __init therm_pm72_init(void) return -ENODEV; } } - of_dev = of_platform_device_create(np, "temperature"); + of_dev = of_platform_device_create(np, "temperature", NULL); if (of_dev == NULL) { printk(KERN_ERR "Can't register FCU platform device !\n"); return -ENODEV; diff --git a/drivers/macintosh/therm_windtunnel.c b/drivers/macintosh/therm_windtunnel.c index cbb72eb0426..6aaa1df1a64 100644 --- a/drivers/macintosh/therm_windtunnel.c +++ b/drivers/macintosh/therm_windtunnel.c @@ -504,7 +504,7 @@ g4fan_init( void ) } if( !(np=of_find_node_by_name(NULL, "fan")) ) return -ENODEV; - x.of_dev = of_platform_device_create( np, "temperature" ); + x.of_dev = of_platform_device_create(np, "temperature", NULL); of_node_put( np ); if( !x.of_dev ) { diff --git a/include/asm-ppc/macio.h b/include/asm-ppc/macio.h index a481b772d15..b553dd4b139 100644 --- a/include/asm-ppc/macio.h +++ b/include/asm-ppc/macio.h @@ -1,7 +1,6 @@ #ifndef __MACIO_ASIC_H__ #define __MACIO_ASIC_H__ -#include #include extern struct bus_type macio_bus_type; diff --git a/include/asm-ppc/of_device.h b/include/asm-ppc/of_device.h index 4b264cfd399..575bce418f8 100644 --- a/include/asm-ppc/of_device.h +++ b/include/asm-ppc/of_device.h @@ -2,6 +2,7 @@ #define __OF_DEVICE_H__ #include +#include #include /* @@ -55,7 +56,9 @@ extern int of_register_driver(struct of_platform_driver *drv); extern void of_unregister_driver(struct of_platform_driver *drv); extern int of_device_register(struct of_device *ofdev); extern void of_device_unregister(struct of_device *ofdev); -extern struct of_device *of_platform_device_create(struct device_node *np, const char *bus_id); +extern struct of_device *of_platform_device_create(struct device_node *np, + const char *bus_id, + struct device *parent); extern void of_release_dev(struct device *dev); #endif /* __OF_DEVICE_H__ */ diff --git a/include/asm-ppc64/smu.h b/include/asm-ppc64/smu.h index 10b4397af9a..dee8eefe47b 100644 --- a/include/asm-ppc64/smu.h +++ b/include/asm-ppc64/smu.h @@ -1,22 +1,379 @@ +#ifndef _SMU_H +#define _SMU_H + /* * Definitions for talking to the SMU chip in newer G5 PowerMacs */ #include +#include /* - * Basic routines for use by architecture. To be extended as - * we understand more of the chip + * Known SMU commands + * + * Most of what is below comes from looking at the Open Firmware driver, + * though this is still incomplete and could use better documentation here + * or there... + */ + + +/* + * Partition info commands + * + * I do not know what those are for at this point + */ +#define SMU_CMD_PARTITION_COMMAND 0x3e + + +/* + * Fan control + * + * This is a "mux" for fan control commands, first byte is the + * "sub" command. + */ +#define SMU_CMD_FAN_COMMAND 0x4a + + +/* + * Battery access + * + * Same command number as the PMU, could it be same syntax ? + */ +#define SMU_CMD_BATTERY_COMMAND 0x6f +#define SMU_CMD_GET_BATTERY_INFO 0x00 + +/* + * Real time clock control + * + * This is a "mux", first data byte contains the "sub" command. + * The "RTC" part of the SMU controls the date, time, powerup + * timer, but also a PRAM + * + * Dates are in BCD format on 7 bytes: + * [sec] [min] [hour] [weekday] [month day] [month] [year] + * with month being 1 based and year minus 100 + */ +#define SMU_CMD_RTC_COMMAND 0x8e +#define SMU_CMD_RTC_SET_PWRUP_TIMER 0x00 /* i: 7 bytes date */ +#define SMU_CMD_RTC_GET_PWRUP_TIMER 0x01 /* o: 7 bytes date */ +#define SMU_CMD_RTC_STOP_PWRUP_TIMER 0x02 +#define SMU_CMD_RTC_SET_PRAM_BYTE_ACC 0x20 /* i: 1 byte (address?) */ +#define SMU_CMD_RTC_SET_PRAM_AUTOINC 0x21 /* i: 1 byte (data?) */ +#define SMU_CMD_RTC_SET_PRAM_LO_BYTES 0x22 /* i: 10 bytes */ +#define SMU_CMD_RTC_SET_PRAM_HI_BYTES 0x23 /* i: 10 bytes */ +#define SMU_CMD_RTC_GET_PRAM_BYTE 0x28 /* i: 1 bytes (address?) */ +#define SMU_CMD_RTC_GET_PRAM_LO_BYTES 0x29 /* o: 10 bytes */ +#define SMU_CMD_RTC_GET_PRAM_HI_BYTES 0x2a /* o: 10 bytes */ +#define SMU_CMD_RTC_SET_DATETIME 0x80 /* i: 7 bytes date */ +#define SMU_CMD_RTC_GET_DATETIME 0x81 /* o: 7 bytes date */ + + /* + * i2c commands + * + * To issue an i2c command, first is to send a parameter block to the + * the SMU. This is a command of type 0x9a with 9 bytes of header + * eventually followed by data for a write: + * + * 0: bus number (from device-tree usually, SMU has lots of busses !) + * 1: transfer type/format (see below) + * 2: device address. For combined and combined4 type transfers, this + * is the "write" version of the address (bit 0x01 cleared) + * 3: subaddress length (0..3) + * 4: subaddress byte 0 (or only byte for subaddress length 1) + * 5: subaddress byte 1 + * 6: subaddress byte 2 + * 7: combined address (device address for combined mode data phase) + * 8: data length + * + * The transfer types are the same good old Apple ones it seems, + * that is: + * - 0x00: Simple transfer + * - 0x01: Subaddress transfer (addr write + data tx, no restart) + * - 0x02: Combined transfer (addr write + restart + data tx) + * + * This is then followed by actual data for a write. + * + * At this point, the OF driver seems to have a limitation on transfer + * sizes of 0xd bytes on reads and 0x5 bytes on writes. I do not know + * wether this is just an OF limit due to some temporary buffer size + * or if this is an SMU imposed limit. This driver has the same limitation + * for now as I use a 0x10 bytes temporary buffer as well + * + * Once that is completed, a response is expected from the SMU. This is + * obtained via a command of type 0x9a with a length of 1 byte containing + * 0 as the data byte. OF also fills the rest of the data buffer with 0xff's + * though I can't tell yet if this is actually necessary. Once this command + * is complete, at this point, all I can tell is what OF does. OF tests + * byte 0 of the reply: + * - on read, 0xfe or 0xfc : bus is busy, wait (see below) or nak ? + * - on read, 0x00 or 0x01 : reply is in buffer (after the byte 0) + * - on write, < 0 -> failure (immediate exit) + * - else, OF just exists (without error, weird) + * + * So on read, there is this wait-for-busy thing when getting a 0xfc or + * 0xfe result. OF does a loop of up to 64 retries, waiting 20ms and + * doing the above again until either the retries expire or the result + * is no longer 0xfe or 0xfc + * + * The Darwin I2C driver is less subtle though. On any non-success status + * from the response command, it waits 5ms and tries again up to 20 times, + * it doesn't differenciate between fatal errors or "busy" status. + * + * This driver provides an asynchronous paramblock based i2c command + * interface to be used either directly by low level code or by a higher + * level driver interfacing to the linux i2c layer. The current + * implementation of this relies on working timers & timer interrupts + * though, so be careful of calling context for now. This may be "fixed" + * in the future by adding a polling facility. + */ +#define SMU_CMD_I2C_COMMAND 0x9a + /* transfer types */ +#define SMU_I2C_TRANSFER_SIMPLE 0x00 +#define SMU_I2C_TRANSFER_STDSUB 0x01 +#define SMU_I2C_TRANSFER_COMBINED 0x02 + +/* + * Power supply control + * + * The "sub" command is an ASCII string in the data, the + * data lenght is that of the string. + * + * The VSLEW command can be used to get or set the voltage slewing. + * - lenght 5 (only "VSLEW") : it returns "DONE" and 3 bytes of + * reply at data offset 6, 7 and 8. + * - lenght 8 ("VSLEWxyz") has 3 additional bytes appended, and is + * used to set the voltage slewing point. The SMU replies with "DONE" + * I yet have to figure out their exact meaning of those 3 bytes in + * both cases. + * + */ +#define SMU_CMD_POWER_COMMAND 0xaa +#define SMU_CMD_POWER_RESTART "RESTART" +#define SMU_CMD_POWER_SHUTDOWN "SHUTDOWN" +#define SMU_CMD_POWER_VOLTAGE_SLEW "VSLEW" + +/* Misc commands + * + * This command seem to be a grab bag of various things + */ +#define SMU_CMD_MISC_df_COMMAND 0xdf +#define SMU_CMD_MISC_df_SET_DISPLAY_LIT 0x02 /* i: 1 byte */ +#define SMU_CMD_MISC_df_NMI_OPTION 0x04 + +/* + * Version info commands + * + * I haven't quite tried to figure out how these work + */ +#define SMU_CMD_VERSION_COMMAND 0xea + + +/* + * Misc commands + * + * This command seem to be a grab bag of various things + */ +#define SMU_CMD_MISC_ee_COMMAND 0xee +#define SMU_CMD_MISC_ee_GET_DATABLOCK_REC 0x02 +#define SMU_CMD_MISC_ee_LEDS_CTRL 0x04 /* i: 00 (00,01) [00] */ +#define SMU_CMD_MISC_ee_GET_DATA 0x05 /* i: 00 , o: ?? */ + + + +/* + * - Kernel side interface - + */ + +#ifdef __KERNEL__ + +/* + * Asynchronous SMU commands + * + * Fill up this structure and submit it via smu_queue_command(), + * and get notified by the optional done() callback, or because + * status becomes != 1 + */ + +struct smu_cmd; + +struct smu_cmd +{ + /* public */ + u8 cmd; /* command */ + int data_len; /* data len */ + int reply_len; /* reply len */ + void *data_buf; /* data buffer */ + void *reply_buf; /* reply buffer */ + int status; /* command status */ + void (*done)(struct smu_cmd *cmd, void *misc); + void *misc; + + /* private */ + struct list_head link; +}; + +/* + * Queues an SMU command, all fields have to be initialized + */ +extern int smu_queue_cmd(struct smu_cmd *cmd); + +/* + * Simple command wrapper. This structure embeds a small buffer + * to ease sending simple SMU commands from the stack + */ +struct smu_simple_cmd +{ + struct smu_cmd cmd; + u8 buffer[16]; +}; + +/* + * Queues a simple command. All fields will be initialized by that + * function + */ +extern int smu_queue_simple(struct smu_simple_cmd *scmd, u8 command, + unsigned int data_len, + void (*done)(struct smu_cmd *cmd, void *misc), + void *misc, + ...); + +/* + * Completion helper. Pass it to smu_queue_simple or as 'done' + * member to smu_queue_cmd, it will call complete() on the struct + * completion passed in the "misc" argument + */ +extern void smu_done_complete(struct smu_cmd *cmd, void *misc); + +/* + * Synchronous helpers. Will spin-wait for completion of a command + */ +extern void smu_spinwait_cmd(struct smu_cmd *cmd); + +static inline void smu_spinwait_simple(struct smu_simple_cmd *scmd) +{ + smu_spinwait_cmd(&scmd->cmd); +} + +/* + * Poll routine to call if blocked with irqs off + */ +extern void smu_poll(void); + + +/* + * Init routine, presence check.... */ extern int smu_init(void); extern int smu_present(void); +struct of_device; +extern struct of_device *smu_get_ofdev(void); + + +/* + * Common command wrappers + */ extern void smu_shutdown(void); extern void smu_restart(void); -extern int smu_get_rtc_time(struct rtc_time *time); -extern int smu_set_rtc_time(struct rtc_time *time); +struct rtc_time; +extern int smu_get_rtc_time(struct rtc_time *time, int spinwait); +extern int smu_set_rtc_time(struct rtc_time *time, int spinwait); /* * SMU command buffer absolute address, exported by pmac_setup, * this is allocated very early during boot. */ extern unsigned long smu_cmdbuf_abs; + + +/* + * Kenrel asynchronous i2c interface + */ + +/* SMU i2c header, exactly matches i2c header on wire */ +struct smu_i2c_param +{ + u8 bus; /* SMU bus ID (from device tree) */ + u8 type; /* i2c transfer type */ + u8 devaddr; /* device address (includes direction) */ + u8 sublen; /* subaddress length */ + u8 subaddr[3]; /* subaddress */ + u8 caddr; /* combined address, filled by SMU driver */ + u8 datalen; /* length of transfer */ + u8 data[7]; /* data */ +}; + +#define SMU_I2C_READ_MAX 0x0d +#define SMU_I2C_WRITE_MAX 0x05 + +struct smu_i2c_cmd +{ + /* public */ + struct smu_i2c_param info; + void (*done)(struct smu_i2c_cmd *cmd, void *misc); + void *misc; + int status; /* 1 = pending, 0 = ok, <0 = fail */ + + /* private */ + struct smu_cmd scmd; + int read; + int stage; + int retries; + u8 pdata[0x10]; + struct list_head link; +}; + +/* + * Call this to queue an i2c command to the SMU. You must fill info, + * including info.data for a write, done and misc. + * For now, no polling interface is provided so you have to use completion + * callback. + */ +extern int smu_queue_i2c(struct smu_i2c_cmd *cmd); + + +#endif /* __KERNEL__ */ + +/* + * - Userland interface - + */ + +/* + * A given instance of the device can be configured for 2 different + * things at the moment: + * + * - sending SMU commands (default at open() time) + * - receiving SMU events (not yet implemented) + * + * Commands are written with write() of a command block. They can be + * "driver" commands (for example to switch to event reception mode) + * or real SMU commands. They are made of a header followed by command + * data if any. + * + * For SMU commands (not for driver commands), you can then read() back + * a reply. The reader will be blocked or not depending on how the device + * file is opened. poll() isn't implemented yet. The reply will consist + * of a header as well, followed by the reply data if any. You should + * always provide a buffer large enough for the maximum reply data, I + * recommand one page. + * + * It is illegal to send SMU commands through a file descriptor configured + * for events reception + * + */ +struct smu_user_cmd_hdr +{ + __u32 cmdtype; +#define SMU_CMDTYPE_SMU 0 /* SMU command */ +#define SMU_CMDTYPE_WANTS_EVENTS 1 /* switch fd to events mode */ + + __u8 cmd; /* SMU command byte */ + __u32 data_len; /* Lenght of data following */ +}; + +struct smu_user_reply_hdr +{ + __u32 status; /* Command status */ + __u32 reply_len; /* Lenght of data follwing */ +}; + +#endif /* _SMU_H */ From e3fc4cc17ac4dbadab0ba0d5c8233d3b18170d4a Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Thu, 22 Sep 2005 21:44:07 -0700 Subject: [PATCH 161/204] [PATCH] corrections to top-level README Corrections to the recent top-level README changes. Signed-off-by: Randy Dunlap Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- README | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/README b/README index 2b5844d8cfa..d1edcc7adab 100644 --- a/README +++ b/README @@ -151,7 +151,7 @@ CONFIGURING the kernel: your existing ./.config file. "make silentoldconfig" Like above, but avoids cluttering the screen - with question already answered. + with questions already answered. NOTES on "make config": - having unnecessary drivers will make the kernel bigger, and can @@ -199,9 +199,9 @@ COMPILING the kernel: are installing a new kernel with the same version number as your working kernel, make a backup of your modules directory before you do a "make modules_install". - In alternative, before compiling, edit your Makefile and change the - "EXTRAVERSION" line - its content is appended to the regular kernel - version. + Alternatively, before compiling, use the kernel config option + "LOCALVERSION" to append a unique suffix to the regular kernel version. + LOCALVERSION can be set in the "General Setup" menu. - In order to boot your new kernel, you'll need to copy the kernel image (e.g. .../linux/arch/i386/boot/bzImage after compilation) From 98b4f5929ce5e616d0d0e13c71baf50083ab7397 Mon Sep 17 00:00:00 2001 From: Chuck Ebbert <76306.1226@compuserve.com> Date: Thu, 22 Sep 2005 21:44:08 -0700 Subject: [PATCH 162/204] [PATCH] atiixp_modem printk fixes Correctly identify atiixp_modem in its error messages. Signed-off-by: Chuck Ebbert <76306.1226@compuserve.com> Cc: Vojtech Pavlik Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- sound/pci/atiixp_modem.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/sound/pci/atiixp_modem.c b/sound/pci/atiixp_modem.c index 8a59598167f..c1a239a4dac 100644 --- a/sound/pci/atiixp_modem.c +++ b/sound/pci/atiixp_modem.c @@ -405,7 +405,7 @@ static int snd_atiixp_acquire_codec(atiixp_t *chip) while (atiixp_read(chip, PHYS_OUT_ADDR) & ATI_REG_PHYS_OUT_ADDR_EN) { if (! timeout--) { - snd_printk(KERN_WARNING "atiixp: codec acquire timeout\n"); + snd_printk(KERN_WARNING "atiixp-modem: codec acquire timeout\n"); return -EBUSY; } udelay(1); @@ -436,7 +436,7 @@ static unsigned short snd_atiixp_codec_read(atiixp_t *chip, unsigned short codec } while (--timeout); /* time out may happen during reset */ if (reg < 0x7c) - snd_printk(KERN_WARNING "atiixp: codec read timeout (reg %x)\n", reg); + snd_printk(KERN_WARNING "atiixp-modem: codec read timeout (reg %x)\n", reg); return 0xffff; } @@ -498,7 +498,7 @@ static int snd_atiixp_aclink_reset(atiixp_t *chip) do_delay(); atiixp_update(chip, CMD, ATI_REG_CMD_AC_RESET, ATI_REG_CMD_AC_RESET); if (--timeout) { - snd_printk(KERN_ERR "atiixp: codec reset timeout\n"); + snd_printk(KERN_ERR "atiixp-modem: codec reset timeout\n"); break; } } @@ -552,7 +552,7 @@ static int snd_atiixp_codec_detect(atiixp_t *chip) atiixp_write(chip, IER, 0); /* disable irqs */ if ((chip->codec_not_ready_bits & ALL_CODEC_NOT_READY) == ALL_CODEC_NOT_READY) { - snd_printk(KERN_ERR "atiixp: no codec detected!\n"); + snd_printk(KERN_ERR "atiixp-modem: no codec detected!\n"); return -ENXIO; } return 0; @@ -635,7 +635,7 @@ static void snd_atiixp_xrun_dma(atiixp_t *chip, atiixp_dma_t *dma) { if (! dma->substream || ! dma->running) return; - snd_printdd("atiixp: XRUN detected (DMA %d)\n", dma->ops->type); + snd_printdd("atiixp-modem: XRUN detected (DMA %d)\n", dma->ops->type); snd_pcm_stop(dma->substream, SNDRV_PCM_STATE_XRUN); } @@ -1081,14 +1081,14 @@ static int __devinit snd_atiixp_mixer_new(atiixp_t *chip, int clock) ac97.scaps = AC97_SCAP_SKIP_AUDIO; if ((err = snd_ac97_mixer(pbus, &ac97, &chip->ac97[i])) < 0) { chip->ac97[i] = NULL; /* to be sure */ - snd_printdd("atiixp: codec %d not available for modem\n", i); + snd_printdd("atiixp-modem: codec %d not available for modem\n", i); continue; } codec_count++; } if (! codec_count) { - snd_printk(KERN_ERR "atiixp: no codec available\n"); + snd_printk(KERN_ERR "atiixp-modem: no codec available\n"); return -ENODEV; } @@ -1159,7 +1159,7 @@ static void __devinit snd_atiixp_proc_init(atiixp_t *chip) { snd_info_entry_t *entry; - if (! snd_card_proc_new(chip->card, "atiixp", &entry)) + if (! snd_card_proc_new(chip->card, "atiixp-modem", &entry)) snd_info_set_text_ops(entry, chip, 1024, snd_atiixp_proc_read); } From 0cdc82ee1ae36c1ae4631fc3a782faeb0b90df38 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Thu, 22 Sep 2005 21:44:09 -0700 Subject: [PATCH 163/204] [PATCH] mesh scsi: fix error handling The PowerMac mesh SCSI driver had some missing error handling which would trigger warnings due to lack of handling of return value from scsi_add_host. This fixes it. Signed-off-by: Benjamin Herrenschmidt Cc: James Bottomley Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/scsi/mesh.c | 29 +++++++++++++++++++++-------- 1 file changed, 21 insertions(+), 8 deletions(-) diff --git a/drivers/scsi/mesh.c b/drivers/scsi/mesh.c index a4857db4f9b..b235556b7b6 100644 --- a/drivers/scsi/mesh.c +++ b/drivers/scsi/mesh.c @@ -1959,22 +1959,35 @@ static int mesh_probe(struct macio_dev *mdev, const struct of_device_id *match) /* Set it up */ mesh_init(ms); - /* XXX FIXME: error should be fatal */ - if (request_irq(ms->meshintr, do_mesh_interrupt, 0, "MESH", ms)) + /* Request interrupt */ + if (request_irq(ms->meshintr, do_mesh_interrupt, 0, "MESH", ms)) { printk(KERN_ERR "MESH: can't get irq %d\n", ms->meshintr); + goto out_shutdown; + } - /* XXX FIXME: handle failure */ - scsi_add_host(mesh_host, &mdev->ofdev.dev); + /* Add scsi host & scan */ + if (scsi_add_host(mesh_host, &mdev->ofdev.dev)) + goto out_release_irq; scsi_scan_host(mesh_host); return 0; -out_unmap: + out_release_irq: + free_irq(ms->meshintr, ms); + out_shutdown: + /* shutdown & reset bus in case of error or macos can be confused + * at reboot if the bus was set to synchronous mode already + */ + mesh_shutdown(mdev); + set_mesh_power(ms, 0); + pci_free_consistent(macio_get_pci_dev(mdev), ms->dma_cmd_size, + ms->dma_cmd_space, ms->dma_cmd_bus); + out_unmap: iounmap(ms->dma); iounmap(ms->mesh); -out_free: + out_free: scsi_host_put(mesh_host); -out_release: + out_release: macio_release_resources(mdev); return -ENODEV; @@ -2001,7 +2014,7 @@ static int mesh_remove(struct macio_dev *mdev) /* Free DMA commands memory */ pci_free_consistent(macio_get_pci_dev(mdev), ms->dma_cmd_size, - ms->dma_cmd_space, ms->dma_cmd_bus); + ms->dma_cmd_space, ms->dma_cmd_bus); /* Release memory resources */ macio_release_resources(mdev); From eafb42707b21beb42bba4eae7b742f837ee9d2e0 Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Thu, 22 Sep 2005 21:44:10 -0700 Subject: [PATCH 164/204] [PATCH] __kmalloc: Generate BUG if size requested is too large. I had an issue on ia64 where I got a bug in kernel/workqueue because kzalloc returned a NULL pointer due to the task structure getting too big for the slab allocator. Usually these cases are caught by the kmalloc macro in include/linux/slab.h. Compilation will fail if a too big value is passed to kmalloc. However, kzalloc uses __kmalloc which has no check for that. This patch makes __kmalloc bug if a too large entity is requested. Signed-off-by: Christoph Lameter Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/slab.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/mm/slab.c b/mm/slab.c index c9adfce0040..ad113c431e3 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -2907,8 +2907,7 @@ void *__kmalloc(size_t size, unsigned int __nocast flags) * functions. */ cachep = __find_general_cachep(size, flags); - if (unlikely(cachep == NULL)) - return NULL; + BUG_ON(!cachep); /* Allocation size too large for kmalloc */ return __cache_alloc(cachep, flags); } EXPORT_SYMBOL(__kmalloc); From 8686bcd0a5f5e3f599ed9f1028ec9e449e7b87e3 Mon Sep 17 00:00:00 2001 From: Pavel Machek Date: Thu, 22 Sep 2005 21:44:11 -0700 Subject: [PATCH 165/204] [PATCH] swsusp: fix comments Fix comments in swsusp. Signed-off-by: Pavel Machek Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/power/power.h | 2 +- kernel/power/swsusp.c | 10 +++++++--- 2 files changed, 8 insertions(+), 4 deletions(-) diff --git a/kernel/power/power.h b/kernel/power/power.h index cd6a3493cc0..9c9167d910d 100644 --- a/kernel/power/power.h +++ b/kernel/power/power.h @@ -1,7 +1,7 @@ #include #include -/* With SUSPEND_CONSOLE defined, it suspend looks *really* cool, but +/* With SUSPEND_CONSOLE defined suspend looks *really* cool, but we probably do not take enough locks for switching consoles, etc, so bad things might happen. */ diff --git a/kernel/power/swsusp.c b/kernel/power/swsusp.c index 353f37f1ba2..1cc9ff25e47 100644 --- a/kernel/power/swsusp.c +++ b/kernel/power/swsusp.c @@ -363,7 +363,7 @@ static void lock_swapdevices(void) } /** - * write_swap_page - Write one page to a fresh swap location. + * write_page - Write one page to a fresh swap location. * @addr: Address we're writing. * @loc: Place to store the entry we used. * @@ -863,6 +863,9 @@ static int alloc_image_pages(void) return 0; } +/* Free pages we allocated for suspend. Suspend pages are alocated + * before atomic copy, so we need to free them after resume. + */ void swsusp_free(void) { BUG_ON(PageNosave(virt_to_page(pagedir_save))); @@ -1213,8 +1216,9 @@ static struct pbe * swsusp_pagedir_relocate(struct pbe *pblist) free_pagedir(pblist); free_eaten_memory(); pblist = NULL; - } - else + /* Is this even worth handling? It should never ever happen, and we + have just lost user's state, anyway... */ + } else printk("swsusp: Relocated %d pages\n", rel); return pblist; From 3a02d6c051cf3c214aab38e4330b4bcec5f6e3f8 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Thu, 22 Sep 2005 21:44:12 -0700 Subject: [PATCH 166/204] [PATCH] uml: don't remove umid files in conflict case Only remove the UML pidfile and management socket if we created them. Currently in case two UMLs are started with the same umid, the second will remove the first's ones. Probably we should also panic() at that point, not sure however. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/kernel/umid.c | 30 +++++++++++++++++++----------- 1 file changed, 19 insertions(+), 11 deletions(-) diff --git a/arch/um/kernel/umid.c b/arch/um/kernel/umid.c index 186c2888501..251d277fff2 100644 --- a/arch/um/kernel/umid.c +++ b/arch/um/kernel/umid.c @@ -31,6 +31,8 @@ static char *uml_dir = UML_DIR; /* Changed by set_umid */ static int umid_is_random = 1; static int umid_inited = 0; +/* Have we created the files? Should we remove them? */ +static int umid_owned = 0; static int make_umid(int (*printer)(const char *fmt, ...)); @@ -82,20 +84,21 @@ int __init umid_file_name(char *name, char *buf, int len) extern int tracing_pid; -static int __init create_pid_file(void) +static void __init create_pid_file(void) { char file[strlen(uml_dir) + UMID_LEN + sizeof("/pid\0")]; char pid[sizeof("nnnnn\0")]; int fd, n; - if(umid_file_name("pid", file, sizeof(file))) return 0; + if(umid_file_name("pid", file, sizeof(file))) + return; fd = os_open_file(file, of_create(of_excl(of_rdwr(OPENFLAGS()))), 0644); if(fd < 0){ printf("Open of machine pid file \"%s\" failed: %s\n", file, strerror(-fd)); - return 0; + return; } sprintf(pid, "%d\n", os_getpid()); @@ -103,7 +106,6 @@ static int __init create_pid_file(void) if(n != strlen(pid)) printf("Write of pid file failed - err = %d\n", -n); os_close_file(fd); - return 0; } static int actually_do_remove(char *dir) @@ -147,7 +149,8 @@ static int actually_do_remove(char *dir) void remove_umid_dir(void) { char dir[strlen(uml_dir) + UMID_LEN + 1]; - if(!umid_inited) return; + if (!umid_owned) + return; sprintf(dir, "%s%s", uml_dir, umid); actually_do_remove(dir); @@ -155,11 +158,12 @@ void remove_umid_dir(void) char *get_umid(int only_if_set) { - if(only_if_set && umid_is_random) return(NULL); - return(umid); + if(only_if_set && umid_is_random) + return NULL; + return umid; } -int not_dead_yet(char *dir) +static int not_dead_yet(char *dir) { char file[strlen(uml_dir) + UMID_LEN + sizeof("/pid\0")]; char pid[sizeof("nnnnn\0")], *end; @@ -193,7 +197,8 @@ int not_dead_yet(char *dir) (p == CHOOSE_MODE(tracing_pid, os_getpid()))) dead = 1; } - if(!dead) return(1); + if(!dead) + return(1); return(actually_do_remove(dir)); } @@ -286,6 +291,7 @@ static int __init make_umid(int (*printer)(const char *fmt, ...)) if(errno == EEXIST){ if(not_dead_yet(tmp)){ (*printer)("umid '%s' is in use\n", umid); + umid_owned = 0; return(-1); } err = mkdir(tmp, 0777); @@ -296,7 +302,8 @@ static int __init make_umid(int (*printer)(const char *fmt, ...)) return(-1); } - return(0); + umid_owned = 1; + return 0; } __uml_setup("uml_dir=", set_uml_dir, @@ -309,7 +316,8 @@ static int __init make_umid_setup(void) /* one function with the ordering we need ... */ make_uml_dir(); make_umid(printf); - return create_pid_file(); + create_pid_file(); + return 0; } __uml_postsetup(make_umid_setup); From a8bfb94c58238666df0d6856861d18f0f52fc752 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Thu, 22 Sep 2005 21:44:14 -0700 Subject: [PATCH 167/204] [PATCH] strlcat: use for uml umid.c Simplify the code by using strlcat() instead of strncat() and manual appending. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/include/user.h | 4 +++- arch/um/kernel/umid.c | 11 ++++------- 2 files changed, 7 insertions(+), 8 deletions(-) diff --git a/arch/um/include/user.h b/arch/um/include/user.h index 57ee9e26122..0f865ef4691 100644 --- a/arch/um/include/user.h +++ b/arch/um/include/user.h @@ -14,7 +14,9 @@ extern void *um_kmalloc_atomic(int size); extern void kfree(void *ptr); extern int in_aton(char *str); extern int open_gdb_chan(void); -extern int strlcpy(char *, const char *, int); +/* These use size_t, however unsigned long is correct on both i386 and x86_64. */ +extern unsigned long strlcpy(char *, const char *, unsigned long); +extern unsigned long strlcat(char *, const char *, unsigned long); extern void *um_vmalloc(int size); extern void vfree(void *ptr); diff --git a/arch/um/kernel/umid.c b/arch/um/kernel/umid.c index 251d277fff2..0b21d59ba0c 100644 --- a/arch/um/kernel/umid.c +++ b/arch/um/kernel/umid.c @@ -237,16 +237,13 @@ static int __init make_uml_dir(void) strlcpy(dir, home, sizeof(dir)); uml_dir++; } + strlcat(dir, uml_dir, sizeof(dir)); len = strlen(dir); - strncat(dir, uml_dir, sizeof(dir) - len); - len = strlen(dir); - if((len > 0) && (len < sizeof(dir) - 1) && (dir[len - 1] != '/')){ - dir[len] = '/'; - dir[len + 1] = '\0'; - } + if (len > 0 && dir[len - 1] != '/') + strlcat(dir, "/", sizeof(dir)); uml_dir = malloc(strlen(dir) + 1); - if(uml_dir == NULL){ + if (uml_dir == NULL) { printf("make_uml_dir : malloc failed, errno = %d\n", errno); exit(1); } From 69e1e688f5698287b45fbff22a01de91b20804cd Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Thu, 22 Sep 2005 21:44:15 -0700 Subject: [PATCH 168/204] [PATCH] uml: don't redundantly mark pte as newpage in pte_modify pte_modify marks a page as needing flush, which is redundant because the resulting PTE is still set with set_pte, which already handles that. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/asm-um/pgtable.h | 1 - 1 file changed, 1 deletion(-) diff --git a/include/asm-um/pgtable.h b/include/asm-um/pgtable.h index ed06170e0ed..616d02b57ea 100644 --- a/include/asm-um/pgtable.h +++ b/include/asm-um/pgtable.h @@ -346,7 +346,6 @@ static inline void set_pte(pte_t *pteptr, pte_t pteval) static inline pte_t pte_modify(pte_t pte, pgprot_t newprot) { pte_set_val(pte, (pte_val(pte) & _PAGE_CHG_MASK), newprot); - if(pte_present(pte)) pte = pte_mknewpage(pte_mknewprot(pte)); return pte; } From 546fe1cbf91d4d62e3849517c31a2327c992e5c5 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Thu, 22 Sep 2005 21:44:16 -0700 Subject: [PATCH 169/204] [PATCH] uml: fix hang in TT mode on fault The current code doesn't handle well general protection faults on the host - it thinks that cr2 is always the address of a page fault. While actually, on general protection faults, that address is not accessible, so we'd better assume we couldn't satisfy the fault. Currently instead we think we've fixed it, so we go back, retry the instruction and fault again endlessly. This leads to the kernel hanging when doing copy_from_user(dest, -1, ...) in TT mode, since reading *(-1) causes a GFP, and we don't support kernel preemption. Thanks to Luo Xin for testing UML with LTP and reporting the failures he got. Cc: Luo Xin Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/kernel/trap_kern.c | 11 ++++++++++- arch/um/kernel/tt/uaccess_user.c | 11 +++++++++-- 2 files changed, 19 insertions(+), 3 deletions(-) diff --git a/arch/um/kernel/trap_kern.c b/arch/um/kernel/trap_kern.c index 87cc6fd76ce..90690b9b1f3 100644 --- a/arch/um/kernel/trap_kern.c +++ b/arch/um/kernel/trap_kern.c @@ -18,6 +18,7 @@ #include "asm/a.out.h" #include "asm/current.h" #include "asm/irq.h" +#include "sysdep/sigcontext.h" #include "user_util.h" #include "kern_util.h" #include "kern.h" @@ -125,7 +126,15 @@ unsigned long segv(struct faultinfo fi, unsigned long ip, int is_user, void *sc) } else if(current->mm == NULL) panic("Segfault with no mm"); - err = handle_page_fault(address, ip, is_write, is_user, &si.si_code); + + if (SEGV_IS_FIXABLE(&fi)) + err = handle_page_fault(address, ip, is_write, is_user, &si.si_code); + else { + err = -EFAULT; + /* A thread accessed NULL, we get a fault, but CR2 is invalid. + * This code is used in __do_copy_from_user() of TT mode. */ + address = 0; + } catcher = current->thread.fault_catcher; if(!err) diff --git a/arch/um/kernel/tt/uaccess_user.c b/arch/um/kernel/tt/uaccess_user.c index f01475512ec..8c220f054b6 100644 --- a/arch/um/kernel/tt/uaccess_user.c +++ b/arch/um/kernel/tt/uaccess_user.c @@ -22,8 +22,15 @@ int __do_copy_from_user(void *to, const void *from, int n, __do_copy, &faulted); TASK_REGS(get_current())->tt = save; - if(!faulted) return(0); - else return(n - (fault - (unsigned long) from)); + if(!faulted) + return 0; + else if (fault) + return n - (fault - (unsigned long) from); + else + /* In case of a general protection fault, we don't have the + * fault address, so NULL is used instead. Pretend we didn't + * copy anything. */ + return n; } static void __do_strncpy(void *dst, const void *src, int count) From dc36abf34e13aabbe06cdd060c434b3cdd80e7da Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Thu, 22 Sep 2005 21:44:18 -0700 Subject: [PATCH 170/204] [PATCH] uml: fix condition in tlb flush Avoid setting w = 0 twice. Spotted this (trivial) thing which is needed for another patch. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/kernel/tlb.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/arch/um/kernel/tlb.c b/arch/um/kernel/tlb.c index 0a562c3c0fd..f5b0636f9ad 100644 --- a/arch/um/kernel/tlb.c +++ b/arch/um/kernel/tlb.c @@ -193,12 +193,12 @@ void fix_range_common(struct mm_struct *mm, unsigned long start_addr, r = pte_read(*npte); w = pte_write(*npte); x = pte_exec(*npte); - if(!pte_dirty(*npte)) - w = 0; - if(!pte_young(*npte)){ - r = 0; - w = 0; - } + if (!pte_young(*npte)) { + r = 0; + w = 0; + } else if (!pte_dirty(*npte)) { + w = 0; + } if(force || pte_newpage(*npte)){ if(pte_present(*npte)) ret = add_mmap(addr, From 12ebcd73e40e09f0dfddf89e465cc0541e0ff8b1 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Thu, 22 Sep 2005 21:44:19 -0700 Subject: [PATCH 171/204] [PATCH] uml: run mconsole "sysrq" in process context Things are breaking horribly with sysrq called in interrupt context. I want to try to fix it, but probably this is simpler. To tell the truth, sysrq is normally run in interrupt context, so there shouldn't be any problem. There's also a warning from the fault handler because it's run in atomic context (I have a patch for that, only I deferred it). This is why I'm doing this. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/drivers/mconsole_user.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/um/drivers/mconsole_user.c b/arch/um/drivers/mconsole_user.c index 310c1f823f2..04383f98f4d 100644 --- a/arch/um/drivers/mconsole_user.c +++ b/arch/um/drivers/mconsole_user.c @@ -23,7 +23,7 @@ static struct mconsole_command commands[] = { { "reboot", mconsole_reboot, MCONSOLE_PROC }, { "config", mconsole_config, MCONSOLE_PROC }, { "remove", mconsole_remove, MCONSOLE_PROC }, - { "sysrq", mconsole_sysrq, MCONSOLE_INTR }, + { "sysrq", mconsole_sysrq, MCONSOLE_PROC }, { "help", mconsole_help, MCONSOLE_INTR }, { "cad", mconsole_cad, MCONSOLE_INTR }, { "stop", mconsole_stop, MCONSOLE_PROC }, From fea03cb40a15e3c6647c0927300202e062e6792a Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Thu, 22 Sep 2005 21:44:20 -0700 Subject: [PATCH 172/204] [PATCH] uml: avoid fixing faults while atomic Following i386, we should maybe refuse trying to fault in pages when we're doing atomic operations, because to handle the fault we could need to take already taken spinlocks. Also, if we're doing an atomic operation (in the sense of in_atomic()) we're surely in kernel mode and we're surely going to handle adequately the failed fault, so it's safe to behave this way. Currently, on UML SMP is rarely used, and we don't support PREEMPT, so this is unlikely to create problems right now, but it might in the future. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/kernel/trap_kern.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/arch/um/kernel/trap_kern.c b/arch/um/kernel/trap_kern.c index 90690b9b1f3..d297429ac36 100644 --- a/arch/um/kernel/trap_kern.c +++ b/arch/um/kernel/trap_kern.c @@ -40,6 +40,12 @@ int handle_page_fault(unsigned long address, unsigned long ip, int err = -EFAULT; *code_out = SEGV_MAPERR; + + /* If the fault was during atomic operation, don't take the fault, just + * fail. */ + if (in_atomic()) + goto out_nosemaphore; + down_read(&mm->mmap_sem); vma = find_vma(mm, address); if(!vma) @@ -90,6 +96,7 @@ survive: flush_tlb_page(vma, address); out: up_read(&mm->mmap_sem); +out_nosemaphore: return(err); /* From 46db4a42dd1190a311c2fb45106dfd0842c65a94 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Thu, 22 Sep 2005 21:44:20 -0700 Subject: [PATCH 173/204] [PATCH] uml: Fix GFP_ flags usage GFP_ATOMIC | GFP_KERNEL is meaningless and won't work. Actually it never worked, even in 2.4. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/kernel/process_kern.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/um/kernel/process_kern.c b/arch/um/kernel/process_kern.c index 39cf568ccfa..ea65db679e9 100644 --- a/arch/um/kernel/process_kern.c +++ b/arch/um/kernel/process_kern.c @@ -82,7 +82,8 @@ unsigned long alloc_stack(int order, int atomic) unsigned long page; int flags = GFP_KERNEL; - if(atomic) flags |= GFP_ATOMIC; + if (atomic) + flags = GFP_ATOMIC; page = __get_free_pages(flags, order); if(page == 0) return(0); From 79ae2cb856ed6a8c48f455d52d5ed5960c671e67 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Thu, 22 Sep 2005 21:44:21 -0700 Subject: [PATCH 174/204] [PATCH] uml: use GFP_ATOMIC for allocations under spinlocks. setup_initial_poll is only called with sigio_lock() held, so use appropriate allocation. Also, parse_chan() can also be called when holding a spinlock (see line_open() -> parse_chan_pair()). I have sporadic problems (spinlock taken twice, with spinlock debugging on UP) which could be caused by a sequence like "take spinlock, alloc and go to sleep, take again the spinlock in the other thread". Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/drivers/chan_kern.c | 2 +- arch/um/kernel/sigio_user.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/um/drivers/chan_kern.c b/arch/um/drivers/chan_kern.c index 14a12d6b3df..244e54a2153 100644 --- a/arch/um/drivers/chan_kern.c +++ b/arch/um/drivers/chan_kern.c @@ -465,7 +465,7 @@ static struct chan *parse_chan(char *str, int pri, int device, data = (*ops->init)(str, device, opts); if(data == NULL) return(NULL); - chan = kmalloc(sizeof(*chan), GFP_KERNEL); + chan = kmalloc(sizeof(*chan), GFP_ATOMIC); if(chan == NULL) return(NULL); *chan = ((struct chan) { .list = LIST_HEAD_INIT(chan->list), .primary = 1, diff --git a/arch/um/kernel/sigio_user.c b/arch/um/kernel/sigio_user.c index e89218958f3..a52751108aa 100644 --- a/arch/um/kernel/sigio_user.c +++ b/arch/um/kernel/sigio_user.c @@ -340,7 +340,7 @@ static int setup_initial_poll(int fd) { struct pollfd *p; - p = um_kmalloc(sizeof(struct pollfd)); + p = um_kmalloc_atomic(sizeof(struct pollfd)); if(p == NULL){ printk("setup_initial_poll : failed to allocate poll\n"); return(-1); From fac97ae0b1a206e2952baf1f9eb46305d673adc6 Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Thu, 22 Sep 2005 21:44:22 -0700 Subject: [PATCH 175/204] [PATCH] uml: replace printk with "stack-friendly" printf - to report console failure User get *a lot* confused when consoles don't work but we don't report anything. And, as reported in the comment, using printk to report "your console doesn't work" isn't likely to go that far. Fix the problem on the base of this: stack consumption by host printf(). Use kernel sprintf() and os_write_file, using a wild guess that one page will be enough for the message, to preallocate the buffer with kmalloc(). Signed-off-by: Paolo 'Blaisorblade' Giarrusso Cc: Jeff Dike Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/um/drivers/chan_kern.c | 58 +++++++++++++++++++++++++++---------- 1 file changed, 42 insertions(+), 16 deletions(-) diff --git a/arch/um/drivers/chan_kern.c b/arch/um/drivers/chan_kern.c index 244e54a2153..16e7dc89f61 100644 --- a/arch/um/drivers/chan_kern.c +++ b/arch/um/drivers/chan_kern.c @@ -19,18 +19,44 @@ #include "line.h" #include "os.h" -#ifdef CONFIG_NOCONFIG_CHAN +/* XXX: could well be moved to somewhere else, if needed. */ +static int my_printf(const char * fmt, ...) + __attribute__ ((format (printf, 1, 2))); -/* The printk's here are wrong because we are complaining that there is no - * output device, but printk is printing to that output device. The user will - * never see the error. printf would be better, except it can't run on a - * kernel stack because it will overflow it. - * Use printk for now since that will avoid crashing. - */ +static int my_printf(const char * fmt, ...) +{ + /* Yes, can be called on atomic context.*/ + char *buf = kmalloc(4096, GFP_ATOMIC); + va_list args; + int r; + + if (!buf) { + /* We print directly fmt. + * Yes, yes, yes, feel free to complain. */ + r = strlen(fmt); + } else { + va_start(args, fmt); + r = vsprintf(buf, fmt, args); + va_end(args); + fmt = buf; + } + + if (r) + r = os_write_file(1, fmt, r); + return r; + +} + +#ifdef CONFIG_NOCONFIG_CHAN +/* Despite its name, there's no added trailing newline. */ +static int my_puts(const char * buf) +{ + return os_write_file(1, buf, strlen(buf)); +} static void *not_configged_init(char *str, int device, struct chan_opts *opts) { - printk(KERN_ERR "Using a channel type which is configured out of " + my_puts("Using a channel type which is configured out of " "UML\n"); return(NULL); } @@ -38,27 +64,27 @@ static void *not_configged_init(char *str, int device, struct chan_opts *opts) static int not_configged_open(int input, int output, int primary, void *data, char **dev_out) { - printk(KERN_ERR "Using a channel type which is configured out of " + my_puts("Using a channel type which is configured out of " "UML\n"); return(-ENODEV); } static void not_configged_close(int fd, void *data) { - printk(KERN_ERR "Using a channel type which is configured out of " + my_puts("Using a channel type which is configured out of " "UML\n"); } static int not_configged_read(int fd, char *c_out, void *data) { - printk(KERN_ERR "Using a channel type which is configured out of " + my_puts("Using a channel type which is configured out of " "UML\n"); return(-EIO); } static int not_configged_write(int fd, const char *buf, int len, void *data) { - printk(KERN_ERR "Using a channel type which is configured out of " + my_puts("Using a channel type which is configured out of " "UML\n"); return(-EIO); } @@ -66,7 +92,7 @@ static int not_configged_write(int fd, const char *buf, int len, void *data) static int not_configged_console_write(int fd, const char *buf, int len, void *data) { - printk(KERN_ERR "Using a channel type which is configured out of " + my_puts("Using a channel type which is configured out of " "UML\n"); return(-EIO); } @@ -74,14 +100,14 @@ static int not_configged_console_write(int fd, const char *buf, int len, static int not_configged_window_size(int fd, void *data, unsigned short *rows, unsigned short *cols) { - printk(KERN_ERR "Using a channel type which is configured out of " + my_puts("Using a channel type which is configured out of " "UML\n"); return(-ENODEV); } static void not_configged_free(void *data) { - printf(KERN_ERR "Using a channel type which is configured out of " + my_puts("Using a channel type which is configured out of " "UML\n"); } @@ -457,7 +483,7 @@ static struct chan *parse_chan(char *str, int pri, int device, } } if(ops == NULL){ - printk(KERN_ERR "parse_chan couldn't parse \"%s\"\n", + my_printf("parse_chan couldn't parse \"%s\"\n", str); return(NULL); } From 288a60cf4d7cc35f84f46cd8ffd0b34f9d8e7346 Mon Sep 17 00:00:00 2001 From: Chris Zankel Date: Thu, 22 Sep 2005 21:44:23 -0700 Subject: [PATCH 176/204] [PATCH] xtensa: remove io_remap_page_range and minor clean-ups Remove io_remap_page_range() from all of Linux 2.6.x (as requested and suggested by Randy Dunlap) and minor clean-ups. Signed-off-by: Chris Zankel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/xtensa/kernel/pci.c | 4 +-- arch/xtensa/kernel/platform.c | 2 +- arch/xtensa/kernel/process.c | 2 +- arch/xtensa/kernel/setup.c | 2 +- arch/xtensa/kernel/signal.c | 2 +- arch/xtensa/kernel/time.c | 2 +- arch/xtensa/mm/init.c | 2 +- include/asm-xtensa/atomic.h | 2 +- include/asm-xtensa/bitops.h | 2 +- include/asm-xtensa/hardirq.h | 1 + include/asm-xtensa/semaphore.h | 49 ++++++++-------------------------- include/asm-xtensa/system.h | 16 ----------- 12 files changed, 22 insertions(+), 64 deletions(-) diff --git a/arch/xtensa/kernel/pci.c b/arch/xtensa/kernel/pci.c index 09887c96e9a..de19501aa80 100644 --- a/arch/xtensa/kernel/pci.c +++ b/arch/xtensa/kernel/pci.c @@ -402,8 +402,8 @@ int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, __pci_mmap_set_flags(dev, vma, mmap_state); __pci_mmap_set_pgprot(dev, vma, mmap_state, write_combine); - ret = io_remap_page_range(vma, vma->vm_start, vma->vm_pgoff<vm_end - vma->vm_start, vma->vm_page_prot); + ret = io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, + vma->vm_end - vma->vm_start,vma->vm_page_prot); return ret; } diff --git a/arch/xtensa/kernel/platform.c b/arch/xtensa/kernel/platform.c index cf136278444..03674daabc6 100644 --- a/arch/xtensa/kernel/platform.c +++ b/arch/xtensa/kernel/platform.c @@ -39,7 +39,7 @@ _F(int, pcibios_fixup, (void), { return 0; }); _F(int, get_rtc_time, (time_t* t), { return 0; }); _F(int, set_rtc_time, (time_t t), { return 0; }); -#if CONFIG_XTENSA_CALIBRATE_CCOUNT +#ifdef CONFIG_XTENSA_CALIBRATE_CCOUNT _F(void, calibrate_ccount, (void), { printk ("ERROR: Cannot calibrate cpu frequency! Assuming 100MHz.\n"); diff --git a/arch/xtensa/kernel/process.c b/arch/xtensa/kernel/process.c index c83bb0d4178..08ef6d82ee5 100644 --- a/arch/xtensa/kernel/process.c +++ b/arch/xtensa/kernel/process.c @@ -457,7 +457,7 @@ int dump_task_fpu(struct pt_regs *regs, struct task_struct *task, elf_fpregset_t *r) { /* see asm/coprocessor.h for this magic number 16 */ -#if TOTAL_CPEXTRA_SIZE > 16 +#if XTENSA_CP_EXTRA_SIZE > 16 do_save_fpregs (r, regs, task); /* For now, bit 16 means some extra state may be present: */ diff --git a/arch/xtensa/kernel/setup.c b/arch/xtensa/kernel/setup.c index 1f5bf5d624e..513ed8d6776 100644 --- a/arch/xtensa/kernel/setup.c +++ b/arch/xtensa/kernel/setup.c @@ -304,7 +304,7 @@ void __init setup_arch(char **cmdline_p) # endif #endif -#if CONFIG_PCI +#ifdef CONFIG_PCI platform_pcibios_init(); #endif } diff --git a/arch/xtensa/kernel/signal.c b/arch/xtensa/kernel/signal.c index dc42cede939..e252b61e45a 100644 --- a/arch/xtensa/kernel/signal.c +++ b/arch/xtensa/kernel/signal.c @@ -182,7 +182,7 @@ restore_cpextra (struct _cpstate *buf) struct task_struct *tsk = current; release_all_cp(tsk); - return __copy_from_user(tsk->thread.cpextra, buf, TOTAL_CPEXTRA_SIZE); + return __copy_from_user(tsk->thread.cpextra, buf, XTENSA_CP_EXTRA_SIZE); #endif return 0; } diff --git a/arch/xtensa/kernel/time.c b/arch/xtensa/kernel/time.c index 1ac7d5ce745..8e423d1335c 100644 --- a/arch/xtensa/kernel/time.c +++ b/arch/xtensa/kernel/time.c @@ -68,7 +68,7 @@ void __init time_init(void) * speed for the CALIBRATE. */ -#if CONFIG_XTENSA_CALIBRATE_CCOUNT +#ifdef CONFIG_XTENSA_CALIBRATE_CCOUNT printk("Calibrating CPU frequency "); platform_calibrate_ccount(); printk("%d.%02d MHz\n", (int)ccount_per_jiffy/(1000000/HZ), diff --git a/arch/xtensa/mm/init.c b/arch/xtensa/mm/init.c index 56aace84aae..5a91d6c9e66 100644 --- a/arch/xtensa/mm/init.c +++ b/arch/xtensa/mm/init.c @@ -239,7 +239,7 @@ void __init mem_init(void) high_memory = (void *) __va(max_mapnr << PAGE_SHIFT); highmemsize = 0; -#if CONFIG_HIGHMEM +#ifdef CONFIG_HIGHMEM #error HIGHGMEM not implemented in init.c #endif diff --git a/include/asm-xtensa/atomic.h b/include/asm-xtensa/atomic.h index 24f86f0e43c..12b5732dc6e 100644 --- a/include/asm-xtensa/atomic.h +++ b/include/asm-xtensa/atomic.h @@ -22,7 +22,7 @@ typedef struct { volatile int counter; } atomic_t; #include #include -#define ATOMIC_INIT(i) ( (atomic_t) { (i) } ) +#define ATOMIC_INIT(i) { (i) } /* * This Xtensa implementation assumes that the right mechanism diff --git a/include/asm-xtensa/bitops.h b/include/asm-xtensa/bitops.h index d395ef226c3..e76ee889e21 100644 --- a/include/asm-xtensa/bitops.h +++ b/include/asm-xtensa/bitops.h @@ -174,7 +174,7 @@ static __inline__ int test_bit(int nr, const volatile void *addr) return 1UL & (((const volatile unsigned int *)addr)[nr>>5] >> (nr&31)); } -#if XCHAL_HAVE_NSAU +#if XCHAL_HAVE_NSA static __inline__ int __cntlz (unsigned long x) { diff --git a/include/asm-xtensa/hardirq.h b/include/asm-xtensa/hardirq.h index e07c76c36b9..aa9c1adf68d 100644 --- a/include/asm-xtensa/hardirq.h +++ b/include/asm-xtensa/hardirq.h @@ -23,6 +23,7 @@ typedef struct { unsigned int __nmi_count; /* arch dependent */ } ____cacheline_aligned irq_cpustat_t; +void ack_bad_irq(unsigned int irq); #include /* Standard mappings for irq_cpustat_t above */ #endif /* _XTENSA_HARDIRQ_H */ diff --git a/include/asm-xtensa/semaphore.h b/include/asm-xtensa/semaphore.h index db740b8bc6f..09e89ab3eb6 100644 --- a/include/asm-xtensa/semaphore.h +++ b/include/asm-xtensa/semaphore.h @@ -20,28 +20,19 @@ struct semaphore { atomic_t count; int sleepers; wait_queue_head_t wait; -#if WAITQUEUE_DEBUG - long __magic; -#endif }; -#if WAITQUEUE_DEBUG -# define __SEM_DEBUG_INIT(name) \ - , (int)&(name).__magic -#else -# define __SEM_DEBUG_INIT(name) -#endif +#define __SEMAPHORE_INITIALIZER(name,n) \ +{ \ + .count = ATOMIC_INIT(n), \ + .sleepers = 0, \ + .wait = __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ +} -#define __SEMAPHORE_INITIALIZER(name,count) \ - { ATOMIC_INIT(count), \ - 0, \ - __WAIT_QUEUE_HEAD_INITIALIZER((name).wait) \ - __SEM_DEBUG_INIT(name) } - -#define __MUTEX_INITIALIZER(name) \ +#define __MUTEX_INITIALIZER(name) \ __SEMAPHORE_INITIALIZER(name, 1) -#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ +#define __DECLARE_SEMAPHORE_GENERIC(name,count) \ struct semaphore name = __SEMAPHORE_INITIALIZER(name,count) #define DECLARE_MUTEX(name) __DECLARE_SEMAPHORE_GENERIC(name,1) @@ -49,17 +40,8 @@ struct semaphore { static inline void sema_init (struct semaphore *sem, int val) { -/* - * *sem = (struct semaphore)__SEMAPHORE_INITIALIZER((*sem),val); - * - * i'd rather use the more flexible initialization above, but sadly - * GCC 2.7.2.3 emits a bogus warning. EGCS doesnt. Oh well. - */ atomic_set(&sem->count, val); init_waitqueue_head(&sem->wait); -#if WAITQUEUE_DEBUG - sem->__magic = (int)&sem->__magic; -#endif } static inline void init_MUTEX (struct semaphore *sem) @@ -81,9 +63,7 @@ extern spinlock_t semaphore_wake_lock; static inline void down(struct semaphore * sem) { -#if WAITQUEUE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif + might_sleep(); if (atomic_sub_return(1, &sem->count) < 0) __down(sem); @@ -92,9 +72,8 @@ static inline void down(struct semaphore * sem) static inline int down_interruptible(struct semaphore * sem) { int ret = 0; -#if WAITQUEUE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif + + might_sleep(); if (atomic_sub_return(1, &sem->count) < 0) ret = __down_interruptible(sem); @@ -104,9 +83,6 @@ static inline int down_interruptible(struct semaphore * sem) static inline int down_trylock(struct semaphore * sem) { int ret = 0; -#if WAITQUEUE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif if (atomic_sub_return(1, &sem->count) < 0) ret = __down_trylock(sem); @@ -119,9 +95,6 @@ static inline int down_trylock(struct semaphore * sem) */ static inline void up(struct semaphore * sem) { -#if WAITQUEUE_DEBUG - CHECK_MAGIC(sem->__magic); -#endif if (atomic_add_return(1, &sem->count) <= 0) __up(sem); } diff --git a/include/asm-xtensa/system.h b/include/asm-xtensa/system.h index f09393232e5..9284867f1cb 100644 --- a/include/asm-xtensa/system.h +++ b/include/asm-xtensa/system.h @@ -189,20 +189,6 @@ static inline unsigned long xchg_u32(volatile int * m, unsigned long val) #define tas(ptr) (xchg((ptr),1)) -#if ( __XCC__ == 1 ) - -/* xt-xcc processes __inline__ differently than xt-gcc and decides to - * insert an out-of-line copy of function __xchg. This presents the - * unresolved symbol at link time of __xchg_called_with_bad_pointer, - * even though such a function would never be called at run-time. - * xt-gcc always inlines __xchg, and optimizes away the undefined - * bad_pointer function. - */ - -#define xchg(ptr,x) xchg_u32(ptr,x) - -#else /* assume xt-gcc */ - #define xchg(ptr,x) ((__typeof__(*(ptr)))__xchg((unsigned long)(x),(ptr),sizeof(*(ptr)))) /* @@ -224,8 +210,6 @@ __xchg(unsigned long x, volatile void * ptr, int size) return x; } -#endif - extern void set_except_vector(int n, void *addr); static inline void spill_registers(void) From f65d9bea8692d7343e4636763f291408f025834f Mon Sep 17 00:00:00 2001 From: Clemens Buchacher Date: Thu, 22 Sep 2005 21:44:24 -0700 Subject: [PATCH 177/204] [PATCH] oss: don't concatenate __FUNCTION__ with strings It's deprecated. Use "%s", __FUNCTION__ instead. Signed-off-by: Clemens Buchacher Signed-off-by: Maximilian Attems Signed-off-by: Domen Puncer Signed-off-by: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- sound/oss/au1000.c | 2 +- sound/oss/ite8172.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/oss/au1000.c b/sound/oss/au1000.c index 4491733c9e4..2c2ae2ee01a 100644 --- a/sound/oss/au1000.c +++ b/sound/oss/au1000.c @@ -1295,7 +1295,7 @@ static int au1000_mmap(struct file *file, struct vm_area_struct *vma) unsigned long size; int ret = 0; - dbg(__FUNCTION__); + dbg("%s", __FUNCTION__); lock_kernel(); down(&s->sem); diff --git a/sound/oss/ite8172.c b/sound/oss/ite8172.c index 58f879fda97..26e5944b6ba 100644 --- a/sound/oss/ite8172.c +++ b/sound/oss/ite8172.c @@ -1859,7 +1859,7 @@ static int it8172_release(struct inode *inode, struct file *file) struct it8172_state *s = (struct it8172_state *)file->private_data; #ifdef IT8172_VERBOSE_DEBUG - dbg(__FUNCTION__); + dbg("%s", __FUNCTION__); #endif lock_kernel(); if (file->f_mode & FMODE_WRITE) From 8bdac5d1ed892da54b6b2069e50a47b3aa39460f Mon Sep 17 00:00:00 2001 From: Glauber de Oliveira Costa Date: Thu, 22 Sep 2005 21:44:26 -0700 Subject: [PATCH 178/204] [PATCH] ext3: EXT3_DEBUG build fixes Fix some warnings and a build error when EXT3_DEBUG is enabled. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/ext3/balloc.c | 6 +++--- fs/ext3/resize.c | 6 +++--- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/fs/ext3/balloc.c b/fs/ext3/balloc.c index e463dca008e..0213db4911a 100644 --- a/fs/ext3/balloc.c +++ b/fs/ext3/balloc.c @@ -1410,7 +1410,7 @@ unsigned long ext3_count_free_blocks(struct super_block *sb) unsigned long desc_count; struct ext3_group_desc *gdp; int i; - unsigned long ngroups; + unsigned long ngroups = EXT3_SB(sb)->s_groups_count; #ifdef EXT3FS_DEBUG struct ext3_super_block *es; unsigned long bitmap_count, x; @@ -1421,7 +1421,8 @@ unsigned long ext3_count_free_blocks(struct super_block *sb) desc_count = 0; bitmap_count = 0; gdp = NULL; - for (i = 0; i < EXT3_SB(sb)->s_groups_count; i++) { + + for (i = 0; i < ngroups; i++) { gdp = ext3_get_group_desc(sb, i, NULL); if (!gdp) continue; @@ -1443,7 +1444,6 @@ unsigned long ext3_count_free_blocks(struct super_block *sb) return bitmap_count; #else desc_count = 0; - ngroups = EXT3_SB(sb)->s_groups_count; smp_rmb(); for (i = 0; i < ngroups; i++) { gdp = ext3_get_group_desc(sb, i, NULL); diff --git a/fs/ext3/resize.c b/fs/ext3/resize.c index 2c9f81278d5..57f79106267 100644 --- a/fs/ext3/resize.c +++ b/fs/ext3/resize.c @@ -242,7 +242,7 @@ static int setup_new_group_blocks(struct super_block *sb, i < sbi->s_itb_per_group; i++, bit++, block++) { struct buffer_head *it; - ext3_debug("clear inode block %#04x (+%ld)\n", block, bit); + ext3_debug("clear inode block %#04lx (+%d)\n", block, bit); if (IS_ERR(it = bclean(handle, sb, block))) { err = PTR_ERR(it); goto exit_bh; @@ -643,8 +643,8 @@ static void update_backups(struct super_block *sb, break; bh = sb_getblk(sb, group * bpg + blk_off); - ext3_debug(sb, __FUNCTION__, "update metadata backup %#04lx\n", - bh->b_blocknr); + ext3_debug("update metadata backup %#04lx\n", + (unsigned long)bh->b_blocknr); if ((err = ext3_journal_get_write_access(handle, bh))) break; lock_buffer(bh); From f7b3a4359b25311fb5894c8809478a2df6bed480 Mon Sep 17 00:00:00 2001 From: Rob Landley Date: Thu, 22 Sep 2005 21:44:27 -0700 Subject: [PATCH 179/204] [PATCH] Fix bd_claim() error code. Problem: In some circumstances, bd_claim() is returning the wrong error code. If we try to swapon an unused block device that isn't swap formatted, we get -EINVAL. But if that same block device is already mounted, we instead get -EBUSY, even though it still isn't a valid swap device. This issue came up on the busybox list trying to get the error message from "swapon -a" right. If a swap device is already enabled, we get -EBUSY, and we shouldn't report this as an error. But we can't distinguish the two -EBUSY conditions, which are very different errors. In the code, bd_claim() returns either 0 or -EBUSY, but in this case busy means "somebody other than sys_swapon has already claimed this", and _that_ means this block device can't be a valid swap device. So return -EINVAL there. Signed-off-by: Rob Landley Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/swapfile.c | 1 + 1 file changed, 1 insertion(+) diff --git a/mm/swapfile.c b/mm/swapfile.c index 0184f510aac..1dcaeda039f 100644 --- a/mm/swapfile.c +++ b/mm/swapfile.c @@ -1381,6 +1381,7 @@ asmlinkage long sys_swapon(const char __user * specialfile, int swap_flags) error = bd_claim(bdev, sys_swapon); if (error < 0) { bdev = NULL; + error = -EINVAL; goto bad_swap; } p->old_block_size = block_size(bdev); From 10d2c46f9408d404bffef89d5052953a3b1d9288 Mon Sep 17 00:00:00 2001 From: Nick Wilson Date: Thu, 22 Sep 2005 21:44:28 -0700 Subject: [PATCH 180/204] [PATCH] NFS: fix client oops when debugging is on nfs_readpage_release() causes an oops while accessing a file with NFS debugging turned on (echo 32767 > /proc/sys/sunrpc/nfs_debug) and a kernel built with CONFIG_DEBUG_SLAB. This patch moves the debugging statement above nfs_release_request() to avoid accessing freed memory. Signed-off-by: Nick Wilson Cc: Trond Myklebust Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/nfs/read.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 6ceb1d471f2..9758ebd4990 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -184,14 +184,13 @@ static void nfs_readpage_release(struct nfs_page *req) { unlock_page(req->wb_page); - nfs_clear_request(req); - nfs_release_request(req); - dprintk("NFS: read done (%s/%Ld %d@%Ld)\n", req->wb_context->dentry->d_inode->i_sb->s_id, (long long)NFS_FILEID(req->wb_context->dentry->d_inode), req->wb_bytes, (long long)req_offset(req)); + nfs_clear_request(req); + nfs_release_request(req); } /* From e484585ec3ee66cd07a627d3a9e2364640a3807f Mon Sep 17 00:00:00 2001 From: Paolo 'Blaisorblade' Giarrusso Date: Thu, 22 Sep 2005 21:44:29 -0700 Subject: [PATCH 181/204] [PATCH] Add dm-snapshot tutorial in Documentation I've recently discovered the real functionality of device-mapper snapshots, and since they are not well known, I've decided to write some docs for them. Signed-off-by: Paolo 'Blaisorblade' Giarrusso Signed-off-by: Alasdair G Kergon Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/device-mapper/snapshot.txt | 73 ++++++++++++++++++++++++ 1 file changed, 73 insertions(+) create mode 100644 Documentation/device-mapper/snapshot.txt diff --git a/Documentation/device-mapper/snapshot.txt b/Documentation/device-mapper/snapshot.txt new file mode 100644 index 00000000000..dca274ff400 --- /dev/null +++ b/Documentation/device-mapper/snapshot.txt @@ -0,0 +1,73 @@ +Device-mapper snapshot support +============================== + +Device-mapper allows you, without massive data copying: + +*) To create snapshots of any block device i.e. mountable, saved states of +the block device which are also writable without interfering with the +original content; +*) To create device "forks", i.e. multiple different versions of the +same data stream. + + +In both cases, dm copies only the chunks of data that get changed and +uses a separate copy-on-write (COW) block device for storage. + + +There are two dm targets available: snapshot and snapshot-origin. + +*) snapshot-origin + +which will normally have one or more snapshots based on it. +You must create the snapshot-origin device before you can create snapshots. +Reads will be mapped directly to the backing device. For each write, the +original data will be saved in the of each snapshot to keep +its visible content unchanged, at least until the fills up. + + +*) snapshot + +A snapshot is created of the block device. Changed chunks of + sectors will be stored on the . Writes will +only go to the . Reads will come from the or +from for unchanged data. will often be +smaller than the origin and if it fills up the snapshot will become +useless and be disabled, returning errors. So it is important to monitor +the amount of free space and expand the before it fills up. + + is P (Persistent) or N (Not persistent - will not survive +after reboot). + + +How this is used by LVM2 +======================== +When you create the first LVM2 snapshot of a volume, four dm devices are used: + +1) a device containing the original mapping table of the source volume; +2) a device used as the ; +3) a "snapshot" device, combining #1 and #2, which is the visible snapshot + volume; +4) the "original" volume (which uses the device number used by the original + source volume), whose table is replaced by a "snapshot-origin" mapping + from device #1. + +A fixed naming scheme is used, so with the following commands: + +lvcreate -L 1G -n base volumeGroup +lvcreate -L 100M --snapshot -n snap volumeGroup/base + +we'll have this situation (with volumes in above order): + +# dmsetup table|grep volumeGroup + +volumeGroup-base-real: 0 2097152 linear 8:19 384 +volumeGroup-snap-cow: 0 204800 linear 8:19 2097536 +volumeGroup-snap: 0 2097152 snapshot 254:11 254:12 P 16 +volumeGroup-base: 0 2097152 snapshot-origin 254:11 + +# ls -lL /dev/mapper/volumeGroup-* +brw------- 1 root root 254, 11 29 ago 18:15 /dev/mapper/volumeGroup-base-real +brw------- 1 root root 254, 12 29 ago 18:15 /dev/mapper/volumeGroup-snap-cow +brw------- 1 root root 254, 13 29 ago 18:15 /dev/mapper/volumeGroup-snap +brw------- 1 root root 254, 10 29 ago 18:14 /dev/mapper/volumeGroup-base + From 83ca28befc43e93849e79c564cda10e39d983e75 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Thu, 22 Sep 2005 23:32:56 -0700 Subject: [PATCH 182/204] [TCP]: Adjust Reno SACK estimate in tcp_fragment Since the introduction of TSO pcount a year ago, it has been possible for tcp_fragment() to cause packets_out to decrease. Prior to that, tcp_retrans_try_collapse() was the only way for that to happen on the retransmission path. When this happens with Reno, it is possible for sasked_out to become invalid because it is only an estimate and not tied to any particular packet on the retransmission queue. Therefore we need to adjust sacked_out as well as left_out in the Reno case. The following patch does exactly that. This bug is pretty difficult to trigger in practice though since you need a SACKless peer with a retransmission that occurs just as the cached MTU value expires. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller --- net/ipv4/tcp_output.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index 5dd6dd7d091..d6e3d269e90 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -509,7 +509,16 @@ int tcp_fragment(struct sock *sk, struct sk_buff *skb, u32 len, unsigned int mss tp->lost_out -= diff; tp->left_out -= diff; } + if (diff > 0) { + /* Adjust Reno SACK estimate. */ + if (!tp->rx_opt.sack_ok) { + tp->sacked_out -= diff; + if ((int)tp->sacked_out < 0) + tp->sacked_out = 0; + tcp_sync_left_out(tp); + } + tp->fackets_out -= diff; if ((int)tp->fackets_out < 0) tp->fackets_out = 0; From 2a7bc3c94cbca32d3d6e03e7d5a71662f591d28a Mon Sep 17 00:00:00 2001 From: Vlad Drukker Date: Thu, 22 Sep 2005 23:35:34 -0700 Subject: [PATCH 183/204] [BRIDGE]: TSO fix in br_dev_queue_push_xmit Signed-off-by: Vlad Drukker Acked-by: Stephen Hemminger Signed-off-by: David S. Miller --- net/bridge/br_forward.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/bridge/br_forward.c b/net/bridge/br_forward.c index 069253f830c..2d24fb400e0 100644 --- a/net/bridge/br_forward.c +++ b/net/bridge/br_forward.c @@ -31,7 +31,8 @@ static inline int should_deliver(const struct net_bridge_port *p, int br_dev_queue_push_xmit(struct sk_buff *skb) { - if (skb->len > skb->dev->mtu) + /* drop mtu oversized packets except tso */ + if (skb->len > skb->dev->mtu && !skb_shinfo(skb)->tso_size) kfree_skb(skb); else { #ifdef CONFIG_BRIDGE_NETFILTER From 0ae5d253adcc467b1c52b512bbca9419eb438409 Mon Sep 17 00:00:00 2001 From: Harald Welte Date: Thu, 22 Sep 2005 23:44:58 -0700 Subject: [PATCH 184/204] [NETFILTER] fix DEBUG statement in PPTP helper As noted by Alexey Dobriyan, the DEBUGP statement prints the wrong callID. Signed-off-by: Harald Welte Signed-off-by: David S. Miller --- net/ipv4/netfilter/ip_conntrack_helper_pptp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c index 79db5b70d5f..11758797870 100644 --- a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c +++ b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c @@ -485,7 +485,7 @@ pptp_inbound_pkt(struct sk_buff **pskb, if (info->pns_call_id != ntohs(*pcid)) { DEBUGP("%s for unknown CallID %u\n", - pptp_msg_name[msg], ntohs(*cid)); + pptp_msg_name[msg], ntohs(*pcid)); break; } From 67497205b12e3cb408259cc09b50c3a9d12cd935 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Thu, 22 Sep 2005 23:45:24 -0700 Subject: [PATCH 185/204] [NETFILTER] Fix sparse endian warnings in pptp helper Signed-off-by: Alexey Dobriyan Signed-off-by: Harald Welte Signed-off-by: David S. Miller --- .../linux/netfilter_ipv4/ip_conntrack_pptp.h | 118 +++++++++--------- .../linux/netfilter_ipv4/ip_conntrack_tuple.h | 6 +- net/ipv4/netfilter/ip_conntrack_helper_pptp.c | 14 ++- 3 files changed, 70 insertions(+), 68 deletions(-) diff --git a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h index 389e3851d52..50a761db5a0 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h +++ b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h @@ -60,8 +60,8 @@ struct ip_ct_pptp_expect { struct pptp_pkt_hdr { __u16 packetLength; - __u16 packetType; - __u32 magicCookie; + __be16 packetType; + __be32 magicCookie; }; /* PptpControlMessageType values */ @@ -93,7 +93,7 @@ struct pptp_pkt_hdr { #define PPTP_REMOVE_DEVICE_ERROR 6 struct PptpControlHeader { - __u16 messageType; + __be16 messageType; __u16 reserved; }; @@ -106,13 +106,13 @@ struct PptpControlHeader { #define PPTP_BEARER_CAP_DIGITAL 0x2 struct PptpStartSessionRequest { - __u16 protocolVersion; + __be16 protocolVersion; __u8 reserved1; __u8 reserved2; - __u32 framingCapability; - __u32 bearerCapability; - __u16 maxChannels; - __u16 firmwareRevision; + __be32 framingCapability; + __be32 bearerCapability; + __be16 maxChannels; + __be16 firmwareRevision; __u8 hostName[64]; __u8 vendorString[64]; }; @@ -125,13 +125,13 @@ struct PptpStartSessionRequest { #define PPTP_START_UNKNOWN_PROTOCOL 5 struct PptpStartSessionReply { - __u16 protocolVersion; + __be16 protocolVersion; __u8 resultCode; __u8 generalErrorCode; - __u32 framingCapability; - __u32 bearerCapability; - __u16 maxChannels; - __u16 firmwareRevision; + __be32 framingCapability; + __be32 bearerCapability; + __be16 maxChannels; + __be16 firmwareRevision; __u8 hostName[64]; __u8 vendorString[64]; }; @@ -155,7 +155,7 @@ struct PptpStopSessionReply { }; struct PptpEchoRequest { - __u32 identNumber; + __be32 identNumber; }; /* PptpEchoReplyResultCode */ @@ -163,7 +163,7 @@ struct PptpEchoRequest { #define PPTP_ECHO_GENERAL_ERROR 2 struct PptpEchoReply { - __u32 identNumber; + __be32 identNumber; __u8 resultCode; __u8 generalErrorCode; __u16 reserved; @@ -180,16 +180,16 @@ struct PptpEchoReply { #define PPTP_DONT_CARE_BEARER_TYPE 3 struct PptpOutCallRequest { - __u16 callID; - __u16 callSerialNumber; - __u32 minBPS; - __u32 maxBPS; - __u32 bearerType; - __u32 framingType; - __u16 packetWindow; - __u16 packetProcDelay; + __be16 callID; + __be16 callSerialNumber; + __be32 minBPS; + __be32 maxBPS; + __be32 bearerType; + __be32 framingType; + __be16 packetWindow; + __be16 packetProcDelay; __u16 reserved1; - __u16 phoneNumberLength; + __be16 phoneNumberLength; __u16 reserved2; __u8 phoneNumber[64]; __u8 subAddress[64]; @@ -205,24 +205,24 @@ struct PptpOutCallRequest { #define PPTP_OUTCALL_DONT_ACCEPT 7 struct PptpOutCallReply { - __u16 callID; - __u16 peersCallID; + __be16 callID; + __be16 peersCallID; __u8 resultCode; __u8 generalErrorCode; - __u16 causeCode; - __u32 connectSpeed; - __u16 packetWindow; - __u16 packetProcDelay; - __u32 physChannelID; + __be16 causeCode; + __be32 connectSpeed; + __be16 packetWindow; + __be16 packetProcDelay; + __be32 physChannelID; }; struct PptpInCallRequest { - __u16 callID; - __u16 callSerialNumber; - __u32 callBearerType; - __u32 physChannelID; - __u16 dialedNumberLength; - __u16 dialingNumberLength; + __be16 callID; + __be16 callSerialNumber; + __be32 callBearerType; + __be32 physChannelID; + __be16 dialedNumberLength; + __be16 dialingNumberLength; __u8 dialedNumber[64]; __u8 dialingNumber[64]; __u8 subAddress[64]; @@ -234,54 +234,54 @@ struct PptpInCallRequest { #define PPTP_INCALL_DONT_ACCEPT 3 struct PptpInCallReply { - __u16 callID; - __u16 peersCallID; + __be16 callID; + __be16 peersCallID; __u8 resultCode; __u8 generalErrorCode; - __u16 packetWindow; - __u16 packetProcDelay; + __be16 packetWindow; + __be16 packetProcDelay; __u16 reserved; }; struct PptpInCallConnected { - __u16 peersCallID; + __be16 peersCallID; __u16 reserved; - __u32 connectSpeed; - __u16 packetWindow; - __u16 packetProcDelay; - __u32 callFramingType; + __be32 connectSpeed; + __be16 packetWindow; + __be16 packetProcDelay; + __be32 callFramingType; }; struct PptpClearCallRequest { - __u16 callID; + __be16 callID; __u16 reserved; }; struct PptpCallDisconnectNotify { - __u16 callID; + __be16 callID; __u8 resultCode; __u8 generalErrorCode; - __u16 causeCode; + __be16 causeCode; __u16 reserved; __u8 callStatistics[128]; }; struct PptpWanErrorNotify { - __u16 peersCallID; + __be16 peersCallID; __u16 reserved; - __u32 crcErrors; - __u32 framingErrors; - __u32 hardwareOverRuns; - __u32 bufferOverRuns; - __u32 timeoutErrors; - __u32 alignmentErrors; + __be32 crcErrors; + __be32 framingErrors; + __be32 hardwareOverRuns; + __be32 bufferOverRuns; + __be32 timeoutErrors; + __be32 alignmentErrors; }; struct PptpSetLinkInfo { - __u16 peersCallID; + __be16 peersCallID; __u16 reserved; - __u32 sendAccm; - __u32 recvAccm; + __be32 sendAccm; + __be32 recvAccm; }; diff --git a/include/linux/netfilter_ipv4/ip_conntrack_tuple.h b/include/linux/netfilter_ipv4/ip_conntrack_tuple.h index 14dc0f7b655..20e43f018b7 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack_tuple.h +++ b/include/linux/netfilter_ipv4/ip_conntrack_tuple.h @@ -17,7 +17,7 @@ union ip_conntrack_manip_proto u_int16_t all; struct { - u_int16_t port; + __be16 port; } tcp; struct { u_int16_t port; @@ -29,7 +29,7 @@ union ip_conntrack_manip_proto u_int16_t port; } sctp; struct { - u_int16_t key; /* key is 32bit, pptp only uses 16 */ + __be16 key; /* key is 32bit, pptp only uses 16 */ } gre; }; @@ -65,7 +65,7 @@ struct ip_conntrack_tuple u_int16_t port; } sctp; struct { - u_int16_t key; /* key is 32bit, + __be16 key; /* key is 32bit, * pptp only uses 16 */ } gre; } u; diff --git a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c index 11758797870..8236ee0fb09 100644 --- a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c +++ b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c @@ -223,8 +223,8 @@ static void pptp_destroy_siblings(struct ip_conntrack *ct) static inline int exp_gre(struct ip_conntrack *master, u_int32_t seq, - u_int16_t callid, - u_int16_t peer_callid) + __be16 callid, + __be16 peer_callid) { struct ip_conntrack_tuple inv_tuple; struct ip_conntrack_tuple exp_tuples[] = { @@ -263,7 +263,7 @@ exp_gre(struct ip_conntrack *master, exp_orig->mask.src.ip = 0xffffffff; exp_orig->mask.src.u.all = 0; exp_orig->mask.dst.u.all = 0; - exp_orig->mask.dst.u.gre.key = 0xffff; + exp_orig->mask.dst.u.gre.key = htons(0xffff); exp_orig->mask.dst.ip = 0xffffffff; exp_orig->mask.dst.protonum = 0xff; @@ -340,7 +340,8 @@ pptp_inbound_pkt(struct sk_buff **pskb, unsigned int reqlen; union pptp_ctrl_union _pptpReq, *pptpReq; struct ip_ct_pptp_master *info = &ct->help.ct_pptp_info; - u_int16_t msg, *cid, *pcid; + u_int16_t msg; + __be16 *cid, *pcid; u_int32_t seq; ctlh = skb_header_pointer(*pskb, nexthdr_off, sizeof(_ctlh), &_ctlh); @@ -551,7 +552,8 @@ pptp_outbound_pkt(struct sk_buff **pskb, unsigned int reqlen; union pptp_ctrl_union _pptpReq, *pptpReq; struct ip_ct_pptp_master *info = &ct->help.ct_pptp_info; - u_int16_t msg, *cid, *pcid; + u_int16_t msg; + __be16 *cid, *pcid; ctlh = skb_header_pointer(*pskb, nexthdr_off, sizeof(_ctlh), &_ctlh); if (!ctlh) @@ -755,7 +757,7 @@ static struct ip_conntrack_helper pptp = { } }, .mask = { .src = { .ip = 0, - .u = { .tcp = { .port = 0xffff } } + .u = { .tcp = { .port = __constant_htons(0xffff) } } }, .dst = { .ip = 0, .u = { .all = 0 }, From a82b748930fce0dab22c64075c38c830ae116904 Mon Sep 17 00:00:00 2001 From: Harald Welte Date: Thu, 22 Sep 2005 23:45:44 -0700 Subject: [PATCH 186/204] [NETFILTER] remove unneeded structure definition from conntrack helper Signed-off-by: Harald Welte Signed-off-by: David S. Miller --- include/linux/netfilter_ipv4/ip_conntrack_pptp.h | 7 ------- 1 file changed, 7 deletions(-) diff --git a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h index 50a761db5a0..816144c75de 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack_pptp.h +++ b/include/linux/netfilter_ipv4/ip_conntrack_pptp.h @@ -284,13 +284,6 @@ struct PptpSetLinkInfo { __be32 recvAccm; }; - -struct pptp_priv_data { - __u16 call_id; - __u16 mcall_id; - __u16 pcall_id; -}; - union pptp_ctrl_union { struct PptpStartSessionRequest sreq; struct PptpStartSessionReply srep; From 1dfbab59498d6f227c91988bab6c71af049a5333 Mon Sep 17 00:00:00 2001 From: Harald Welte Date: Thu, 22 Sep 2005 23:46:57 -0700 Subject: [PATCH 187/204] [NETFILTER] Fix conntrack event cache deadlock/oops This patch fixes a number of bugs. It cannot be reasonably split up in multiple fixes, since all bugs interact with each other and affect the same function: Bug #1: The event cache code cannot be called while a lock is held. Therefore, the call to ip_conntrack_event_cache() within ip_ct_refresh_acct() needs to be moved outside of the locked section. This fixes a number of 2.6.14-rcX oops and deadlock reports. Bug #2: We used to call ct_add_counters() for unconfirmed connections without holding a lock. Since the add operations are not atomic, we could race with another CPU. Bug #3: ip_ct_refresh_acct() lost REFRESH events in some cases where refresh (and the corresponding event) are desired, but no accounting shall be performed. Both, evenst and accounting implicitly depended on the skb parameter bein non-null. We now re-introduce a non-accounting "ip_ct_refresh()" variant to explicitly state the desired behaviour. Signed-off-by: Harald Welte Signed-off-by: David S. Miller --- include/linux/netfilter_ipv4/ip_conntrack.h | 25 ++++++++-- net/ipv4/netfilter/ip_conntrack_amanda.c | 2 +- net/ipv4/netfilter/ip_conntrack_core.c | 49 ++++++++++--------- net/ipv4/netfilter/ip_conntrack_helper_pptp.c | 1 - net/ipv4/netfilter/ip_conntrack_netbios_ns.c | 2 +- net/ipv4/netfilter/ip_conntrack_standalone.c | 2 +- 6 files changed, 49 insertions(+), 32 deletions(-) diff --git a/include/linux/netfilter_ipv4/ip_conntrack.h b/include/linux/netfilter_ipv4/ip_conntrack.h index bace72a76cc..4ced3873681 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack.h +++ b/include/linux/netfilter_ipv4/ip_conntrack.h @@ -332,11 +332,28 @@ extern void need_ip_conntrack(void); extern int invert_tuplepr(struct ip_conntrack_tuple *inverse, const struct ip_conntrack_tuple *orig); +extern void __ip_ct_refresh_acct(struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + const struct sk_buff *skb, + unsigned long extra_jiffies, + int do_acct); + +/* Refresh conntrack for this many jiffies and do accounting */ +static inline void ip_ct_refresh_acct(struct ip_conntrack *ct, + enum ip_conntrack_info ctinfo, + const struct sk_buff *skb, + unsigned long extra_jiffies) +{ + __ip_ct_refresh_acct(ct, ctinfo, skb, extra_jiffies, 1); +} + /* Refresh conntrack for this many jiffies */ -extern void ip_ct_refresh_acct(struct ip_conntrack *ct, - enum ip_conntrack_info ctinfo, - const struct sk_buff *skb, - unsigned long extra_jiffies); +static inline void ip_ct_refresh(struct ip_conntrack *ct, + const struct sk_buff *skb, + unsigned long extra_jiffies) +{ + __ip_ct_refresh_acct(ct, 0, skb, extra_jiffies, 0); +} /* These are for NAT. Icky. */ /* Update TCP window tracking data when NAT mangles the packet */ diff --git a/net/ipv4/netfilter/ip_conntrack_amanda.c b/net/ipv4/netfilter/ip_conntrack_amanda.c index dc20881004b..fa3f914117e 100644 --- a/net/ipv4/netfilter/ip_conntrack_amanda.c +++ b/net/ipv4/netfilter/ip_conntrack_amanda.c @@ -65,7 +65,7 @@ static int help(struct sk_buff **pskb, /* increase the UDP timeout of the master connection as replies from * Amanda clients to the server can be quite delayed */ - ip_ct_refresh_acct(ct, ctinfo, NULL, master_timeout * HZ); + ip_ct_refresh(ct, *pskb, master_timeout * HZ); /* No data? */ dataoff = (*pskb)->nh.iph->ihl*4 + sizeof(struct udphdr); diff --git a/net/ipv4/netfilter/ip_conntrack_core.c b/net/ipv4/netfilter/ip_conntrack_core.c index c1f82e0c81c..ea65dd3e517 100644 --- a/net/ipv4/netfilter/ip_conntrack_core.c +++ b/net/ipv4/netfilter/ip_conntrack_core.c @@ -1112,45 +1112,46 @@ void ip_conntrack_helper_unregister(struct ip_conntrack_helper *me) synchronize_net(); } -static inline void ct_add_counters(struct ip_conntrack *ct, - enum ip_conntrack_info ctinfo, - const struct sk_buff *skb) -{ -#ifdef CONFIG_IP_NF_CT_ACCT - if (skb) { - ct->counters[CTINFO2DIR(ctinfo)].packets++; - ct->counters[CTINFO2DIR(ctinfo)].bytes += - ntohs(skb->nh.iph->tot_len); - } -#endif -} - -/* Refresh conntrack for this many jiffies and do accounting (if skb != NULL) */ -void ip_ct_refresh_acct(struct ip_conntrack *ct, +/* Refresh conntrack for this many jiffies and do accounting if do_acct is 1 */ +void __ip_ct_refresh_acct(struct ip_conntrack *ct, enum ip_conntrack_info ctinfo, const struct sk_buff *skb, - unsigned long extra_jiffies) + unsigned long extra_jiffies, + int do_acct) { + int do_event = 0; + IP_NF_ASSERT(ct->timeout.data == (unsigned long)ct); + IP_NF_ASSERT(skb); + + write_lock_bh(&ip_conntrack_lock); /* If not in hash table, timer will not be active yet */ if (!is_confirmed(ct)) { ct->timeout.expires = extra_jiffies; - ct_add_counters(ct, ctinfo, skb); + do_event = 1; } else { - write_lock_bh(&ip_conntrack_lock); /* Need del_timer for race avoidance (may already be dying). */ if (del_timer(&ct->timeout)) { ct->timeout.expires = jiffies + extra_jiffies; add_timer(&ct->timeout); - /* FIXME: We loose some REFRESH events if this function - * is called without an skb. I'll fix this later -HW */ - if (skb) - ip_conntrack_event_cache(IPCT_REFRESH, skb); + do_event = 1; } - ct_add_counters(ct, ctinfo, skb); - write_unlock_bh(&ip_conntrack_lock); } + +#ifdef CONFIG_IP_NF_CT_ACCT + if (do_acct) { + ct->counters[CTINFO2DIR(ctinfo)].packets++; + ct->counters[CTINFO2DIR(ctinfo)].bytes += + ntohs(skb->nh.iph->tot_len); + } +#endif + + write_unlock_bh(&ip_conntrack_lock); + + /* must be unlocked when calling event cache */ + if (do_event) + ip_conntrack_event_cache(IPCT_REFRESH, skb); } #if defined(CONFIG_IP_NF_CONNTRACK_NETLINK) || \ diff --git a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c index 8236ee0fb09..926a6684643 100644 --- a/net/ipv4/netfilter/ip_conntrack_helper_pptp.c +++ b/net/ipv4/netfilter/ip_conntrack_helper_pptp.c @@ -172,7 +172,6 @@ static int destroy_sibling_or_exp(const struct ip_conntrack_tuple *t) DEBUGP("setting timeout of conntrack %p to 0\n", sibling); sibling->proto.gre.timeout = 0; sibling->proto.gre.stream_timeout = 0; - /* refresh_acct will not modify counters if skb == NULL */ if (del_timer(&sibling->timeout)) sibling->timeout.function((unsigned long)sibling); ip_conntrack_put(sibling); diff --git a/net/ipv4/netfilter/ip_conntrack_netbios_ns.c b/net/ipv4/netfilter/ip_conntrack_netbios_ns.c index 71ef19d126d..577bac22dcc 100644 --- a/net/ipv4/netfilter/ip_conntrack_netbios_ns.c +++ b/net/ipv4/netfilter/ip_conntrack_netbios_ns.c @@ -91,7 +91,7 @@ static int help(struct sk_buff **pskb, ip_conntrack_expect_related(exp); ip_conntrack_expect_put(exp); - ip_ct_refresh_acct(ct, ctinfo, NULL, timeout * HZ); + ip_ct_refresh(ct, *pskb, timeout * HZ); out: return NF_ACCEPT; } diff --git a/net/ipv4/netfilter/ip_conntrack_standalone.c b/net/ipv4/netfilter/ip_conntrack_standalone.c index d3c7808010e..dd476b191f4 100644 --- a/net/ipv4/netfilter/ip_conntrack_standalone.c +++ b/net/ipv4/netfilter/ip_conntrack_standalone.c @@ -989,7 +989,7 @@ EXPORT_SYMBOL(need_ip_conntrack); EXPORT_SYMBOL(ip_conntrack_helper_register); EXPORT_SYMBOL(ip_conntrack_helper_unregister); EXPORT_SYMBOL(ip_ct_iterate_cleanup); -EXPORT_SYMBOL(ip_ct_refresh_acct); +EXPORT_SYMBOL(__ip_ct_refresh_acct); EXPORT_SYMBOL(ip_conntrack_expect_alloc); EXPORT_SYMBOL(ip_conntrack_expect_put); From eb0e0076878a4f9e8e6e7e524ded0d6f7d4a6130 Mon Sep 17 00:00:00 2001 From: Sridhar Samudrala Date: Thu, 22 Sep 2005 23:48:38 -0700 Subject: [PATCH 188/204] [SCTP]: Fix SCTP_SHUTDOWN notifications. Fix to allow SCTP_SHUTDOWN notifications to be received on 1-1 style SCTP SOCK_STREAM sockets. Add SCTP_SHUTDOWN notification to the receive queue before updating the state of the association. Signed-off-by: Sridhar Samudrala Signed-off-by: Neil Horman Signed-off-by: David S. Miller --- net/sctp/sm_statefuns.c | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c index 86073df418f..505c7de10c5 100644 --- a/net/sctp/sm_statefuns.c +++ b/net/sctp/sm_statefuns.c @@ -2414,6 +2414,17 @@ sctp_disposition_t sctp_sf_do_9_2_shutdown(const struct sctp_endpoint *ep, skb_pull(chunk->skb, sizeof(sctp_shutdownhdr_t)); chunk->subh.shutdown_hdr = sdh; + /* API 5.3.1.5 SCTP_SHUTDOWN_EVENT + * When a peer sends a SHUTDOWN, SCTP delivers this notification to + * inform the application that it should cease sending data. + */ + ev = sctp_ulpevent_make_shutdown_event(asoc, 0, GFP_ATOMIC); + if (!ev) { + disposition = SCTP_DISPOSITION_NOMEM; + goto out; + } + sctp_add_cmd_sf(commands, SCTP_CMD_EVENT_ULP, SCTP_ULPEVENT(ev)); + /* Upon the reception of the SHUTDOWN, the peer endpoint shall * - enter the SHUTDOWN-RECEIVED state, * - stop accepting new data from its SCTP user @@ -2439,17 +2450,6 @@ sctp_disposition_t sctp_sf_do_9_2_shutdown(const struct sctp_endpoint *ep, sctp_add_cmd_sf(commands, SCTP_CMD_PROCESS_CTSN, SCTP_U32(chunk->subh.shutdown_hdr->cum_tsn_ack)); - /* API 5.3.1.5 SCTP_SHUTDOWN_EVENT - * When a peer sends a SHUTDOWN, SCTP delivers this notification to - * inform the application that it should cease sending data. - */ - ev = sctp_ulpevent_make_shutdown_event(asoc, 0, GFP_ATOMIC); - if (!ev) { - disposition = SCTP_DISPOSITION_NOMEM; - goto out; - } - sctp_add_cmd_sf(commands, SCTP_CMD_EVENT_ULP, SCTP_ULPEVENT(ev)); - out: return disposition; } From 9389d79fbf9a0167ff2de87e8796c6bb803219bf Mon Sep 17 00:00:00 2001 From: Paul Gortmaker Date: Fri, 23 Sep 2005 05:18:45 -0400 Subject: [PATCH 189/204] [PATCH] 8390 Tx fix for non i386 machines While this is true, E8390_CMD is zero on i386, and thus there should be no effect for these machines. Machines like Mac, Amiga etc. which use Alan's clever register mapping may have a non-zero E8390_CMD and result in bogus "transmitter busy" type messages from this bug. Fixes BUG# 3991. --- drivers/net/8390.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/8390.c b/drivers/net/8390.c index 6d76f3a99b1..f8702742008 100644 --- a/drivers/net/8390.c +++ b/drivers/net/8390.c @@ -1094,7 +1094,7 @@ static void NS8390_trigger_send(struct net_device *dev, unsigned int length, outb_p(E8390_NODMA+E8390_PAGE0, e8390_base+E8390_CMD); - if (inb_p(e8390_base) & E8390_TRANS) + if (inb_p(e8390_base + E8390_CMD) & E8390_TRANS) { printk(KERN_WARNING "%s: trigger_send() called with the transmitter busy.\n", dev->name); From 12f44f46bc9c6dc79255e661b085797af395d8da Mon Sep 17 00:00:00 2001 From: Ivan Kokshaysky Date: Thu, 22 Sep 2005 21:06:31 -0700 Subject: [PATCH 190/204] [PATCH] pci: fixup parent subordinate busnr I believe the change that broke things is introduction of pci_fixup_parent_subordinate_busnr(). The patch here does two things: - hunk #1 should fix the problems you've seen when you boot without additional "pci" kernel options; - hunk #2 supposedly fixes boot with "pci=assign-busses" option which otherwise hangs Acer TM81xx machines as reported. Please try this with and without "pci=assign-busses". If it boots, I'd like to see 'lspci -vvx' for both cases. Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/pci/probe.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c index c77d5b1bbff..005786416bb 100644 --- a/drivers/pci/probe.c +++ b/drivers/pci/probe.c @@ -402,6 +402,12 @@ static void pci_enable_crs(struct pci_dev *dev) static void __devinit pci_fixup_parent_subordinate_busnr(struct pci_bus *child, int max) { struct pci_bus *parent = child->parent; + + /* Attempts to fix that up are really dangerous unless + we're going to re-assign all bus numbers. */ + if (!pcibios_assign_all_busses()) + return; + while (parent->parent && parent->subordinate < max) { parent->subordinate = max; pci_write_config_byte(parent->self, PCI_SUBORDINATE_BUS, max); @@ -478,8 +484,18 @@ int __devinit pci_scan_bridge(struct pci_bus *bus, struct pci_dev * dev, int max * We need to assign a number to this bus which we always * do in the second pass. */ - if (!pass) + if (!pass) { + if (pcibios_assign_all_busses()) + /* Temporarily disable forwarding of the + configuration cycles on all bridges in + this bus segment to avoid possible + conflicts in the second pass between two + bridges programmed with overlapping + bus ranges. */ + pci_write_config_dword(dev, PCI_PRIMARY_BUS, + buses & ~0xffffff); return max; + } /* Clear errors */ pci_write_config_word(dev, PCI_STATUS, 0xffff); From ede1327ea4ca8019ec6df24b3e837def091c26b8 Mon Sep 17 00:00:00 2001 From: Steve French Date: Tue, 30 Aug 2005 20:10:14 -0700 Subject: [PATCH 191/204] [PATCH] cifs: Add support for suspend cifsd had been preventing software suspend from completing. Signed-off-by: pavel@suse.de Signed-off-by: Steve French lightly modified Signed-off-by: Linus Torvalds --- fs/cifs/cifsfs.c | 2 ++ fs/cifs/connect.c | 2 ++ 2 files changed, 4 insertions(+) diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 8cc23e7d0d5..1ebf7dafc1d 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -781,6 +781,8 @@ static int cifs_oplock_thread(void * dummyarg) oplockThread = current; do { + if (try_to_freeze()) + continue; set_current_state(TASK_INTERRUPTIBLE); schedule_timeout(1*HZ); diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 2335f14a158..47360156cc5 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -344,6 +344,8 @@ cifs_demultiplex_thread(struct TCP_Server_Info *server) } while (server->tcpStatus != CifsExiting) { + if (try_to_freeze()) + continue; if (bigbuf == NULL) { bigbuf = cifs_buf_get(); if(bigbuf == NULL) { From 188a1eafa03aaa5e5fe6f53e637e704cd2c31c7c Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Fri, 23 Sep 2005 13:22:21 -0700 Subject: [PATCH 192/204] Make sure SIGKILL gets proper respect Bhavesh P. Davda noticed that SIGKILL wouldn't properly kill a process under just the right cicumstances: a stopped task that already had another signal queued would get the SIGKILL queued onto the shared queue, and there it would remain until SIGCONT. This simplifies the signal acceptance logic, and fixes the bug in the process. Losely based on an earlier patch by Bhavesh. Signed-off-by: Linus Torvalds --- kernel/signal.c | 31 ++++++++++++++----------------- 1 file changed, 14 insertions(+), 17 deletions(-) diff --git a/kernel/signal.c b/kernel/signal.c index b92c3c9f8b9..5a274705ba1 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -936,34 +936,31 @@ force_sig_specific(int sig, struct task_struct *t) * as soon as they're available, so putting the signal on the shared queue * will be equivalent to sending it to one such thread. */ -#define wants_signal(sig, p, mask) \ - (!sigismember(&(p)->blocked, sig) \ - && !((p)->state & mask) \ - && !((p)->flags & PF_EXITING) \ - && (task_curr(p) || !signal_pending(p))) - +static inline int wants_signal(int sig, struct task_struct *p) +{ + if (sigismember(&p->blocked, sig)) + return 0; + if (p->flags & PF_EXITING) + return 0; + if (sig == SIGKILL) + return 1; + if (p->state & (TASK_STOPPED | TASK_TRACED)) + return 0; + return task_curr(p) || !signal_pending(p); +} static void __group_complete_signal(int sig, struct task_struct *p) { - unsigned int mask; struct task_struct *t; - /* - * Don't bother traced and stopped tasks (but - * SIGKILL will punch through that). - */ - mask = TASK_STOPPED | TASK_TRACED; - if (sig == SIGKILL) - mask = 0; - /* * Now find a thread we can wake up to take the signal off the queue. * * If the main thread wants the signal, it gets first crack. * Probably the least surprising to the average bear. */ - if (wants_signal(sig, p, mask)) + if (wants_signal(sig, p)) t = p; else if (thread_group_empty(p)) /* @@ -981,7 +978,7 @@ __group_complete_signal(int sig, struct task_struct *p) t = p->signal->curr_target = p; BUG_ON(t->tgid != p->tgid); - while (!wants_signal(sig, t, mask)) { + while (!wants_signal(sig, t)) { t = next_thread(t); if (t == p->signal->curr_target) /* From a3b5f0d43863402b03b3af9a7eaec0037a823374 Mon Sep 17 00:00:00 2001 From: Karsten Keil Date: Fri, 23 Sep 2005 13:24:05 -0700 Subject: [PATCH 193/204] [PATCH] hisax: remove URB_ASYNC_UNLINK usb_unlink_urb is always async now, so URB_ASYNC_UNLINK was removed from core USB and we must do as well. Signed-off-by: Karsten Keil Cc: Greg KH Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/isdn/hisax/st5481_b.c | 2 -- drivers/isdn/hisax/st5481_usb.c | 2 -- 2 files changed, 4 deletions(-) diff --git a/drivers/isdn/hisax/st5481_b.c b/drivers/isdn/hisax/st5481_b.c index 0a2536d6240..657817a591f 100644 --- a/drivers/isdn/hisax/st5481_b.c +++ b/drivers/isdn/hisax/st5481_b.c @@ -209,9 +209,7 @@ static void st5481B_mode(struct st5481_bcs *bcs, int mode) bcs->mode = mode; // Cancel all USB transfers on this B channel - b_out->urb[0]->transfer_flags |= URB_ASYNC_UNLINK; usb_unlink_urb(b_out->urb[0]); - b_out->urb[1]->transfer_flags |= URB_ASYNC_UNLINK; usb_unlink_urb(b_out->urb[1]); b_out->busy = 0; diff --git a/drivers/isdn/hisax/st5481_usb.c b/drivers/isdn/hisax/st5481_usb.c index ffd5b2d4555..89fbeb58485 100644 --- a/drivers/isdn/hisax/st5481_usb.c +++ b/drivers/isdn/hisax/st5481_usb.c @@ -645,9 +645,7 @@ void st5481_in_mode(struct st5481_in *in, int mode) in->mode = mode; - in->urb[0]->transfer_flags |= URB_ASYNC_UNLINK; usb_unlink_urb(in->urb[0]); - in->urb[1]->transfer_flags |= URB_ASYNC_UNLINK; usb_unlink_urb(in->urb[1]); if (in->mode != L1_MODE_NULL) { From 2601c2e278863cd48c01bce1377b4c9747893025 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Fri, 23 Sep 2005 13:24:06 -0700 Subject: [PATCH 194/204] [PATCH] atyfb c99 fix - fix this: drivers/video/aty/xlinit.c: In function `atyfb_xl_init': drivers/video/aty/xlinit.c:256: warning: ISO C90 forbids mixed declarations and code - repair some kooky coding style - Use ARRAY_SIZE() Cc: Tom 'spot' Callaway Cc: "Antonino A. Daplas" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/video/aty/xlinit.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/video/aty/xlinit.c b/drivers/video/aty/xlinit.c index 0bea0d8d782..a085cbf74ec 100644 --- a/drivers/video/aty/xlinit.c +++ b/drivers/video/aty/xlinit.c @@ -253,9 +253,11 @@ int atyfb_xl_init(struct fb_info *info) aty_st_le32(0xFC, 0x00000000, par); #if defined (CONFIG_FB_ATY_GENERIC_LCD) - int i; - for (i=0; i Date: Fri, 23 Sep 2005 13:24:07 -0700 Subject: [PATCH 195/204] [PATCH] ppc64: Fix huge pages MMU mapping bug Current kernel has a couple of sneaky bugs in the ppc64 hugetlb code that cause huge pages to be potentially left stale in the hash table and TLBs (improperly invalidated), with all the nasty consequences that can have. One is that we forgot to set the "secondary" bit in the hash PTEs when hashing a huge page in the secondary bucket (fortunately very rare). The other one is on non-LPAR machines (like Apple G5s), flush_hash_range() which is used to flush a batch of PTEs simply did not work for huge pages. Historically, our huge page code didn't batch, but this was changed without fixing this routine. This patch fixes both. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/ppc64/mm/hash_native.c | 5 ++--- arch/ppc64/mm/hugetlbpage.c | 7 +++++-- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/arch/ppc64/mm/hash_native.c b/arch/ppc64/mm/hash_native.c index 7626bb59954..eb1bbb5b6c1 100644 --- a/arch/ppc64/mm/hash_native.c +++ b/arch/ppc64/mm/hash_native.c @@ -343,9 +343,7 @@ static void native_flush_hash_range(unsigned long context, hpte_t *hptep; unsigned long hpte_v; struct ppc64_tlb_batch *batch = &__get_cpu_var(ppc64_tlb_batch); - - /* XXX fix for large ptes */ - unsigned long large = 0; + unsigned long large; local_irq_save(flags); @@ -358,6 +356,7 @@ static void native_flush_hash_range(unsigned long context, va = (vsid << 28) | (batch->addr[i] & 0x0fffffff); batch->vaddr[j] = va; + large = pte_huge(batch->pte[i]); if (large) vpn = va >> HPAGE_SHIFT; else diff --git a/arch/ppc64/mm/hugetlbpage.c b/arch/ppc64/mm/hugetlbpage.c index 338771ec70d..0ea0994ed97 100644 --- a/arch/ppc64/mm/hugetlbpage.c +++ b/arch/ppc64/mm/hugetlbpage.c @@ -710,10 +710,13 @@ repeat: hpte_group = ((~hash & htab_hash_mask) * HPTES_PER_GROUP) & ~0x7UL; slot = ppc_md.hpte_insert(hpte_group, va, prpn, - HPTE_V_LARGE, rflags); + HPTE_V_LARGE | + HPTE_V_SECONDARY, + rflags); if (slot == -1) { if (mftb() & 0x1) - hpte_group = ((hash & htab_hash_mask) * HPTES_PER_GROUP) & ~0x7UL; + hpte_group = ((hash & htab_hash_mask) * + HPTES_PER_GROUP)&~0x7UL; ppc_md.hpte_remove(hpte_group); goto repeat; From e7c965084392e23a3494f0247c75795360a3fa0d Mon Sep 17 00:00:00 2001 From: Hironobu Ishii Date: Fri, 23 Sep 2005 13:24:08 -0700 Subject: [PATCH 196/204] [PATCH] ipmi_msghandler: inconsistent spin_lock usage I found an inconsistent spin_lock usage in ipmi_smi_msg_received. Signed-off-by: Hironobu Ishii Cc: Corey Minyard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/char/ipmi/ipmi_msghandler.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/char/ipmi/ipmi_msghandler.c b/drivers/char/ipmi/ipmi_msghandler.c index 463351d4f94..32fa82c78c7 100644 --- a/drivers/char/ipmi/ipmi_msghandler.c +++ b/drivers/char/ipmi/ipmi_msghandler.c @@ -2620,7 +2620,7 @@ void ipmi_smi_msg_received(ipmi_smi_t intf, spin_lock_irqsave(&(intf->waiting_msgs_lock), flags); if (!list_empty(&(intf->waiting_msgs))) { list_add_tail(&(msg->link), &(intf->waiting_msgs)); - spin_unlock(&(intf->waiting_msgs_lock)); + spin_unlock_irqrestore(&(intf->waiting_msgs_lock), flags); goto out_unlock; } spin_unlock_irqrestore(&(intf->waiting_msgs_lock), flags); @@ -2629,9 +2629,9 @@ void ipmi_smi_msg_received(ipmi_smi_t intf, if (rv > 0) { /* Could not handle the message now, just add it to a list to handle later. */ - spin_lock(&(intf->waiting_msgs_lock)); + spin_lock_irqsave(&(intf->waiting_msgs_lock), flags); list_add_tail(&(msg->link), &(intf->waiting_msgs)); - spin_unlock(&(intf->waiting_msgs_lock)); + spin_unlock_irqrestore(&(intf->waiting_msgs_lock), flags); } else if (rv == 0) { ipmi_free_smi_msg(msg); } From 86513e726b494796175b6c4fdd705797f01b0ca2 Mon Sep 17 00:00:00 2001 From: Harald Welte Date: Fri, 23 Sep 2005 13:24:10 -0700 Subject: [PATCH 197/204] [PATCH] documentation: sparse no longer uses bk, but git Signed-off-by: Harald Welte Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/sparse.txt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/Documentation/sparse.txt b/Documentation/sparse.txt index 5df44dc894e..1829009db77 100644 --- a/Documentation/sparse.txt +++ b/Documentation/sparse.txt @@ -51,9 +51,9 @@ or you don't get any checking at all. Where to get sparse ~~~~~~~~~~~~~~~~~~~ -With BK, you can just get it from +With git, you can just get it from - bk://sparse.bkbits.net/sparse + rsync://rsync.kernel.org/pub/scm/devel/sparse/sparse.git and DaveJ has tar-balls at From dbdb90450059e17e8e005ebd3ce0a1fd6008a0c8 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Fri, 23 Sep 2005 13:24:10 -0700 Subject: [PATCH 198/204] [PATCH] revert oversized kmalloc check As davem points out, this wasn't such a great idea. There may be some code which does: size = 1024*1024; while (kmalloc(size, ...) == 0) size /= 2; which will now explode. Cc: "David S. Miller" Cc: Christoph Lameter Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/slab.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/mm/slab.c b/mm/slab.c index ad113c431e3..c9adfce0040 100644 --- a/mm/slab.c +++ b/mm/slab.c @@ -2907,7 +2907,8 @@ void *__kmalloc(size_t size, unsigned int __nocast flags) * functions. */ cachep = __find_general_cachep(size, flags); - BUG_ON(!cachep); /* Allocation size too large for kmalloc */ + if (unlikely(cachep == NULL)) + return NULL; return __cache_alloc(cachep, flags); } EXPORT_SYMBOL(__kmalloc); From 73394322a4bf4108847ba3c2eee33a81b1b44450 Mon Sep 17 00:00:00 2001 From: Russell King Date: Fri, 23 Sep 2005 21:49:58 +0100 Subject: [PATCH 199/204] [ARM] Fix context switch with ARMv6 + TLS We accidentally corrupted the TLS value when clearing out the ARMv6 exclusive monitor. Avoid doing so. Signed-off-by: Russell King --- arch/arm/kernel/entry-armv.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index 7152bfbee58..93b5e8e5292 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -537,7 +537,7 @@ ENTRY(__switch_to) #ifdef CONFIG_CPU_MPCORE clrex #else - strex r3, r4, [ip] @ Clear exclusive monitor + strex r5, r4, [ip] @ Clear exclusive monitor #endif #endif #if defined(CONFIG_CPU_XSCALE) && !defined(CONFIG_IWMMXT) From c8868611389aa28e0e5e0d63f468727781eac68c Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Fri, 23 Sep 2005 09:08:30 -0700 Subject: [PATCH 200/204] [PATCH] skge: fix Yukon-Lite A0 workaround This is one of those workarounds sucked over from sk98lin driver. The skge driver needs to detect the Yukon-Lite A0 chip properly, and turn of Rx FIFO Flush. Signed-off-by: Stephen Hemminger Signed-off-by: Jeff Garzik --- drivers/net/skge.c | 22 ++++++++++++++++++++-- 1 file changed, 20 insertions(+), 2 deletions(-) diff --git a/drivers/net/skge.c b/drivers/net/skge.c index ae1996a3bc5..fd398da4993 100644 --- a/drivers/net/skge.c +++ b/drivers/net/skge.c @@ -1643,6 +1643,22 @@ static void yukon_reset(struct skge_hw *hw, int port) | GM_RXCR_UCF_ENA | GM_RXCR_MCF_ENA); } +/* Apparently, early versions of Yukon-Lite had wrong chip_id? */ +static int is_yukon_lite_a0(struct skge_hw *hw) +{ + u32 reg; + int ret; + + if (hw->chip_id != CHIP_ID_YUKON) + return 0; + + reg = skge_read32(hw, B2_FAR); + skge_write8(hw, B2_FAR + 3, 0xff); + ret = (skge_read8(hw, B2_FAR + 3) != 0); + skge_write32(hw, B2_FAR, reg); + return ret; +} + static void yukon_mac_init(struct skge_hw *hw, int port) { struct skge_port *skge = netdev_priv(hw->dev[port]); @@ -1758,9 +1774,11 @@ static void yukon_mac_init(struct skge_hw *hw, int port) /* Configure Rx MAC FIFO */ skge_write16(hw, SK_REG(port, RX_GMF_FL_MSK), RX_FF_FL_DEF_MSK); reg = GMF_OPER_ON | GMF_RX_F_FL_ON; - if (hw->chip_id == CHIP_ID_YUKON_LITE && - hw->chip_rev >= CHIP_REV_YU_LITE_A3) + + /* disable Rx GMAC FIFO Flush for YUKON-Lite Rev. A0 only */ + if (is_yukon_lite_a0(hw)) reg &= ~GMF_RX_F_FL_ON; + skge_write8(hw, SK_REG(port, RX_GMF_CTRL_T), GMF_RST_CLR); skge_write16(hw, SK_REG(port, RX_GMF_CTRL_T), reg); /* From f993724210744a9b07cf5496808828c4da6ab6df Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 24 Sep 2005 10:12:47 +0100 Subject: [PATCH 201/204] [SERIAL] Remove unused variable in clps711x.c Signed-off-by: Russell King --- drivers/serial/clps711x.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/serial/clps711x.c b/drivers/serial/clps711x.c index 78c1f36ad9b..87ef368384f 100644 --- a/drivers/serial/clps711x.c +++ b/drivers/serial/clps711x.c @@ -98,7 +98,7 @@ static irqreturn_t clps711xuart_int_rx(int irq, void *dev_id, struct pt_regs *re { struct uart_port *port = dev_id; struct tty_struct *tty = port->info->tty; - unsigned int status, ch, flg, ignored = 0; + unsigned int status, ch, flg; status = clps_readl(SYSFLG(port)); while (!(status & SYSFLG_URXFE)) { From 6b9ea4213e3a9c9b8de29a001a2b8d8799cbc1c9 Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 24 Sep 2005 10:24:37 +0100 Subject: [PATCH 202/204] [MFD] Fix "bious one-bit signed bitfield" errors ucb1x00-ts declared a couple of one-bit 'int' bitfields. Make them unsigned. Signed-off-by: Russell King --- drivers/mfd/ucb1x00-ts.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/mfd/ucb1x00-ts.c b/drivers/mfd/ucb1x00-ts.c index a851d65c7cf..a260f83bcb0 100644 --- a/drivers/mfd/ucb1x00-ts.c +++ b/drivers/mfd/ucb1x00-ts.c @@ -48,8 +48,8 @@ struct ucb1x00_ts { u16 x_res; u16 y_res; - int restart:1; - int adcsync:1; + unsigned int restart:1; + unsigned int adcsync:1; }; static int adcsync; From d2f607484f8210cf19b1730dee26d426a5a770a5 Mon Sep 17 00:00:00 2001 From: Russell King Date: Sat, 24 Sep 2005 10:42:06 +0100 Subject: [PATCH 203/204] [ARM] Fix compiler warnings for memcpy_toio/memcpy_fromio/memset_io Add 'volatile' to the __iomem pointers for these functions as per x86. Signed-off-by: Russell King --- arch/arm/kernel/io.c | 6 +++--- include/asm-arm/io.h | 6 +++--- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/arch/arm/kernel/io.c b/arch/arm/kernel/io.c index 6c20c1188b6..1f6822dfae7 100644 --- a/arch/arm/kernel/io.c +++ b/arch/arm/kernel/io.c @@ -7,7 +7,7 @@ * Copy data from IO memory space to "real" memory space. * This needs to be optimized. */ -void _memcpy_fromio(void *to, void __iomem *from, size_t count) +void _memcpy_fromio(void *to, const volatile void __iomem *from, size_t count) { unsigned char *t = to; while (count) { @@ -22,7 +22,7 @@ void _memcpy_fromio(void *to, void __iomem *from, size_t count) * Copy data from "real" memory space to IO memory space. * This needs to be optimized. */ -void _memcpy_toio(void __iomem *to, const void *from, size_t count) +void _memcpy_toio(volatile void __iomem *to, const void *from, size_t count) { const unsigned char *f = from; while (count) { @@ -37,7 +37,7 @@ void _memcpy_toio(void __iomem *to, const void *from, size_t count) * "memset" on IO memory space. * This needs to be optimized. */ -void _memset_io(void __iomem *dst, int c, size_t count) +void _memset_io(volatile void __iomem *dst, int c, size_t count) { while (count) { count--; diff --git a/include/asm-arm/io.h b/include/asm-arm/io.h index cfa71a0dffb..5c4ae8f5dbb 100644 --- a/include/asm-arm/io.h +++ b/include/asm-arm/io.h @@ -136,9 +136,9 @@ extern void __readwrite_bug(const char *fn); /* * String version of IO memory access ops: */ -extern void _memcpy_fromio(void *, void __iomem *, size_t); -extern void _memcpy_toio(void __iomem *, const void *, size_t); -extern void _memset_io(void __iomem *, int, size_t); +extern void _memcpy_fromio(void *, const volatile void __iomem *, size_t); +extern void _memcpy_toio(volatile void __iomem *, const void *, size_t); +extern void _memset_io(volatile void __iomem *, int, size_t); #define mmiowb() From ff13f98b97cb6c5a2d201dfa4db0a2fd0b4d9d50 Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Sat, 24 Sep 2005 02:02:29 -0500 Subject: [PATCH 204/204] [PATCH] Input: check switch bitmap when matching handlers The wwitch bitmap was added to input_device_id structure and we should check it when matching handlers and input devices. Signed-off-by: Dmitry Torokhov Signed-off-by: Linus Torvalds --- drivers/input/input.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/input/input.c b/drivers/input/input.c index 88636a20452..14ae5583e19 100644 --- a/drivers/input/input.c +++ b/drivers/input/input.c @@ -308,6 +308,7 @@ static struct input_device_id *input_match_device(struct input_device_id *id, st MATCH_BIT(ledbit, LED_MAX); MATCH_BIT(sndbit, SND_MAX); MATCH_BIT(ffbit, FF_MAX); + MATCH_BIT(swbit, SW_MAX); return id; }