From 1311c24fad3b2acad5c6545d2107c226774f02e8 Mon Sep 17 00:00:00 2001 From: Adrian Bunk Date: Fri, 10 Feb 2006 11:56:30 +0100 Subject: [PATCH 001/102] [AGPGART] help text updates This patch contains help text updates including the following: - XFree86 * -> X - there is no need for repeating part of the help text of the AGP option and having "If unsure, say Y/N." in the chip specific options. Signed-off-by: Adrian Bunk Signed-off-by: Dave Jones --- drivers/char/agp/Kconfig | 55 +++++++++++++--------------------------- 1 file changed, 17 insertions(+), 38 deletions(-) diff --git a/drivers/char/agp/Kconfig b/drivers/char/agp/Kconfig index 486ed8a11b5..a4d425d2dce 100644 --- a/drivers/char/agp/Kconfig +++ b/drivers/char/agp/Kconfig @@ -15,22 +15,23 @@ config AGP due to kernel allocation issues), you could use PCI accesses and have up to a couple gigs of texture space. - Note that this is the only means to have XFree4/GLX use + Note that this is the only means to have X/GLX use write-combining with MTRR support on the AGP bus. Without it, OpenGL direct rendering will be a lot slower but still faster than PIO. - You should say Y here if you use XFree86 3.3.6 or 4.x and want to - use GLX or DRI. If unsure, say N. - To compile this driver as a module, choose M here: the module will be called agpgart. + You should say Y here if you want to use GLX or DRI. + + If unsure, say N. + config AGP_ALI tristate "ALI chipset support" depends on AGP && X86_32 ---help--- This option gives you AGP support for the GLX component of - XFree86 4.x on the following ALi chipsets. The supported chipsets + X on the following ALi chipsets. The supported chipsets include M1541, M1621, M1631, M1632, M1641,M1647,and M1651. For the ALi-chipset question, ALi suggests you refer to . @@ -40,28 +41,19 @@ config AGP_ALI timing issues, this chipset cannot do AGP 2x with the G200. This is a hardware limitation. AGP 1x seems to be fine, though. - You should say Y here if you use XFree86 3.3.6 or 4.x and want to - use GLX or DRI. If unsure, say N. - config AGP_ATI tristate "ATI chipset support" depends on AGP && X86_32 ---help--- - This option gives you AGP support for the GLX component of - XFree86 4.x on the ATI RadeonIGP family of chipsets. - - You should say Y here if you use XFree86 3.3.6 or 4.x and want to - use GLX or DRI. If unsure, say N. + This option gives you AGP support for the GLX component of + X on the ATI RadeonIGP family of chipsets. config AGP_AMD tristate "AMD Irongate, 761, and 762 chipset support" depends on AGP && X86_32 help This option gives you AGP support for the GLX component of - XFree86 4.x on AMD Irongate, 761, and 762 chipsets. - - You should say Y here if you use XFree86 3.3.6 or 4.x and want to - use GLX or DRI. If unsure, say N. + X on AMD Irongate, 761, and 762 chipsets. config AGP_AMD64 tristate "AMD Opteron/Athlon64 on-CPU GART support" if !GART_IOMMU @@ -69,45 +61,38 @@ config AGP_AMD64 default y if GART_IOMMU help This option gives you AGP support for the GLX component of - XFree86 4.x using the on-CPU northbridge of the AMD Athlon64/Opteron CPUs. + X using the on-CPU northbridge of the AMD Athlon64/Opteron CPUs. You still need an external AGP bridge like the AMD 8151, VIA K8T400M, SiS755. It may also support other AGP bridges when loaded with agp_try_unsupported=1. - You should say Y here if you use XFree86 3.3.6 or 4.x and want to - use GLX or DRI. If unsure, say Y config AGP_INTEL tristate "Intel 440LX/BX/GX, I8xx and E7x05 chipset support" depends on AGP && X86 help - This option gives you AGP support for the GLX component of XFree86 4.x + This option gives you AGP support for the GLX component of X on Intel 440LX/BX/GX, 815, 820, 830, 840, 845, 850, 860, 875, - E7205 and E7505 chipsets and full support for the 810, 815, 830M, 845G, - 852GM, 855GM, 865G and I915 integrated graphics chipsets. + E7205 and E7505 chipsets and full support for the 810, 815, 830M, + 845G, 852GM, 855GM, 865G and I915 integrated graphics chipsets. + - You should say Y here if you use XFree86 3.3.6 or 4.x and want to - use GLX or DRI, or if you have any Intel integrated graphics - chipsets. If unsure, say Y. config AGP_NVIDIA tristate "NVIDIA nForce/nForce2 chipset support" depends on AGP && X86_32 help This option gives you AGP support for the GLX component of - XFree86 4.x on the following NVIDIA chipsets. The supported chipsets - include nForce and nForce2 + X on NVIDIA chipsets including nForce and nForce2 config AGP_SIS tristate "SiS chipset support" depends on AGP && X86_32 help This option gives you AGP support for the GLX component of - XFree86 4.x on Silicon Integrated Systems [SiS] chipsets. + X on Silicon Integrated Systems [SiS] chipsets. Note that 5591/5592 AGP chipsets are NOT supported. - You should say Y here if you use XFree86 3.3.6 or 4.x and want to - use GLX or DRI. If unsure, say N. config AGP_SWORKS tristate "Serverworks LE/HE chipset support" @@ -121,10 +106,7 @@ config AGP_VIA depends on AGP && X86_32 help This option gives you AGP support for the GLX component of - XFree86 4.x on VIA MVP3/Apollo Pro chipsets. - - You should say Y here if you use XFree86 3.3.6 or 4.x and want to - use GLX or DRI. If unsure, say N. + X on VIA MVP3/Apollo Pro chipsets. config AGP_I460 tristate "Intel 460GX chipset support" @@ -159,9 +141,6 @@ config AGP_EFFICEON This option gives you AGP support for the Transmeta Efficeon series processors with integrated northbridges. - You should say Y here if you use XFree86 3.3.6 or 4.x and want to - use GLX or DRI. If unsure, say Y. - config AGP_SGI_TIOCA tristate "SGI TIO chipset AGP support" depends on AGP && (IA64_SGI_SN2 || IA64_GENERIC) From cf5e40221bc509e13e22dc83c77c0c115eab531f Mon Sep 17 00:00:00 2001 From: Dave Jones Date: Sun, 12 Feb 2006 21:05:32 -0500 Subject: [PATCH 002/102] [AGPGART] Improve the error message shown when we detect a ServerWorks CNB20HE This chipset is unsupported, and likely to remain that way. Signed-off-by: Dave Jones --- drivers/char/agp/sworks-agp.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/char/agp/sworks-agp.c b/drivers/char/agp/sworks-agp.c index 268f78d926d..efef9999f1c 100644 --- a/drivers/char/agp/sworks-agp.c +++ b/drivers/char/agp/sworks-agp.c @@ -468,9 +468,7 @@ static int __devinit agp_serverworks_probe(struct pci_dev *pdev, switch (pdev->device) { case 0x0006: - /* ServerWorks CNB20HE - Fail silently.*/ - printk (KERN_ERR PFX "Detected ServerWorks CNB20HE chipset: No AGP present.\n"); + printk (KERN_ERR PFX "ServerWorks CNB20HE is unsupported due to lack of documentation.\n"); return -ENODEV; case PCI_DEVICE_ID_SERVERWORKS_HE: From 91e3738ebc6d858e784090382e02afeae5a93b08 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Sat, 18 Feb 2006 15:17:04 +1100 Subject: [PATCH 003/102] drm: fixup i915 interrupt on X server exit Fixes: IRQ disabled (i915?) when switchig between gnome themes (gnome-theme-manager) Signed-off-by: Dave Airlie --- drivers/char/drm/i915_irq.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/char/drm/i915_irq.c b/drivers/char/drm/i915_irq.c index a1381c61aa6..d3879ac9970 100644 --- a/drivers/char/drm/i915_irq.c +++ b/drivers/char/drm/i915_irq.c @@ -202,10 +202,15 @@ void i915_driver_irq_postinstall(drm_device_t * dev) void i915_driver_irq_uninstall(drm_device_t * dev) { drm_i915_private_t *dev_priv = (drm_i915_private_t *) dev->dev_private; + u16 temp; + if (!dev_priv) return; I915_WRITE16(I915REG_HWSTAM, 0xffff); I915_WRITE16(I915REG_INT_MASK_R, 0xffff); I915_WRITE16(I915REG_INT_ENABLE_R, 0x0); + + temp = I915_READ16(I915REG_INT_IDENTITY_R); + I915_WRITE16(I915REG_INT_IDENTITY_R, temp); } From 4e5e2e2560aa1d1d01f7af97af2f72706f61da27 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Sat, 18 Feb 2006 15:51:35 +1100 Subject: [PATCH 004/102] drm: radeon add r300 TX_CNTL and verify bitblt packets The Xgl on r300 doesn't work unless you add a verify bitblt function to the DRM, and we need to pass TX_CNTL to flush texture caches. Signed-off-by: Dave Airlie --- drivers/char/drm/r300_cmdbuf.c | 50 ++++++++++++++++++++++++++++++++++ drivers/char/drm/r300_reg.h | 3 ++ drivers/char/drm/radeon_drv.h | 3 +- 3 files changed, 55 insertions(+), 1 deletion(-) diff --git a/drivers/char/drm/r300_cmdbuf.c b/drivers/char/drm/r300_cmdbuf.c index 291dbf4c818..6dd21754ba6 100644 --- a/drivers/char/drm/r300_cmdbuf.c +++ b/drivers/char/drm/r300_cmdbuf.c @@ -161,6 +161,7 @@ void r300_init_reg_flags(void) ADD_RANGE(R300_VAP_PVS_CNTL_1, 3); ADD_RANGE(R300_GB_ENABLE, 1); ADD_RANGE(R300_GB_MSPOS0, 5); + ADD_RANGE(R300_TX_CNTL, 1); ADD_RANGE(R300_TX_ENABLE, 1); ADD_RANGE(0x4200, 4); ADD_RANGE(0x4214, 1); @@ -489,6 +490,52 @@ static __inline__ int r300_emit_3d_load_vbpntr(drm_radeon_private_t *dev_priv, return 0; } +static __inline__ int r300_emit_bitblt_multi(drm_radeon_private_t *dev_priv, + drm_radeon_kcmd_buffer_t *cmdbuf) +{ + u32 *cmd = (u32 *) cmdbuf->buf; + int count, ret; + RING_LOCALS; + + count=(cmd[0]>>16) & 0x3fff; + + if (cmd[0] & 0x8000) { + u32 offset; + + if (cmd[1] & (RADEON_GMC_SRC_PITCH_OFFSET_CNTL + | RADEON_GMC_DST_PITCH_OFFSET_CNTL)) { + offset = cmd[2] << 10; + ret = r300_check_offset(dev_priv, offset); + if (ret) + { + DRM_ERROR("Invalid bitblt first offset is %08X\n", offset); + return DRM_ERR(EINVAL); + } + } + + if ((cmd[1] & RADEON_GMC_SRC_PITCH_OFFSET_CNTL) && + (cmd[1] & RADEON_GMC_DST_PITCH_OFFSET_CNTL)) { + offset = cmd[3] << 10; + ret = r300_check_offset(dev_priv, offset); + if (ret) + { + DRM_ERROR("Invalid bitblt second offset is %08X\n", offset); + return DRM_ERR(EINVAL); + } + + } + } + + BEGIN_RING(count+2); + OUT_RING(cmd[0]); + OUT_RING_TABLE((int *)(cmdbuf->buf + 4), count + 1); + ADVANCE_RING(); + + cmdbuf->buf += (count+2)*4; + cmdbuf->bufsz -= (count+2)*4; + + return 0; +} static __inline__ int r300_emit_raw_packet3(drm_radeon_private_t *dev_priv, drm_radeon_kcmd_buffer_t *cmdbuf) @@ -527,6 +574,9 @@ static __inline__ int r300_emit_raw_packet3(drm_radeon_private_t *dev_priv, case RADEON_3D_LOAD_VBPNTR: /* load vertex array pointers */ return r300_emit_3d_load_vbpntr(dev_priv, cmdbuf, header); + case RADEON_CNTL_BITBLT_MULTI: + return r300_emit_bitblt_multi(dev_priv, cmdbuf); + case RADEON_CP_3D_DRAW_IMMD_2: /* triggers drawing using in-packet vertex data */ case RADEON_CP_3D_DRAW_VBUF_2: /* triggers drawing of vertex buffers setup elsewhere */ case RADEON_CP_3D_DRAW_INDX_2: /* triggers drawing using indices to vertex buffer */ diff --git a/drivers/char/drm/r300_reg.h b/drivers/char/drm/r300_reg.h index a0ed20e2522..d1e19954406 100644 --- a/drivers/char/drm/r300_reg.h +++ b/drivers/char/drm/r300_reg.h @@ -451,6 +451,9 @@ I am fairly certain that they are correct unless stated otherwise in comments. /* END */ /* gap */ +/* Zero to flush caches. */ +#define R300_TX_CNTL 0x4100 + /* The upper enable bits are guessed, based on fglrx reported limits. */ #define R300_TX_ENABLE 0x4104 # define R300_TX_ENABLE_0 (1 << 0) diff --git a/drivers/char/drm/radeon_drv.h b/drivers/char/drm/radeon_drv.h index 498b19b1d64..1f7d2ab8c4f 100644 --- a/drivers/char/drm/radeon_drv.h +++ b/drivers/char/drm/radeon_drv.h @@ -90,9 +90,10 @@ * 1.19- Add support for gart table in FB memory and PCIE r300 * 1.20- Add support for r300 texrect * 1.21- Add support for card type getparam + * 1.22- Add support for texture cache flushes (R300_TX_CNTL) */ #define DRIVER_MAJOR 1 -#define DRIVER_MINOR 21 +#define DRIVER_MINOR 22 #define DRIVER_PATCHLEVEL 0 /* From 73d72cffe53407e447df0cbb0bf15a2c931108b3 Mon Sep 17 00:00:00 2001 From: Dave Airlie Date: Sat, 18 Feb 2006 16:30:54 +1100 Subject: [PATCH 005/102] drm: fix brace placement Signed-off-by: Dave Airlie --- drivers/char/drm/r300_cmdbuf.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/drivers/char/drm/r300_cmdbuf.c b/drivers/char/drm/r300_cmdbuf.c index 6dd21754ba6..c08fa5076f0 100644 --- a/drivers/char/drm/r300_cmdbuf.c +++ b/drivers/char/drm/r300_cmdbuf.c @@ -506,8 +506,7 @@ static __inline__ int r300_emit_bitblt_multi(drm_radeon_private_t *dev_priv, | RADEON_GMC_DST_PITCH_OFFSET_CNTL)) { offset = cmd[2] << 10; ret = r300_check_offset(dev_priv, offset); - if (ret) - { + if (ret) { DRM_ERROR("Invalid bitblt first offset is %08X\n", offset); return DRM_ERR(EINVAL); } @@ -517,8 +516,7 @@ static __inline__ int r300_emit_bitblt_multi(drm_radeon_private_t *dev_priv, (cmd[1] & RADEON_GMC_DST_PITCH_OFFSET_CNTL)) { offset = cmd[3] << 10; ret = r300_check_offset(dev_priv, offset); - if (ret) - { + if (ret) { DRM_ERROR("Invalid bitblt second offset is %08X\n", offset); return DRM_ERR(EINVAL); } From ef20c8c197df9b8d5bd4af0679123826da028861 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 18 Feb 2006 15:41:50 -0500 Subject: [PATCH 006/102] [PATCH] GFP_KERNEL allocations in atomic (auditsc) audit_log_exit() is called from atomic contexts and gets explicit gfp_mask argument; it should use it for all allocations rather than doing some with gfp_mask and some with GFP_KERNEL. Signed-off-by: Al Viro --- kernel/auditsc.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/kernel/auditsc.c b/kernel/auditsc.c index 685c25175d9..d7e7e637b92 100644 --- a/kernel/auditsc.c +++ b/kernel/auditsc.c @@ -841,7 +841,7 @@ static void audit_log_exit(struct audit_context *context, gfp_t gfp_mask) for (aux = context->aux; aux; aux = aux->next) { - ab = audit_log_start(context, GFP_KERNEL, aux->type); + ab = audit_log_start(context, gfp_mask, aux->type); if (!ab) continue; /* audit_panic has been called */ @@ -878,14 +878,14 @@ static void audit_log_exit(struct audit_context *context, gfp_t gfp_mask) } if (context->pwd && context->pwdmnt) { - ab = audit_log_start(context, GFP_KERNEL, AUDIT_CWD); + ab = audit_log_start(context, gfp_mask, AUDIT_CWD); if (ab) { audit_log_d_path(ab, "cwd=", context->pwd, context->pwdmnt); audit_log_end(ab); } } for (i = 0; i < context->name_count; i++) { - ab = audit_log_start(context, GFP_KERNEL, AUDIT_PATH); + ab = audit_log_start(context, gfp_mask, AUDIT_PATH); if (!ab) continue; /* audit_panic has been called */ From e30809fde59d591809f00caa1a4c960cca5916af Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 8 Feb 2006 14:09:00 -0500 Subject: [PATCH 007/102] [PATCH] don't mangle INQUIRY if cmddt or evpd bits are set sbp2.c mangles INQUIRY response in a way that only applies to standard inquiry data (i.e. when both cmddt and evpd bits are 0). Leave other cases alone; e.g. when asking for VPD the length of reply is in byte 3, not 4 and byte 4 is the first byte of device serial number. Signed-off-by: Al Viro --- drivers/ieee1394/sbp2.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) diff --git a/drivers/ieee1394/sbp2.c b/drivers/ieee1394/sbp2.c index 18d7eda3885..c2c776fbda0 100644 --- a/drivers/ieee1394/sbp2.c +++ b/drivers/ieee1394/sbp2.c @@ -2082,9 +2082,7 @@ static void sbp2_check_sbp2_response(struct scsi_id_instance_data *scsi_id, SBP2_DEBUG("sbp2_check_sbp2_response"); - switch (SCpnt->cmnd[0]) { - - case INQUIRY: + if (SCpnt->cmnd[0] == INQUIRY && (SCpnt->cmnd[1] & 3) == 0) { /* * Make sure data length is ok. Minimum length is 36 bytes */ @@ -2097,13 +2095,7 @@ static void sbp2_check_sbp2_response(struct scsi_id_instance_data *scsi_id, */ scsi_buf[2] |= 2; scsi_buf[3] = (scsi_buf[3] & 0xf0) | 2; - - break; - - default: - break; } - return; } /* From 76b6159ba094544e003a237cedcf555d82fa3bfe Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 8 Feb 2006 14:37:40 -0500 Subject: [PATCH 008/102] [PATCH] fix handling of st_nlink on procfs root 1) it should use nr_processes(), not nr_threads; otherwise we are getting very confused find(1) and friends, among other things. 2) better do that at stat() time than at every damn lookup in procfs root. Patch had been sitting in FC4 kernels for many months now... Signed-off-by: Al Viro --- fs/proc/inode.c | 4 ---- fs/proc/root.c | 17 +++++++++-------- 2 files changed, 9 insertions(+), 12 deletions(-) diff --git a/fs/proc/inode.c b/fs/proc/inode.c index 6573f31f1fd..075d3e94560 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -204,10 +204,6 @@ int proc_fill_super(struct super_block *s, void *data, int silent) root_inode = proc_get_inode(s, PROC_ROOT_INO, &proc_root); if (!root_inode) goto out_no_root; - /* - * Fixup the root inode's nlink value - */ - root_inode->i_nlink += nr_processes(); root_inode->i_uid = 0; root_inode->i_gid = 0; s->s_root = d_alloc_root(root_inode); diff --git a/fs/proc/root.c b/fs/proc/root.c index 68896283c8a..c3fd3611112 100644 --- a/fs/proc/root.c +++ b/fs/proc/root.c @@ -80,16 +80,16 @@ void __init proc_root_init(void) proc_bus = proc_mkdir("bus", NULL); } +static int proc_root_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat +) +{ + generic_fillattr(dentry->d_inode, stat); + stat->nlink = proc_root.nlink + nr_processes(); + return 0; +} + static struct dentry *proc_root_lookup(struct inode * dir, struct dentry * dentry, struct nameidata *nd) { - /* - * nr_threads is actually protected by the tasklist_lock; - * however, it's conventional to do reads, especially for - * reporting, without any locking whatsoever. - */ - if (dir->i_ino == PROC_ROOT_INO) /* check for safety... */ - dir->i_nlink = proc_root.nlink + nr_threads; - if (!proc_lookup(dir, dentry, nd)) { return NULL; } @@ -134,6 +134,7 @@ static struct file_operations proc_root_operations = { */ static struct inode_operations proc_root_inode_operations = { .lookup = proc_root_lookup, + .getattr = proc_root_getattr, }; /* From 00fc00df9e7b637cd13fe1f163da0a2957273947 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 18 Jan 2006 21:22:55 -0500 Subject: [PATCH 009/102] [PATCH] m68k: restore disable_irq_nosync() Patch claiming to remove enable_irq_nosync() had left it alive but killed disable_irq_nosync() instead... Signed-off-by: Al Viro --- include/asm-m68k/irq.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/asm-m68k/irq.h b/include/asm-m68k/irq.h index 325c86f8512..9ac047c400c 100644 --- a/include/asm-m68k/irq.h +++ b/include/asm-m68k/irq.h @@ -79,7 +79,7 @@ static __inline__ int irq_canonicalize(int irq) extern void (*enable_irq)(unsigned int); extern void (*disable_irq)(unsigned int); -#define enable_irq_nosync enable_irq +#define disable_irq_nosync disable_irq struct pt_regs; From cc6cdac0cf11955dc81d6c50b6738d05e1dca12b Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 18 Feb 2006 16:02:18 -0500 Subject: [PATCH 010/102] [PATCH] missing ntohs() in ip6_tunnel ->payload_len is net-endian Signed-off-by: Al Viro --- net/ipv6/ip6_tunnel.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 92ead3cf956..faea8a120ee 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -458,7 +458,7 @@ ip6ip6_err(struct sk_buff *skb, struct inet6_skb_parm *opt, mtu = IPV6_MIN_MTU; t->dev->mtu = mtu; - if ((len = sizeof (*ipv6h) + ipv6h->payload_len) > mtu) { + if ((len = sizeof (*ipv6h) + ntohs(ipv6h->payload_len)) > mtu) { rel_type = ICMPV6_PKT_TOOBIG; rel_code = 0; rel_info = mtu; From cead14da59fc261534fa749886c12c16757711fd Mon Sep 17 00:00:00 2001 From: Al Viro Date: Wed, 18 Jan 2006 19:41:35 -0500 Subject: [PATCH 011/102] [PATCH] m68k: pm_power_off() breakage Signed-off-by: Al Viro --- arch/m68k/kernel/process.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/m68k/kernel/process.c b/arch/m68k/kernel/process.c index 3f9cb55d035..2d8ad0727b6 100644 --- a/arch/m68k/kernel/process.c +++ b/arch/m68k/kernel/process.c @@ -129,6 +129,9 @@ void machine_power_off(void) for (;;); } +void (*pm_power_off)(void) = machine_power_off; +EXPORT_SYMBOL(pm_power_off); + void show_regs(struct pt_regs * regs) { printk("\n"); From ad6b97fc929e5844bfd1d708ab1d74d131d7960d Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 3 Feb 2006 02:06:42 -0500 Subject: [PATCH 012/102] [PATCH] iomap_copy fallout (m68k) added __raw_writel(), sanitized include order in iomap_copy.c Signed-off-by: Al Viro --- include/asm-m68k/raw_io.h | 1 + lib/iomap_copy.c | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/include/asm-m68k/raw_io.h b/include/asm-m68k/raw_io.h index 5439bcaa57c..811ccd25d4a 100644 --- a/include/asm-m68k/raw_io.h +++ b/include/asm-m68k/raw_io.h @@ -336,6 +336,7 @@ static inline void raw_outsw_swapw(volatile u16 __iomem *port, const u16 *buf, : "d0", "a0", "a1", "d6"); } +#define __raw_writel raw_outl #endif /* __KERNEL__ */ diff --git a/lib/iomap_copy.c b/lib/iomap_copy.c index a6b1e271d53..351045f4f63 100644 --- a/lib/iomap_copy.c +++ b/lib/iomap_copy.c @@ -15,8 +15,8 @@ * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA. */ -#include #include +#include /** * __iowrite32_copy - copy data to MMIO space, in 32-bit units From 092b8f3488a3e50a4ab5f2f3f7c8bbf56b3144e1 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Mon, 20 Feb 2006 10:38:56 +1100 Subject: [PATCH 013/102] powerpc: Keep xtime and gettimeofday in sync This fixes a regression which was introduced by moving ppc32 to use the same sort of lockless gettimeofday as ppc64 has been using for some time. This involves getting the timebase and performing some simple arithmetic to convert it to seconds and microseconds. However, the factor and offset used there weren't being updated when NTP varied the tick length using adjtimex. 64-bit didn't notice the problem because it had a hook in the 32-bit adjtimex compat routine that attempted to work out what the generic timekeeping code would do and alter the factor and offset to match. However, that code was very complex and it wasn't clear that it still matched what the generic code would do. Now we use the generic current_tick_length() routine that was recently added to check that the current tick will be as long as we expect; if not we recompute the factor and offset. This keeps gettimeofday and xtime in sync. In addition we check that gettimeofday hasn't got ahead of xtime on each timer interrupt; if it has, we resync. Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/sys_ppc32.c | 4 - arch/powerpc/kernel/time.c | 282 +++++++++++--------------------- 2 files changed, 99 insertions(+), 187 deletions(-) diff --git a/arch/powerpc/kernel/sys_ppc32.c b/arch/powerpc/kernel/sys_ppc32.c index 475249dc235..cd75ab2908f 100644 --- a/arch/powerpc/kernel/sys_ppc32.c +++ b/arch/powerpc/kernel/sys_ppc32.c @@ -176,7 +176,6 @@ struct timex32 { }; extern int do_adjtimex(struct timex *); -extern void ppc_adjtimex(void); asmlinkage long compat_sys_adjtimex(struct timex32 __user *utp) { @@ -209,9 +208,6 @@ asmlinkage long compat_sys_adjtimex(struct timex32 __user *utp) ret = do_adjtimex(&txc); - /* adjust the conversion of TB to time of day to track adjtimex */ - ppc_adjtimex(); - if(put_user(txc.modes, &utp->modes) || __put_user(txc.offset, &utp->offset) || __put_user(txc.freq, &utp->freq) || diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c index 1886045a2fd..2a7ddc57937 100644 --- a/arch/powerpc/kernel/time.c +++ b/arch/powerpc/kernel/time.c @@ -50,6 +50,7 @@ #include #include #include +#include #include #include @@ -99,7 +100,15 @@ EXPORT_SYMBOL(tb_ticks_per_usec); unsigned long tb_ticks_per_sec; u64 tb_to_xs; unsigned tb_to_us; -unsigned long processor_freq; + +#define TICKLEN_SCALE (SHIFT_SCALE - 10) +u64 last_tick_len; /* units are ns / 2^TICKLEN_SCALE */ +u64 ticklen_to_xs; /* 0.64 fraction */ + +/* If last_tick_len corresponds to about 1/HZ seconds, then + last_tick_len << TICKLEN_SHIFT will be about 2^63. */ +#define TICKLEN_SHIFT (63 - 30 - TICKLEN_SCALE + SHIFT_HZ) + DEFINE_SPINLOCK(rtc_lock); EXPORT_SYMBOL_GPL(rtc_lock); @@ -113,10 +122,6 @@ extern unsigned long wall_jiffies; extern struct timezone sys_tz; static long timezone_offset; -void ppc_adjtimex(void); - -static unsigned adjusting_time = 0; - unsigned long ppc_proc_freq; unsigned long ppc_tb_freq; @@ -178,8 +183,7 @@ static __inline__ void timer_check_rtc(void) */ if (ppc_md.set_rtc_time && ntp_synced() && xtime.tv_sec - last_rtc_update >= 659 && - abs((xtime.tv_nsec/1000) - (1000000-1000000/HZ)) < 500000/HZ && - jiffies - wall_jiffies == 1) { + abs((xtime.tv_nsec/1000) - (1000000-1000000/HZ)) < 500000/HZ) { struct rtc_time tm; to_tm(xtime.tv_sec + 1 + timezone_offset, &tm); tm.tm_year -= 1900; @@ -226,15 +230,14 @@ void do_gettimeofday(struct timeval *tv) if (__USE_RTC()) { /* do this the old way */ unsigned long flags, seq; - unsigned int sec, nsec, usec, lost; + unsigned int sec, nsec, usec; do { seq = read_seqbegin_irqsave(&xtime_lock, flags); sec = xtime.tv_sec; nsec = xtime.tv_nsec + tb_ticks_since(tb_last_stamp); - lost = jiffies - wall_jiffies; } while (read_seqretry_irqrestore(&xtime_lock, seq, flags)); - usec = nsec / 1000 + lost * (1000000 / HZ); + usec = nsec / 1000; while (usec >= 1000000) { usec -= 1000000; ++sec; @@ -248,23 +251,6 @@ void do_gettimeofday(struct timeval *tv) EXPORT_SYMBOL(do_gettimeofday); -/* Synchronize xtime with do_gettimeofday */ - -static inline void timer_sync_xtime(unsigned long cur_tb) -{ -#ifdef CONFIG_PPC64 - /* why do we do this? */ - struct timeval my_tv; - - __do_gettimeofday(&my_tv, cur_tb); - - if (xtime.tv_sec <= my_tv.tv_sec) { - xtime.tv_sec = my_tv.tv_sec; - xtime.tv_nsec = my_tv.tv_usec * 1000; - } -#endif -} - /* * There are two copies of tb_to_xs and stamp_xsec so that no * lock is needed to access and use these values in @@ -323,15 +309,30 @@ static __inline__ void timer_recalc_offset(u64 cur_tb) { unsigned long offset; u64 new_stamp_xsec; + u64 tlen, t2x; if (__USE_RTC()) return; + tlen = current_tick_length(); offset = cur_tb - do_gtod.varp->tb_orig_stamp; - if ((offset & 0x80000000u) == 0) - return; - new_stamp_xsec = do_gtod.varp->stamp_xsec - + mulhdu(offset, do_gtod.varp->tb_to_xs); - update_gtod(cur_tb, new_stamp_xsec, do_gtod.varp->tb_to_xs); + if (tlen == last_tick_len && offset < 0x80000000u) { + /* check that we're still in sync; if not, resync */ + struct timeval tv; + __do_gettimeofday(&tv, cur_tb); + if (tv.tv_sec <= xtime.tv_sec && + (tv.tv_sec < xtime.tv_sec || + tv.tv_usec * 1000 <= xtime.tv_nsec)) + return; + } + if (tlen != last_tick_len) { + t2x = mulhdu(tlen << TICKLEN_SHIFT, ticklen_to_xs); + last_tick_len = tlen; + } else + t2x = do_gtod.varp->tb_to_xs; + new_stamp_xsec = (u64) xtime.tv_nsec * XSEC_PER_SEC; + do_div(new_stamp_xsec, 1000000000); + new_stamp_xsec += (u64) xtime.tv_sec * XSEC_PER_SEC; + update_gtod(cur_tb, new_stamp_xsec, t2x); } #ifdef CONFIG_SMP @@ -462,13 +463,10 @@ void timer_interrupt(struct pt_regs * regs) write_seqlock(&xtime_lock); tb_last_jiffy += tb_ticks_per_jiffy; tb_last_stamp = per_cpu(last_jiffy, cpu); - timer_recalc_offset(tb_last_jiffy); do_timer(regs); - timer_sync_xtime(tb_last_jiffy); + timer_recalc_offset(tb_last_jiffy); timer_check_rtc(); write_sequnlock(&xtime_lock); - if (adjusting_time && (time_adjust == 0)) - ppc_adjtimex(); } next_dec = tb_ticks_per_jiffy - ticks; @@ -492,16 +490,18 @@ void timer_interrupt(struct pt_regs * regs) void wakeup_decrementer(void) { - int i; + unsigned long ticks; - set_dec(tb_ticks_per_jiffy); /* - * We don't expect this to be called on a machine with a 601, - * so using get_tbl is fine. + * The timebase gets saved on sleep and restored on wakeup, + * so all we need to do is to reset the decrementer. */ - tb_last_stamp = tb_last_jiffy = get_tb(); - for_each_cpu(i) - per_cpu(last_jiffy, i) = tb_last_stamp; + ticks = tb_ticks_since(__get_cpu_var(last_jiffy)); + if (ticks < tb_ticks_per_jiffy) + ticks = tb_ticks_per_jiffy - ticks; + else + ticks = 1; + set_dec(ticks); } #ifdef CONFIG_SMP @@ -541,8 +541,8 @@ int do_settimeofday(struct timespec *tv) time_t wtm_sec, new_sec = tv->tv_sec; long wtm_nsec, new_nsec = tv->tv_nsec; unsigned long flags; - long int tb_delta; - u64 new_xsec, tb_delta_xs; + u64 new_xsec; + unsigned long tb_delta; if ((unsigned long)tv->tv_nsec >= NSEC_PER_SEC) return -EINVAL; @@ -563,9 +563,19 @@ int do_settimeofday(struct timespec *tv) first_settimeofday = 0; } #endif + + /* + * Subtract off the number of nanoseconds since the + * beginning of the last tick. + * Note that since we don't increment jiffies_64 anywhere other + * than in do_timer (since we don't have a lost tick problem), + * wall_jiffies will always be the same as jiffies, + * and therefore the (jiffies - wall_jiffies) computation + * has been removed. + */ tb_delta = tb_ticks_since(tb_last_stamp); - tb_delta += (jiffies - wall_jiffies) * tb_ticks_per_jiffy; - tb_delta_xs = mulhdu(tb_delta, do_gtod.varp->tb_to_xs); + tb_delta = mulhdu(tb_delta, do_gtod.varp->tb_to_xs); /* in xsec */ + new_nsec -= SCALE_XSEC(tb_delta, 1000000000); wtm_sec = wall_to_monotonic.tv_sec + (xtime.tv_sec - new_sec); wtm_nsec = wall_to_monotonic.tv_nsec + (xtime.tv_nsec - new_nsec); @@ -580,12 +590,12 @@ int do_settimeofday(struct timespec *tv) ntp_clear(); - new_xsec = 0; - if (new_nsec != 0) { - new_xsec = (u64)new_nsec * XSEC_PER_SEC; + new_xsec = xtime.tv_nsec; + if (new_xsec != 0) { + new_xsec *= XSEC_PER_SEC; do_div(new_xsec, NSEC_PER_SEC); } - new_xsec += (u64)new_sec * XSEC_PER_SEC - tb_delta_xs; + new_xsec += (u64)xtime.tv_sec * XSEC_PER_SEC; update_gtod(tb_last_jiffy, new_xsec, do_gtod.varp->tb_to_xs); vdso_data->tz_minuteswest = sys_tz.tz_minuteswest; @@ -671,7 +681,7 @@ void __init time_init(void) unsigned long flags; unsigned long tm = 0; struct div_result res; - u64 scale; + u64 scale, x; unsigned shift; if (ppc_md.time_init != NULL) @@ -693,11 +703,36 @@ void __init time_init(void) } tb_ticks_per_jiffy = ppc_tb_freq / HZ; - tb_ticks_per_sec = tb_ticks_per_jiffy * HZ; + tb_ticks_per_sec = ppc_tb_freq; tb_ticks_per_usec = ppc_tb_freq / 1000000; tb_to_us = mulhwu_scale_factor(ppc_tb_freq, 1000000); - div128_by_32(1024*1024, 0, tb_ticks_per_sec, &res); - tb_to_xs = res.result_low; + + /* + * Calculate the length of each tick in ns. It will not be + * exactly 1e9/HZ unless ppc_tb_freq is divisible by HZ. + * We compute 1e9 * tb_ticks_per_jiffy / ppc_tb_freq, + * rounded up. + */ + x = (u64) NSEC_PER_SEC * tb_ticks_per_jiffy + ppc_tb_freq - 1; + do_div(x, ppc_tb_freq); + tick_nsec = x; + last_tick_len = x << TICKLEN_SCALE; + + /* + * Compute ticklen_to_xs, which is a factor which gets multiplied + * by (last_tick_len << TICKLEN_SHIFT) to get a tb_to_xs value. + * It is computed as: + * ticklen_to_xs = 2^N / (tb_ticks_per_jiffy * 1e9) + * where N = 64 + 20 - TICKLEN_SCALE - TICKLEN_SHIFT + * so as to give the result as a 0.64 fixed-point fraction. + */ + div128_by_32(1ULL << (64 + 20 - TICKLEN_SCALE - TICKLEN_SHIFT), 0, + tb_ticks_per_jiffy, &res); + div128_by_32(res.result_high, res.result_low, NSEC_PER_SEC, &res); + ticklen_to_xs = res.result_low; + + /* Compute tb_to_xs from tick_nsec */ + tb_to_xs = mulhdu(last_tick_len << TICKLEN_SHIFT, ticklen_to_xs); /* * Compute scale factor for sched_clock. @@ -724,6 +759,14 @@ void __init time_init(void) tm = get_boot_time(); write_seqlock_irqsave(&xtime_lock, flags); + + /* If platform provided a timezone (pmac), we correct the time */ + if (timezone_offset) { + sys_tz.tz_minuteswest = -timezone_offset / 60; + sys_tz.tz_dsttime = 0; + tm -= timezone_offset; + } + xtime.tv_sec = tm; xtime.tv_nsec = 0; do_gtod.varp = &do_gtod.vars[0]; @@ -738,18 +781,11 @@ void __init time_init(void) vdso_data->tb_orig_stamp = tb_last_jiffy; vdso_data->tb_update_count = 0; vdso_data->tb_ticks_per_sec = tb_ticks_per_sec; - vdso_data->stamp_xsec = xtime.tv_sec * XSEC_PER_SEC; + vdso_data->stamp_xsec = (u64) xtime.tv_sec * XSEC_PER_SEC; vdso_data->tb_to_xs = tb_to_xs; time_freq = 0; - /* If platform provided a timezone (pmac), we correct the time */ - if (timezone_offset) { - sys_tz.tz_minuteswest = -timezone_offset / 60; - sys_tz.tz_dsttime = 0; - xtime.tv_sec -= timezone_offset; - } - last_rtc_update = xtime.tv_sec; set_normalized_timespec(&wall_to_monotonic, -xtime.tv_sec, -xtime.tv_nsec); @@ -759,126 +795,6 @@ void __init time_init(void) set_dec(tb_ticks_per_jiffy); } -/* - * After adjtimex is called, adjust the conversion of tb ticks - * to microseconds to keep do_gettimeofday synchronized - * with ntpd. - * - * Use the time_adjust, time_freq and time_offset computed by adjtimex to - * adjust the frequency. - */ - -/* #define DEBUG_PPC_ADJTIMEX 1 */ - -void ppc_adjtimex(void) -{ -#ifdef CONFIG_PPC64 - unsigned long den, new_tb_ticks_per_sec, tb_ticks, old_xsec, - new_tb_to_xs, new_xsec, new_stamp_xsec; - unsigned long tb_ticks_per_sec_delta; - long delta_freq, ltemp; - struct div_result divres; - unsigned long flags; - long singleshot_ppm = 0; - - /* - * Compute parts per million frequency adjustment to - * accomplish the time adjustment implied by time_offset to be - * applied over the elapsed time indicated by time_constant. - * Use SHIFT_USEC to get it into the same units as - * time_freq. - */ - if ( time_offset < 0 ) { - ltemp = -time_offset; - ltemp <<= SHIFT_USEC - SHIFT_UPDATE; - ltemp >>= SHIFT_KG + time_constant; - ltemp = -ltemp; - } else { - ltemp = time_offset; - ltemp <<= SHIFT_USEC - SHIFT_UPDATE; - ltemp >>= SHIFT_KG + time_constant; - } - - /* If there is a single shot time adjustment in progress */ - if ( time_adjust ) { -#ifdef DEBUG_PPC_ADJTIMEX - printk("ppc_adjtimex: "); - if ( adjusting_time == 0 ) - printk("starting "); - printk("single shot time_adjust = %ld\n", time_adjust); -#endif - - adjusting_time = 1; - - /* - * Compute parts per million frequency adjustment - * to match time_adjust - */ - singleshot_ppm = tickadj * HZ; - /* - * The adjustment should be tickadj*HZ to match the code in - * linux/kernel/timer.c, but experiments show that this is too - * large. 3/4 of tickadj*HZ seems about right - */ - singleshot_ppm -= singleshot_ppm / 4; - /* Use SHIFT_USEC to get it into the same units as time_freq */ - singleshot_ppm <<= SHIFT_USEC; - if ( time_adjust < 0 ) - singleshot_ppm = -singleshot_ppm; - } - else { -#ifdef DEBUG_PPC_ADJTIMEX - if ( adjusting_time ) - printk("ppc_adjtimex: ending single shot time_adjust\n"); -#endif - adjusting_time = 0; - } - - /* Add up all of the frequency adjustments */ - delta_freq = time_freq + ltemp + singleshot_ppm; - - /* - * Compute a new value for tb_ticks_per_sec based on - * the frequency adjustment - */ - den = 1000000 * (1 << (SHIFT_USEC - 8)); - if ( delta_freq < 0 ) { - tb_ticks_per_sec_delta = ( tb_ticks_per_sec * ( (-delta_freq) >> (SHIFT_USEC - 8))) / den; - new_tb_ticks_per_sec = tb_ticks_per_sec + tb_ticks_per_sec_delta; - } - else { - tb_ticks_per_sec_delta = ( tb_ticks_per_sec * ( delta_freq >> (SHIFT_USEC - 8))) / den; - new_tb_ticks_per_sec = tb_ticks_per_sec - tb_ticks_per_sec_delta; - } - -#ifdef DEBUG_PPC_ADJTIMEX - printk("ppc_adjtimex: ltemp = %ld, time_freq = %ld, singleshot_ppm = %ld\n", ltemp, time_freq, singleshot_ppm); - printk("ppc_adjtimex: tb_ticks_per_sec - base = %ld new = %ld\n", tb_ticks_per_sec, new_tb_ticks_per_sec); -#endif - - /* - * Compute a new value of tb_to_xs (used to convert tb to - * microseconds) and a new value of stamp_xsec which is the - * time (in 1/2^20 second units) corresponding to - * tb_orig_stamp. This new value of stamp_xsec compensates - * for the change in frequency (implied by the new tb_to_xs) - * which guarantees that the current time remains the same. - */ - write_seqlock_irqsave( &xtime_lock, flags ); - tb_ticks = get_tb() - do_gtod.varp->tb_orig_stamp; - div128_by_32(1024*1024, 0, new_tb_ticks_per_sec, &divres); - new_tb_to_xs = divres.result_low; - new_xsec = mulhdu(tb_ticks, new_tb_to_xs); - - old_xsec = mulhdu(tb_ticks, do_gtod.varp->tb_to_xs); - new_stamp_xsec = do_gtod.varp->stamp_xsec + old_xsec - new_xsec; - - update_gtod(do_gtod.varp->tb_orig_stamp, new_stamp_xsec, new_tb_to_xs); - - write_sequnlock_irqrestore( &xtime_lock, flags ); -#endif /* CONFIG_PPC64 */ -} - #define FEBRUARY 2 #define STARTOFTIME 1970 From 0728a2f99ef6efd1984f9e0ed59834c1cc602e6f Mon Sep 17 00:00:00 2001 From: Olaf Hering Date: Sat, 11 Feb 2006 18:21:47 +0100 Subject: [PATCH 014/102] [PATCH] powerpc: remove duplicate exports A few symbols are exported twice, remove them from ppc_ksyms.c Remove users of sys_ctrler in arch/ppc/ WARNING: vmlinux: duplicate symbol '__delay' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol '__up' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol '__down' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol '__down_interruptible' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol 'sys_ctrler' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol 'strncat' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol 'strncmp' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol 'strchr' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol 'strrchr' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol 'strnlen' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol 'strpbrk' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol 'memscan' previous definition was in vmlinux WARNING: vmlinux: duplicate symbol 'strstr' previous definition was in vmlinux Signed-off-by: Olaf Hering Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/ppc_ksyms.c | 16 ---------------- arch/ppc/kernel/ppc_ksyms.c | 8 -------- arch/ppc/xmon/start.c | 15 +-------------- include/asm-ppc/machdep.h | 13 ------------- 4 files changed, 1 insertion(+), 51 deletions(-) diff --git a/arch/powerpc/kernel/ppc_ksyms.c b/arch/powerpc/kernel/ppc_ksyms.c index d9a459c144d..8a731ea877b 100644 --- a/arch/powerpc/kernel/ppc_ksyms.c +++ b/arch/powerpc/kernel/ppc_ksyms.c @@ -79,15 +79,8 @@ EXPORT_SYMBOL(sys_sigreturn); EXPORT_SYMBOL(strcpy); EXPORT_SYMBOL(strncpy); EXPORT_SYMBOL(strcat); -EXPORT_SYMBOL(strncat); -EXPORT_SYMBOL(strchr); -EXPORT_SYMBOL(strrchr); -EXPORT_SYMBOL(strpbrk); -EXPORT_SYMBOL(strstr); EXPORT_SYMBOL(strlen); -EXPORT_SYMBOL(strnlen); EXPORT_SYMBOL(strcmp); -EXPORT_SYMBOL(strncmp); EXPORT_SYMBOL(strcasecmp); EXPORT_SYMBOL(csum_partial); @@ -185,9 +178,6 @@ EXPORT_SYMBOL(adb_try_handler_change); EXPORT_SYMBOL(cuda_request); EXPORT_SYMBOL(cuda_poll); #endif /* CONFIG_ADB_CUDA */ -#ifdef CONFIG_PPC_PMAC -EXPORT_SYMBOL(sys_ctrler); -#endif #ifdef CONFIG_VT EXPORT_SYMBOL(kd_mksound); #endif @@ -205,7 +195,6 @@ EXPORT_SYMBOL(__lshrdi3); EXPORT_SYMBOL(memcpy); EXPORT_SYMBOL(memset); EXPORT_SYMBOL(memmove); -EXPORT_SYMBOL(memscan); EXPORT_SYMBOL(memcmp); EXPORT_SYMBOL(memchr); @@ -214,7 +203,6 @@ EXPORT_SYMBOL(screen_info); #endif #ifdef CONFIG_PPC32 -EXPORT_SYMBOL(__delay); EXPORT_SYMBOL(timer_interrupt); EXPORT_SYMBOL(irq_desc); EXPORT_SYMBOL(tb_ticks_per_jiffy); @@ -222,10 +210,6 @@ EXPORT_SYMBOL(console_drivers); EXPORT_SYMBOL(cacheable_memcpy); #endif -EXPORT_SYMBOL(__up); -EXPORT_SYMBOL(__down); -EXPORT_SYMBOL(__down_interruptible); - #ifdef CONFIG_8xx EXPORT_SYMBOL(cpm_install_handler); EXPORT_SYMBOL(cpm_free_handler); diff --git a/arch/ppc/kernel/ppc_ksyms.c b/arch/ppc/kernel/ppc_ksyms.c index 15bd9b448a4..82adb460134 100644 --- a/arch/ppc/kernel/ppc_ksyms.c +++ b/arch/ppc/kernel/ppc_ksyms.c @@ -93,15 +93,8 @@ EXPORT_SYMBOL(test_and_change_bit); EXPORT_SYMBOL(strcpy); EXPORT_SYMBOL(strncpy); EXPORT_SYMBOL(strcat); -EXPORT_SYMBOL(strncat); -EXPORT_SYMBOL(strchr); -EXPORT_SYMBOL(strrchr); -EXPORT_SYMBOL(strpbrk); -EXPORT_SYMBOL(strstr); EXPORT_SYMBOL(strlen); -EXPORT_SYMBOL(strnlen); EXPORT_SYMBOL(strcmp); -EXPORT_SYMBOL(strncmp); EXPORT_SYMBOL(strcasecmp); EXPORT_SYMBOL(__div64_32); @@ -253,7 +246,6 @@ EXPORT_SYMBOL(memcpy); EXPORT_SYMBOL(cacheable_memcpy); EXPORT_SYMBOL(memset); EXPORT_SYMBOL(memmove); -EXPORT_SYMBOL(memscan); EXPORT_SYMBOL(memcmp); EXPORT_SYMBOL(memchr); diff --git a/arch/ppc/xmon/start.c b/arch/ppc/xmon/start.c index 4344cbe9b5c..484f5bb1aa3 100644 --- a/arch/ppc/xmon/start.c +++ b/arch/ppc/xmon/start.c @@ -146,19 +146,6 @@ xmon_map_scc(void) static int scc_initialized = 0; void xmon_init_scc(void); -extern void cuda_poll(void); - -static inline void do_poll_adb(void) -{ -#ifdef CONFIG_ADB_PMU - if (sys_ctrler == SYS_CTRLER_PMU) - pmu_poll_adb(); -#endif /* CONFIG_ADB_PMU */ -#ifdef CONFIG_ADB_CUDA - if (sys_ctrler == SYS_CTRLER_CUDA) - cuda_poll(); -#endif /* CONFIG_ADB_CUDA */ -} int xmon_write(void *handle, void *ptr, int nb) @@ -189,7 +176,7 @@ xmon_write(void *handle, void *ptr, int nb) ct = 0; for (i = 0; i < nb; ++i) { while ((*sccc & TXRDY) == 0) - do_poll_adb(); + ; c = p[i]; if (c == '\n' && !ct) { c = '\r'; diff --git a/include/asm-ppc/machdep.h b/include/asm-ppc/machdep.h index 39200def8d1..a3e8a45e45a 100644 --- a/include/asm-ppc/machdep.h +++ b/include/asm-ppc/machdep.h @@ -154,19 +154,6 @@ extern char cmd_line[COMMAND_LINE_SIZE]; extern void setup_pci_ptrs(void); -/* - * Power macintoshes have either a CUDA or a PMU controlling - * system reset, power, NVRAM, RTC. - */ -typedef enum sys_ctrler_kind { - SYS_CTRLER_UNKNOWN = 0, - SYS_CTRLER_CUDA = 1, - SYS_CTRLER_PMU = 2, - SYS_CTRLER_SMU = 3, -} sys_ctrler_t; - -extern sys_ctrler_t sys_ctrler; - #ifdef CONFIG_SMP struct smp_ops_t { void (*message_pass)(int target, int msg); From 2b9a32edba3af9ad4ccb23574bea0cc34455dc43 Mon Sep 17 00:00:00 2001 From: Olof Johansson Date: Wed, 15 Feb 2006 21:40:44 -0600 Subject: [PATCH 015/102] [PATCH] powerpc: Fix OOPS in lparcfg on G5 Fallback gracefully when reading /proc/ppc64/lparcfg when the /rtas device node can't be found. Signed-off-by: Olof Johansson Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/lparcfg.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/kernel/lparcfg.c b/arch/powerpc/kernel/lparcfg.c index 1ae96a8ed7e..e789fef4eb8 100644 --- a/arch/powerpc/kernel/lparcfg.c +++ b/arch/powerpc/kernel/lparcfg.c @@ -341,7 +341,7 @@ static int lparcfg_data(struct seq_file *m, void *v) const char *system_id = ""; unsigned int *lp_index_ptr, lp_index = 0; struct device_node *rtas_node; - int *lrdrp; + int *lrdrp = NULL; rootdn = find_path_device("/"); if (rootdn) { @@ -362,7 +362,9 @@ static int lparcfg_data(struct seq_file *m, void *v) seq_printf(m, "partition_id=%d\n", (int)lp_index); rtas_node = find_path_device("/rtas"); - lrdrp = (int *)get_property(rtas_node, "ibm,lrdr-capacity", NULL); + if (rtas_node) + lrdrp = (int *)get_property(rtas_node, "ibm,lrdr-capacity", + NULL); if (lrdrp == NULL) { partition_potential_processors = vdso_data->processorCount; From f018b36f3e1f21318066de8d01740d30e38b03d5 Mon Sep 17 00:00:00 2001 From: Michael Ellerman Date: Thu, 16 Feb 2006 14:13:50 +1100 Subject: [PATCH 016/102] [PATCH] powerpc: Don't start secondary CPUs in a UP && KEXEC kernel Because smp_release_cpus() is built for SMP || KEXEC, it's not safe to unconditionally call it from setup_system(). On a UP && KEXEC kernel we'll start up the secondary CPUs which will then go beserk and we die. Simple fix is to conditionally call smp_release_cpus() in setup_system(). With that in place we don't need the dummy definition of smp_release_cpus() because all call sites are #ifdef'ed either SMP or KEXEC. Signed-off-by: Michael Ellerman Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/setup_64.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c index a717dff695e..f96c49b03ba 100644 --- a/arch/powerpc/kernel/setup_64.c +++ b/arch/powerpc/kernel/setup_64.c @@ -311,8 +311,6 @@ void smp_release_cpus(void) DBG(" <- smp_release_cpus()\n"); } -#else -#define smp_release_cpus() #endif /* CONFIG_SMP || CONFIG_KEXEC */ /* @@ -473,10 +471,12 @@ void __init setup_system(void) check_smt_enabled(); smp_setup_cpu_maps(); +#ifdef CONFIG_SMP /* Release secondary cpus out of their spinloops at 0x60 now that * we can map physical -> logical CPU ids */ smp_release_cpus(); +#endif printk("Starting Linux PPC64 %s\n", system_utsname.version); From 8fca92705ef462f39e7db5a0f7100bcaae91bfd3 Mon Sep 17 00:00:00 2001 From: Michael Ellerman Date: Thu, 16 Feb 2006 14:13:51 +1100 Subject: [PATCH 017/102] [PATCH] powerpc: Make UP -> SMP kexec work again For UP to SMP kexec to work we need to jump into pSeries_secondary_smp_init event on a UP + KEXEC kernel. The secondary cpus will not find their hw_cpu_id in the paca and so they'll jump into kexec_wait, ready for a kexec. Signed-off-by: Michael Ellerman Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/head_64.S | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/arch/powerpc/kernel/head_64.S b/arch/powerpc/kernel/head_64.S index 41565962939..2b03a09fe5e 100644 --- a/arch/powerpc/kernel/head_64.S +++ b/arch/powerpc/kernel/head_64.S @@ -157,8 +157,7 @@ _GLOBAL(__secondary_hold) SET_REG_IMMEDIATE(r4, .hmt_init) mtctr r4 bctr -#else -#ifdef CONFIG_SMP +#elif defined(CONFIG_SMP) || defined(CONFIG_KEXEC) LOAD_REG_IMMEDIATE(r4, .pSeries_secondary_smp_init) mtctr r4 mr r3,r24 @@ -166,7 +165,6 @@ _GLOBAL(__secondary_hold) #else BUG_OPCODE #endif -#endif /* This value is used to mark exception frames on the stack. */ .section ".toc","aw" From 496b7a5159b8366b003bbc17f8c4e27f69b6779e Mon Sep 17 00:00:00 2001 From: Michael Ellerman Date: Thu, 16 Feb 2006 14:13:53 +1100 Subject: [PATCH 018/102] [PATCH] powerpc: Fix bug in spinup of renumbered secondary threads If the logical and physical cpu ids of a secondary thread don't match, we will fail to spin the thread up on pSeries machines due to a bug in pseries/smp.c We call the RTAS "start-cpu" method with the physical cpu id, the address of pSeries_secondary_smp_init and the value to pass that function in r3. Currently we pass "lcpu", the logical cpu id, but pSeries_secondary_smp_init expects the physical cpu id in r3. We should be passing pcpu instead. Signed-off-by: Michael Ellerman Signed-off-by: Paul Mackerras --- arch/powerpc/platforms/pseries/smp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/platforms/pseries/smp.c b/arch/powerpc/platforms/pseries/smp.c index 8e6b1ed1396..8d710af5075 100644 --- a/arch/powerpc/platforms/pseries/smp.c +++ b/arch/powerpc/platforms/pseries/smp.c @@ -292,7 +292,7 @@ static inline int __devinit smp_startup_cpu(unsigned int lcpu) if (start_cpu == RTAS_UNKNOWN_SERVICE) return 1; - status = rtas_call(start_cpu, 3, 1, NULL, pcpu, start_here, lcpu); + status = rtas_call(start_cpu, 3, 1, NULL, pcpu, start_here, pcpu); if (status != 0) { printk(KERN_ERR "start-cpu failed: %i\n", status); return 0; From 995110143880fd9cb255fa5df05f8950c56fb43a Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Sun, 19 Feb 2006 22:11:50 -0800 Subject: [PATCH 019/102] [XFRM]: Fix policy double put The policy is put once immediately and once at the error label, which results in the following Oops: kernel BUG at net/xfrm/xfrm_policy.c:250! invalid opcode: 0000 [#2] PREEMPT [...] CPU: 0 EIP: 0060:[] Not tainted VLI EFLAGS: 00210246 (2.6.16-rc3 #39) EIP is at __xfrm_policy_destroy+0xf/0x46 eax: d49f2000 ebx: d49f2000 ecx: f74bd880 edx: f74bd280 esi: d49f2000 edi: 00000001 ebp: cd506dcc esp: cd506dc8 ds: 007b es: 007b ss: 0068 Process ssh (pid: 31970, threadinfo=cd506000 task=cfb04a70) Stack: <0>cd506000 cd506e34 c028e92b ebde7280 cd506e58 cd506ec0 f74bd280 00000000 00000214 0000000a 0000000a 00000000 00000002 f7ae6000 00000000 cd506e58 cd506e14 c0299e36 f74bd280 e873fe00 c02943fd cd506ec0 ebde7280 f271f440 Call Trace: [] show_stack_log_lvl+0xaa/0xb5 [] show_registers+0x126/0x18c [] die+0x14e/0x1db [] do_trap+0x7c/0x96 [] do_invalid_op+0x89/0x93 [] error_code+0x4f/0x54 [] xfrm_lookup+0x349/0x3c2 [] ip6_datagram_connect+0x317/0x452 [] inet_dgram_connect+0x49/0x54 [] sys_connect+0x51/0x68 [] sys_socketcall+0x6f/0x166 [] syscall_call+0x7/0xb Signed-off-by: Patrick McHardy Signed-off-by: David S. Miller --- net/xfrm/xfrm_policy.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c index 98ec53bd3ac..5e6b05ac126 100644 --- a/net/xfrm/xfrm_policy.c +++ b/net/xfrm/xfrm_policy.c @@ -885,8 +885,6 @@ restart: * We can't enlist stable bundles either. */ write_unlock_bh(&policy->lock); - - xfrm_pol_put(policy); if (dst) dst_free(dst); From bc6e14b6f0b06fe93d809d22e257ddd275feeda9 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Sun, 19 Feb 2006 22:26:40 -0800 Subject: [PATCH 020/102] [NETFILTER]: Fix NAT PMTUD problems ICMP errors are only SNATed when their source matches the source of the connection they are related to, otherwise the source address is not changed. This creates problems with ICMP frag. required messages originating from a router behind the NAT, if private IPs are used the packet has a good change of getting dropped on the path to its destination. Always NAT ICMP errors similar to the original connection. Based on report by Al Viro. Signed-off-by: Patrick McHardy Signed-off-by: David S. Miller --- net/ipv4/netfilter/ip_nat_core.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/net/ipv4/netfilter/ip_nat_core.c b/net/ipv4/netfilter/ip_nat_core.c index c1a61462507..1741d555ad0 100644 --- a/net/ipv4/netfilter/ip_nat_core.c +++ b/net/ipv4/netfilter/ip_nat_core.c @@ -434,6 +434,7 @@ int ip_nat_icmp_reply_translation(struct sk_buff **pskb, } *inside; struct ip_conntrack_tuple inner, target; int hdrlen = (*pskb)->nh.iph->ihl * 4; + unsigned long statusbit; if (!skb_make_writable(pskb, hdrlen + sizeof(*inside))) return 0; @@ -495,17 +496,16 @@ int ip_nat_icmp_reply_translation(struct sk_buff **pskb, /* Change outer to look the reply to an incoming packet * (proto 0 means don't invert per-proto part). */ + if (manip == IP_NAT_MANIP_SRC) + statusbit = IPS_SRC_NAT; + else + statusbit = IPS_DST_NAT; - /* Obviously, we need to NAT destination IP, but source IP - should be NAT'ed only if it is from a NAT'd host. + /* Invert if this is reply dir. */ + if (dir == IP_CT_DIR_REPLY) + statusbit ^= IPS_NAT_MASK; - Explanation: some people use NAT for anonymizing. Also, - CERT recommends dropping all packets from private IP - addresses (although ICMP errors from internal links with - such addresses are not too uncommon, as Alan Cox points - out) */ - if (manip != IP_NAT_MANIP_SRC - || ((*pskb)->nh.iph->saddr == ct->tuplehash[dir].tuple.src.ip)) { + if (ct->status & statusbit) { invert_tuplepr(&target, &ct->tuplehash[!dir].tuple); if (!manip_pkt(0, pskb, 0, &target, manip)) return 0; From 669d32a293a348e692c365ddac2b23f3b907fcf1 Mon Sep 17 00:00:00 2001 From: Jean Tourrilhes Date: Sun, 19 Feb 2006 22:28:25 -0800 Subject: [PATCH 021/102] [IRDA]: irda-usb bug fixes This patch fixes 2 bugs in the USB-IrDA code. The first one is a buffer overrun in the RX path. We are now using IRDA_SKB_MAX_MTU when initializing the Rx URB. The second one is a potential stack recursion when unplugging the USB dongle. It seems that first we get the Rx URB with a generic error code, and after a while the Rx URB comes again with a "disconnect" error code. Since we are resubmitting the Rx URB immediately after receiving the first error one, we might enter an endless loop. When getting an error Rx URB, the patch defers the Rx URB resubmitting so that it gives us a chance to catch the disconnect one, in case the dongle has juts been unplugged. Tested against 2.6.16-rc2. Patch from Jean Tourrilhes Signed-off-by: Jean Tourrilhes Signed-off-by: Samuel Ortiz Signed-off-by: David S. Miller --- drivers/net/irda/irda-usb.c | 90 +++++++++++++++++++++++++++++++------ drivers/net/irda/irda-usb.h | 7 ++- 2 files changed, 79 insertions(+), 18 deletions(-) diff --git a/drivers/net/irda/irda-usb.c b/drivers/net/irda/irda-usb.c index fa176ffb4ad..8936058a3cc 100644 --- a/drivers/net/irda/irda-usb.c +++ b/drivers/net/irda/irda-usb.c @@ -108,6 +108,7 @@ static void irda_usb_close(struct irda_usb_cb *self); static void speed_bulk_callback(struct urb *urb, struct pt_regs *regs); static void write_bulk_callback(struct urb *urb, struct pt_regs *regs); static void irda_usb_receive(struct urb *urb, struct pt_regs *regs); +static void irda_usb_rx_defer_expired(unsigned long data); static int irda_usb_net_open(struct net_device *dev); static int irda_usb_net_close(struct net_device *dev); static int irda_usb_net_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); @@ -677,6 +678,12 @@ static void irda_usb_net_timeout(struct net_device *netdev) * on the interrupt pipe and hang the Rx URB only when an interrupt is * received. * Jean II + * + * Note : don't read the above as what we are currently doing, but as + * something we could do with KC dongle. Also don't forget that the + * interrupt pipe is not part of the original standard, so this would + * need to be optional... + * Jean II */ /*------------------------------------------------------------------*/ @@ -704,10 +711,8 @@ static void irda_usb_submit(struct irda_usb_cb *self, struct sk_buff *skb, struc /* Reinitialize URB */ usb_fill_bulk_urb(urb, self->usbdev, usb_rcvbulkpipe(self->usbdev, self->bulk_in_ep), - skb->data, skb->truesize, + skb->data, IRDA_SKB_MAX_MTU, irda_usb_receive, skb); - /* Note : unlink *must* be synchronous because of the code in - * irda_usb_net_close() -> free the skb - Jean II */ urb->status = 0; /* Can be called from irda_usb_receive (irq handler) -> GFP_ATOMIC */ @@ -734,6 +739,7 @@ static void irda_usb_receive(struct urb *urb, struct pt_regs *regs) struct irda_skb_cb *cb; struct sk_buff *newskb; struct sk_buff *dataskb; + struct urb *next_urb; int docopy; IRDA_DEBUG(2, "%s(), len=%d\n", __FUNCTION__, urb->actual_length); @@ -755,20 +761,37 @@ static void irda_usb_receive(struct urb *urb, struct pt_regs *regs) if (urb->status != 0) { switch (urb->status) { case -EILSEQ: - self->stats.rx_errors++; self->stats.rx_crc_errors++; - break; + /* Also precursor to a hot-unplug on UHCI. */ + /* Fallthrough... */ case -ECONNRESET: /* -104 */ - IRDA_DEBUG(0, "%s(), Connection Reset (-104), transfer_flags 0x%04X \n", __FUNCTION__, urb->transfer_flags); + /* Random error, if I remember correctly */ /* uhci_cleanup_unlink() is going to kill the Rx * URB just after we return. No problem, at this * point the URB will be idle ;-) - Jean II */ - break; + case -ESHUTDOWN: /* -108 */ + /* That's usually a hot-unplug. Submit will fail... */ + case -ETIMEDOUT: /* -110 */ + /* Usually precursor to a hot-unplug on OHCI. */ default: - IRDA_DEBUG(0, "%s(), RX status %d,transfer_flags 0x%04X \n", __FUNCTION__, urb->status, urb->transfer_flags); + self->stats.rx_errors++; + IRDA_DEBUG(0, "%s(), RX status %d, transfer_flags 0x%04X \n", __FUNCTION__, urb->status, urb->transfer_flags); break; } - goto done; + /* If we received an error, we don't want to resubmit the + * Rx URB straight away but to give the USB layer a little + * bit of breathing room. + * We are in the USB thread context, therefore there is a + * danger of recursion (new URB we submit fails, we come + * back here). + * With recent USB stack (2.6.15+), I'm seeing that on + * hot unplug of the dongle... + * Lowest effective timer is 10ms... + * Jean II */ + self->rx_defer_timer.function = &irda_usb_rx_defer_expired; + self->rx_defer_timer.data = (unsigned long) urb; + mod_timer(&self->rx_defer_timer, jiffies + (10 * HZ / 1000)); + return; } /* Check for empty frames */ @@ -845,13 +868,45 @@ done: * idle slot.... * Jean II */ /* Note : with this scheme, we could submit the idle URB before - * processing the Rx URB. Another time... Jean II */ + * processing the Rx URB. I don't think it would buy us anything as + * we are running in the USB thread context. Jean II */ + next_urb = self->idle_rx_urb; - /* Submit the idle URB to replace the URB we've just received */ - irda_usb_submit(self, skb, self->idle_rx_urb); /* Recycle Rx URB : Now, the idle URB is the present one */ urb->context = NULL; self->idle_rx_urb = urb; + + /* Submit the idle URB to replace the URB we've just received. + * Do it last to avoid race conditions... Jean II */ + irda_usb_submit(self, skb, next_urb); +} + +/*------------------------------------------------------------------*/ +/* + * In case of errors, we want the USB layer to have time to recover. + * Now, it is time to resubmit ouur Rx URB... + */ +static void irda_usb_rx_defer_expired(unsigned long data) +{ + struct urb *urb = (struct urb *) data; + struct sk_buff *skb = (struct sk_buff *) urb->context; + struct irda_usb_cb *self; + struct irda_skb_cb *cb; + struct urb *next_urb; + + IRDA_DEBUG(2, "%s()\n", __FUNCTION__); + + /* Find ourselves */ + cb = (struct irda_skb_cb *) skb->cb; + IRDA_ASSERT(cb != NULL, return;); + self = (struct irda_usb_cb *) cb->context; + IRDA_ASSERT(self != NULL, return;); + + /* Same stuff as when Rx is done, see above... */ + next_urb = self->idle_rx_urb; + urb->context = NULL; + self->idle_rx_urb = urb; + irda_usb_submit(self, skb, next_urb); } /*------------------------------------------------------------------*/ @@ -990,6 +1045,9 @@ static int irda_usb_net_close(struct net_device *netdev) /* Stop network Tx queue */ netif_stop_queue(netdev); + /* Kill defered Rx URB */ + del_timer(&self->rx_defer_timer); + /* Deallocate all the Rx path buffers (URBs and skb) */ for (i = 0; i < IU_MAX_RX_URBS; i++) { struct urb *urb = self->rx_urb[i]; @@ -1365,6 +1423,7 @@ static int irda_usb_probe(struct usb_interface *intf, self = net->priv; self->netdev = net; spin_lock_init(&self->lock); + init_timer(&self->rx_defer_timer); /* Create all of the needed urbs */ for (i = 0; i < IU_MAX_RX_URBS; i++) { @@ -1498,6 +1557,9 @@ static void irda_usb_disconnect(struct usb_interface *intf) * This will stop/desactivate the Tx path. - Jean II */ self->present = 0; + /* Kill defered Rx URB */ + del_timer(&self->rx_defer_timer); + /* We need to have irq enabled to unlink the URBs. That's OK, * at this point the Tx path is gone - Jean II */ spin_unlock_irqrestore(&self->lock, flags); @@ -1507,11 +1569,11 @@ static void irda_usb_disconnect(struct usb_interface *intf) /* Accept no more transmissions */ /*netif_device_detach(self->netdev);*/ netif_stop_queue(self->netdev); - /* Stop all the receive URBs */ + /* Stop all the receive URBs. Must be synchronous. */ for (i = 0; i < IU_MAX_RX_URBS; i++) usb_kill_urb(self->rx_urb[i]); /* Cancel Tx and speed URB. - * Toggle flags to make sure it's synchronous. */ + * Make sure it's synchronous to avoid races. */ usb_kill_urb(self->tx_urb); usb_kill_urb(self->speed_urb); } diff --git a/drivers/net/irda/irda-usb.h b/drivers/net/irda/irda-usb.h index bd8f6654232..4026af42dd4 100644 --- a/drivers/net/irda/irda-usb.h +++ b/drivers/net/irda/irda-usb.h @@ -136,8 +136,6 @@ struct irda_usb_cb { __u16 bulk_out_mtu; /* Max Tx packet size in bytes */ __u8 bulk_int_ep; /* Interrupt Endpoint assignments */ - wait_queue_head_t wait_q; /* for timeouts */ - struct urb *rx_urb[IU_MAX_RX_URBS]; /* URBs used to receive data frames */ struct urb *idle_rx_urb; /* Pointer to idle URB in Rx path */ struct urb *tx_urb; /* URB used to send data frames */ @@ -147,17 +145,18 @@ struct irda_usb_cb { struct net_device_stats stats; struct irlap_cb *irlap; /* The link layer we are binded to */ struct qos_info qos; - hashbin_t *tx_list; /* Queued transmit skb's */ char *speed_buff; /* Buffer for speed changes */ struct timeval stamp; struct timeval now; - spinlock_t lock; /* For serializing operations */ + spinlock_t lock; /* For serializing Tx operations */ __u16 xbofs; /* Current xbofs setting */ __s16 new_xbofs; /* xbofs we need to set */ __u32 speed; /* Current speed */ __s32 new_speed; /* speed we need to set */ + + struct timer_list rx_defer_timer; /* Wait for Rx error to clear */ }; From 8e249f088131cde5f77fd073bf0b0e8b3e9ea4ac Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Sun, 19 Feb 2006 22:29:47 -0800 Subject: [PATCH 022/102] [NETFILTER]: Fix outgoing redirects to loopback When redirecting an outgoing packet to loopback, it keeps the original conntrack reference and information from the outgoing path, which falsely triggers the check for DNAT on input and the dst_entry is released to trigger rerouting. ip_route_input refuses to route the packet because it has a local source address and it is dropped. Look at the packet itself to dermine if it was NATed. Also fix a missing inversion that causes unneccesary xfrm lookups. Signed-off-by: Patrick McHardy Signed-off-by: David S. Miller --- net/ipv4/netfilter/ip_nat_standalone.c | 16 +++++----------- 1 file changed, 5 insertions(+), 11 deletions(-) diff --git a/net/ipv4/netfilter/ip_nat_standalone.c b/net/ipv4/netfilter/ip_nat_standalone.c index 7c3f7d38024..ab1f88fa21e 100644 --- a/net/ipv4/netfilter/ip_nat_standalone.c +++ b/net/ipv4/netfilter/ip_nat_standalone.c @@ -200,20 +200,14 @@ ip_nat_in(unsigned int hooknum, const struct net_device *out, int (*okfn)(struct sk_buff *)) { - struct ip_conntrack *ct; - enum ip_conntrack_info ctinfo; unsigned int ret; + u_int32_t daddr = (*pskb)->nh.iph->daddr; ret = ip_nat_fn(hooknum, pskb, in, out, okfn); if (ret != NF_DROP && ret != NF_STOLEN - && (ct = ip_conntrack_get(*pskb, &ctinfo)) != NULL) { - enum ip_conntrack_dir dir = CTINFO2DIR(ctinfo); - - if (ct->tuplehash[dir].tuple.dst.ip != - ct->tuplehash[!dir].tuple.src.ip) { - dst_release((*pskb)->dst); - (*pskb)->dst = NULL; - } + && daddr != (*pskb)->nh.iph->daddr) { + dst_release((*pskb)->dst); + (*pskb)->dst = NULL; } return ret; } @@ -276,7 +270,7 @@ ip_nat_local_fn(unsigned int hooknum, ct->tuplehash[!dir].tuple.src.ip #ifdef CONFIG_XFRM || ct->tuplehash[dir].tuple.dst.u.all != - ct->tuplehash[dir].tuple.src.u.all + ct->tuplehash[!dir].tuple.src.u.all #endif ) return ip_route_me_harder(pskb) == 0 ? ret : NF_DROP; From a8372f035aa2f6717123eb30679a08b619321dd4 Mon Sep 17 00:00:00 2001 From: YOSHIFUJI Hideaki Date: Sun, 19 Feb 2006 22:32:06 -0800 Subject: [PATCH 023/102] [NET]: NETFILTER: remove duplicated lines and fix order in skb_clone(). Some of netfilter-related members are initalized / copied twice in skb_clone(). Remove one. Pointed out by Olivier MATZ . And this patch also fixes order of copying / clearing members. Signed-off-by: YOSHIFUJI Hideaki Signed-off-by: David S. Miller --- net/core/skbuff.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 6766f118f07..2144952d1c6 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -411,6 +411,9 @@ struct sk_buff *skb_clone(struct sk_buff *skb, gfp_t gfp_mask) C(pkt_type); C(ip_summed); C(priority); +#if defined(CONFIG_IP_VS) || defined(CONFIG_IP_VS_MODULE) + C(ipvs_property); +#endif C(protocol); n->destructor = NULL; #ifdef CONFIG_NETFILTER @@ -422,13 +425,6 @@ struct sk_buff *skb_clone(struct sk_buff *skb, gfp_t gfp_mask) C(nfct_reasm); nf_conntrack_get_reasm(skb->nfct_reasm); #endif -#if defined(CONFIG_IP_VS) || defined(CONFIG_IP_VS_MODULE) - C(ipvs_property); -#endif -#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE) - C(nfct_reasm); - nf_conntrack_get_reasm(skb->nfct_reasm); -#endif #ifdef CONFIG_BRIDGE_NETFILTER C(nf_bridge); nf_bridge_get(skb->nf_bridge); From 9ae61c6cb69f5251d160576c324948805f97e901 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 20 Feb 2006 23:48:37 +0900 Subject: [PATCH 024/102] [PATCH] libata: fix WARN_ON() condition in *_fill_sg() For ATAPI commands, padding can reduce qc->n_elem by one and thus to zero making assert(qc->n_elem > 0)'s in ata_fill_sg() and qs_fill_sg() fail for legal commands. This patch fixes the assert()'s to take qc->pad_len into account. Although the condition check seems a bit excessive, as this part of code isn't still stable yet, I think it's worth to keep those. Signed-off-by: Tejun Heo Signed-off-by: Jeff Garzik --- drivers/scsi/libata-core.c | 2 +- drivers/scsi/sata_qstor.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index 7ddd5a69352..bbac87a13d5 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -2570,7 +2570,7 @@ static void ata_fill_sg(struct ata_queued_cmd *qc) unsigned int idx; assert(qc->__sg != NULL); - assert(qc->n_elem > 0); + assert(qc->n_elem > 0 || qc->pad_len > 0); idx = 0; ata_for_each_sg(sg, qc) { diff --git a/drivers/scsi/sata_qstor.c b/drivers/scsi/sata_qstor.c index de05e2883f9..80480f0fb2b 100644 --- a/drivers/scsi/sata_qstor.c +++ b/drivers/scsi/sata_qstor.c @@ -277,7 +277,7 @@ static unsigned int qs_fill_sg(struct ata_queued_cmd *qc) u8 *prd = pp->pkt + QS_CPB_BYTES; assert(qc->__sg != NULL); - assert(qc->n_elem > 0); + assert(qc->n_elem > 0 || qc->pad_len > 0); nelem = 0; ata_for_each_sg(sg, qc) { From cc1887f3d8ae8ea61efa1a75af8ec0467b9dd546 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 20 Feb 2006 23:48:38 +0900 Subject: [PATCH 025/102] [PATCH] libata: fix qc->n_elem == 0 case handling in ata_qc_next_sg This patch makes ata_for_each_sg() start with pad_sgent when qc->n_elem is zero. Previously, ata_for_each_sg() unconditionally started with qc->__sg, handling the first sg to fill_sg() routines even when the entry was invalid. And while at it, unwind ?: in ata_qc_next_sg() into if statement. Signed-off-by: Tejun Heo Signed-off-by: Jeff Garzik --- include/linux/libata.h | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) diff --git a/include/linux/libata.h b/include/linux/libata.h index 9e5db2949c5..c91be5e64ed 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -556,6 +556,16 @@ ata_sg_is_last(struct scatterlist *sg, struct ata_queued_cmd *qc) return 0; } +static inline struct scatterlist * +ata_qc_first_sg(struct ata_queued_cmd *qc) +{ + if (qc->n_elem) + return qc->__sg; + if (qc->pad_len) + return &qc->pad_sgent; + return NULL; +} + static inline struct scatterlist * ata_qc_next_sg(struct scatterlist *sg, struct ata_queued_cmd *qc) { @@ -563,11 +573,13 @@ ata_qc_next_sg(struct scatterlist *sg, struct ata_queued_cmd *qc) return NULL; if (++sg - qc->__sg < qc->n_elem) return sg; - return qc->pad_len ? &qc->pad_sgent : NULL; + if (qc->pad_len) + return &qc->pad_sgent; + return NULL; } #define ata_for_each_sg(sg, qc) \ - for (sg = qc->__sg; sg; sg = ata_qc_next_sg(sg, qc)) + for (sg = ata_qc_first_sg(qc); sg; sg = ata_qc_next_sg(sg, qc)) static inline unsigned int ata_tag_valid(unsigned int tag) { From 2e242fa994428bd1a40b6a7e97430413246d0a16 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 20 Feb 2006 23:48:38 +0900 Subject: [PATCH 026/102] [PATCH] libata: make ata_sg_setup_one() trim zero length sg This patch makes ata_sg_setup_one() trim sg entry (thus making qc->n_elem zero) if padding results in zero length sg entry. Signed-off-by: Tejun Heo Signed-off-by: Jeff Garzik --- drivers/scsi/libata-core.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/scsi/libata-core.c b/drivers/scsi/libata-core.c index bbac87a13d5..5f1d7580218 100644 --- a/drivers/scsi/libata-core.c +++ b/drivers/scsi/libata-core.c @@ -2514,7 +2514,7 @@ static void ata_sg_clean(struct ata_queued_cmd *qc) assert(sg != NULL); if (qc->flags & ATA_QCFLAG_SINGLE) - assert(qc->n_elem == 1); + assert(qc->n_elem <= 1); VPRINTK("unmapping %u sg elements\n", qc->n_elem); @@ -2537,7 +2537,7 @@ static void ata_sg_clean(struct ata_queued_cmd *qc) kunmap_atomic(addr, KM_IRQ0); } } else { - if (sg_dma_len(&sg[0]) > 0) + if (qc->n_elem) dma_unmap_single(ap->host_set->dev, sg_dma_address(&sg[0]), sg_dma_len(&sg[0]), dir); @@ -2715,6 +2715,7 @@ static int ata_sg_setup_one(struct ata_queued_cmd *qc) int dir = qc->dma_dir; struct scatterlist *sg = qc->__sg; dma_addr_t dma_address; + int trim_sg = 0; /* we must lengthen transfers to end on a 32-bit boundary */ qc->pad_len = sg->length & 3; @@ -2734,13 +2735,15 @@ static int ata_sg_setup_one(struct ata_queued_cmd *qc) sg_dma_len(psg) = ATA_DMA_PAD_SZ; /* trim sg */ sg->length -= qc->pad_len; + if (sg->length == 0) + trim_sg = 1; DPRINTK("padding done, sg->length=%u pad_len=%u\n", sg->length, qc->pad_len); } - if (!sg->length) { - sg_dma_address(sg) = 0; + if (trim_sg) { + qc->n_elem--; goto skip_map; } @@ -2753,9 +2756,9 @@ static int ata_sg_setup_one(struct ata_queued_cmd *qc) } sg_dma_address(sg) = dma_address; -skip_map: sg_dma_len(sg) = sg->length; +skip_map: DPRINTK("mapped buffer of %d bytes for %s\n", sg_dma_len(sg), qc->tf.flags & ATA_TFLAG_WRITE ? "write" : "read"); From b41c82eb5fb49912ce26c51ec221ba35e06c7d9b Mon Sep 17 00:00:00 2001 From: Dave Jones Date: Mon, 20 Feb 2006 18:34:37 -0500 Subject: [PATCH 027/102] [AGPGART] Add some informational printk to nforce GART failure path. Signed-off-by: Dave Jones --- drivers/char/agp/amd64-agp.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/char/agp/amd64-agp.c b/drivers/char/agp/amd64-agp.c index 9964c508c11..1251b2515bb 100644 --- a/drivers/char/agp/amd64-agp.c +++ b/drivers/char/agp/amd64-agp.c @@ -516,8 +516,10 @@ static int __devinit nforce3_agp_init(struct pci_dev *pdev) pci_read_config_dword (hammers[0], AMD64_GARTAPERTUREBASE, &apbase); /* if x86-64 aperture base is beyond 4G, exit here */ - if ( (apbase & 0x7fff) >> (32 - 25) ) - return -ENODEV; + if ( (apbase & 0x7fff) >> (32 - 25) ) { + printk(KERN_INFO PFX "aperture base > 4G\n"); + return -ENODEV; + } apbase = (apbase & 0x7fff) << 25; From 9827b781f20828e5ceb911b879f268f78fe90815 Mon Sep 17 00:00:00 2001 From: Kurt Garloff Date: Mon, 20 Feb 2006 18:27:51 -0800 Subject: [PATCH 028/102] [PATCH] OOM kill: children accounting In the badness() calculation, there's currently this piece of code: /* * Processes which fork a lot of child processes are likely * a good choice. We add the vmsize of the children if they * have an own mm. This prevents forking servers to flood the * machine with an endless amount of children */ list_for_each(tsk, &p->children) { struct task_struct *chld; chld = list_entry(tsk, struct task_struct, sibling); if (chld->mm = p->mm && chld->mm) points += chld->mm->total_vm; } The intention is clear: If some server (apache) keeps spawning new children and we run OOM, we want to kill the father rather than picking a child. This -- to some degree -- also helps a bit with getting fork bombs under control, though I'd consider this a desirable side-effect rather than a feature. There's one problem with this: No matter how many or few children there are, if just one of them misbehaves, and all others (including the father) do everything right, we still always kill the whole family. This hits in real life; whether it's javascript in konqueror resulting in kdeinit (and thus the whole KDE session) being hit or just a classical server that spawns children. Sidenote: The killer does kill all direct children as well, not only the selected father, see oom_kill_process(). The idea in attached patch is that we do want to account the memory consumption of the (direct) children to the father -- however not fully. This maintains the property that fathers with too many children will still very likely be picked, whereas a single misbehaving child has the chance to be picked by the OOM killer. In the patch I account only half (rounded up) of the children's vm_size to the parent. This means that if one child eats more mem than the rest of the family, it will be picked, otherwise it's still the father and thus the whole family that gets selected. This is heuristics -- we could debate whether accounting for a fourth would be better than for half of it. Or -- if people would consider it worth the trouble -- make it a sysctl. For now I sticked to accounting for half, which should IMHO be a significant improvement. The patch does one more thing: As users tend to be irritated by the choice of killed processes (mainly because the children are killed first, despite some of them having a very low OOM score), I added some more output: The selected (father) process will be reported first and it's oom_score printed to syslog. Description: Only account for half of children's vm size in oom score calculation This should still give the parent enough point in case of fork bombs. If any child however has more than 50% of the vm size of all children together, it'll get a higher score and be elected. This patch also makes the kernel display the oom_score. Signed-off-by: Kurt Garloff Cc: Rik van Riel Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/oom_kill.c | 26 ++++++++++++++++---------- 1 file changed, 16 insertions(+), 10 deletions(-) diff --git a/mm/oom_kill.c b/mm/oom_kill.c index b05ab8f2a56..949eba1d5ba 100644 --- a/mm/oom_kill.c +++ b/mm/oom_kill.c @@ -58,15 +58,17 @@ unsigned long badness(struct task_struct *p, unsigned long uptime) /* * Processes which fork a lot of child processes are likely - * a good choice. We add the vmsize of the children if they + * a good choice. We add half the vmsize of the children if they * have an own mm. This prevents forking servers to flood the - * machine with an endless amount of children + * machine with an endless amount of children. In case a single + * child is eating the vast majority of memory, adding only half + * to the parents will make the child our kill candidate of choice. */ list_for_each(tsk, &p->children) { struct task_struct *chld; chld = list_entry(tsk, struct task_struct, sibling); if (chld->mm != p->mm && chld->mm) - points += chld->mm->total_vm; + points += chld->mm->total_vm/2 + 1; } /* @@ -136,12 +138,12 @@ unsigned long badness(struct task_struct *p, unsigned long uptime) * * (not docbooked, we don't want this one cluttering up the manual) */ -static struct task_struct * select_bad_process(void) +static struct task_struct *select_bad_process(unsigned long *ppoints) { - unsigned long maxpoints = 0; struct task_struct *g, *p; struct task_struct *chosen = NULL; struct timespec uptime; + *ppoints = 0; do_posix_clock_monotonic_gettime(&uptime); do_each_thread(g, p) { @@ -169,9 +171,9 @@ static struct task_struct * select_bad_process(void) return p; points = badness(p, uptime.tv_sec); - if (points > maxpoints || !chosen) { + if (points > *ppoints || !chosen) { chosen = p; - maxpoints = points; + *ppoints = points; } } while_each_thread(g, p); return chosen; @@ -237,12 +239,15 @@ static struct mm_struct *oom_kill_task(task_t *p) return mm; } -static struct mm_struct *oom_kill_process(struct task_struct *p) +static struct mm_struct *oom_kill_process(struct task_struct *p, + unsigned long points) { struct mm_struct *mm; struct task_struct *c; struct list_head *tsk; + printk(KERN_ERR "Out of Memory: Kill process %d (%s) score %li and " + "children.\n", p->pid, p->comm, points); /* Try to kill a child first */ list_for_each(tsk, &p->children) { c = list_entry(tsk, struct task_struct, sibling); @@ -267,6 +272,7 @@ void out_of_memory(gfp_t gfp_mask, int order) { struct mm_struct *mm = NULL; task_t * p; + unsigned long points; if (printk_ratelimit()) { printk("oom-killer: gfp_mask=0x%x, order=%d\n", @@ -278,7 +284,7 @@ void out_of_memory(gfp_t gfp_mask, int order) cpuset_lock(); read_lock(&tasklist_lock); retry: - p = select_bad_process(); + p = select_bad_process(&points); if (PTR_ERR(p) == -1UL) goto out; @@ -290,7 +296,7 @@ retry: panic("Out of memory and no killable processes...\n"); } - mm = oom_kill_process(p); + mm = oom_kill_process(p, points); if (!mm) goto retry; From 9b0f8b040acd8dfd23860754c0d09ff4f44e2cbc Mon Sep 17 00:00:00 2001 From: Christoph Lameter Date: Mon, 20 Feb 2006 18:27:52 -0800 Subject: [PATCH 029/102] [PATCH] Terminate process that fails on a constrained allocation Some allocations are restricted to a limited set of nodes (due to memory policies or cpuset constraints). If the page allocator is not able to find enough memory then that does not mean that overall system memory is low. In particular going postal and more or less randomly shooting at processes is not likely going to help the situation but may just lead to suicide (the whole system coming down). It is better to signal to the process that no memory exists given the constraints that the process (or the configuration of the process) has placed on the allocation behavior. The process may be killed but then the sysadmin or developer can investigate the situation. The solution is similar to what we do when running out of hugepages. This patch adds a check before we kill processes. At that point performance considerations do not matter much so we just scan the zonelist and reconstruct a list of nodes. If the list of nodes does not contain all online nodes then this is a constrained allocation and we should kill the current process. Signed-off-by: Christoph Lameter Cc: Nick Piggin Cc: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/char/sysrq.c | 2 +- include/linux/swap.h | 2 +- mm/oom_kill.c | 103 ++++++++++++++++++++++++++++++++----------- mm/page_alloc.c | 2 +- 4 files changed, 81 insertions(+), 28 deletions(-) diff --git a/drivers/char/sysrq.c b/drivers/char/sysrq.c index 5765f672e85..d58f8231885 100644 --- a/drivers/char/sysrq.c +++ b/drivers/char/sysrq.c @@ -243,7 +243,7 @@ static struct sysrq_key_op sysrq_term_op = { static void moom_callback(void *ignored) { - out_of_memory(GFP_KERNEL, 0); + out_of_memory(&NODE_DATA(0)->node_zonelists[ZONE_NORMAL], GFP_KERNEL, 0); } static DECLARE_WORK(moom_work, moom_callback, NULL); diff --git a/include/linux/swap.h b/include/linux/swap.h index f3e17d5963c..d572b19afb7 100644 --- a/include/linux/swap.h +++ b/include/linux/swap.h @@ -147,7 +147,7 @@ struct swap_list_t { #define vm_swap_full() (nr_swap_pages*2 < total_swap_pages) /* linux/mm/oom_kill.c */ -extern void out_of_memory(gfp_t gfp_mask, int order); +extern void out_of_memory(struct zonelist *zonelist, gfp_t gfp_mask, int order); /* linux/mm/memory.c */ extern void swapin_readahead(swp_entry_t, unsigned long, struct vm_area_struct *); diff --git a/mm/oom_kill.c b/mm/oom_kill.c index 949eba1d5ba..8123fad5a48 100644 --- a/mm/oom_kill.c +++ b/mm/oom_kill.c @@ -132,6 +132,36 @@ unsigned long badness(struct task_struct *p, unsigned long uptime) return points; } +/* + * Types of limitations to the nodes from which allocations may occur + */ +#define CONSTRAINT_NONE 1 +#define CONSTRAINT_MEMORY_POLICY 2 +#define CONSTRAINT_CPUSET 3 + +/* + * Determine the type of allocation constraint. + */ +static inline int constrained_alloc(struct zonelist *zonelist, gfp_t gfp_mask) +{ +#ifdef CONFIG_NUMA + struct zone **z; + nodemask_t nodes = node_online_map; + + for (z = zonelist->zones; *z; z++) + if (cpuset_zone_allowed(*z, gfp_mask)) + node_clear((*z)->zone_pgdat->node_id, + nodes); + else + return CONSTRAINT_CPUSET; + + if (!nodes_empty(nodes)) + return CONSTRAINT_MEMORY_POLICY; +#endif + + return CONSTRAINT_NONE; +} + /* * Simple selection loop. We chose the process with the highest * number of 'points'. We expect the caller will lock the tasklist. @@ -184,7 +214,7 @@ static struct task_struct *select_bad_process(unsigned long *ppoints) * CAP_SYS_RAW_IO set, send SIGTERM instead (but it's unlikely that * we select a process with CAP_SYS_RAW_IO set). */ -static void __oom_kill_task(task_t *p) +static void __oom_kill_task(task_t *p, const char *message) { if (p->pid == 1) { WARN_ON(1); @@ -200,8 +230,8 @@ static void __oom_kill_task(task_t *p) return; } task_unlock(p); - printk(KERN_ERR "Out of Memory: Killed process %d (%s).\n", - p->pid, p->comm); + printk(KERN_ERR "%s: Killed process %d (%s).\n", + message, p->pid, p->comm); /* * We give our sacrificial lamb high priority and access to @@ -214,7 +244,7 @@ static void __oom_kill_task(task_t *p) force_sig(SIGKILL, p); } -static struct mm_struct *oom_kill_task(task_t *p) +static struct mm_struct *oom_kill_task(task_t *p, const char *message) { struct mm_struct *mm = get_task_mm(p); task_t * g, * q; @@ -226,21 +256,21 @@ static struct mm_struct *oom_kill_task(task_t *p) return NULL; } - __oom_kill_task(p); + __oom_kill_task(p, message); /* * kill all processes that share the ->mm (i.e. all threads), * but are in a different thread group */ do_each_thread(g, q) if (q->mm == mm && q->tgid != p->tgid) - __oom_kill_task(q); + __oom_kill_task(q, message); while_each_thread(g, q); return mm; } static struct mm_struct *oom_kill_process(struct task_struct *p, - unsigned long points) + unsigned long points, const char *message) { struct mm_struct *mm; struct task_struct *c; @@ -253,11 +283,11 @@ static struct mm_struct *oom_kill_process(struct task_struct *p, c = list_entry(tsk, struct task_struct, sibling); if (c->mm == p->mm) continue; - mm = oom_kill_task(c); + mm = oom_kill_task(c, message); if (mm) return mm; } - return oom_kill_task(p); + return oom_kill_task(p, message); } /** @@ -268,10 +298,10 @@ static struct mm_struct *oom_kill_process(struct task_struct *p, * OR try to be smart about which process to kill. Note that we * don't have to be perfect here, we just have to be good. */ -void out_of_memory(gfp_t gfp_mask, int order) +void out_of_memory(struct zonelist *zonelist, gfp_t gfp_mask, int order) { struct mm_struct *mm = NULL; - task_t * p; + task_t *p; unsigned long points; if (printk_ratelimit()) { @@ -283,25 +313,48 @@ void out_of_memory(gfp_t gfp_mask, int order) cpuset_lock(); read_lock(&tasklist_lock); + + /* + * Check if there were limitations on the allocation (only relevant for + * NUMA) that may require different handling. + */ + switch (constrained_alloc(zonelist, gfp_mask)) { + case CONSTRAINT_MEMORY_POLICY: + mm = oom_kill_process(current, points, + "No available memory (MPOL_BIND)"); + break; + + case CONSTRAINT_CPUSET: + mm = oom_kill_process(current, points, + "No available memory in cpuset"); + break; + + case CONSTRAINT_NONE: retry: - p = select_bad_process(&points); + /* + * Rambo mode: Shoot down a process and hope it solves whatever + * issues we may have. + */ + p = select_bad_process(&points); - if (PTR_ERR(p) == -1UL) - goto out; + if (PTR_ERR(p) == -1UL) + goto out; - /* Found nothing?!?! Either we hang forever, or we panic. */ - if (!p) { - read_unlock(&tasklist_lock); - cpuset_unlock(); - panic("Out of memory and no killable processes...\n"); + /* Found nothing?!?! Either we hang forever, or we panic. */ + if (!p) { + read_unlock(&tasklist_lock); + cpuset_unlock(); + panic("Out of memory and no killable processes...\n"); + } + + mm = oom_kill_process(p, points, "Out of memory"); + if (!mm) + goto retry; + + break; } - mm = oom_kill_process(p, points); - if (!mm) - goto retry; - - out: - read_unlock(&tasklist_lock); +out: cpuset_unlock(); if (mm) mmput(mm); diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 208812b2559..791690d7d3f 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1015,7 +1015,7 @@ rebalance: if (page) goto got_pg; - out_of_memory(gfp_mask, order); + out_of_memory(zonelist, gfp_mask, order); goto restart; } From 7d4c8e56109e0799ab9fb644c08a8daf4a026675 Mon Sep 17 00:00:00 2001 From: Daniel Yeisley Date: Mon, 20 Feb 2006 18:27:54 -0800 Subject: [PATCH 030/102] [PATCH] i386: need to pass virtual address to smp_read_mpc() I'm seeing a kernel panic on an ES7000-600 when booting in virtual wire mode. The panic happens because smp_read_mpc() is passed a physical address, and it should be virtual. I tested the attached patch on the ES7000-600 and on a 2 cpu Dell box, and saw no problems on either. Signed-off-by: Dan Yeisley Acked-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/i386/kernel/mpparse.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/i386/kernel/mpparse.c b/arch/i386/kernel/mpparse.c index 0102f3d50e5..e7609abf379 100644 --- a/arch/i386/kernel/mpparse.c +++ b/arch/i386/kernel/mpparse.c @@ -710,7 +710,7 @@ void __init get_smp_config (void) * Read the physical hardware table. Anything here will * override the defaults. */ - if (!smp_read_mpc((void *)mpf->mpf_physptr)) { + if (!smp_read_mpc(phys_to_virt(mpf->mpf_physptr))) { smp_found_config = 0; printk(KERN_ERR "BIOS bug, MP table errors detected!...\n"); printk(KERN_ERR "... disabling SMP support. (tell your hw vendor)\n"); From d86d43706a27bb87c2873de369f94a10f8758063 Mon Sep 17 00:00:00 2001 From: Alexey Korolev Date: Mon, 20 Feb 2006 18:27:55 -0800 Subject: [PATCH 031/102] [PATCH] cfi_cmdset_0001: fix range for cache invalidation I found an issue in cfi_cmdset0001.c. It is related to cache region invalidation in the buffered write procedure. The code performs cache invalidation from "cmd_addr" to "cmd_adr + len" in do_write_buffer() while we modify region from "adr" to "adr+len". This issue affects writes + reads of data by small chunks. Signed-off-by: Nicolas Pitre Acked-by: Thomas Gleixner Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/mtd/chips/cfi_cmdset_0001.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/mtd/chips/cfi_cmdset_0001.c b/drivers/mtd/chips/cfi_cmdset_0001.c index 69c04945591..ded2c33f5b8 100644 --- a/drivers/mtd/chips/cfi_cmdset_0001.c +++ b/drivers/mtd/chips/cfi_cmdset_0001.c @@ -1019,8 +1019,8 @@ static void __xipram xip_udelay(struct map_info *map, struct flchip *chip, #define XIP_INVAL_CACHED_RANGE(map, from, size) \ INVALIDATE_CACHED_RANGE(map, from, size) -#define INVALIDATE_CACHE_UDELAY(map, chip, adr, len, usec) \ - UDELAY(map, chip, adr, usec) +#define INVALIDATE_CACHE_UDELAY(map, chip, cmd_adr, adr, len, usec) \ + UDELAY(map, chip, cmd_adr, usec) /* * Extra notes: @@ -1052,7 +1052,7 @@ do { \ spin_lock(chip->mutex); \ } while (0) -#define INVALIDATE_CACHE_UDELAY(map, chip, adr, len, usec) \ +#define INVALIDATE_CACHE_UDELAY(map, chip, cmd_adr, adr, len, usec) \ do { \ spin_unlock(chip->mutex); \ INVALIDATE_CACHED_RANGE(map, adr, len); \ @@ -1284,7 +1284,7 @@ static int __xipram do_write_oneword(struct map_info *map, struct flchip *chip, map_write(map, datum, adr); chip->state = mode; - INVALIDATE_CACHE_UDELAY(map, chip, + INVALIDATE_CACHE_UDELAY(map, chip, adr, adr, map_bankwidth(map), chip->word_write_time); @@ -1572,8 +1572,8 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, map_write(map, CMD(0xd0), cmd_adr); chip->state = FL_WRITING; - INVALIDATE_CACHE_UDELAY(map, chip, - cmd_adr, len, + INVALIDATE_CACHE_UDELAY(map, chip, cmd_adr, + adr, len, chip->buffer_write_time); timeo = jiffies + (HZ/2); @@ -1744,7 +1744,7 @@ static int __xipram do_erase_oneblock(struct map_info *map, struct flchip *chip, chip->state = FL_ERASING; chip->erase_suspended = 0; - INVALIDATE_CACHE_UDELAY(map, chip, + INVALIDATE_CACHE_UDELAY(map, chip, adr, adr, len, chip->erase_time*1000/2); From d2799f083dcad0413ad1a396e9bc32d9afb70535 Mon Sep 17 00:00:00 2001 From: Stephen Street Date: Mon, 20 Feb 2006 18:27:56 -0800 Subject: [PATCH 032/102] [PATCH] spi: Fix modular master driver remove and device suspend/remove Fix two problems in the spi subsystem: 1) spi subsystem core dumps when modular spi master is unloaded. 2) spi subsystem core dumps when spi slave device is suspended/resumed and module slave driver is not loaded. Signed-off-by: Stephen Street Signed-off-by: David Brownell Cc: Greg KH Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/spi/spi.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c index 791c4dc550a..94f5e8ed83a 100644 --- a/drivers/spi/spi.c +++ b/drivers/spi/spi.c @@ -90,7 +90,7 @@ static int spi_suspend(struct device *dev, pm_message_t message) int value; struct spi_driver *drv = to_spi_driver(dev->driver); - if (!drv->suspend) + if (!drv || !drv->suspend) return 0; /* suspend will stop irqs and dma; no more i/o */ @@ -105,7 +105,7 @@ static int spi_resume(struct device *dev) int value; struct spi_driver *drv = to_spi_driver(dev->driver); - if (!drv->resume) + if (!drv || !drv->resume) return 0; /* resume may restart the i/o queue */ @@ -449,7 +449,6 @@ void spi_unregister_master(struct spi_master *master) { (void) device_for_each_child(master->cdev.dev, NULL, __unregister); class_device_unregister(&master->cdev); - master->cdev.dev = NULL; } EXPORT_SYMBOL_GPL(spi_unregister_master); From 2e2b42636614f3c152672b5da67947ccbcbe0d32 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Mon, 20 Feb 2006 18:27:57 -0800 Subject: [PATCH 033/102] [PATCH] x86_64: Don't set CONFIG_DEBUG_INFO in defconfig Undo setting of CONFIG_DEBUG_INFO in the previous defconfig update. It will make every build much slower and need more disk space and isn't a good default. Signed-off-by: Andi Kleen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/x86_64/defconfig | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/arch/x86_64/defconfig b/arch/x86_64/defconfig index b337136f28b..ce4de61ed85 100644 --- a/arch/x86_64/defconfig +++ b/arch/x86_64/defconfig @@ -1,7 +1,7 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.16-rc3 -# Mon Feb 13 22:31:24 2006 +# Linux kernel version: 2.6.16-rc3-git9 +# Sat Feb 18 00:27:03 2006 # CONFIG_X86_64=y CONFIG_64BIT=y @@ -1317,7 +1317,7 @@ CONFIG_DETECT_SOFTLOCKUP=y # CONFIG_DEBUG_SPINLOCK is not set # CONFIG_DEBUG_SPINLOCK_SLEEP is not set # CONFIG_DEBUG_KOBJECT is not set -CONFIG_DEBUG_INFO=y +# CONFIG_DEBUG_INFO is not set CONFIG_DEBUG_FS=y # CONFIG_DEBUG_VM is not set # CONFIG_FRAME_POINTER is not set From 6303dbf570e410067380daec670fdb4137ac0d1d Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 20 Feb 2006 18:27:58 -0800 Subject: [PATCH 034/102] [PATCH] cpu hotplug documentation fix Looks like there was a merge conflict when patches 8f8b1138fc9f65e3591aac83a4ee394fef34ac1d and 255acee706b333b79f593dd366f16e1f107cccc3 were applied which wasn't properly resolved. Fix this and add some additional description. Signed-off-by: Heiko Carstens Cc: Ashok Raj Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/cpu-hotplug.txt | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/Documentation/cpu-hotplug.txt b/Documentation/cpu-hotplug.txt index e71bc6cbbc5..57a09f99ecb 100644 --- a/Documentation/cpu-hotplug.txt +++ b/Documentation/cpu-hotplug.txt @@ -46,10 +46,12 @@ maxcpus=n Restrict boot time cpus to n. Say if you have 4 cpus, using maxcpus=2 will only boot 2. You can choose to bring the other cpus later online, read FAQ's for more info. -additional_cpus=n [x86_64, s390 only] use this to limit hotpluggable cpus. - This option sets +additional_cpus*=n Use this to limit hotpluggable cpus. This option sets cpu_possible_map = cpu_present_map + additional_cpus +(*) Option valid only for following architectures +- x86_64, ia64, s390 + ia64 and x86_64 use the number of disabled local apics in ACPI tables MADT to determine the number of potentially hot-pluggable cpus. The implementation should only rely on this to count the #of cpus, but *MUST* not rely on the @@ -57,6 +59,9 @@ apicid values in those tables for disabled apics. In the event BIOS doesnt mark such hot-pluggable cpus as disabled entries, one could use this parameter "additional_cpus=x" to represent those cpus in the cpu_possible_map. +s390 uses the number of cpus it detects at IPL time to also the number of bits +in cpu_possible_map. If it is desired to add additional cpus at a later time +the number should be specified using this option or the possible_cpus option. possible_cpus=n [s390 only] use this to set hotpluggable cpus. This option sets possible_cpus bits in From c255d844dd73616f23e4b4733edcc2e5fa4042b2 Mon Sep 17 00:00:00 2001 From: Pavel Machek Date: Mon, 20 Feb 2006 18:27:58 -0800 Subject: [PATCH 035/102] [PATCH] suspend-to-ram: allow video options to be set at runtime Currently, acpi video options can only be set on kernel command line. That's little inflexible; I'd like userland s2ram application that just works, and modifying kernel command line according to whitelist is not fun. It is better to just allow s2ram application to set video options just before suspend (according to the whitelist). This implements sysctl to allow setting suspend video options without reboot. (akpm: Documentation updates for this new sysctl are pending..) Signed-off-by: Pavel Machek Cc: "Brown, Len" Cc: "Antonino A. Daplas" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/sysctl/kernel.txt | 10 ++++++++++ include/linux/acpi.h | 3 ++- include/linux/sysctl.h | 1 + kernel/sysctl.c | 16 ++++++++++++---- 4 files changed, 25 insertions(+), 5 deletions(-) diff --git a/Documentation/sysctl/kernel.txt b/Documentation/sysctl/kernel.txt index 9f11d36a8c1..b0c7ab93dcb 100644 --- a/Documentation/sysctl/kernel.txt +++ b/Documentation/sysctl/kernel.txt @@ -16,6 +16,7 @@ before actually making adjustments. Currently, these files might (depending on your configuration) show up in /proc/sys/kernel: +- acpi_video_flags - acct - core_pattern - core_uses_pid @@ -57,6 +58,15 @@ show up in /proc/sys/kernel: ============================================================== +acpi_video_flags: + +flags + +See Doc*/kernel/power/video.txt, it allows mode of video boot to be +set during run time. + +============================================================== + acct: highwater lowwater frequency diff --git a/include/linux/acpi.h b/include/linux/acpi.h index 84d3d9f034c..d3bc25e6d27 100644 --- a/include/linux/acpi.h +++ b/include/linux/acpi.h @@ -427,7 +427,8 @@ extern int acpi_mp_config; extern struct acpi_table_mcfg_config *pci_mmcfg_config; extern int pci_mmcfg_config_num; -extern int sbf_port ; +extern int sbf_port; +extern unsigned long acpi_video_flags; #else /* !CONFIG_ACPI */ diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h index 32a4139c4ad..0e92bf7ec28 100644 --- a/include/linux/sysctl.h +++ b/include/linux/sysctl.h @@ -146,6 +146,7 @@ enum KERN_RANDOMIZE=68, /* int: randomize virtual address space */ KERN_SETUID_DUMPABLE=69, /* int: behaviour of dumps for setuid core */ KERN_SPIN_RETRY=70, /* int: number of spinlock retries */ + KERN_ACPI_VIDEO_FLAGS=71, /* int: flags for setting up video after ACPI sleep */ }; diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 7654d55c47f..ebc41bf22f1 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -44,14 +44,12 @@ #include #include #include +#include +#include #include #include -#ifdef CONFIG_ROOT_NFS -#include -#endif - #if defined(CONFIG_SYSCTL) /* External variables not in a header file. */ @@ -655,6 +653,16 @@ static ctl_table kern_table[] = { .mode = 0644, .proc_handler = &proc_dointvec, }, +#endif +#ifdef CONFIG_ACPI_SLEEP + { + .ctl_name = KERN_ACPI_VIDEO_FLAGS, + .procname = "acpi_video_flags", + .data = &acpi_video_flags, + .maxlen = sizeof (unsigned long), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, #endif { .ctl_name = 0 } }; From a9c930bac163c5e616ca0ba9378e7dc746c93227 Mon Sep 17 00:00:00 2001 From: Andi Kleen Date: Mon, 20 Feb 2006 18:27:59 -0800 Subject: [PATCH 036/102] [PATCH] Fix units in mbind check maxnode is a bit index and can't be directly compared against a byte length like PAGE_SIZE Signed-off-by: Andi Kleen Cc: Chris Wright Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/mempolicy.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/mempolicy.c b/mm/mempolicy.c index bedfa4f09c8..6422fe47811 100644 --- a/mm/mempolicy.c +++ b/mm/mempolicy.c @@ -808,7 +808,7 @@ static int get_nodes(nodemask_t *nodes, const unsigned long __user *nmask, nodes_clear(*nodes); if (maxnode == 0 || !nmask) return 0; - if (maxnode > PAGE_SIZE) + if (maxnode > PAGE_SIZE*BITS_PER_BYTE) return -EINVAL; nlongs = BITS_TO_LONGS(maxnode); From aa657ca9245a06fa435e00332a13da1fce182abc Mon Sep 17 00:00:00 2001 From: Juergen Kreileder Date: Mon, 20 Feb 2006 18:28:00 -0800 Subject: [PATCH 037/102] [PATCH] Fix snd-usb-audio in 32-bit compat environment I'm getting oopses with snd-usb-audio in 32-bit compat environments: control_compat.c:get_ctl_type() doesn't initialize 'info', so 'itemlist[uinfo->value.enumerated.item]' in usbmixer.c:mixer_ctl_selector_info() might access random memory (The 'if ((int)uinfo->value.enumerated.item >= cval->max)' doesn't fix all problems because of the unsigned -> signed conversion.) Signed-off-by: Juergen Kreileder Cc: Jaroslav Kysela Acked-by: Takashi Iwai Cc: Greg KH Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- sound/core/control_compat.c | 16 +++++++++++----- 1 file changed, 11 insertions(+), 5 deletions(-) diff --git a/sound/core/control_compat.c b/sound/core/control_compat.c index 418c6d4e5da..a529b62972b 100644 --- a/sound/core/control_compat.c +++ b/sound/core/control_compat.c @@ -167,7 +167,7 @@ static int get_ctl_type(struct snd_card *card, struct snd_ctl_elem_id *id, int *countp) { struct snd_kcontrol *kctl; - struct snd_ctl_elem_info info; + struct snd_ctl_elem_info *info; int err; down_read(&card->controls_rwsem); @@ -176,13 +176,19 @@ static int get_ctl_type(struct snd_card *card, struct snd_ctl_elem_id *id, up_read(&card->controls_rwsem); return -ENXIO; } - info.id = *id; - err = kctl->info(kctl, &info); + info = kzalloc(sizeof(*info), GFP_KERNEL); + if (info == NULL) { + up_read(&card->controls_rwsem); + return -ENOMEM; + } + info->id = *id; + err = kctl->info(kctl, info); up_read(&card->controls_rwsem); if (err >= 0) { - err = info.type; - *countp = info.count; + err = info->type; + *countp = info->count; } + kfree(info); return err; } From cef289633a9321cd99dd5f6cc935657dc487e9f0 Mon Sep 17 00:00:00 2001 From: Peter Osterlund Date: Mon, 20 Feb 2006 18:28:01 -0800 Subject: [PATCH 038/102] [PATCH] pktcdvd: Correctly set rq->cmd_len in pkt_generic_packet() It looks like the code in pkt_generic_packet() worked by luck in the past, but after commit 186d330e682210100c671355580a8592e4a21692 leaving rq->cmd_len uninitialized doesn't work any more. Signed-off-by: Peter Osterlund Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/block/pktcdvd.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c index 93e44d0292a..f40310a42ba 100644 --- a/drivers/block/pktcdvd.c +++ b/drivers/block/pktcdvd.c @@ -58,6 +58,7 @@ #include #include #include +#include #include @@ -380,6 +381,7 @@ static int pkt_generic_packet(struct pktcdvd_device *pd, struct packet_command * memcpy(rq->cmd, cgc->cmd, CDROM_PACKET_SIZE); if (sizeof(rq->cmd) > CDROM_PACKET_SIZE) memset(rq->cmd + CDROM_PACKET_SIZE, 0, sizeof(rq->cmd) - CDROM_PACKET_SIZE); + rq->cmd_len = COMMAND_SIZE(rq->cmd[0]); rq->ref_count++; rq->flags |= REQ_NOMERGE; From 7c613d593370292d1685f5794c743a2323be3a09 Mon Sep 17 00:00:00 2001 From: Peter Osterlund Date: Mon, 20 Feb 2006 18:28:02 -0800 Subject: [PATCH 039/102] [PATCH] pktcdvd: Rename functions and make their return values sane Boolean functions should return non-zero when they mean "true", otherwise the calling code looks weird. (As suggested by Linus.) Signed-off-by: Peter Osterlund Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/block/pktcdvd.c | 36 ++++++++++++++++++------------------ 1 file changed, 18 insertions(+), 18 deletions(-) diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c index f40310a42ba..fb2c31edca2 100644 --- a/drivers/block/pktcdvd.c +++ b/drivers/block/pktcdvd.c @@ -1497,9 +1497,9 @@ static int pkt_set_write_settings(struct pktcdvd_device *pd) } /* - * 0 -- we can write to this track, 1 -- we can't + * 1 -- we can write to this track, 0 -- we can't */ -static int pkt_good_track(track_information *ti) +static int pkt_writable_track(track_information *ti) { /* * only good for CD-RW at the moment, not DVD-RW @@ -1509,28 +1509,28 @@ static int pkt_good_track(track_information *ti) * FIXME: only for FP */ if (ti->fp == 0) - return 0; + return 1; /* * "good" settings as per Mt Fuji. */ if (ti->rt == 0 && ti->blank == 0 && ti->packet == 1) - return 0; + return 1; if (ti->rt == 0 && ti->blank == 1 && ti->packet == 1) - return 0; + return 1; if (ti->rt == 1 && ti->blank == 0 && ti->packet == 1) - return 0; + return 1; printk("pktcdvd: bad state %d-%d-%d\n", ti->rt, ti->blank, ti->packet); - return 1; + return 0; } /* - * 0 -- we can write to this disc, 1 -- we can't + * 1 -- we can write to this disc, 0 -- we can't */ -static int pkt_good_disc(struct pktcdvd_device *pd, disc_information *di) +static int pkt_writable_disc(struct pktcdvd_device *pd, disc_information *di) { switch (pd->mmc3_profile) { case 0x0a: /* CD-RW */ @@ -1539,10 +1539,10 @@ static int pkt_good_disc(struct pktcdvd_device *pd, disc_information *di) case 0x1a: /* DVD+RW */ case 0x13: /* DVD-RW */ case 0x12: /* DVD-RAM */ - return 0; + return 1; default: VPRINTK("pktcdvd: Wrong disc profile (%x)\n", pd->mmc3_profile); - return 1; + return 0; } /* @@ -1551,25 +1551,25 @@ static int pkt_good_disc(struct pktcdvd_device *pd, disc_information *di) */ if (di->disc_type == 0xff) { printk("pktcdvd: Unknown disc. No track?\n"); - return 1; + return 0; } if (di->disc_type != 0x20 && di->disc_type != 0) { printk("pktcdvd: Wrong disc type (%x)\n", di->disc_type); - return 1; + return 0; } if (di->erasable == 0) { printk("pktcdvd: Disc not erasable\n"); - return 1; + return 0; } if (di->border_status == PACKET_SESSION_RESERVED) { printk("pktcdvd: Can't write to last track (reserved)\n"); - return 1; + return 0; } - return 0; + return 1; } static int pkt_probe_settings(struct pktcdvd_device *pd) @@ -1594,7 +1594,7 @@ static int pkt_probe_settings(struct pktcdvd_device *pd) return ret; } - if (pkt_good_disc(pd, &di)) + if (!pkt_writable_disc(pd, &di)) return -ENXIO; switch (pd->mmc3_profile) { @@ -1619,7 +1619,7 @@ static int pkt_probe_settings(struct pktcdvd_device *pd) return ret; } - if (pkt_good_track(&ti)) { + if (!pkt_writable_track(&ti)) { printk("pktcdvd: can't write to this track\n"); return -ENXIO; } From 3b4828047d4ebe3703dedbfc739958c319ff0b24 Mon Sep 17 00:00:00 2001 From: Peter Osterlund Date: Mon, 20 Feb 2006 18:28:03 -0800 Subject: [PATCH 040/102] [PATCH] pktcdvd: Remove useless printk statements Writing the detected disc type in the kernel log is not useful during normal use of the driver, so remove the printk statements. Signed-off-by: Peter Osterlund Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/block/pktcdvd.c | 14 -------------- 1 file changed, 14 deletions(-) diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c index fb2c31edca2..fb08ae6de40 100644 --- a/drivers/block/pktcdvd.c +++ b/drivers/block/pktcdvd.c @@ -1597,20 +1597,6 @@ static int pkt_probe_settings(struct pktcdvd_device *pd) if (!pkt_writable_disc(pd, &di)) return -ENXIO; - switch (pd->mmc3_profile) { - case 0x1a: /* DVD+RW */ - printk("pktcdvd: inserted media is DVD+RW\n"); - break; - case 0x13: /* DVD-RW */ - printk("pktcdvd: inserted media is DVD-RW\n"); - break; - case 0x12: /* DVD-RAM */ - printk("pktcdvd: inserted media is DVD-RAM\n"); - break; - default: - printk("pktcdvd: inserted media is CD-R%s\n", di.erasable ? "W" : ""); - break; - } pd->type = di.erasable ? PACKET_CDRW : PACKET_CDR; track = 1; /* (di.last_track_msb << 8) | di.last_track_lsb; */ From ab863ec342cf148d02ed180b8ecf3a71a024e4be Mon Sep 17 00:00:00 2001 From: Peter Osterlund Date: Mon, 20 Feb 2006 18:28:04 -0800 Subject: [PATCH 041/102] [PATCH] pktcdvd: Fix the logic in the pkt_writable_track function Fix the pkt_writable_track() function to make it work correctly for all types of CD/DVD discs. Signed-off-by: Peter Osterlund Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/block/pktcdvd.c | 28 +++++++++++++++------------- 1 file changed, 15 insertions(+), 13 deletions(-) diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c index fb08ae6de40..eb83197cc36 100644 --- a/drivers/block/pktcdvd.c +++ b/drivers/block/pktcdvd.c @@ -1499,28 +1499,30 @@ static int pkt_set_write_settings(struct pktcdvd_device *pd) /* * 1 -- we can write to this track, 0 -- we can't */ -static int pkt_writable_track(track_information *ti) +static int pkt_writable_track(struct pktcdvd_device *pd, track_information *ti) { - /* - * only good for CD-RW at the moment, not DVD-RW - */ + switch (pd->mmc3_profile) { + case 0x1a: /* DVD+RW */ + case 0x12: /* DVD-RAM */ + /* The track is always writable on DVD+RW/DVD-RAM */ + return 1; + default: + break; + } - /* - * FIXME: only for FP - */ - if (ti->fp == 0) - return 1; + if (!ti->packet || !ti->fp) + return 0; /* * "good" settings as per Mt Fuji. */ - if (ti->rt == 0 && ti->blank == 0 && ti->packet == 1) + if (ti->rt == 0 && ti->blank == 0) return 1; - if (ti->rt == 0 && ti->blank == 1 && ti->packet == 1) + if (ti->rt == 0 && ti->blank == 1) return 1; - if (ti->rt == 1 && ti->blank == 0 && ti->packet == 1) + if (ti->rt == 1 && ti->blank == 0) return 1; printk("pktcdvd: bad state %d-%d-%d\n", ti->rt, ti->blank, ti->packet); @@ -1605,7 +1607,7 @@ static int pkt_probe_settings(struct pktcdvd_device *pd) return ret; } - if (!pkt_writable_track(&ti)) { + if (!pkt_writable_track(pd, &ti)) { printk("pktcdvd: can't write to this track\n"); return -ENXIO; } From 9db91546570ca1b3bc90b4c2d25d5bb74a44be24 Mon Sep 17 00:00:00 2001 From: Peter Osterlund Date: Mon, 20 Feb 2006 18:28:04 -0800 Subject: [PATCH 042/102] [PATCH] pktcdvd: Only return -EROFS when appropriate When attempting to open the device for writing, only return -EROFS if the disc appears to be readable but not writable. Signed-off-by: Peter Osterlund Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/block/pktcdvd.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/block/pktcdvd.c b/drivers/block/pktcdvd.c index eb83197cc36..bc9b2bcd7db 100644 --- a/drivers/block/pktcdvd.c +++ b/drivers/block/pktcdvd.c @@ -1597,7 +1597,7 @@ static int pkt_probe_settings(struct pktcdvd_device *pd) } if (!pkt_writable_disc(pd, &di)) - return -ENXIO; + return -EROFS; pd->type = di.erasable ? PACKET_CDRW : PACKET_CDR; @@ -1609,7 +1609,7 @@ static int pkt_probe_settings(struct pktcdvd_device *pd) if (!pkt_writable_track(pd, &ti)) { printk("pktcdvd: can't write to this track\n"); - return -ENXIO; + return -EROFS; } /* @@ -1623,7 +1623,7 @@ static int pkt_probe_settings(struct pktcdvd_device *pd) } if (pd->settings.size > PACKET_MAX_SECTORS) { printk("pktcdvd: packet size is too big\n"); - return -ENXIO; + return -EROFS; } pd->settings.fp = ti.fp; pd->offset = (be32_to_cpu(ti.track_start) << 2) & (pd->settings.size - 1); @@ -1665,7 +1665,7 @@ static int pkt_probe_settings(struct pktcdvd_device *pd) break; default: printk("pktcdvd: unknown data mode\n"); - return 1; + return -EROFS; } return 0; } @@ -1876,7 +1876,7 @@ static int pkt_open_write(struct pktcdvd_device *pd) if ((ret = pkt_probe_settings(pd))) { VPRINTK("pktcdvd: %s failed probe\n", pd->name); - return -EROFS; + return ret; } if ((ret = pkt_set_write_settings(pd))) { From e1c92117558261d5504c59712751f6c7925ff3ba Mon Sep 17 00:00:00 2001 From: Eric Van Hensbergen Date: Mon, 20 Feb 2006 18:28:05 -0800 Subject: [PATCH 043/102] [PATCH] v9fs: update documentation and fix debug flag Minor updates to the documentation to bring them into sync with current websites and available features. The debug flag was switched back to hex to match the documentation. Signed-off-by: Eric Van Hensbergen Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/filesystems/v9fs.txt | 16 ++++++++++------ fs/9p/v9fs.c | 2 +- 2 files changed, 11 insertions(+), 7 deletions(-) diff --git a/Documentation/filesystems/v9fs.txt b/Documentation/filesystems/v9fs.txt index 4e92feb6b50..24c7a9c41f0 100644 --- a/Documentation/filesystems/v9fs.txt +++ b/Documentation/filesystems/v9fs.txt @@ -57,8 +57,6 @@ OPTIONS port=n port to connect to on the remote server - timeout=n request timeouts (in ms) (default 60000ms) - noextend force legacy mode (no 9P2000.u semantics) uid attempt to mount as a particular uid @@ -74,10 +72,16 @@ OPTIONS RESOURCES ========= -The Linux version of the 9P server, along with some client-side utilities -can be found at http://v9fs.sf.net (along with a CVS repository of the -development branch of this module). There are user and developer mailing -lists here, as well as a bug-tracker. +The Linux version of the 9P server is now maintained under the npfs project +on sourceforge (http://sourceforge.net/projects/npfs). + +There are user and developer mailing lists available through the v9fs project +on sourceforge (http://sourceforge.net/projects/v9fs). + +News and other information is maintained on SWiK (http://swik.net/v9fs). + +Bug reports may be issued through the kernel.org bugzilla +(http://bugzilla.kernel.org) For more information on the Plan 9 Operating System check out http://plan9.bell-labs.com/plan9 diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index 5250c428fc1..ef338654914 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -66,7 +66,7 @@ static match_table_t tokens = { {Opt_afid, "afid=%u"}, {Opt_rfdno, "rfdno=%u"}, {Opt_wfdno, "wfdno=%u"}, - {Opt_debug, "debug=%u"}, + {Opt_debug, "debug=%x"}, {Opt_name, "name=%s"}, {Opt_remotename, "aname=%s"}, {Opt_unix, "proto=unix"}, From c8b8b1f2e0eeb91cca22211950742b5f51564672 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Mon, 20 Feb 2006 18:28:06 -0800 Subject: [PATCH 044/102] [PATCH] powermac: Fix loss of ethernet PHY on sleep Some recent PowerBook models tend to lose the ethernet PHY on suspend/resume. It -seems- that they use a combo ethernet-firewire PHY chip and the firewire PHY seems to die the same way when that happens. Not trying to toggle the firewire cable power appears to fix it. So this patch disables changes to the firewire cable power control GPIO on those models. Signed-off-by: Benjamin Herrenschmidt Cc: Jeff Garzik Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/powerpc/platforms/powermac/feature.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/arch/powerpc/platforms/powermac/feature.c b/arch/powerpc/platforms/powermac/feature.c index 558dd069209..34714d3ea69 100644 --- a/arch/powerpc/platforms/powermac/feature.c +++ b/arch/powerpc/platforms/powermac/feature.c @@ -1646,10 +1646,10 @@ static void intrepid_shutdown(struct macio_chip *macio, int sleep_mode) KL0_SCC_CELL_ENABLE); MACIO_BIC(KEYLARGO_FCR1, - /*KL1_USB2_CELL_ENABLE |*/ KL1_I2S0_CELL_ENABLE | KL1_I2S0_CLK_ENABLE_BIT | KL1_I2S0_ENABLE | KL1_I2S1_CELL_ENABLE | - KL1_I2S1_CLK_ENABLE_BIT | KL1_I2S1_ENABLE); + KL1_I2S1_CLK_ENABLE_BIT | KL1_I2S1_ENABLE | + KL1_EIDE0_ENABLE); if (pmac_mb.board_flags & PMAC_MB_MOBILE) MACIO_BIC(KEYLARGO_FCR1, KL1_UIDE_RESET_N); @@ -2183,7 +2183,7 @@ static struct pmac_mb_def pmac_mb_defs[] = { }, { "PowerMac10,1", "Mac mini", PMAC_TYPE_UNKNOWN_INTREPID, intrepid_features, - PMAC_MB_MAY_SLEEP | PMAC_MB_HAS_FW_POWER, + PMAC_MB_MAY_SLEEP, }, { "iMac,1", "iMac (first generation)", PMAC_TYPE_ORIG_IMAC, paddington_features, @@ -2295,11 +2295,11 @@ static struct pmac_mb_def pmac_mb_defs[] = { }, { "PowerBook5,8", "PowerBook G4 15\"", PMAC_TYPE_UNKNOWN_INTREPID, intrepid_features, - PMAC_MB_MAY_SLEEP | PMAC_MB_HAS_FW_POWER | PMAC_MB_MOBILE, + PMAC_MB_MAY_SLEEP | PMAC_MB_MOBILE, }, { "PowerBook5,9", "PowerBook G4 17\"", PMAC_TYPE_UNKNOWN_INTREPID, intrepid_features, - PMAC_MB_MAY_SLEEP | PMAC_MB_HAS_FW_POWER | PMAC_MB_MOBILE, + PMAC_MB_MAY_SLEEP | PMAC_MB_MOBILE, }, { "PowerBook6,1", "PowerBook G4 12\"", PMAC_TYPE_UNKNOWN_INTREPID, intrepid_features, From 7a9166e3b037296366cea6f3c97f705d33e209e6 Mon Sep 17 00:00:00 2001 From: Luke Yang Date: Mon, 20 Feb 2006 18:28:07 -0800 Subject: [PATCH 045/102] [PATCH] Fix undefined symbols for nommu architecture Signed-off-by: Luke Yang Acked-by: David Howells Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/mm.h | 4 ++++ kernel/sysctl.c | 2 ++ mm/nommu.c | 2 ++ 3 files changed, 8 insertions(+) diff --git a/include/linux/mm.h b/include/linux/mm.h index 26e1663a5cb..498ff8778fb 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -1051,7 +1051,11 @@ int shrink_slab(unsigned long scanned, gfp_t gfp_mask, void drop_pagecache(void); void drop_slab(void); +#ifndef CONFIG_MMU +#define randomize_va_space 0 +#else extern int randomize_va_space; +#endif #endif /* __KERNEL__ */ #endif /* _LINUX_MM_H */ diff --git a/kernel/sysctl.c b/kernel/sysctl.c index ebc41bf22f1..c05a2b7125e 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -636,6 +636,7 @@ static ctl_table kern_table[] = { .proc_handler = &proc_dointvec, }, #endif +#if defined(CONFIG_MMU) { .ctl_name = KERN_RANDOMIZE, .procname = "randomize_va_space", @@ -644,6 +645,7 @@ static ctl_table kern_table[] = { .mode = 0644, .proc_handler = &proc_dointvec, }, +#endif #if defined(CONFIG_S390) && defined(CONFIG_SMP) { .ctl_name = KERN_SPIN_RETRY, diff --git a/mm/nommu.c b/mm/nommu.c index c10262d6823..99d21020ec9 100644 --- a/mm/nommu.c +++ b/mm/nommu.c @@ -57,6 +57,8 @@ EXPORT_SYMBOL(vmalloc); EXPORT_SYMBOL(vfree); EXPORT_SYMBOL(vmalloc_to_page); EXPORT_SYMBOL(vmalloc_32); +EXPORT_SYMBOL(vmap); +EXPORT_SYMBOL(vunmap); /* * Handle all mappings that got truncated by a "truncate()" From 7fd105e758c8d746d57ab7e77f100e096bf153c8 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Mon, 20 Feb 2006 18:28:08 -0800 Subject: [PATCH 046/102] [PATCH] Fix compile for CONFIG_SYSVIPC=n or CONFIG_SYSCTL=n The compat syscalls are added to sys_ni.c since they are not defined if the above CONFIG options are off. Also, nfs would not build with CONFIG_SYSCTL off. Noticed by Arthur Othieno. Signed-off-by: Stephen Rothwell Cc: "David S. Miller" Cc: Trond Myklebust Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/nfs_fs.h | 2 +- kernel/sys_ni.c | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index 547d649b274..b4dc6e2e10c 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -398,7 +398,7 @@ extern struct inode_operations nfs_symlink_inode_operations; extern int nfs_register_sysctl(void); extern void nfs_unregister_sysctl(void); #else -#define nfs_register_sysctl() do { } while(0) +#define nfs_register_sysctl() 0 #define nfs_unregister_sysctl() do { } while(0) #endif diff --git a/kernel/sys_ni.c b/kernel/sys_ni.c index 17313b99e53..1067090db6b 100644 --- a/kernel/sys_ni.c +++ b/kernel/sys_ni.c @@ -104,6 +104,8 @@ cond_syscall(sys_setreuid16); cond_syscall(sys_setuid16); cond_syscall(sys_vm86old); cond_syscall(sys_vm86); +cond_syscall(compat_sys_ipc); +cond_syscall(compat_sys_sysctl); /* arch-specific weak syscall entries */ cond_syscall(sys_pciconfig_read); From 1dd31b6c89611ee91c0ff309c8733c0af61579e8 Mon Sep 17 00:00:00 2001 From: Zhu Yi Date: Mon, 20 Feb 2006 18:28:09 -0800 Subject: [PATCH 047/102] [PATCH] ipw2200: Suppress warning message The following message will be only printed if DEBUG_NOTIF is on. "Unknown notification: subtype=40,flags=0xa0,size=40" Signed-off-by: Zhu Yi Cc: James Ketrenos Cc: Jeff Garzik Cc: "John W. Linville" Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/net/wireless/ipw2200.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/net/wireless/ipw2200.c b/drivers/net/wireless/ipw2200.c index 14beab4bc91..287676ad80d 100644 --- a/drivers/net/wireless/ipw2200.c +++ b/drivers/net/wireless/ipw2200.c @@ -4616,9 +4616,9 @@ static void ipw_rx_notification(struct ipw_priv *priv, } default: - IPW_ERROR("Unknown notification: " - "subtype=%d,flags=0x%2x,size=%d\n", - notif->subtype, notif->flags, notif->size); + IPW_DEBUG_NOTIF("Unknown notification: " + "subtype=%d,flags=0x%2x,size=%d\n", + notif->subtype, notif->flags, notif->size); } } From fcab6f351305029fc5e3c632209d45cae57e4835 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Mon, 20 Feb 2006 18:28:10 -0800 Subject: [PATCH 048/102] [PATCH] mm/mempolicy.c: fix 'if ();' typo [akpm; it happens that the code was still correct, only inefficient ] Signed-off-by: Alexey Dobriyan Cc: Christoph Lameter Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/mempolicy.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/mempolicy.c b/mm/mempolicy.c index 6422fe47811..880831bd300 100644 --- a/mm/mempolicy.c +++ b/mm/mempolicy.c @@ -587,7 +587,7 @@ redo: } list_add(&page->lru, &newlist); nr_pages++; - if (nr_pages > MIGRATE_CHUNK_SIZE); + if (nr_pages > MIGRATE_CHUNK_SIZE) break; } err = migrate_pages(pagelist, &newlist, &moved, &failed); From ec33b5fe1a6f68c0a494aab476b9667945e029c4 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Mon, 20 Feb 2006 18:28:10 -0800 Subject: [PATCH 049/102] [PATCH] drivers/fc4/fc.c: memset correct length Signed-off-by: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/fc4/fc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/fc4/fc.c b/drivers/fc4/fc.c index 5c8943509cc..66d03f242d3 100644 --- a/drivers/fc4/fc.c +++ b/drivers/fc4/fc.c @@ -1053,7 +1053,7 @@ static int fc_do_els(fc_channel *fc, unsigned int alpa, void *data, int len) int i; fcmd = &_fcmd; - memset(fcmd, 0, sizeof(fcmd)); + memset(fcmd, 0, sizeof(fcp_cmnd)); FCD(("PLOGI SID %d DID %d\n", fc->sid, alpa)) fch = &fcmd->fch; FILL_FCHDR_RCTL_DID(fch, R_CTL_ELS_REQ, alpa); From a1909e631caa3a02c25e493ac4004cd67984e0e0 Mon Sep 17 00:00:00 2001 From: Carl-Daniel Hailfinger Date: Mon, 20 Feb 2006 18:28:11 -0800 Subject: [PATCH 050/102] [PATCH] radeonfb: resume support for Samsung P35 laptops Make resume from suspend-to-ram possible for Samsung P35 laptops. The radeon mobility 9700 chip on Samsung P35 laptops locks up everything on resume from suspend-to-ram if it is not reinitialized. VGA compatible controller: ATI Technologies Inc RV350 [Mobility Radeon 9600 M10] Class 0300: 1002:4e50 Subsystem: 144d:c00c Unfortunately, the DMI strings are mostly identical for all Samsung laptops. So we match the PCI ID and subsystem ID of the graphics card which is unique for each Samsung laptop model. Signed-off-by: Carl-Daniel Hailfinger Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/video/aty/radeon_pm.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) diff --git a/drivers/video/aty/radeon_pm.c b/drivers/video/aty/radeon_pm.c index 556895e9964..1f8d805c61e 100644 --- a/drivers/video/aty/radeon_pm.c +++ b/drivers/video/aty/radeon_pm.c @@ -1321,8 +1321,6 @@ static void radeon_pm_full_reset_sdram(struct radeonfb_info *rinfo) mdelay( 15); } -#ifdef CONFIG_PPC_OF - static void radeon_pm_reset_pad_ctlr_strength(struct radeonfb_info *rinfo) { u32 tmp, tmp2; @@ -1836,6 +1834,8 @@ static void radeon_reinitialize_M10(struct radeonfb_info *rinfo) radeon_pm_m10_enable_lvds_spread_spectrum(rinfo); } +#ifdef CONFIG_PPC_OF + static void radeon_pm_m9p_reconfigure_mc(struct radeonfb_info *rinfo) { OUTREG(MC_CNTL, rinfo->save_regs[46]); @@ -2728,13 +2728,23 @@ void radeonfb_pm_init(struct radeonfb_info *rinfo, int dynclk) printk("radeonfb: Dynamic Clock Power Management disabled\n"); } +#if defined(CONFIG_PM) /* Check if we can power manage on suspend/resume. We can do * D2 on M6, M7 and M9, and we can resume from D3 cold a few other * "Mac" cards, but that's all. We need more infos about what the * BIOS does tho. Right now, all this PM stuff is pmac-only for that * reason. --BenH */ -#if defined(CONFIG_PM) && defined(CONFIG_PPC_PMAC) + /* Special case for Samsung P35 laptops + */ + if ((rinfo->pdev->vendor == PCI_VENDOR_ID_ATI) && + (rinfo->pdev->device == PCI_CHIP_RV350_NP) && + (rinfo->pdev->subsystem_vendor == PCI_VENDOR_ID_SAMSUNG) && + (rinfo->pdev->subsystem_device == 0xc00c)) { + rinfo->reinit_func = radeon_reinitialize_M10; + rinfo->pm_mode |= radeon_pm_off; + } +#if defined(CONFIG_PPC_PMAC) if (_machine == _MACH_Pmac && rinfo->of_node) { if (rinfo->is_mobility && rinfo->pm_reg && rinfo->family <= CHIP_FAMILY_RV250) @@ -2778,7 +2788,8 @@ void radeonfb_pm_init(struct radeonfb_info *rinfo, int dynclk) OUTREG(TV_DAC_CNTL, INREG(TV_DAC_CNTL) | 0x07000000); #endif } -#endif /* defined(CONFIG_PM) && defined(CONFIG_PPC_PMAC) */ +#endif /* defined(CONFIG_PPC_PMAC) */ +#endif /* defined(CONFIG_PM) */ } void radeonfb_pm_exit(struct radeonfb_info *rinfo) From 15c73691780252a5571bfa7902b4dc227ec66c84 Mon Sep 17 00:00:00 2001 From: Frank Pavlic Date: Mon, 20 Feb 2006 18:28:12 -0800 Subject: [PATCH 051/102] [PATCH] s390: V=V qdio fixes Using FCP devices with V=V support, the input queue stalled when CCQ 97 had been returned in qdio_do_eqbs. When this happen we have to reissue the eqbs instruction. Another bug was when V=V was enabled we checked if hardware has SIGA-sync support. If not we returned with 0 from tiqdio_is_inbound_q_done. Thus qdio lost initiative on FCP devices and input queue stalled. Running devices in V=V there is no SIGA-sync support but nevertheless we have to process tiqdio_is_inbound_q_done either. Signed-off-by: Frank Pavlic Signed-off-by: Heiko Carstens Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/s390/cio/qdio.c | 13 +++++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/drivers/s390/cio/qdio.c b/drivers/s390/cio/qdio.c index 45ce032772f..9ed37dc9a1b 100644 --- a/drivers/s390/cio/qdio.c +++ b/drivers/s390/cio/qdio.c @@ -165,8 +165,13 @@ qdio_do_eqbs(struct qdio_q *q, unsigned char *state, q_no = q->q_no; if(!q->is_input_q) q_no += irq->no_input_qs; +again: ccq = do_eqbs(irq->sch_token, state, q_no, start, cnt); rc = qdio_check_ccq(q, ccq); + if (rc == 1) { + QDIO_DBF_TEXT5(1,trace,"eqAGAIN"); + goto again; + } if (rc < 0) { QDIO_DBF_TEXT2(1,trace,"eqberr"); sprintf(dbf_text,"%2x,%2x,%d,%d",tmp_cnt, *cnt, ccq, q_no); @@ -195,8 +200,13 @@ qdio_do_sqbs(struct qdio_q *q, unsigned char state, q_no = q->q_no; if(!q->is_input_q) q_no += irq->no_input_qs; +again: ccq = do_sqbs(irq->sch_token, state, q_no, start, cnt); rc = qdio_check_ccq(q, ccq); + if (rc == 1) { + QDIO_DBF_TEXT5(1,trace,"sqAGAIN"); + goto again; + } if (rc < 0) { QDIO_DBF_TEXT3(1,trace,"sqberr"); sprintf(dbf_text,"%2x,%2x,%d,%d",tmp_cnt,*cnt,ccq,q_no); @@ -1187,8 +1197,7 @@ tiqdio_is_inbound_q_done(struct qdio_q *q) if (!no_used) return 1; - - if (!q->siga_sync) + if (!q->siga_sync && !irq->is_qebsm) /* we'll check for more primed buffers in qeth_stop_polling */ return 0; if (irq->is_qebsm) { From aa88861fc3184a7d830954661dd281de4ae8d2ba Mon Sep 17 00:00:00 2001 From: Peter Oberparleiter Date: Mon, 20 Feb 2006 18:28:13 -0800 Subject: [PATCH 052/102] [PATCH] s390: dasd reference counting When using the dasd diag discipline, the base discipline module (eckd or fba) can be unloaded, even though the dasd driver requires both discipline modules (base and diag) to work correctly. Implement reference counting for both base and diag discipline modules in order to fix this. Signed-off-by: Peter Oberparleiter Signed-off-by: Heiko Carstens Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/s390/block/dasd.c | 21 ++++++++++++++++++++- drivers/s390/block/dasd_int.h | 1 + 2 files changed, 21 insertions(+), 1 deletion(-) diff --git a/drivers/s390/block/dasd.c b/drivers/s390/block/dasd.c index 08c88fcd896..06bb992a4c6 100644 --- a/drivers/s390/block/dasd.c +++ b/drivers/s390/block/dasd.c @@ -156,7 +156,12 @@ dasd_state_known_to_new(struct dasd_device * device) /* disable extended error reporting for this device */ dasd_disable_eer(device); /* Forget the discipline information. */ + if (device->discipline) + module_put(device->discipline->owner); device->discipline = NULL; + if (device->base_discipline) + module_put(device->base_discipline->owner); + device->base_discipline = NULL; device->state = DASD_STATE_NEW; dasd_free_queue(device); @@ -1880,9 +1885,10 @@ dasd_generic_remove (struct ccw_device *cdev) */ int dasd_generic_set_online (struct ccw_device *cdev, - struct dasd_discipline *discipline) + struct dasd_discipline *base_discipline) { + struct dasd_discipline *discipline; struct dasd_device *device; int rc; @@ -1890,6 +1896,7 @@ dasd_generic_set_online (struct ccw_device *cdev, if (IS_ERR(device)) return PTR_ERR(device); + discipline = base_discipline; if (device->features & DASD_FEATURE_USEDIAG) { if (!dasd_diag_discipline_pointer) { printk (KERN_WARNING @@ -1901,6 +1908,16 @@ dasd_generic_set_online (struct ccw_device *cdev, } discipline = dasd_diag_discipline_pointer; } + if (!try_module_get(base_discipline->owner)) { + dasd_delete_device(device); + return -EINVAL; + } + if (!try_module_get(discipline->owner)) { + module_put(base_discipline->owner); + dasd_delete_device(device); + return -EINVAL; + } + device->base_discipline = base_discipline; device->discipline = discipline; rc = discipline->check_device(device); @@ -1909,6 +1926,8 @@ dasd_generic_set_online (struct ccw_device *cdev, "dasd_generic couldn't online device %s " "with discipline %s rc=%i\n", cdev->dev.bus_id, discipline->name, rc); + module_put(discipline->owner); + module_put(base_discipline->owner); dasd_delete_device(device); return rc; } diff --git a/drivers/s390/block/dasd_int.h b/drivers/s390/block/dasd_int.h index d1b08fa13fd..5efac1b97ec 100644 --- a/drivers/s390/block/dasd_int.h +++ b/drivers/s390/block/dasd_int.h @@ -321,6 +321,7 @@ struct dasd_device { /* Device discipline stuff. */ struct dasd_discipline *discipline; + struct dasd_discipline *base_discipline; char *private; /* Device state and target state. */ From 49d9c81a699b57a5b6488f3a761669d05e116588 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Mon, 20 Feb 2006 18:28:14 -0800 Subject: [PATCH 053/102] [PATCH] s390: revert dasd eer module Revert dasd eer module until we have a common understanding of how the interface should be. Signed-off-by: Heiko Carstens Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/s390/block/Kconfig | 14 +- drivers/s390/block/Makefile | 2 - drivers/s390/block/dasd.c | 76 +- drivers/s390/block/dasd_3990_erp.c | 3 - drivers/s390/block/dasd_eckd.h | 1 - drivers/s390/block/dasd_eer.c | 1090 ---------------------------- drivers/s390/block/dasd_int.h | 37 - include/asm-s390/dasd.h | 13 +- 8 files changed, 5 insertions(+), 1231 deletions(-) delete mode 100644 drivers/s390/block/dasd_eer.c diff --git a/drivers/s390/block/Kconfig b/drivers/s390/block/Kconfig index 6912399d093..6f50cc9323d 100644 --- a/drivers/s390/block/Kconfig +++ b/drivers/s390/block/Kconfig @@ -55,21 +55,13 @@ config DASD_DIAG Disks under VM. If you are not running under VM or unsure what it is, say "N". -config DASD_EER - tristate "Extended error reporting (EER)" - depends on DASD - help - This driver provides a character device interface to the - DASD extended error reporting. This is only needed if you want to - use applications written for the EER facility. - config DASD_CMB tristate "Compatibility interface for DASD channel measurement blocks" depends on DASD help - This driver provides an additional interface to the channel - measurement facility, which is normally accessed though sysfs, with - a set of ioctl functions specific to the dasd driver. + This driver provides an additional interface to the channel measurement + facility, which is normally accessed though sysfs, with a set of + ioctl functions specific to the dasd driver. This is only needed if you want to use applications written for linux-2.4 dasd channel measurement facility interface. diff --git a/drivers/s390/block/Makefile b/drivers/s390/block/Makefile index 0c0d871e8f5..58c6780134f 100644 --- a/drivers/s390/block/Makefile +++ b/drivers/s390/block/Makefile @@ -5,7 +5,6 @@ dasd_eckd_mod-objs := dasd_eckd.o dasd_3990_erp.o dasd_9343_erp.o dasd_fba_mod-objs := dasd_fba.o dasd_3370_erp.o dasd_9336_erp.o dasd_diag_mod-objs := dasd_diag.o -dasd_eer_mod-objs := dasd_eer.o dasd_mod-objs := dasd.o dasd_ioctl.o dasd_proc.o dasd_devmap.o \ dasd_genhd.o dasd_erp.o @@ -14,6 +13,5 @@ obj-$(CONFIG_DASD_DIAG) += dasd_diag_mod.o obj-$(CONFIG_DASD_ECKD) += dasd_eckd_mod.o obj-$(CONFIG_DASD_FBA) += dasd_fba_mod.o obj-$(CONFIG_DASD_CMB) += dasd_cmb.o -obj-$(CONFIG_DASD_EER) += dasd_eer.o obj-$(CONFIG_BLK_DEV_XPRAM) += xpram.o obj-$(CONFIG_DCSSBLK) += dcssblk.o diff --git a/drivers/s390/block/dasd.c b/drivers/s390/block/dasd.c index 06bb992a4c6..af1d5b404ce 100644 --- a/drivers/s390/block/dasd.c +++ b/drivers/s390/block/dasd.c @@ -18,7 +18,6 @@ #include #include #include -#include #include #include @@ -58,7 +57,6 @@ static void dasd_int_handler(struct ccw_device *, unsigned long, struct irb *); static void dasd_flush_ccw_queue(struct dasd_device *, int); static void dasd_tasklet(struct dasd_device *); static void do_kick_device(void *data); -static void dasd_disable_eer(struct dasd_device *device); /* * SECTION: Operations on the device structure. @@ -153,8 +151,6 @@ dasd_state_new_to_known(struct dasd_device *device) static inline void dasd_state_known_to_new(struct dasd_device * device) { - /* disable extended error reporting for this device */ - dasd_disable_eer(device); /* Forget the discipline information. */ if (device->discipline) module_put(device->discipline->owner); @@ -876,9 +872,6 @@ dasd_handle_state_change_pending(struct dasd_device *device) struct dasd_ccw_req *cqr; struct list_head *l, *n; - /* first of all call extended error reporting */ - dasd_write_eer_trigger(DASD_EER_STATECHANGE, device, NULL); - device->stopped &= ~DASD_STOPPED_PENDING; /* restart all 'running' IO on queue */ @@ -1098,19 +1091,6 @@ restart: } goto restart; } - - /* first of all call extended error reporting */ - if (device->eer && cqr->status == DASD_CQR_FAILED) { - dasd_write_eer_trigger(DASD_EER_FATALERROR, - device, cqr); - - /* restart request */ - cqr->status = DASD_CQR_QUEUED; - cqr->retries = 255; - device->stopped |= DASD_STOPPED_QUIESCE; - goto restart; - } - /* Process finished ERP request. */ if (cqr->refers) { __dasd_process_erp(device, cqr); @@ -1248,8 +1228,7 @@ __dasd_start_head(struct dasd_device * device) cqr = list_entry(device->ccw_queue.next, struct dasd_ccw_req, list); /* check FAILFAST */ if (device->stopped & ~DASD_STOPPED_PENDING && - test_bit(DASD_CQR_FLAGS_FAILFAST, &cqr->flags) && - (!device->eer)) { + test_bit(DASD_CQR_FLAGS_FAILFAST, &cqr->flags)) { cqr->status = DASD_CQR_FAILED; dasd_schedule_bh(device); } @@ -2005,9 +1984,6 @@ dasd_generic_notify(struct ccw_device *cdev, int event) switch (event) { case CIO_GONE: case CIO_NO_PATH: - /* first of all call extended error reporting */ - dasd_write_eer_trigger(DASD_EER_NOPATH, device, NULL); - if (device->state < DASD_STATE_BASIC) break; /* Device is active. We want to keep it. */ @@ -2065,51 +2041,6 @@ dasd_generic_auto_online (struct ccw_driver *dasd_discipline_driver) put_driver(drv); } -/* - * notifications for extended error reports - */ -static struct notifier_block *dasd_eer_chain; - -int -dasd_register_eer_notifier(struct notifier_block *nb) -{ - return notifier_chain_register(&dasd_eer_chain, nb); -} - -int -dasd_unregister_eer_notifier(struct notifier_block *nb) -{ - return notifier_chain_unregister(&dasd_eer_chain, nb); -} - -/* - * Notify the registered error reporting module of a problem - */ -void -dasd_write_eer_trigger(unsigned int id, struct dasd_device *device, - struct dasd_ccw_req *cqr) -{ - if (device->eer) { - struct dasd_eer_trigger temp; - temp.id = id; - temp.device = device; - temp.cqr = cqr; - notifier_call_chain(&dasd_eer_chain, DASD_EER_TRIGGER, - (void *)&temp); - } -} - -/* - * Tell the registered error reporting module to disable error reporting for - * a given device and to cleanup any private data structures on that device. - */ -static void -dasd_disable_eer(struct dasd_device *device) -{ - notifier_call_chain(&dasd_eer_chain, DASD_EER_DISABLE, (void *)device); -} - - static int __init dasd_init(void) { @@ -2191,11 +2122,6 @@ EXPORT_SYMBOL_GPL(dasd_generic_set_online); EXPORT_SYMBOL_GPL(dasd_generic_set_offline); EXPORT_SYMBOL_GPL(dasd_generic_auto_online); -EXPORT_SYMBOL(dasd_register_eer_notifier); -EXPORT_SYMBOL(dasd_unregister_eer_notifier); -EXPORT_SYMBOL(dasd_write_eer_trigger); - - /* * Overrides for Emacs so that we follow Linus's tabbing style. * Emacs will notice this stuff at the end of the file and automatically diff --git a/drivers/s390/block/dasd_3990_erp.c b/drivers/s390/block/dasd_3990_erp.c index c811380b907..4ee0f934e32 100644 --- a/drivers/s390/block/dasd_3990_erp.c +++ b/drivers/s390/block/dasd_3990_erp.c @@ -1108,9 +1108,6 @@ dasd_3990_handle_env_data(struct dasd_ccw_req * erp, char *sense) case 0x0B: DEV_MESSAGE(KERN_WARNING, device, "%s", "FORMAT F - Volume is suspended duplex"); - /* call extended error reporting (EER) */ - dasd_write_eer_trigger(DASD_EER_PPRCSUSPEND, device, - erp->refers); break; case 0x0C: DEV_MESSAGE(KERN_WARNING, device, "%s", diff --git a/drivers/s390/block/dasd_eckd.h b/drivers/s390/block/dasd_eckd.h index e15dd797805..bc3823d3522 100644 --- a/drivers/s390/block/dasd_eckd.h +++ b/drivers/s390/block/dasd_eckd.h @@ -29,7 +29,6 @@ #define DASD_ECKD_CCW_PSF 0x27 #define DASD_ECKD_CCW_RSSD 0x3e #define DASD_ECKD_CCW_LOCATE_RECORD 0x47 -#define DASD_ECKD_CCW_SNSS 0x54 #define DASD_ECKD_CCW_DEFINE_EXTENT 0x63 #define DASD_ECKD_CCW_WRITE_MT 0x85 #define DASD_ECKD_CCW_READ_MT 0x86 diff --git a/drivers/s390/block/dasd_eer.c b/drivers/s390/block/dasd_eer.c deleted file mode 100644 index f70cd7716b2..00000000000 --- a/drivers/s390/block/dasd_eer.c +++ /dev/null @@ -1,1090 +0,0 @@ -/* - * character device driver for extended error reporting - * - * - * Copyright (C) 2005 IBM Corporation - * extended error reporting for DASD ECKD devices - * Author(s): Stefan Weinhuber - * - */ - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#include -#include -#include -#include - -#include "dasd_int.h" -#include "dasd_eckd.h" - - -MODULE_LICENSE("GPL"); - -MODULE_AUTHOR("Stefan Weinhuber "); -MODULE_DESCRIPTION("DASD extended error reporting module"); - - -#ifdef PRINTK_HEADER -#undef PRINTK_HEADER -#endif /* PRINTK_HEADER */ -#define PRINTK_HEADER "dasd(eer):" - - - - - -/*****************************************************************************/ -/* the internal buffer */ -/*****************************************************************************/ - -/* - * The internal buffer is meant to store obaque blobs of data, so it doesn't - * know of higher level concepts like triggers. - * It consists of a number of pages that are used as a ringbuffer. Each data - * blob is stored in a simple record that consists of an integer, which - * contains the size of the following data, and the data bytes themselfes. - * - * To allow for multiple independent readers we create one internal buffer - * each time the device is opened and destroy the buffer when the file is - * closed again. - * - * One record can be written to a buffer by using the functions - * - dasd_eer_start_record (one time per record to write the size to the buffer - * and reserve the space for the data) - * - dasd_eer_write_buffer (one or more times per record to write the data) - * The data can be written in several steps but you will have to compute - * the total size up front for the invocation of dasd_eer_start_record. - * If the ringbuffer is full, dasd_eer_start_record will remove the required - * number of old records. - * - * A record is typically read in two steps, first read the integer that - * specifies the size of the following data, then read the data. - * Both can be done by - * - dasd_eer_read_buffer - * - * For all mentioned functions you need to get the bufferlock first and keep it - * until a complete record is written or read. - */ - - -/* - * Alle information necessary to keep track of an internal buffer is kept in - * a struct eerbuffer. The buffer specific to a file pointer is strored in - * the private_data field of that file. To be able to write data to all - * existing buffers, each buffer is also added to the bufferlist. - * If the user doesn't want to read a complete record in one go, we have to - * keep track of the rest of the record. residual stores the number of bytes - * that are still to deliver. If the rest of the record is invalidated between - * two reads then residual will be set to -1 so that the next read will fail. - * All entries in the eerbuffer structure are protected with the bufferlock. - * To avoid races between writing to a buffer on the one side and creating - * and destroying buffers on the other side, the bufferlock must also be used - * to protect the bufferlist. - */ - -struct eerbuffer { - struct list_head list; - char **buffer; - int buffersize; - int buffer_page_count; - int head; - int tail; - int residual; -}; - -LIST_HEAD(bufferlist); - -static spinlock_t bufferlock = SPIN_LOCK_UNLOCKED; - -DECLARE_WAIT_QUEUE_HEAD(dasd_eer_read_wait_queue); - -/* - * How many free bytes are available on the buffer. - * needs to be called with bufferlock held - */ -static int -dasd_eer_get_free_bytes(struct eerbuffer *eerb) -{ - if (eerb->head < eerb->tail) { - return eerb->tail - eerb->head - 1; - } else - return eerb->buffersize - eerb->head + eerb->tail -1; -} - -/* - * How many bytes of buffer space are used. - * needs to be called with bufferlock held - */ -static int -dasd_eer_get_filled_bytes(struct eerbuffer *eerb) -{ - - if (eerb->head >= eerb->tail) { - return eerb->head - eerb->tail; - } else - return eerb->buffersize - eerb->tail + eerb->head; -} - -/* - * The dasd_eer_write_buffer function just copies count bytes of data - * to the buffer. Make sure to call dasd_eer_start_record first, to - * make sure that enough free space is available. - * needs to be called with bufferlock held - */ -static void -dasd_eer_write_buffer(struct eerbuffer *eerb, int count, char *data) -{ - - unsigned long headindex,localhead; - unsigned long rest, len; - char *nextdata; - - nextdata = data; - rest = count; - while (rest > 0) { - headindex = eerb->head / PAGE_SIZE; - localhead = eerb->head % PAGE_SIZE; - len = min(rest, (PAGE_SIZE - localhead)); - memcpy(eerb->buffer[headindex]+localhead, nextdata, len); - nextdata += len; - rest -= len; - eerb->head += len; - if ( eerb->head == eerb->buffersize ) - eerb->head = 0; /* wrap around */ - if (eerb->head > eerb->buffersize) { - MESSAGE(KERN_ERR, "%s", "runaway buffer head."); - BUG(); - } - } -} - -/* - * needs to be called with bufferlock held - */ -static int -dasd_eer_read_buffer(struct eerbuffer *eerb, int count, char *data) -{ - - unsigned long tailindex,localtail; - unsigned long rest, len, finalcount; - char *nextdata; - - finalcount = min(count, dasd_eer_get_filled_bytes(eerb)); - nextdata = data; - rest = finalcount; - while (rest > 0) { - tailindex = eerb->tail / PAGE_SIZE; - localtail = eerb->tail % PAGE_SIZE; - len = min(rest, (PAGE_SIZE - localtail)); - memcpy(nextdata, eerb->buffer[tailindex]+localtail, len); - nextdata += len; - rest -= len; - eerb->tail += len; - if ( eerb->tail == eerb->buffersize ) - eerb->tail = 0; /* wrap around */ - if (eerb->tail > eerb->buffersize) { - MESSAGE(KERN_ERR, "%s", "runaway buffer tail."); - BUG(); - } - } - return finalcount; -} - -/* - * Whenever you want to write a blob of data to the internal buffer you - * have to start by using this function first. It will write the number - * of bytes that will be written to the buffer. If necessary it will remove - * old records to make room for the new one. - * needs to be called with bufferlock held - */ -static int -dasd_eer_start_record(struct eerbuffer *eerb, int count) -{ - int tailcount; - if (count + sizeof(count) > eerb->buffersize) - return -ENOMEM; - while (dasd_eer_get_free_bytes(eerb) < count + sizeof(count)) { - if (eerb->residual > 0) { - eerb->tail += eerb->residual; - if (eerb->tail >= eerb->buffersize) - eerb->tail -= eerb->buffersize; - eerb->residual = -1; - } - dasd_eer_read_buffer(eerb, sizeof(tailcount), - (char*)(&tailcount)); - eerb->tail += tailcount; - if (eerb->tail >= eerb->buffersize) - eerb->tail -= eerb->buffersize; - } - dasd_eer_write_buffer(eerb, sizeof(count), (char*)(&count)); - - return 0; -}; - -/* - * release pages that are not used anymore - */ -static void -dasd_eer_free_buffer_pages(char **buf, int no_pages) -{ - int i; - - for (i = 0; i < no_pages; ++i) { - free_page((unsigned long)buf[i]); - } -} - -/* - * allocate a new set of memory pages - */ -static int -dasd_eer_allocate_buffer_pages(char **buf, int no_pages) -{ - int i; - - for (i = 0; i < no_pages; ++i) { - buf[i] = (char *) get_zeroed_page(GFP_KERNEL); - if (!buf[i]) { - dasd_eer_free_buffer_pages(buf, i); - return -ENOMEM; - } - } - return 0; -} - -/* - * empty the buffer by resetting head and tail - * In case there is a half read data blob in the buffer, we set residual - * to -1 to indicate that the remainder of the blob is lost. - */ -static void -dasd_eer_purge_buffer(struct eerbuffer *eerb) -{ - unsigned long flags; - - spin_lock_irqsave(&bufferlock, flags); - if (eerb->residual > 0) - eerb->residual = -1; - eerb->tail=0; - eerb->head=0; - spin_unlock_irqrestore(&bufferlock, flags); -} - -/* - * set the size of the buffer, newsize is the new number of pages to be used - * we don't try to copy any data back an forth, so any resize will also purge - * the buffer - */ -static int -dasd_eer_resize_buffer(struct eerbuffer *eerb, int newsize) -{ - int i, oldcount, reuse; - char **new; - char **old; - unsigned long flags; - - if (newsize < 1) - return -EINVAL; - if (eerb->buffer_page_count == newsize) { - /* documented behaviour is that any successfull invocation - * will purge all records */ - dasd_eer_purge_buffer(eerb); - return 0; - } - new = kmalloc(newsize*sizeof(char*), GFP_KERNEL); - if (!new) - return -ENOMEM; - - reuse=min(eerb->buffer_page_count, newsize); - for (i = 0; i < reuse; ++i) { - new[i] = eerb->buffer[i]; - } - if (eerb->buffer_page_count < newsize) { - if (dasd_eer_allocate_buffer_pages( - &new[eerb->buffer_page_count], - newsize - eerb->buffer_page_count)) { - kfree(new); - return -ENOMEM; - } - } - - spin_lock_irqsave(&bufferlock, flags); - old = eerb->buffer; - eerb->buffer = new; - if (eerb->residual > 0) - eerb->residual = -1; - eerb->tail = 0; - eerb->head = 0; - oldcount = eerb->buffer_page_count; - eerb->buffer_page_count = newsize; - spin_unlock_irqrestore(&bufferlock, flags); - - if (oldcount > newsize) { - for (i = newsize; i < oldcount; ++i) { - free_page((unsigned long)old[i]); - } - } - kfree(old); - - return 0; -} - - -/*****************************************************************************/ -/* The extended error reporting functionality */ -/*****************************************************************************/ - -/* - * When a DASD device driver wants to report an error, it calls the - * function dasd_eer_write_trigger (via a notifier mechanism) and gives the - * respective trigger ID as parameter. - * Currently there are four kinds of triggers: - * - * DASD_EER_FATALERROR: all kinds of unrecoverable I/O problems - * DASD_EER_PPRCSUSPEND: PPRC was suspended - * DASD_EER_NOPATH: There is no path to the device left. - * DASD_EER_STATECHANGE: The state of the device has changed. - * - * For the first three triggers all required information can be supplied by - * the caller. For these triggers a record is written by the function - * dasd_eer_write_standard_trigger. - * - * When dasd_eer_write_trigger is called to write a DASD_EER_STATECHANGE - * trigger, we have to gather the necessary sense data first. We cannot queue - * the necessary SNSS (sense subsystem status) request immediatly, since we - * are likely to run in a deadlock situation. Instead, we schedule a - * work_struct that calls the function dasd_eer_sense_subsystem_status to - * create and start an SNSS request asynchronously. - * - * To avoid memory allocations at runtime, the necessary memory is allocated - * when the extended error reporting is enabled for a device (by - * dasd_eer_probe). There is one private eer data structure for each eer - * enabled DASD device. It contains memory for the work_struct, one SNSS cqr - * and a flags field that is used to coordinate the use of the cqr. The call - * to write a state change trigger can come in at any time, so we have one flag - * CQR_IN_USE that protects the cqr itself. When this flag indicates that the - * cqr is currently in use, dasd_eer_sense_subsystem_status cannot start a - * second request but sets the SNSS_REQUESTED flag instead. - * - * When the request is finished, the callback function dasd_eer_SNSS_cb - * is called. This function will invoke the function - * dasd_eer_write_SNSS_trigger to finally write the trigger. It will also - * check the SNSS_REQUESTED flag and if it is set it will call - * dasd_eer_sense_subsystem_status again. - * - * To avoid race conditions during the handling of the lock, the flags must - * be protected by the snsslock. - */ - -struct dasd_eer_private { - struct dasd_ccw_req *cqr; - unsigned long flags; - struct work_struct worker; -}; - -static void dasd_eer_destroy(struct dasd_device *device, - struct dasd_eer_private *eer); -static int -dasd_eer_write_trigger(struct dasd_eer_trigger *trigger); -static void dasd_eer_sense_subsystem_status(void *data); -static int dasd_eer_notify(struct notifier_block *self, - unsigned long action, void *data); - -struct workqueue_struct *dasd_eer_workqueue; - -#define SNSS_DATA_SIZE 44 -static spinlock_t snsslock = SPIN_LOCK_UNLOCKED; - -#define DASD_EER_BUSID_SIZE 10 -struct dasd_eer_header { - __u32 total_size; - __u32 trigger; - __u64 tv_sec; - __u64 tv_usec; - char busid[DASD_EER_BUSID_SIZE]; -} __attribute__ ((packed)); - -static struct notifier_block dasd_eer_nb = { - .notifier_call = dasd_eer_notify, -}; - -/* - * flags for use with dasd_eer_private - */ -#define CQR_IN_USE 0 -#define SNSS_REQUESTED 1 - -/* - * This function checks if extended error reporting is available for a given - * dasd_device. If yes, then it creates and returns a struct dasd_eer, - * otherwise it returns an -EPERM error pointer. - */ -struct dasd_eer_private * -dasd_eer_probe(struct dasd_device *device) -{ - struct dasd_eer_private *private; - - if (!(device && device->discipline - && !strcmp(device->discipline->name, "ECKD"))) { - return ERR_PTR(-EPERM); - } - /* allocate the private data structure */ - private = (struct dasd_eer_private *)kmalloc( - sizeof(struct dasd_eer_private), GFP_KERNEL); - if (!private) { - return ERR_PTR(-ENOMEM); - } - INIT_WORK(&private->worker, dasd_eer_sense_subsystem_status, - (void *)device); - private->cqr = dasd_kmalloc_request("ECKD", - 1 /* SNSS */ , - SNSS_DATA_SIZE , - device); - if (!private->cqr) { - kfree(private); - return ERR_PTR(-ENOMEM); - } - private->flags = 0; - return private; -}; - -/* - * If our private SNSS request is queued, remove it from the - * dasd ccw queue so we can free the requests memory. - */ -static void -dasd_eer_dequeue_SNSS_request(struct dasd_device *device, - struct dasd_eer_private *eer) -{ - struct list_head *lst, *nxt; - struct dasd_ccw_req *cqr, *erpcqr; - dasd_erp_fn_t erp_fn; - - spin_lock_irq(get_ccwdev_lock(device->cdev)); - list_for_each_safe(lst, nxt, &device->ccw_queue) { - cqr = list_entry(lst, struct dasd_ccw_req, list); - /* we are looking for two kinds or requests */ - /* first kind: our SNSS request: */ - if (cqr == eer->cqr) { - if (cqr->status == DASD_CQR_IN_IO) - device->discipline->term_IO(cqr); - list_del(&cqr->list); - break; - } - /* second kind: ERP requests for our SNSS request */ - if (cqr->refers) { - /* If this erp request chain ends in our cqr, then */ - /* cal the erp_postaction to clean it up */ - erpcqr = cqr; - while (erpcqr->refers) { - erpcqr = erpcqr->refers; - } - if (erpcqr == eer->cqr) { - erp_fn = device->discipline->erp_postaction( - cqr); - erp_fn(cqr); - } - continue; - } - } - spin_unlock_irq(get_ccwdev_lock(device->cdev)); -} - -/* - * This function dismantles a struct dasd_eer that was created by - * dasd_eer_probe. Since we want to free our private data structure, - * we must make sure that the memory is not in use anymore. - * We have to flush the work queue and remove a possible SNSS request - * from the dasd queue. - */ -static void -dasd_eer_destroy(struct dasd_device *device, struct dasd_eer_private *eer) -{ - flush_workqueue(dasd_eer_workqueue); - dasd_eer_dequeue_SNSS_request(device, eer); - dasd_kfree_request(eer->cqr, device); - kfree(eer); -}; - -/* - * enable the extended error reporting for a particular device - */ -static int -dasd_eer_enable_on_device(struct dasd_device *device) -{ - void *eer; - if (!device) - return -ENODEV; - if (device->eer) - return 0; - if (!try_module_get(THIS_MODULE)) { - return -EINVAL; - } - eer = (void *)dasd_eer_probe(device); - if (IS_ERR(eer)) { - module_put(THIS_MODULE); - return PTR_ERR(eer); - } - device->eer = eer; - return 0; -} - -/* - * enable the extended error reporting for a particular device - */ -static int -dasd_eer_disable_on_device(struct dasd_device *device) -{ - struct dasd_eer_private *eer = device->eer; - - if (!device) - return -ENODEV; - if (!device->eer) - return 0; - device->eer = NULL; - dasd_eer_destroy(device,eer); - module_put(THIS_MODULE); - - return 0; -} - -/* - * Set extended error reporting (eer) - * Note: This will be registered as a DASD ioctl, to be called on DASD devices. - */ -static int -dasd_ioctl_set_eer(struct block_device *bdev, int no, long args) -{ - struct dasd_device *device; - int intval; - - if (!capable(CAP_SYS_ADMIN)) - return -EACCES; - if (bdev != bdev->bd_contains) - /* Error-reporting is not allowed for partitions */ - return -EINVAL; - if (get_user(intval, (int __user *) args)) - return -EFAULT; - device = bdev->bd_disk->private_data; - if (device == NULL) - return -ENODEV; - - intval = (intval != 0); - DEV_MESSAGE (KERN_DEBUG, device, - "set eer on device to %d", intval); - if (intval) - return dasd_eer_enable_on_device(device); - else - return dasd_eer_disable_on_device(device); -} - -/* - * Get value of extended error reporting. - * Note: This will be registered as a DASD ioctl, to be called on DASD devices. - */ -static int -dasd_ioctl_get_eer(struct block_device *bdev, int no, long args) -{ - struct dasd_device *device; - - device = bdev->bd_disk->private_data; - if (device == NULL) - return -ENODEV; - return put_user((device->eer != NULL), (int __user *) args); -} - -/* - * The following function can be used for those triggers that have - * all necessary data available when the function is called. - * If the parameter cqr is not NULL, the chain of requests will be searched - * for valid sense data, and all valid sense data sets will be added to - * the triggers data. - */ -static int -dasd_eer_write_standard_trigger(int trigger, struct dasd_device *device, - struct dasd_ccw_req *cqr) -{ - struct dasd_ccw_req *temp_cqr; - int data_size; - struct timeval tv; - struct dasd_eer_header header; - unsigned long flags; - struct eerbuffer *eerb; - - /* go through cqr chain and count the valid sense data sets */ - temp_cqr = cqr; - data_size = 0; - while (temp_cqr) { - if (temp_cqr->irb.esw.esw0.erw.cons) - data_size += 32; - temp_cqr = temp_cqr->refers; - } - - header.total_size = sizeof(header) + data_size + 4; /* "EOR" */ - header.trigger = trigger; - do_gettimeofday(&tv); - header.tv_sec = tv.tv_sec; - header.tv_usec = tv.tv_usec; - strncpy(header.busid, device->cdev->dev.bus_id, DASD_EER_BUSID_SIZE); - - spin_lock_irqsave(&bufferlock, flags); - list_for_each_entry(eerb, &bufferlist, list) { - dasd_eer_start_record(eerb, header.total_size); - dasd_eer_write_buffer(eerb, sizeof(header), (char*)(&header)); - temp_cqr = cqr; - while (temp_cqr) { - if (temp_cqr->irb.esw.esw0.erw.cons) - dasd_eer_write_buffer(eerb, 32, cqr->irb.ecw); - temp_cqr = temp_cqr->refers; - } - dasd_eer_write_buffer(eerb, 4,"EOR"); - } - spin_unlock_irqrestore(&bufferlock, flags); - - wake_up_interruptible(&dasd_eer_read_wait_queue); - - return 0; -} - -/* - * This function writes a DASD_EER_STATECHANGE trigger. - */ -static void -dasd_eer_write_SNSS_trigger(struct dasd_device *device, - struct dasd_ccw_req *cqr) -{ - int data_size; - int snss_rc; - struct timeval tv; - struct dasd_eer_header header; - unsigned long flags; - struct eerbuffer *eerb; - - snss_rc = (cqr->status == DASD_CQR_FAILED) ? -EIO : 0; - if (snss_rc) - data_size = 0; - else - data_size = SNSS_DATA_SIZE; - - header.total_size = sizeof(header) + data_size + 4; /* "EOR" */ - header.trigger = DASD_EER_STATECHANGE; - do_gettimeofday(&tv); - header.tv_sec = tv.tv_sec; - header.tv_usec = tv.tv_usec; - strncpy(header.busid, device->cdev->dev.bus_id, DASD_EER_BUSID_SIZE); - - spin_lock_irqsave(&bufferlock, flags); - list_for_each_entry(eerb, &bufferlist, list) { - dasd_eer_start_record(eerb, header.total_size); - dasd_eer_write_buffer(eerb, sizeof(header),(char*)(&header)); - if (!snss_rc) - dasd_eer_write_buffer(eerb, SNSS_DATA_SIZE, cqr->data); - dasd_eer_write_buffer(eerb, 4,"EOR"); - } - spin_unlock_irqrestore(&bufferlock, flags); - - wake_up_interruptible(&dasd_eer_read_wait_queue); -} - -/* - * callback function for use with SNSS request - */ -static void -dasd_eer_SNSS_cb(struct dasd_ccw_req *cqr, void *data) -{ - struct dasd_device *device; - struct dasd_eer_private *private; - unsigned long irqflags; - - device = (struct dasd_device *)data; - private = (struct dasd_eer_private *)device->eer; - dasd_eer_write_SNSS_trigger(device, cqr); - spin_lock_irqsave(&snsslock, irqflags); - if(!test_and_clear_bit(SNSS_REQUESTED, &private->flags)) { - clear_bit(CQR_IN_USE, &private->flags); - spin_unlock_irqrestore(&snsslock, irqflags); - return; - }; - clear_bit(CQR_IN_USE, &private->flags); - spin_unlock_irqrestore(&snsslock, irqflags); - dasd_eer_sense_subsystem_status(device); - return; -} - -/* - * clean a used cqr before using it again - */ -static void -dasd_eer_clean_SNSS_request(struct dasd_ccw_req *cqr) -{ - struct ccw1 *cpaddr = cqr->cpaddr; - void *data = cqr->data; - - memset(cqr, 0, sizeof(struct dasd_ccw_req)); - memset(cpaddr, 0, sizeof(struct ccw1)); - memset(data, 0, SNSS_DATA_SIZE); - cqr->cpaddr = cpaddr; - cqr->data = data; - strncpy((char *) &cqr->magic, "ECKD", 4); - ASCEBC((char *) &cqr->magic, 4); - set_bit(DASD_CQR_FLAGS_USE_ERP, &cqr->flags); -} - -/* - * build and start an SNSS request - * This function is called from a work queue so we have to - * pass the dasd_device pointer as a void pointer. - */ -static void -dasd_eer_sense_subsystem_status(void *data) -{ - struct dasd_device *device; - struct dasd_eer_private *private; - struct dasd_ccw_req *cqr; - struct ccw1 *ccw; - unsigned long irqflags; - - device = (struct dasd_device *)data; - private = (struct dasd_eer_private *)device->eer; - if (!private) /* device not eer enabled any more */ - return; - cqr = private->cqr; - spin_lock_irqsave(&snsslock, irqflags); - if(test_and_set_bit(CQR_IN_USE, &private->flags)) { - set_bit(SNSS_REQUESTED, &private->flags); - spin_unlock_irqrestore(&snsslock, irqflags); - return; - }; - spin_unlock_irqrestore(&snsslock, irqflags); - dasd_eer_clean_SNSS_request(cqr); - cqr->device = device; - cqr->retries = 255; - cqr->expires = 10 * HZ; - - ccw = cqr->cpaddr; - ccw->cmd_code = DASD_ECKD_CCW_SNSS; - ccw->count = SNSS_DATA_SIZE; - ccw->flags = 0; - ccw->cda = (__u32)(addr_t)cqr->data; - - cqr->buildclk = get_clock(); - cqr->status = DASD_CQR_FILLED; - cqr->callback = dasd_eer_SNSS_cb; - cqr->callback_data = (void *)device; - dasd_add_request_head(cqr); - - return; -} - -/* - * This function is called for all triggers. It calls the appropriate - * function that writes the actual trigger records. - */ -static int -dasd_eer_write_trigger(struct dasd_eer_trigger *trigger) -{ - int rc; - struct dasd_eer_private *private = trigger->device->eer; - - switch (trigger->id) { - case DASD_EER_FATALERROR: - case DASD_EER_PPRCSUSPEND: - rc = dasd_eer_write_standard_trigger( - trigger->id, trigger->device, trigger->cqr); - break; - case DASD_EER_NOPATH: - rc = dasd_eer_write_standard_trigger( - trigger->id, trigger->device, NULL); - break; - case DASD_EER_STATECHANGE: - if (queue_work(dasd_eer_workqueue, &private->worker)) { - rc=0; - } else { - /* If the work_struct was already queued, it can't - * be queued again. But this is OK since we don't - * need to have it queued twice. - */ - rc = -EBUSY; - } - break; - default: /* unknown trigger, so we write it without any sense data */ - rc = dasd_eer_write_standard_trigger( - trigger->id, trigger->device, NULL); - break; - } - return rc; -} - -/* - * This function is registered with the dasd device driver and gets called - * for all dasd eer notifications. - */ -static int dasd_eer_notify(struct notifier_block *self, - unsigned long action, void *data) -{ - switch (action) { - case DASD_EER_DISABLE: - dasd_eer_disable_on_device((struct dasd_device *)data); - break; - case DASD_EER_TRIGGER: - dasd_eer_write_trigger((struct dasd_eer_trigger *)data); - break; - } - return NOTIFY_OK; -} - - -/*****************************************************************************/ -/* the device operations */ -/*****************************************************************************/ - -/* - * On the one side we need a lock to access our internal buffer, on the - * other side a copy_to_user can sleep. So we need to copy the data we have - * to transfer in a readbuffer, which is protected by the readbuffer_mutex. - */ -static char readbuffer[PAGE_SIZE]; -DECLARE_MUTEX(readbuffer_mutex); - - -static int -dasd_eer_open(struct inode *inp, struct file *filp) -{ - struct eerbuffer *eerb; - unsigned long flags; - - eerb = kmalloc(sizeof(struct eerbuffer), GFP_KERNEL); - eerb->head = 0; - eerb->tail = 0; - eerb->residual = 0; - eerb->buffer_page_count = 1; - eerb->buffersize = eerb->buffer_page_count * PAGE_SIZE; - eerb->buffer = kmalloc(eerb->buffer_page_count*sizeof(char*), - GFP_KERNEL); - if (!eerb->buffer) - return -ENOMEM; - if (dasd_eer_allocate_buffer_pages(eerb->buffer, - eerb->buffer_page_count)) { - kfree(eerb->buffer); - return -ENOMEM; - } - filp->private_data = eerb; - spin_lock_irqsave(&bufferlock, flags); - list_add(&eerb->list, &bufferlist); - spin_unlock_irqrestore(&bufferlock, flags); - - return nonseekable_open(inp,filp); -} - -static int -dasd_eer_close(struct inode *inp, struct file *filp) -{ - struct eerbuffer *eerb; - unsigned long flags; - - eerb = (struct eerbuffer *)filp->private_data; - spin_lock_irqsave(&bufferlock, flags); - list_del(&eerb->list); - spin_unlock_irqrestore(&bufferlock, flags); - dasd_eer_free_buffer_pages(eerb->buffer, eerb->buffer_page_count); - kfree(eerb->buffer); - kfree(eerb); - - return 0; -} - -static long -dasd_eer_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) -{ - int intval; - struct eerbuffer *eerb; - - eerb = (struct eerbuffer *)filp->private_data; - switch (cmd) { - case DASD_EER_PURGE: - dasd_eer_purge_buffer(eerb); - return 0; - case DASD_EER_SETBUFSIZE: - if (get_user(intval, (int __user *)arg)) - return -EFAULT; - return dasd_eer_resize_buffer(eerb, intval); - default: - return -ENOIOCTLCMD; - } -} - -static ssize_t -dasd_eer_read(struct file *filp, char __user *buf, size_t count, loff_t *ppos) -{ - int tc,rc; - int tailcount,effective_count; - unsigned long flags; - struct eerbuffer *eerb; - - eerb = (struct eerbuffer *)filp->private_data; - if(down_interruptible(&readbuffer_mutex)) - return -ERESTARTSYS; - - spin_lock_irqsave(&bufferlock, flags); - - if (eerb->residual < 0) { /* the remainder of this record */ - /* has been deleted */ - eerb->residual = 0; - spin_unlock_irqrestore(&bufferlock, flags); - up(&readbuffer_mutex); - return -EIO; - } else if (eerb->residual > 0) { - /* OK we still have a second half of a record to deliver */ - effective_count = min(eerb->residual, (int)count); - eerb->residual -= effective_count; - } else { - tc = 0; - while (!tc) { - tc = dasd_eer_read_buffer(eerb, - sizeof(tailcount), (char*)(&tailcount)); - if (!tc) { - /* no data available */ - spin_unlock_irqrestore(&bufferlock, flags); - up(&readbuffer_mutex); - if (filp->f_flags & O_NONBLOCK) - return -EAGAIN; - rc = wait_event_interruptible( - dasd_eer_read_wait_queue, - eerb->head != eerb->tail); - if (rc) { - return rc; - } - if(down_interruptible(&readbuffer_mutex)) - return -ERESTARTSYS; - spin_lock_irqsave(&bufferlock, flags); - } - } - WARN_ON(tc != sizeof(tailcount)); - effective_count = min(tailcount,(int)count); - eerb->residual = tailcount - effective_count; - } - - tc = dasd_eer_read_buffer(eerb, effective_count, readbuffer); - WARN_ON(tc != effective_count); - - spin_unlock_irqrestore(&bufferlock, flags); - - if (copy_to_user(buf, readbuffer, effective_count)) { - up(&readbuffer_mutex); - return -EFAULT; - } - - up(&readbuffer_mutex); - return effective_count; -} - -static unsigned int -dasd_eer_poll (struct file *filp, poll_table *ptable) -{ - unsigned int mask; - unsigned long flags; - struct eerbuffer *eerb; - - eerb = (struct eerbuffer *)filp->private_data; - poll_wait(filp, &dasd_eer_read_wait_queue, ptable); - spin_lock_irqsave(&bufferlock, flags); - if (eerb->head != eerb->tail) - mask = POLLIN | POLLRDNORM ; - else - mask = 0; - spin_unlock_irqrestore(&bufferlock, flags); - return mask; -} - -static struct file_operations dasd_eer_fops = { - .open = &dasd_eer_open, - .release = &dasd_eer_close, - .unlocked_ioctl = &dasd_eer_ioctl, - .compat_ioctl = &dasd_eer_ioctl, - .read = &dasd_eer_read, - .poll = &dasd_eer_poll, - .owner = THIS_MODULE, -}; - -static struct miscdevice dasd_eer_dev = { - .minor = MISC_DYNAMIC_MINOR, - .name = "dasd_eer", - .fops = &dasd_eer_fops, -}; - - -/*****************************************************************************/ -/* Init and exit */ -/*****************************************************************************/ - -static int -__init dasd_eer_init(void) -{ - int rc; - - dasd_eer_workqueue = create_singlethread_workqueue("dasd_eer"); - if (!dasd_eer_workqueue) { - MESSAGE(KERN_ERR , "%s", "dasd_eer_init could not " - "create workqueue \n"); - rc = -ENOMEM; - goto out; - } - - rc = dasd_register_eer_notifier(&dasd_eer_nb); - if (rc) { - MESSAGE(KERN_ERR, "%s", "dasd_eer_init could not " - "register error reporting"); - goto queue; - } - - dasd_ioctl_no_register(THIS_MODULE, BIODASDEERSET, dasd_ioctl_set_eer); - dasd_ioctl_no_register(THIS_MODULE, BIODASDEERGET, dasd_ioctl_get_eer); - - /* we don't need our own character device, - * so we just register as misc device */ - rc = misc_register(&dasd_eer_dev); - if (rc) { - MESSAGE(KERN_ERR, "%s", "dasd_eer_init could not " - "register misc device"); - goto unregister; - } - - return 0; - -unregister: - dasd_unregister_eer_notifier(&dasd_eer_nb); - dasd_ioctl_no_unregister(THIS_MODULE, BIODASDEERSET, - dasd_ioctl_set_eer); - dasd_ioctl_no_unregister(THIS_MODULE, BIODASDEERGET, - dasd_ioctl_get_eer); -queue: - destroy_workqueue(dasd_eer_workqueue); -out: - return rc; - -} -module_init(dasd_eer_init); - -static void -__exit dasd_eer_exit(void) -{ - dasd_unregister_eer_notifier(&dasd_eer_nb); - dasd_ioctl_no_unregister(THIS_MODULE, BIODASDEERSET, - dasd_ioctl_set_eer); - dasd_ioctl_no_unregister(THIS_MODULE, BIODASDEERGET, - dasd_ioctl_get_eer); - destroy_workqueue(dasd_eer_workqueue); - - WARN_ON(misc_deregister(&dasd_eer_dev) != 0); -} -module_exit(dasd_eer_exit); diff --git a/drivers/s390/block/dasd_int.h b/drivers/s390/block/dasd_int.h index 5efac1b97ec..0592354cc60 100644 --- a/drivers/s390/block/dasd_int.h +++ b/drivers/s390/block/dasd_int.h @@ -275,34 +275,6 @@ struct dasd_discipline { extern struct dasd_discipline *dasd_diag_discipline_pointer; - -/* - * Notification numbers for extended error reporting notifications: - * The DASD_EER_DISABLE notification is sent before a dasd_device (and it's - * eer pointer) is freed. The error reporting module needs to do all necessary - * cleanup steps. - * The DASD_EER_TRIGGER notification sends the actual error reports (triggers). - */ -#define DASD_EER_DISABLE 0 -#define DASD_EER_TRIGGER 1 - -/* Trigger IDs for extended error reporting DASD_EER_TRIGGER notification */ -#define DASD_EER_FATALERROR 1 -#define DASD_EER_NOPATH 2 -#define DASD_EER_STATECHANGE 3 -#define DASD_EER_PPRCSUSPEND 4 - -/* - * The dasd_eer_trigger structure contains all data that we need to send - * along with an DASD_EER_TRIGGER notification. - */ -struct dasd_eer_trigger { - unsigned int id; - struct dasd_device *device; - struct dasd_ccw_req *cqr; -}; - - struct dasd_device { /* Block device stuff. */ struct gendisk *gdp; @@ -316,9 +288,6 @@ struct dasd_device { unsigned long flags; /* per device flags */ unsigned short features; /* copy of devmap-features (read-only!) */ - /* extended error reporting stuff (eer) */ - void *eer; - /* Device discipline stuff. */ struct dasd_discipline *discipline; struct dasd_discipline *base_discipline; @@ -520,12 +489,6 @@ int dasd_generic_set_online(struct ccw_device *, struct dasd_discipline *); int dasd_generic_set_offline (struct ccw_device *cdev); int dasd_generic_notify(struct ccw_device *, int); void dasd_generic_auto_online (struct ccw_driver *); -int dasd_register_eer_notifier(struct notifier_block *); -int dasd_unregister_eer_notifier(struct notifier_block *); -void dasd_write_eer_trigger(unsigned int , struct dasd_device *, - struct dasd_ccw_req *); - - /* externals in dasd_devmap.c */ extern int dasd_max_devindex; diff --git a/include/asm-s390/dasd.h b/include/asm-s390/dasd.h index c744ff33b1d..1630c26e8f4 100644 --- a/include/asm-s390/dasd.h +++ b/include/asm-s390/dasd.h @@ -204,8 +204,7 @@ typedef struct attrib_data_t { * * Here ist how the ioctl-nr should be used: * 0 - 31 DASD driver itself - * 32 - 229 still open - * 230 - 239 DASD extended error reporting + * 32 - 239 still open * 240 - 255 reserved for EMC *******************************************************************************/ @@ -237,22 +236,12 @@ typedef struct attrib_data_t { #define BIODASDPSRD _IOR(DASD_IOCTL_LETTER,4,dasd_rssd_perf_stats_t) /* Get Attributes (cache operations) */ #define BIODASDGATTR _IOR(DASD_IOCTL_LETTER,5,attrib_data_t) -/* retrieve extended error-reporting value */ -#define BIODASDEERGET _IOR(DASD_IOCTL_LETTER,6,int) /* #define BIODASDFORMAT _IOW(IOCTL_LETTER,0,format_data_t) , deprecated */ #define BIODASDFMT _IOW(DASD_IOCTL_LETTER,1,format_data_t) /* Set Attributes (cache operations) */ #define BIODASDSATTR _IOW(DASD_IOCTL_LETTER,2,attrib_data_t) -/* retrieve extended error-reporting value */ -#define BIODASDEERSET _IOW(DASD_IOCTL_LETTER,3,int) - - -/* remove all records from the eer buffer */ -#define DASD_EER_PURGE _IO(DASD_IOCTL_LETTER,230) -/* set the number of pages that are used for the internal eer buffer */ -#define DASD_EER_SETBUFSIZE _IOW(DASD_IOCTL_LETTER,230,int) #endif /* DASD_H */ From b04ec261bd64f927bf3fce5cf9eeb0225557939d Mon Sep 17 00:00:00 2001 From: Hirokazu Takata Date: Mon, 20 Feb 2006 18:28:15 -0800 Subject: [PATCH 054/102] [PATCH] m32r: __cmpxchg_u32 fix This patch fixes a bug of include/asm-m32r/system.h:__cmpxchg_u32(). static __inline__ unsigned long __cmpxchg_u32(volatile unsigned int *p, unsigned int old, unsigned int new); In __cmpxchg_u32(), the "old" value must not be changed to the previous "*p" value. But the former code modifies the previous "*p" value. A deadlock at _atomic_dec_and_lock sometimes happened due to this bug. Signed-off-by: Hayato Fujiwara Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/asm-m32r/system.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/asm-m32r/system.h b/include/asm-m32r/system.h index 06c12a037cb..d6a2c613be6 100644 --- a/include/asm-m32r/system.h +++ b/include/asm-m32r/system.h @@ -239,7 +239,7 @@ __cmpxchg_u32(volatile unsigned int *p, unsigned int old, unsigned int new) " bra 2f; \n" " .fillinsn \n" "1:" - M32R_UNLOCK" %2, @%1; \n" + M32R_UNLOCK" %0, @%1; \n" " .fillinsn \n" "2:" : "=&r" (retval) From cf535ea52e68e3ee6f4a90cc383faa1ee857f14d Mon Sep 17 00:00:00 2001 From: Hirokazu Takata Date: Mon, 20 Feb 2006 18:28:17 -0800 Subject: [PATCH 055/102] [PATCH] m32r: update sys_tas() routine This patch updates and fixes sys_tas() routine for m32r. In the previous implementation, a lockup rarely caused at sys_tas() routine in SMP environment. > > The problem is that touching *addr will generate an oops if that page isn't > > paged in. If we convert it to use get_user() then that's an improvement, > > but we must not run get_user() under spinlock or local_irq_disable(). I rewrote sys_tas() routine by using "lock -> unlock" instructions, and utilizing the m32r's interrupt handling characteristics; the m32r processor can accept interrupts only at the 32-bit instruction boundary. So, the "unlock" instruction can be executed continuously after the "lock" instruction execution without any interruptions. In addition, to solve such a page_fault problem, I use a fixup code like get_user(). And, as for the kernel lockup problem, we found that a calling do_page_fault() routine with disabling interrupts might cause a lockup at flush_tlb_others(), because we checked a completion of IPI handler's operations in a spin-locked critical section. Therefore, by using "lock -> unlock" code, we can implement the sys_tas() rouitine without disabling interrupts explicitly, then no lockups would happen at flush_tlb_others(), I hope. Compile check and some working test in SMP environment have done. Signed-off-by: Hirokazu Takata Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/m32r/kernel/sys_m32r.c | 61 +++++++++++++++++++++---------------- 1 file changed, 34 insertions(+), 27 deletions(-) diff --git a/arch/m32r/kernel/sys_m32r.c b/arch/m32r/kernel/sys_m32r.c index fe55b28d372..670cb49210a 100644 --- a/arch/m32r/kernel/sys_m32r.c +++ b/arch/m32r/kernel/sys_m32r.c @@ -29,28 +29,7 @@ /* * sys_tas() - test-and-set - * linuxthreads testing version */ -#ifndef CONFIG_SMP -asmlinkage int sys_tas(int *addr) -{ - int oldval; - unsigned long flags; - - if (!access_ok(VERIFY_WRITE, addr, sizeof (int))) - return -EFAULT; - local_irq_save(flags); - oldval = *addr; - if (!oldval) - *addr = 1; - local_irq_restore(flags); - return oldval; -} -#else /* CONFIG_SMP */ -#include - -static DEFINE_SPINLOCK(tas_lock); - asmlinkage int sys_tas(int *addr) { int oldval; @@ -58,15 +37,43 @@ asmlinkage int sys_tas(int *addr) if (!access_ok(VERIFY_WRITE, addr, sizeof (int))) return -EFAULT; - _raw_spin_lock(&tas_lock); - oldval = *addr; - if (!oldval) - *addr = 1; - _raw_spin_unlock(&tas_lock); + /* atomic operation: + * oldval = *addr; *addr = 1; + */ + __asm__ __volatile__ ( + DCACHE_CLEAR("%0", "r4", "%1") + " .fillinsn\n" + "1:\n" + " lock %0, @%1 -> unlock %2, @%1\n" + "2:\n" + /* NOTE: + * The m32r processor can accept interrupts only + * at the 32-bit instruction boundary. + * So, in the above code, the "unlock" instruction + * can be executed continuously after the "lock" + * instruction execution without any interruptions. + */ + ".section .fixup,\"ax\"\n" + " .balign 4\n" + "3: ldi %0, #%3\n" + " seth r14, #high(2b)\n" + " or3 r14, r14, #low(2b)\n" + " jmp r14\n" + ".previous\n" + ".section __ex_table,\"a\"\n" + " .balign 4\n" + " .long 1b,3b\n" + ".previous\n" + : "=&r" (oldval) + : "r" (addr), "r" (1), "i"(-EFAULT) + : "r14", "memory" +#ifdef CONFIG_CHIP_M32700_TS1 + , "r4" +#endif /* CONFIG_CHIP_M32700_TS1 */ + ); return oldval; } -#endif /* CONFIG_SMP */ /* * sys_pipe() is the normal C calling standard for creating From 35e622a67e6d2c5f7e3d3e2da92ebdb2f46db783 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Mon, 20 Feb 2006 12:41:55 +0000 Subject: [PATCH 056/102] [PATCH] H8/300: CONFIG_CONFIG_ doesn't fly. All actual uses of the symbol refer to CONFIG_SH_STANDARD_BIOS so this option could never be activated on H8/300. Signed-off-by: Ralf Baechle Signed-off-by: Linus Torvalds --- arch/h8300/Kconfig.debug | 2 +- arch/h8300/defconfig | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/h8300/Kconfig.debug b/arch/h8300/Kconfig.debug index 55034d08abf..e0e9bcb015a 100644 --- a/arch/h8300/Kconfig.debug +++ b/arch/h8300/Kconfig.debug @@ -34,7 +34,7 @@ config GDB_DEBUG help gdb stub exception support -config CONFIG_SH_STANDARD_BIOS +config SH_STANDARD_BIOS bool "Use gdb protocol serial console" depends on (!H8300H_SIM && !H8S_SIM) help diff --git a/arch/h8300/defconfig b/arch/h8300/defconfig index 9d9b491cfc2..8f1ec329715 100644 --- a/arch/h8300/defconfig +++ b/arch/h8300/defconfig @@ -328,7 +328,7 @@ CONFIG_FULLDEBUG=y CONFIG_NO_KERNEL_MSG=y # CONFIG_SYSCALL_PRINT is not set # CONFIG_GDB_DEBUG is not set -# CONFIG_CONFIG_SH_STANDARD_BIOS is not set +# CONFIG_SH_STANDARD_BIOS is not set # CONFIG_DEFAULT_CMDLINE is not set # CONFIG_BLKDEV_RESERVE is not set From 5914811acf36c3ff091f860a6964808f668f27d0 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Bj=F6rn=20Steinbrink?= Date: Sat, 18 Feb 2006 18:12:43 +0100 Subject: [PATCH 057/102] [PATCH] kjournald keeps reference to namespace MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In daemonize() a new thread gets cleaned up and 'merged' with init_task. The current fs_struct is handled there, but not the current namespace. This adds the namespace part. [ Eric Biederman pointed out the namespace wrappers, and also notes that we can't ever count on using our parents namespace because we already have called exit_fs(), which is the only way to the namespace from a process. ] Signed-off-by: Björn Steinbrink Acked-by: Eric Biederman Signed-off-by: Linus Torvalds --- kernel/exit.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/kernel/exit.c b/kernel/exit.c index 93cee367133..531aadca553 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -360,6 +360,9 @@ void daemonize(const char *name, ...) fs = init_task.fs; current->fs = fs; atomic_inc(&fs->count); + exit_namespace(current); + current->namespace = init_task.namespace; + get_namespace(current->namespace); exit_files(current); current->files = init_task.files; atomic_inc(¤t->files->count); From 36ccf1c0e3917f1f73abc17c38ad704c59f8d1b6 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Tue, 14 Feb 2006 21:04:54 +0000 Subject: [PATCH 058/102] [MIPS] Make integer overflow exceptions in kernel mode fatal. Signed-off-by: Ralf Baechle --- arch/mips/kernel/traps.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/arch/mips/kernel/traps.c b/arch/mips/kernel/traps.c index c9d2b5147ca..005debbfbe8 100644 --- a/arch/mips/kernel/traps.c +++ b/arch/mips/kernel/traps.c @@ -3,7 +3,7 @@ * License. See the file "COPYING" in the main directory of this archive * for more details. * - * Copyright (C) 1994 - 1999, 2000, 01 Ralf Baechle + * Copyright (C) 1994 - 1999, 2000, 01, 06 Ralf Baechle * Copyright (C) 1995, 1996 Paul M. Antoine * Copyright (C) 1998 Ulf Carlsson * Copyright (C) 1999 Silicon Graphics, Inc. @@ -548,6 +548,8 @@ asmlinkage void do_ov(struct pt_regs *regs) { siginfo_t info; + die_if_kernel("Integer overflow", regs); + info.si_code = FPE_INTOVF; info.si_signo = SIGFPE; info.si_errno = 0; From 8ecbbcaf08c13c57d6602472478739d64650ee0e Mon Sep 17 00:00:00 2001 From: Atsushi Nemoto Date: Tue, 14 Feb 2006 15:57:50 +0900 Subject: [PATCH 059/102] [MIPS] Fixes for uaccess.h with gcc >= 4.0.1 It seems current get_user() incorrectly sign-extend an unsigned int value on 64bit kernel. I think this is because '(__typeof__(val))' cast in final assignment. I suppose the cast should be '(__typeof__(*(addr))'. Signed-off-by: Atsushi Nemoto Signed-off-by: Ralf Baechle --- include/asm-mips/uaccess.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/include/asm-mips/uaccess.h b/include/asm-mips/uaccess.h index 7a553e9d44d..b96f3e0f393 100644 --- a/include/asm-mips/uaccess.h +++ b/include/asm-mips/uaccess.h @@ -233,7 +233,7 @@ do { \ #define __get_user_check(x,ptr,size) \ ({ \ long __gu_err = -EFAULT; \ - const void __user * __gu_ptr = (ptr); \ + const __typeof__(*(ptr)) __user * __gu_ptr = (ptr); \ \ if (likely(access_ok(VERIFY_READ, __gu_ptr, size))) \ __get_user_common((x), size, __gu_ptr); \ @@ -258,7 +258,7 @@ do { \ : "=r" (__gu_err), "=r" (__gu_tmp) \ : "0" (0), "o" (__m(addr)), "i" (-EFAULT)); \ \ - (val) = (__typeof__(val)) __gu_tmp; \ + (val) = (__typeof__(*(addr))) __gu_tmp; \ } /* @@ -284,7 +284,7 @@ do { \ " .previous \n" \ : "=r" (__gu_err), "=&r" (__gu_tmp) \ : "0" (0), "r" (addr), "i" (-EFAULT)); \ - (val) = __gu_tmp; \ + (val) = (__typeof__(*(addr))) __gu_tmp; \ } /* From 68fa383f3e58b5adf1d8089c93c83ab8d0d00e8d Mon Sep 17 00:00:00 2001 From: Martin Michlmayr Date: Sat, 18 Feb 2006 14:55:45 +0000 Subject: [PATCH 060/102] [MIPS] Add support for TIF_RESTORE_SIGMASK for signal32 Following the recent implementation of TIF_RESTORE_SIGMASK in arch/mips/kernel/signal.c, 64-bit kernels with 32-bit user-land compatibility oops when starting init. signal32.c needs to be converted to use TIF_RESTORE_SIGMASK too. Signed-off-by: Martin Michlmayr Signed-off-by: Ralf Baechle --- arch/mips/kernel/signal32.c | 68 ++++++++++++++++++++++--------------- 1 file changed, 41 insertions(+), 27 deletions(-) diff --git a/arch/mips/kernel/signal32.c b/arch/mips/kernel/signal32.c index 8a8b8dd9041..c0701957521 100644 --- a/arch/mips/kernel/signal32.c +++ b/arch/mips/kernel/signal32.c @@ -106,8 +106,6 @@ typedef struct compat_siginfo { #define _BLOCKABLE (~(sigmask(SIGKILL) | sigmask(SIGSTOP))) -extern int do_signal32(sigset_t *oldset, struct pt_regs *regs); - /* 32-bit compatibility types */ #define _NSIG_BPW32 32 @@ -198,7 +196,7 @@ __attribute_used__ noinline static int _sys32_sigsuspend(nabi_no_regargs struct pt_regs regs) { compat_sigset_t *uset; - sigset_t newset, saveset; + sigset_t newset; uset = (compat_sigset_t *) regs.regs[4]; if (get_sigset(&newset, uset)) @@ -206,19 +204,15 @@ _sys32_sigsuspend(nabi_no_regargs struct pt_regs regs) sigdelsetmask(&newset, ~_BLOCKABLE); spin_lock_irq(¤t->sighand->siglock); - saveset = current->blocked; + current->saved_sigmask = current->blocked; current->blocked = newset; recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); - regs.regs[2] = EINTR; - regs.regs[7] = 1; - while (1) { - current->state = TASK_INTERRUPTIBLE; - schedule(); - if (do_signal32(&saveset, ®s)) - return -EINTR; - } + current->state = TASK_INTERRUPTIBLE; + schedule(); + set_thread_flag(TIF_RESTORE_SIGMASK); + return -ERESTARTNOHAND; } save_static_function(sys32_rt_sigsuspend); @@ -226,7 +220,7 @@ __attribute_used__ noinline static int _sys32_rt_sigsuspend(nabi_no_regargs struct pt_regs regs) { compat_sigset_t *uset; - sigset_t newset, saveset; + sigset_t newset; size_t sigsetsize; /* XXX Don't preclude handling different sized sigset_t's. */ @@ -240,19 +234,15 @@ _sys32_rt_sigsuspend(nabi_no_regargs struct pt_regs regs) sigdelsetmask(&newset, ~_BLOCKABLE); spin_lock_irq(¤t->sighand->siglock); - saveset = current->blocked; + current->saved_sigmask = current->blocked; current->blocked = newset; recalc_sigpending(); spin_unlock_irq(¤t->sighand->siglock); - regs.regs[2] = EINTR; - regs.regs[7] = 1; - while (1) { - current->state = TASK_INTERRUPTIBLE; - schedule(); - if (do_signal32(&saveset, ®s)) - return -EINTR; - } + current->state = TASK_INTERRUPTIBLE; + schedule(); + set_thread_flag(TIF_RESTORE_SIGMASK); + return -ERESTARTNOHAND; } asmlinkage int sys32_sigaction(int sig, const struct sigaction32 *act, @@ -783,7 +773,7 @@ static inline int handle_signal(unsigned long sig, siginfo_t *info, regs->regs[2] = EINTR; break; case ERESTARTSYS: - if(!(ka->sa.sa_flags & SA_RESTART)) { + if (!(ka->sa.sa_flags & SA_RESTART)) { regs->regs[2] = EINTR; break; } @@ -810,9 +800,10 @@ static inline int handle_signal(unsigned long sig, siginfo_t *info, return ret; } -int do_signal32(sigset_t *oldset, struct pt_regs *regs) +int do_signal32(struct pt_regs *regs) { struct k_sigaction ka; + sigset_t *oldset; siginfo_t info; int signr; @@ -827,12 +818,25 @@ int do_signal32(sigset_t *oldset, struct pt_regs *regs) if (try_to_freeze()) goto no_signal; - if (!oldset) + if (test_thread_flag(TIF_RESTORE_SIGMASK)) + oldset = ¤t->saved_sigmask; + else oldset = ¤t->blocked; signr = get_signal_to_deliver(&info, &ka, regs, NULL); - if (signr > 0) - return handle_signal(signr, &info, &ka, oldset, regs); + if (signr > 0) { + /* Whee! Actually deliver the signal. */ + if (handle_signal(signr, &info, &ka, oldset, regs) == 0) { + /* + * A signal was successfully delivered; the saved + * sigmask will have been stored in the signal frame, + * and will be restored by sigreturn, so we can simply + * clear the TIF_RESTORE_SIGMASK flag. + */ + if (test_thread_flag(TIF_RESTORE_SIGMASK)) + clear_thread_flag(TIF_RESTORE_SIGMASK); + } + } no_signal: /* @@ -853,6 +857,16 @@ no_signal: regs->cp0_epc -= 4; } } + + /* + * If there's no signal to deliver, we just put the saved sigmask + * back + */ + if (test_thread_flag(TIF_RESTORE_SIGMASK)) { + clear_thread_flag(TIF_RESTORE_SIGMASK); + sigprocmask(SIG_SETMASK, ¤t->saved_sigmask, NULL); + } + return 0; } From dda73d0bb1d358e4337d2c4da9c61903873664cf Mon Sep 17 00:00:00 2001 From: Martin Michlmayr Date: Sat, 18 Feb 2006 15:21:30 +0000 Subject: [PATCH 061/102] [MIPS] Make do_signal32 return void. do_signal has been changed to return void since the "return value is ignored everywhere". Convert do_signal32 accordingly. Signed-off-by: Martin Michlmayr Signed-off-by: Ralf Baechle --- arch/mips/kernel/signal32.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/arch/mips/kernel/signal32.c b/arch/mips/kernel/signal32.c index c0701957521..118a0a9d9a2 100644 --- a/arch/mips/kernel/signal32.c +++ b/arch/mips/kernel/signal32.c @@ -4,7 +4,7 @@ * for more details. * * Copyright (C) 1991, 1992 Linus Torvalds - * Copyright (C) 1994 - 2000 Ralf Baechle + * Copyright (C) 1994 - 2000, 2006 Ralf Baechle * Copyright (C) 1999, 2000 Silicon Graphics, Inc. */ #include @@ -800,7 +800,7 @@ static inline int handle_signal(unsigned long sig, siginfo_t *info, return ret; } -int do_signal32(struct pt_regs *regs) +void do_signal32(struct pt_regs *regs) { struct k_sigaction ka; sigset_t *oldset; @@ -813,7 +813,7 @@ int do_signal32(struct pt_regs *regs) * if so. */ if (!user_mode(regs)) - return 1; + return; if (try_to_freeze()) goto no_signal; @@ -866,8 +866,6 @@ no_signal: clear_thread_flag(TIF_RESTORE_SIGMASK); sigprocmask(SIG_SETMASK, ¤t->saved_sigmask, NULL); } - - return 0; } asmlinkage int sys32_rt_sigaction(int sig, const struct sigaction32 *act, From 304416da860b8cd548e61ee7038f852418008a85 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Sat, 18 Feb 2006 18:20:47 +0000 Subject: [PATCH 062/102] [MIPS] Reformat _sys32_rt_sigsuspend with tabs instead of space for consistency. Signed-off-by: Ralf Baechle --- arch/mips/kernel/signal32.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/mips/kernel/signal32.c b/arch/mips/kernel/signal32.c index 118a0a9d9a2..237cd8a2cd3 100644 --- a/arch/mips/kernel/signal32.c +++ b/arch/mips/kernel/signal32.c @@ -221,7 +221,7 @@ _sys32_rt_sigsuspend(nabi_no_regargs struct pt_regs regs) { compat_sigset_t *uset; sigset_t newset; - size_t sigsetsize; + size_t sigsetsize; /* XXX Don't preclude handling different sized sigset_t's. */ sigsetsize = regs.regs[5]; From 82ad93f4a002294820e9a5e6f84beef2222a54b7 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Sat, 18 Feb 2006 22:47:26 +0000 Subject: [PATCH 063/102] [MIPS] N32: Fix N32 rt_sigtimedwait and rt_sigsuspend breakage. Originally found through an oops in the Gentoo N32 userland build; patch based on original patch by Daniel Jacobwitz. Signed-off-by: Ralf Baechle --- arch/mips/kernel/linux32.c | 19 ------------------- arch/mips/kernel/scall64-n32.S | 4 ++-- arch/mips/kernel/signal_n32.c | 33 +++++++++++++++++++++++++++++++++ 3 files changed, 35 insertions(+), 21 deletions(-) diff --git a/arch/mips/kernel/linux32.c b/arch/mips/kernel/linux32.c index 60353f5acc4..9996b6e8458 100644 --- a/arch/mips/kernel/linux32.c +++ b/arch/mips/kernel/linux32.c @@ -1450,25 +1450,6 @@ sys32_timer_create(u32 clock, struct sigevent32 __user *se32, timer_t __user *ti return sys_timer_create(clock, p, timer_id); } -asmlinkage long -sysn32_rt_sigtimedwait(const sigset_t __user *uthese, - siginfo_t __user *uinfo, - const struct compat_timespec __user *uts32, - size_t sigsetsize) -{ - struct timespec __user *uts = NULL; - - if (uts32) { - struct timespec ts; - uts = compat_alloc_user_space(sizeof(struct timespec)); - if (get_user(ts.tv_sec, &uts32->tv_sec) || - get_user(ts.tv_nsec, &uts32->tv_nsec) || - copy_to_user (uts, &ts, sizeof (ts))) - return -EFAULT; - } - return sys_rt_sigtimedwait(uthese, uinfo, uts, sigsetsize); -} - save_static_function(sys32_clone); __attribute_used__ noinline static int _sys32_clone(nabi_no_regargs struct pt_regs regs) diff --git a/arch/mips/kernel/scall64-n32.S b/arch/mips/kernel/scall64-n32.S index bc4980cefc8..d87b5446fa1 100644 --- a/arch/mips/kernel/scall64-n32.S +++ b/arch/mips/kernel/scall64-n32.S @@ -245,9 +245,9 @@ EXPORT(sysn32_call_table) PTR sys_capget PTR sys_capset PTR sys32_rt_sigpending /* 6125 */ - PTR sysn32_rt_sigtimedwait + PTR compat_sys_rt_sigtimedwait PTR sys_rt_sigqueueinfo - PTR sys32_rt_sigsuspend + PTR sysn32_rt_sigsuspend PTR sys32_sigaltstack PTR compat_sys_utime /* 6130 */ PTR sys_mknod diff --git a/arch/mips/kernel/signal_n32.c b/arch/mips/kernel/signal_n32.c index 5a3776096f0..3e168c08a3a 100644 --- a/arch/mips/kernel/signal_n32.c +++ b/arch/mips/kernel/signal_n32.c @@ -81,6 +81,39 @@ struct rt_sigframe_n32 { #endif }; +extern void sigset_from_compat (sigset_t *set, compat_sigset_t *compat); + +save_static_function(sysn32_rt_sigsuspend); +__attribute_used__ noinline static int +_sysn32_rt_sigsuspend(nabi_no_regargs struct pt_regs regs) +{ + compat_sigset_t __user *unewset, uset; + size_t sigsetsize; + sigset_t newset; + + /* XXX Don't preclude handling different sized sigset_t's. */ + sigsetsize = regs.regs[5]; + if (sigsetsize != sizeof(sigset_t)) + return -EINVAL; + + unewset = (compat_sigset_t __user *) regs.regs[4]; + if (copy_from_user(&uset, unewset, sizeof(uset))) + return -EFAULT; + sigset_from_compat (&newset, &uset); + sigdelsetmask(&newset, ~_BLOCKABLE); + + spin_lock_irq(¤t->sighand->siglock); + current->saved_sigmask = current->blocked; + current->blocked = newset; + recalc_sigpending(); + spin_unlock_irq(¤t->sighand->siglock); + + current->state = TASK_INTERRUPTIBLE; + schedule(); + set_thread_flag(TIF_RESTORE_SIGMASK); + return -ERESTARTNOHAND; +} + save_static_function(sysn32_rt_sigreturn); __attribute_used__ noinline static void _sysn32_rt_sigreturn(nabi_no_regargs struct pt_regs regs) From f3468e0c34c8de919062582575a01e2434c8e727 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Sun, 19 Feb 2006 03:42:11 +0000 Subject: [PATCH 064/102] [MIPS] N32: Make sure pointer is good before passing it to sys_waitid(). After all we're calling sys_waitid() with fs set to KERNEL_DS ... Signed-off-by: Ralf Baechle --- arch/mips/kernel/linux32.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/mips/kernel/linux32.c b/arch/mips/kernel/linux32.c index 9996b6e8458..5f68b220c26 100644 --- a/arch/mips/kernel/linux32.c +++ b/arch/mips/kernel/linux32.c @@ -230,6 +230,9 @@ sysn32_waitid(int which, compat_pid_t pid, long ret; mm_segment_t old_fs = get_fs(); + if (!access_ok(VERIFY_WRITE, uinfo, sizeof(*uinfo))) + return -EFAULT; + set_fs (KERNEL_DS); ret = sys_waitid(which, pid, uinfo, options, uru ? (struct rusage __user *) &ru : NULL); From 51939fbb79f66cc471f5bde9845be797ea61ad0b Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Thu, 10 Nov 2005 16:35:03 +0000 Subject: [PATCH 065/102] [MIPS] Sibyte: #if CONFIG_* doesn't fly. Signed-off-by: Ralf Baechle --- arch/mips/mm/cex-sb1.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/mips/mm/cex-sb1.S b/arch/mips/mm/cex-sb1.S index 0e71580774f..e54a62f2807 100644 --- a/arch/mips/mm/cex-sb1.S +++ b/arch/mips/mm/cex-sb1.S @@ -64,7 +64,7 @@ LEAF(except_vec2_sb1) sd k0,0x170($0) sd k1,0x178($0) -#if CONFIG_SB1_CEX_ALWAYS_FATAL +#ifdef CONFIG_SB1_CEX_ALWAYS_FATAL j handle_vec2_sb1 nop #else From 77607635c3f15e6bf6366e6d7db731a5cb209fb1 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Thu, 10 Nov 2005 16:32:14 +0000 Subject: [PATCH 066/102] [MIPS] Sibyte: Config option names shouldn't be prefixed with CONFIG_ Signed-off-by: Ralf Baechle --- arch/mips/sibyte/Kconfig | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/mips/sibyte/Kconfig b/arch/mips/sibyte/Kconfig index de46f62ac46..816aee7fcd2 100644 --- a/arch/mips/sibyte/Kconfig +++ b/arch/mips/sibyte/Kconfig @@ -102,11 +102,11 @@ config SIMULATION Build a kernel suitable for running under the GDB simulator. Primarily adjusts the kernel's notion of time. -config CONFIG_SB1_CEX_ALWAYS_FATAL +config SB1_CEX_ALWAYS_FATAL bool "All cache exceptions considered fatal (no recovery attempted)" depends on SIBYTE_SB1xxx_SOC -config CONFIG_SB1_CERR_STALL +config SB1_CERR_STALL bool "Stall (rather than panic) on fatal cache error" depends on SIBYTE_SB1xxx_SOC From 124273773596cbf8aa9c79304b01091d4693f368 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Tue, 14 Feb 2006 14:22:10 +0000 Subject: [PATCH 067/102] [MIPS] Follow Uli's latest *at syscall changes. (This really is only the half of the patch which was forgotten in 326a625748535c4cdb1c632b1dcb07030989a393 ...) Signed-off-by: Ralf Baechle --- include/asm-mips/unistd.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/include/asm-mips/unistd.h b/include/asm-mips/unistd.h index 769305d2010..b5c78a4a019 100644 --- a/include/asm-mips/unistd.h +++ b/include/asm-mips/unistd.h @@ -313,7 +313,7 @@ #define __NR_mknodat (__NR_Linux + 290) #define __NR_fchownat (__NR_Linux + 291) #define __NR_futimesat (__NR_Linux + 292) -#define __NR_newfstatat (__NR_Linux + 293) +#define __NR_fstatat (__NR_Linux + 293) #define __NR_unlinkat (__NR_Linux + 294) #define __NR_renameat (__NR_Linux + 295) #define __NR_linkat (__NR_Linux + 296) @@ -593,7 +593,7 @@ #define __NR_mknodat (__NR_Linux + 249) #define __NR_fchownat (__NR_Linux + 250) #define __NR_futimesat (__NR_Linux + 251) -#define __NR_newfstatat (__NR_Linux + 252) +#define __NR_fstatat (__NR_Linux + 252) #define __NR_unlinkat (__NR_Linux + 253) #define __NR_renameat (__NR_Linux + 254) #define __NR_linkat (__NR_Linux + 255) @@ -877,7 +877,7 @@ #define __NR_mknodat (__NR_Linux + 253) #define __NR_fchownat (__NR_Linux + 254) #define __NR_futimesat (__NR_Linux + 255) -#define __NR_newfstatat (__NR_Linux + 256) +#define __NR_fstatat (__NR_Linux + 256) #define __NR_unlinkat (__NR_Linux + 257) #define __NR_renameat (__NR_Linux + 258) #define __NR_linkat (__NR_Linux + 259) From 76e1daee7db153400aaed55646e05a312da353b3 Mon Sep 17 00:00:00 2001 From: Martin Michlmayr Date: Mon, 20 Feb 2006 04:57:00 +0000 Subject: [PATCH 068/102] [MIPS] Fix compiler warnings in arch/mips/sibyte/bcm1480/irq.c MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix the following compiler warnings: CC arch/mips/sibyte/bcm1480/irq.o arch/mips/sibyte/bcm1480/irq.c: In function ‘bcm1480_set_affinity’: arch/mips/sibyte/bcm1480/irq.c:168: warning: ISO C90 forbids mixed declarations and code arch/mips/sibyte/bcm1480/irq.c: In function ‘ack_bcm1480_irq’: arch/mips/sibyte/bcm1480/irq.c:230: warning: ISO C90 forbids mixed declarations and code Signed-off-by: Martin Michlmayr Signed-off-by: Ralf Baechle --- arch/mips/sibyte/bcm1480/irq.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/arch/mips/sibyte/bcm1480/irq.c b/arch/mips/sibyte/bcm1480/irq.c index b2a1ba5d23d..9cf7d713b13 100644 --- a/arch/mips/sibyte/bcm1480/irq.c +++ b/arch/mips/sibyte/bcm1480/irq.c @@ -139,7 +139,7 @@ void bcm1480_unmask_irq(int cpu, int irq) #ifdef CONFIG_SMP static void bcm1480_set_affinity(unsigned int irq, cpumask_t mask) { - int i = 0, old_cpu, cpu, int_on; + int i = 0, old_cpu, cpu, int_on, k; u64 cur_ints; irq_desc_t *desc = irq_desc + irq; unsigned long flags; @@ -165,7 +165,6 @@ static void bcm1480_set_affinity(unsigned int irq, cpumask_t mask) irq_dirty -= BCM1480_NR_IRQS_HALF; } - int k; for (k=0; k<2; k++) { /* Loop through high and low interrupt mask register */ cur_ints = ____raw_readq(IOADDR(A_BCM1480_IMR_MAPPER(old_cpu) + R_BCM1480_IMR_INTERRUPT_MASK_H + (k*BCM1480_IMR_HL_SPACING))); int_on = !(cur_ints & (((u64) 1) << irq_dirty)); @@ -216,6 +215,7 @@ static void ack_bcm1480_irq(unsigned int irq) { u64 pending; unsigned int irq_dirty; + int k; /* * If the interrupt was an HT interrupt, now is the time to @@ -227,7 +227,6 @@ static void ack_bcm1480_irq(unsigned int irq) if ((irq_dirty >= BCM1480_NR_IRQS_HALF) && (irq_dirty <= BCM1480_NR_IRQS)) { irq_dirty -= BCM1480_NR_IRQS_HALF; } - int k; for (k=0; k<2; k++) { /* Loop through high and low LDT interrupts */ pending = __raw_readq(IOADDR(A_BCM1480_IMR_REGISTER(bcm1480_irq_owner[irq], R_BCM1480_IMR_LDT_INTERRUPT_H + (k*BCM1480_IMR_HL_SPACING)))); From 1e35aabab82f8a6f7ab884b642e68be260f92f2c Mon Sep 17 00:00:00 2001 From: Rojhalat Ibrahim Date: Mon, 20 Feb 2006 13:35:27 +0000 Subject: [PATCH 069/102] [MIPS] Add topology_init. A recent patch introduced cpu topology in sysfs. When you run a kernel with SMP and sysfs enabled, you now get an Oops on boot. The following patch fixes that by adding topology_init to arch/mips/kernel/smp.c. The code is copied from arch/s390/kernel/smp.c. Signed-off-by: Rojhalat Ibrahim Signed-off-by: Ralf Baechle --- arch/mips/kernel/smp.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/arch/mips/kernel/smp.c b/arch/mips/kernel/smp.c index 25472fcaf71..5e189862e52 100644 --- a/arch/mips/kernel/smp.c +++ b/arch/mips/kernel/smp.c @@ -29,6 +29,7 @@ #include #include #include +#include #include #include @@ -424,6 +425,25 @@ void flush_tlb_one(unsigned long vaddr) local_flush_tlb_one(vaddr); } +static DEFINE_PER_CPU(struct cpu, cpu_devices); + +static int __init topology_init(void) +{ + int cpu; + int ret; + + for_each_cpu(cpu) { + ret = register_cpu(&per_cpu(cpu_devices, cpu), cpu, NULL); + if (ret) + printk(KERN_WARNING "topology_init: register_cpu %d " + "failed (%d)\n", cpu, ret); + } + + return 0; +} + +subsys_initcall(topology_init); + EXPORT_SYMBOL(flush_tlb_page); EXPORT_SYMBOL(flush_tlb_one); EXPORT_SYMBOL(cpu_data); From 909fa64c8bf0ef20722cddaa402404fb1aa14a4a Mon Sep 17 00:00:00 2001 From: Atsushi Nemoto Date: Tue, 21 Feb 2006 01:25:06 +0900 Subject: [PATCH 070/102] [MIPS] jiffies_to_compat_timeval fix The last argument of div_long_long_rem() must be long. Signed-off-by: Atsushi Nemoto Signed-off-by: Ralf Baechle --- arch/mips/kernel/binfmt_elfn32.c | 5 +++-- arch/mips/kernel/binfmt_elfo32.c | 5 +++-- 2 files changed, 6 insertions(+), 4 deletions(-) diff --git a/arch/mips/kernel/binfmt_elfn32.c b/arch/mips/kernel/binfmt_elfn32.c index d8e2674a154..4a9f1ecefaf 100644 --- a/arch/mips/kernel/binfmt_elfn32.c +++ b/arch/mips/kernel/binfmt_elfn32.c @@ -103,8 +103,9 @@ jiffies_to_compat_timeval(unsigned long jiffies, struct compat_timeval *value) * one divide. */ u64 nsec = (u64)jiffies * TICK_NSEC; - value->tv_sec = div_long_long_rem(nsec, NSEC_PER_SEC, &value->tv_usec); - value->tv_usec /= NSEC_PER_USEC; + long rem; + value->tv_sec = div_long_long_rem(nsec, NSEC_PER_SEC, &rem); + value->tv_usec = rem / NSEC_PER_USEC; } #define ELF_CORE_EFLAGS EF_MIPS_ABI2 diff --git a/arch/mips/kernel/binfmt_elfo32.c b/arch/mips/kernel/binfmt_elfo32.c index cec5f327e36..e3181377989 100644 --- a/arch/mips/kernel/binfmt_elfo32.c +++ b/arch/mips/kernel/binfmt_elfo32.c @@ -105,8 +105,9 @@ jiffies_to_compat_timeval(unsigned long jiffies, struct compat_timeval *value) * one divide. */ u64 nsec = (u64)jiffies * TICK_NSEC; - value->tv_sec = div_long_long_rem(nsec, NSEC_PER_SEC, &value->tv_usec); - value->tv_usec /= NSEC_PER_USEC; + long rem; + value->tv_sec = div_long_long_rem(nsec, NSEC_PER_SEC, &rem); + value->tv_usec = rem / NSEC_PER_USEC; } #undef ELF_CORE_COPY_REGS From 1e93e70d035a26c0e108f0a9e8f735dabcf948ac Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Tue, 21 Feb 2006 00:17:50 +0000 Subject: [PATCH 071/102] [MIPS] Yosemite: Fix build damage by dc8f6029cd51af1b148846a32e68d69013a5cc0f. Signed-off-by: Ralf Baechle --- arch/mips/pmc-sierra/yosemite/smp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/mips/pmc-sierra/yosemite/smp.c b/arch/mips/pmc-sierra/yosemite/smp.c index f17f575f58f..7f8fda96219 100644 --- a/arch/mips/pmc-sierra/yosemite/smp.c +++ b/arch/mips/pmc-sierra/yosemite/smp.c @@ -94,7 +94,7 @@ void __init prom_prepare_cpus(unsigned int max_cpus) void prom_boot_secondary(int cpu, struct task_struct *idle) { unsigned long gp = (unsigned long) task_thread_info(idle); - unsigned long sp = __KSTK_TOP(idle); + unsigned long sp = __KSTK_TOS(idle); secondary_sp = sp; secondary_gp = gp; From 8db41685c73ad1893d8571861171e183a551e90d Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Tue, 21 Feb 2006 13:46:01 +0100 Subject: [PATCH 072/102] [MIPS] Disable CONFIG_ISCSI_TCP; it triggers a gcc 3.4 endless loop. Signed-off-by: Ralf Baechle --- arch/mips/configs/ip27_defconfig | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/arch/mips/configs/ip27_defconfig b/arch/mips/configs/ip27_defconfig index e17d3adff02..58c22cd344d 100644 --- a/arch/mips/configs/ip27_defconfig +++ b/arch/mips/configs/ip27_defconfig @@ -1,7 +1,7 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.15-rc2 -# Thu Nov 24 01:06:21 2005 +# Linux kernel version: 2.6.16-rc4 +# Tue Feb 21 13:44:31 2006 # CONFIG_MIPS=y @@ -144,7 +144,6 @@ CONFIG_PREEMPT_BKL=y # Code maturity level options # CONFIG_EXPERIMENTAL=y -CONFIG_CLEAN_COMPILE=y CONFIG_LOCK_KERNEL=y CONFIG_INIT_ENV_ARG_LIMIT=32 @@ -250,6 +249,7 @@ CONFIG_NET=y # # Networking options # +# CONFIG_NETDEBUG is not set CONFIG_PACKET=y CONFIG_PACKET_MMAP=y CONFIG_UNIX=y @@ -289,6 +289,7 @@ CONFIG_TCP_CONG_BIC=y # SCTP Configuration (EXPERIMENTAL) # # CONFIG_IP_SCTP is not set + # CONFIG_ATM is not set # CONFIG_BRIDGE is not set # CONFIG_VLAN_8021Q is not set @@ -448,7 +449,7 @@ CONFIG_SCSI_SAS_ATTRS=m # # SCSI low-level drivers # -CONFIG_ISCSI_TCP=m +# CONFIG_ISCSI_TCP is not set # CONFIG_BLK_DEV_3W_XXXX_RAID is not set # CONFIG_SCSI_3W_9XXX is not set # CONFIG_SCSI_ACARD is not set @@ -773,6 +774,10 @@ CONFIG_USB_ARCH_HAS_OHCI=y # SN Devices # +# +# EDAC - error detection and reporting (RAS) +# + # # File systems # From b00dc3ad74fdb676552d46ee573b88e927240d0c Mon Sep 17 00:00:00 2001 From: Hugh Dickins Date: Tue, 21 Feb 2006 23:49:47 +0000 Subject: [PATCH 073/102] [PATCH] tmpfs: fix mount mpol nodelist parsing I've been dissatisfied with the mpol_nodelist mount option which was added to tmpfs earlier in -rc. Replace it by mpol=policy:nodelist. And it was broken: a nodelist is a comma-separated list of numbers and ranges; the mount options are a comma-separated list of token=values. Whoops, blindly strsep'ing on commas doesn't work so well: since we've no numeric tokens, and unlikely to add them, use that to distinguish. Move the mpol= parsing to shmem_parse_mpol under CONFIG_NUMA, reject all its options as invalid if not NUMA. /proc shows MPOL_PREFERRED as "prefer", so use that name for the policy instead of "preferred". Enforce that mpol=default has no nodelist; that mpol=prefer has one node only; that mpol=bind has a nodelist; but let mpol=interleave use node_online_map if no nodelist given. Describe this in tmpfs.txt. Signed-off-by: Hugh Dickins Acked-by: Robin Holt Acked-by: Andi Kleen Signed-off-by: Linus Torvalds --- Documentation/filesystems/tmpfs.txt | 21 ++++---- mm/shmem.c | 81 ++++++++++++++++++++++++----- 2 files changed, 81 insertions(+), 21 deletions(-) diff --git a/Documentation/filesystems/tmpfs.txt b/Documentation/filesystems/tmpfs.txt index dbe4d87d261..8a155418c70 100644 --- a/Documentation/filesystems/tmpfs.txt +++ b/Documentation/filesystems/tmpfs.txt @@ -79,15 +79,18 @@ that instance in a system with many cpus making intensive use of it. tmpfs has a mount option to set the NUMA memory allocation policy for -all files in that instance: -mpol=interleave prefers to allocate memory from each node in turn -mpol=default prefers to allocate memory from the local node -mpol=bind prefers to allocate from mpol_nodelist -mpol=preferred prefers to allocate from first node in mpol_nodelist +all files in that instance (if CONFIG_NUMA is enabled) - which can be +adjusted on the fly via 'mount -o remount ...' -The following mount option is used in conjunction with mpol=interleave, -mpol=bind or mpol=preferred: -mpol_nodelist: nodelist suitable for parsing with nodelist_parse. +mpol=default prefers to allocate memory from the local node +mpol=prefer:Node prefers to allocate memory from the given Node +mpol=bind:NodeList allocates memory only from nodes in NodeList +mpol=interleave prefers to allocate from each node in turn +mpol=interleave:NodeList allocates from each node of NodeList in turn + +NodeList format is a comma-separated list of decimal numbers and ranges, +a range being two hyphen-separated decimal numbers, the smallest and +largest node numbers in the range. For example, mpol=bind:0-3,5,7,9-15 To specify the initial root directory you can use the following mount @@ -109,4 +112,4 @@ RAM/SWAP in 10240 inodes and it is only accessible by root. Author: Christoph Rohland , 1.12.01 Updated: - Hugh Dickins , 13 March 2005 + Hugh Dickins , 19 February 2006 diff --git a/mm/shmem.c b/mm/shmem.c index f7ac7b812f9..7c455fbaff7 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -45,6 +45,7 @@ #include #include #include +#include #include #include #include @@ -874,6 +875,51 @@ redirty: } #ifdef CONFIG_NUMA +static int shmem_parse_mpol(char *value, int *policy, nodemask_t *policy_nodes) +{ + char *nodelist = strchr(value, ':'); + int err = 1; + + if (nodelist) { + /* NUL-terminate policy string */ + *nodelist++ = '\0'; + if (nodelist_parse(nodelist, *policy_nodes)) + goto out; + } + if (!strcmp(value, "default")) { + *policy = MPOL_DEFAULT; + /* Don't allow a nodelist */ + if (!nodelist) + err = 0; + } else if (!strcmp(value, "prefer")) { + *policy = MPOL_PREFERRED; + /* Insist on a nodelist of one node only */ + if (nodelist) { + char *rest = nodelist; + while (isdigit(*rest)) + rest++; + if (!*rest) + err = 0; + } + } else if (!strcmp(value, "bind")) { + *policy = MPOL_BIND; + /* Insist on a nodelist */ + if (nodelist) + err = 0; + } else if (!strcmp(value, "interleave")) { + *policy = MPOL_INTERLEAVE; + /* Default to nodes online if no nodelist */ + if (!nodelist) + *policy_nodes = node_online_map; + err = 0; + } +out: + /* Restore string for error message */ + if (nodelist) + *--nodelist = ':'; + return err; +} + static struct page *shmem_swapin_async(struct shared_policy *p, swp_entry_t entry, unsigned long idx) { @@ -926,6 +972,11 @@ shmem_alloc_page(gfp_t gfp, struct shmem_inode_info *info, return page; } #else +static inline int shmem_parse_mpol(char *value, int *policy, nodemask_t *policy_nodes) +{ + return 1; +} + static inline struct page * shmem_swapin(struct shmem_inode_info *info,swp_entry_t entry,unsigned long idx) { @@ -1859,7 +1910,23 @@ static int shmem_parse_options(char *options, int *mode, uid_t *uid, { char *this_char, *value, *rest; - while ((this_char = strsep(&options, ",")) != NULL) { + while (options != NULL) { + this_char = options; + for (;;) { + /* + * NUL-terminate this option: unfortunately, + * mount options form a comma-separated list, + * but mpol's nodelist may also contain commas. + */ + options = strchr(options, ','); + if (options == NULL) + break; + options++; + if (!isdigit(*options)) { + options[-1] = '\0'; + break; + } + } if (!*this_char) continue; if ((value = strchr(this_char,'=')) != NULL) { @@ -1910,18 +1977,8 @@ static int shmem_parse_options(char *options, int *mode, uid_t *uid, if (*rest) goto bad_val; } else if (!strcmp(this_char,"mpol")) { - if (!strcmp(value,"default")) - *policy = MPOL_DEFAULT; - else if (!strcmp(value,"preferred")) - *policy = MPOL_PREFERRED; - else if (!strcmp(value,"bind")) - *policy = MPOL_BIND; - else if (!strcmp(value,"interleave")) - *policy = MPOL_INTERLEAVE; - else + if (shmem_parse_mpol(value,policy,policy_nodes)) goto bad_val; - } else if (!strcmp(this_char,"mpol_nodelist")) { - nodelist_parse(value, *policy_nodes); } else { printk(KERN_ERR "tmpfs: Bad mount option %s\n", this_char); From 5bd546aa78b5d74f3162815e41940f862215d9e3 Mon Sep 17 00:00:00 2001 From: Russell King Date: Fri, 17 Feb 2006 20:23:29 +0000 Subject: [PATCH 074/102] [MMC] Fix mmc_cmd_type() mask It's MMC_CMD_MASK not MMC_CMD_TYPE. Signed-off-by: Russell King --- include/linux/mmc/mmc.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/mmc/mmc.h b/include/linux/mmc/mmc.h index f38872abc12..bdc556d8849 100644 --- a/include/linux/mmc/mmc.h +++ b/include/linux/mmc/mmc.h @@ -49,7 +49,7 @@ struct mmc_command { /* * These are the command types. */ -#define mmc_cmd_type(cmd) ((cmd)->flags & MMC_CMD_TYPE) +#define mmc_cmd_type(cmd) ((cmd)->flags & MMC_CMD_MASK) unsigned int retries; /* max number of retries */ unsigned int error; /* command error */ From 31867499b21b2374eb0cc6b3d1ea6b4ade4d1cc2 Mon Sep 17 00:00:00 2001 From: Russell King Date: Sun, 19 Feb 2006 19:53:56 +0000 Subject: [PATCH 075/102] [ARM] Add panic-on-oops support Although you could ask the kernel for panic-on-oops, it remained non-functional because the architecture specific code fragment had not been implemented. Add it, so it works as advertised. Signed-off-by: Russell King --- arch/arm/kernel/traps.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c index 10235b01582..03924bcc612 100644 --- a/arch/arm/kernel/traps.c +++ b/arch/arm/kernel/traps.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include @@ -231,6 +232,13 @@ NORET_TYPE void die(const char *str, struct pt_regs *regs, int err) __die(str, err, thread, regs); bust_spinlocks(0); spin_unlock_irq(&die_lock); + + if (panic_on_oops) { + printk(KERN_EMERG "Fatal exception: panic in 5 seconds\n"); + ssleep(5); + panic("Fatal exception"); + } + do_exit(SIGSEGV); } From a43c7ff8bafe841502cab52b7795e2825c2e42c4 Mon Sep 17 00:00:00 2001 From: Russell King Date: Mon, 20 Feb 2006 10:18:38 +0000 Subject: [PATCH 076/102] [ARM] Update mach-types Signed-off-by: Russell King --- arch/arm/tools/mach-types | 35 +++++++++++++++++++++++++++++++++-- 1 file changed, 33 insertions(+), 2 deletions(-) diff --git a/arch/arm/tools/mach-types b/arch/arm/tools/mach-types index d0f9bb5e902..8ab5300dcb9 100644 --- a/arch/arm/tools/mach-types +++ b/arch/arm/tools/mach-types @@ -12,7 +12,7 @@ # # http://www.arm.linux.org.uk/developer/machines/?action=new # -# Last update: Mon Jan 9 12:56:42 2006 +# Last update: Mon Feb 20 10:18:02 2006 # # machine_is_xxx CONFIG_xxxx MACH_TYPE_xxx number # @@ -904,7 +904,7 @@ wg302v2 MACH_WG302V2 WG302V2 890 eb42x MACH_EB42X EB42X 891 iq331es MACH_IQ331ES IQ331ES 892 cosydsp MACH_COSYDSP COSYDSP 893 -uplat7d MACH_UPLAT7D UPLAT7D 894 +uplat7d_proto MACH_UPLAT7D UPLAT7D 894 ptdavinci MACH_PTDAVINCI PTDAVINCI 895 mbus MACH_MBUS MBUS 896 nadia2vb MACH_NADIA2VB NADIA2VB 897 @@ -938,3 +938,34 @@ auckland MACH_AUCKLAND AUCKLAND 924 ak3220m MACH_AK3320M AK3320M 925 duramax MACH_DURAMAX DURAMAX 926 n35 MACH_N35 N35 927 +pronghorn MACH_PRONGHORN PRONGHORN 928 +fundy MACH_FUNDY FUNDY 929 +logicpd_pxa270 MACH_LOGICPD_PXA270 LOGICPD_PXA270 930 +cpu777 MACH_CPU777 CPU777 931 +simicon9201 MACH_SIMICON9201 SIMICON9201 932 +leap2_hpm MACH_LEAP2_HPM LEAP2_HPM 933 +cm922txa10 MACH_CM922TXA10 CM922TXA10 934 +sandgate MACH_PXA PXA 935 +sandgate2 MACH_SANDGATE2 SANDGATE2 936 +sandgate2g MACH_SANDGATE2G SANDGATE2G 937 +sandgate2p MACH_SANDGATE2P SANDGATE2P 938 +fred_jack MACH_FRED_JACK FRED_JACK 939 +ttg_color1 MACH_TTG_COLOR1 TTG_COLOR1 940 +nxeb500hmi MACH_NXEB500HMI NXEB500HMI 941 +netdcu8 MACH_NETDCU8 NETDCU8 942 +ml675050_cpu_boa MACH_ML675050_CPU_BOA ML675050_CPU_BOA 943 +ng_fvx538 MACH_NG_FVX538 NG_FVX538 944 +ng_fvs338 MACH_NG_FVS338 NG_FVS338 945 +pnx4103 MACH_PNX4103 PNX4103 946 +hesdb MACH_HESDB HESDB 947 +xsilo MACH_XSILO XSILO 948 +espresso MACH_ESPRESSO ESPRESSO 949 +emlc MACH_EMLC EMLC 950 +sisteron MACH_SISTERON SISTERON 951 +rx1950 MACH_RX1950 RX1950 952 +tsc_venus MACH_TSC_VENUS TSC_VENUS 953 +ds101j MACH_DS101J DS101J 954 +mxc300_30ads MACH_MXC30030ADS MXC30030ADS 955 +fujitsu_wimaxsoc MACH_FUJITSU_WIMAXSOC FUJITSU_WIMAXSOC 956 +dualpcmodem MACH_DUALPCMODEM DUALPCMODEM 957 +gesbc9312 MACH_GESBC9312 GESBC9312 958 From 102d60a2d8a6b54a20317a855dca3b598a2fd581 Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Wed, 22 Feb 2006 23:43:40 +1100 Subject: [PATCH 077/102] [PATCH] padlock: Fix typo that broke 256-bit keys A typo crept into the le32_to_cpu patch which broke 256-bit keys in the padlock driver. The following patch based on observations by Michael Heyse fixes the problem. Signed-off-by: Herbert Xu Signed-off-by: Linus Torvalds --- drivers/crypto/padlock-aes.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/crypto/padlock-aes.c b/drivers/crypto/padlock-aes.c index 64819aa7cac..0c08c58252b 100644 --- a/drivers/crypto/padlock-aes.c +++ b/drivers/crypto/padlock-aes.c @@ -348,10 +348,10 @@ aes_set_key(void *ctx_arg, const uint8_t *in_key, unsigned int key_len, uint32_t break; case 32: - E_KEY[4] = le32_to_cpu(in_key[4]); - E_KEY[5] = le32_to_cpu(in_key[5]); - E_KEY[6] = le32_to_cpu(in_key[6]); - t = E_KEY[7] = le32_to_cpu(in_key[7]); + E_KEY[4] = le32_to_cpu(key[4]); + E_KEY[5] = le32_to_cpu(key[5]); + E_KEY[6] = le32_to_cpu(key[6]); + t = E_KEY[7] = le32_to_cpu(key[7]); for (i = 0; i < 7; ++i) loop8 (i); break; From fa675765afed59bb89adba3369094ebd428b930b Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Wed, 22 Feb 2006 09:39:02 -0800 Subject: [PATCH 078/102] Revert mount/umount uevent removal This change reverts the 033b96fd30db52a710d97b06f87d16fc59fee0f1 commit from Kay Sievers that removed the mount/umount uevents from the kernel. Some older versions of HAL still depend on these events to detect when a new device has been mounted. These events are not correctly emitted, and are broken by design, and so, should not be relied upon by any future program. Instead, the /proc/mounts file should be polled to properly detect this kind of event. A feature-removal-schedule.txt entry has been added, noting when this interface will be removed from the kernel. Signed-off-by: Greg Kroah-Hartman --- Documentation/feature-removal-schedule.txt | 9 +++++++++ fs/super.c | 15 ++++++++++++++- include/linux/kobject.h | 6 ++++-- lib/kobject_uevent.c | 4 ++++ 4 files changed, 31 insertions(+), 3 deletions(-) diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt index b730d765b52..be5ae600f53 100644 --- a/Documentation/feature-removal-schedule.txt +++ b/Documentation/feature-removal-schedule.txt @@ -171,3 +171,12 @@ Why: The ISA interface is faster and should be always available. The I2C probing is also known to cause trouble in at least one case (see bug #5889.) Who: Jean Delvare + +--------------------------- + +What: mount/umount uevents +When: February 2007 +Why: These events are not correct, and do not properly let userspace know + when a file system has been mounted or unmounted. Userspace should + poll the /proc/mounts file instead to detect this properly. +Who: Greg Kroah-Hartman diff --git a/fs/super.c b/fs/super.c index 30294218fa6..e20b5580afd 100644 --- a/fs/super.c +++ b/fs/super.c @@ -666,6 +666,16 @@ static int test_bdev_super(struct super_block *s, void *data) return (void *)s->s_bdev == data; } +static void bdev_uevent(struct block_device *bdev, enum kobject_action action) +{ + if (bdev->bd_disk) { + if (bdev->bd_part) + kobject_uevent(&bdev->bd_part->kobj, action); + else + kobject_uevent(&bdev->bd_disk->kobj, action); + } +} + struct super_block *get_sb_bdev(struct file_system_type *fs_type, int flags, const char *dev_name, void *data, int (*fill_super)(struct super_block *, void *, int)) @@ -707,8 +717,10 @@ struct super_block *get_sb_bdev(struct file_system_type *fs_type, up_write(&s->s_umount); deactivate_super(s); s = ERR_PTR(error); - } else + } else { s->s_flags |= MS_ACTIVE; + bdev_uevent(bdev, KOBJ_MOUNT); + } } return s; @@ -724,6 +736,7 @@ void kill_block_super(struct super_block *sb) { struct block_device *bdev = sb->s_bdev; + bdev_uevent(bdev, KOBJ_UMOUNT); generic_shutdown_super(sb); sync_blockdev(bdev); close_bdev_excl(bdev); diff --git a/include/linux/kobject.h b/include/linux/kobject.h index 2a8d8da7096..c374b5fa8d3 100644 --- a/include/linux/kobject.h +++ b/include/linux/kobject.h @@ -41,8 +41,10 @@ enum kobject_action { KOBJ_ADD = (__force kobject_action_t) 0x01, /* exclusive to core */ KOBJ_REMOVE = (__force kobject_action_t) 0x02, /* exclusive to core */ KOBJ_CHANGE = (__force kobject_action_t) 0x03, /* device state change */ - KOBJ_OFFLINE = (__force kobject_action_t) 0x04, /* device offline */ - KOBJ_ONLINE = (__force kobject_action_t) 0x05, /* device online */ + KOBJ_MOUNT = (__force kobject_action_t) 0x04, /* mount event for block devices (broken) */ + KOBJ_UMOUNT = (__force kobject_action_t) 0x05, /* umount event for block devices (broken) */ + KOBJ_OFFLINE = (__force kobject_action_t) 0x06, /* device offline */ + KOBJ_ONLINE = (__force kobject_action_t) 0x07, /* device online */ }; struct kobject { diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c index 1b1985c136e..086a0c6e888 100644 --- a/lib/kobject_uevent.c +++ b/lib/kobject_uevent.c @@ -38,6 +38,10 @@ static char *action_to_string(enum kobject_action action) return "remove"; case KOBJ_CHANGE: return "change"; + case KOBJ_MOUNT: + return "mount"; + case KOBJ_UMOUNT: + return "umount"; case KOBJ_OFFLINE: return "offline"; case KOBJ_ONLINE: From c27a2164a32cfda80dc1647638a940103669af3d Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Wed, 22 Feb 2006 19:51:38 +0000 Subject: [PATCH 079/102] [ARM] 3340/1: Fix the PCI setup for direct master access to SDRAM Patch from Catalin Marinas The initial code did not configure the inbound memory windows for direct master access to the SDRAM. This patch creates a 1:1 mapping between the Versatile/PB PCI memory windows and its SDRAM. Note that an updated FPGA image is needed for Versatile/PB since the original windows were 1MB and not able to cover the whole SDRAM (now extended to 256MB). The patch also fixes the PCI IRQ mapping for slot #2. Signed-off-by: Catalin Marinas Signed-off-by: Russell King --- arch/arm/mach-versatile/pci.c | 93 ++++++++++++++++++----------------- 1 file changed, 48 insertions(+), 45 deletions(-) diff --git a/arch/arm/mach-versatile/pci.c b/arch/arm/mach-versatile/pci.c index b80d57d5169..722fbabc9cf 100644 --- a/arch/arm/mach-versatile/pci.c +++ b/arch/arm/mach-versatile/pci.c @@ -240,6 +240,14 @@ int __init pci_versatile_setup(int nr, struct pci_sys_data *sys) int i; int myslot = -1; unsigned long val; + void __iomem *local_pci_cfg_base; + + val = __raw_readl(SYS_PCICTL); + if (!(val & 1)) { + printk("Not plugged into PCI backplane!\n"); + ret = -EIO; + goto out; + } if (nr == 0) { sys->mem_offset = 0; @@ -253,48 +261,45 @@ int __init pci_versatile_setup(int nr, struct pci_sys_data *sys) goto out; } - __raw_writel(VERSATILE_PCI_MEM_BASE0 >> 28,PCI_IMAP0); - __raw_writel(VERSATILE_PCI_MEM_BASE1 >> 28,PCI_IMAP1); - __raw_writel(VERSATILE_PCI_MEM_BASE2 >> 28,PCI_IMAP2); - - __raw_writel(1, SYS_PCICTL); - - val = __raw_readl(SYS_PCICTL); - if (!(val & 1)) { - printk("Not plugged into PCI backplane!\n"); - ret = -EIO; - goto out; - } - /* * We need to discover the PCI core first to configure itself * before the main PCI probing is performed */ - for (i=0; i<32; i++) { + for (i=0; i<32; i++) if ((__raw_readl(VERSATILE_PCI_VIRT_BASE+(i<<11)+DEVICE_ID_OFFSET) == VP_PCI_DEVICE_ID) && (__raw_readl(VERSATILE_PCI_VIRT_BASE+(i<<11)+CLASS_ID_OFFSET) == VP_PCI_CLASS_ID)) { myslot = i; - - __raw_writel(myslot, PCI_SELFID); - val = __raw_readl(VERSATILE_PCI_CFG_VIRT_BASE+(myslot<<11)+CSR_OFFSET); - val |= (1<<2); - __raw_writel(val, VERSATILE_PCI_CFG_VIRT_BASE+(myslot<<11)+CSR_OFFSET); break; } - } if (myslot == -1) { printk("Cannot find PCI core!\n"); ret = -EIO; - } else { - printk("PCI core found (slot %d)\n",myslot); - /* Do not to map Versatile FPGA PCI device - into memory space as we are short of - mappable memory */ - pci_slot_ignore |= (1 << myslot); - ret = 1; + goto out; } + printk("PCI core found (slot %d)\n",myslot); + + __raw_writel(myslot, PCI_SELFID); + local_pci_cfg_base = (void *) VERSATILE_PCI_CFG_VIRT_BASE + (myslot << 11); + + val = __raw_readl(local_pci_cfg_base + CSR_OFFSET); + val |= PCI_COMMAND_MEMORY | PCI_COMMAND_MASTER | PCI_COMMAND_INVALIDATE; + __raw_writel(val, local_pci_cfg_base + CSR_OFFSET); + + /* + * Configure the PCI inbound memory windows to be 1:1 mapped to SDRAM + */ + __raw_writel(PHYS_OFFSET, local_pci_cfg_base + PCI_BASE_ADDRESS_0); + __raw_writel(PHYS_OFFSET, local_pci_cfg_base + PCI_BASE_ADDRESS_1); + __raw_writel(PHYS_OFFSET, local_pci_cfg_base + PCI_BASE_ADDRESS_2); + + /* + * Do not to map Versatile FPGA PCI device into memory space + */ + pci_slot_ignore |= (1 << myslot); + ret = 1; + out: return ret; } @@ -305,18 +310,18 @@ struct pci_bus *pci_versatile_scan_bus(int nr, struct pci_sys_data *sys) return pci_scan_bus(sys->busnr, &pci_versatile_ops, sys); } -/* - * V3_LB_BASE? - local bus address - * V3_LB_MAP? - pci bus address - */ void __init pci_versatile_preinit(void) { -} + __raw_writel(VERSATILE_PCI_MEM_BASE0 >> 28, PCI_IMAP0); + __raw_writel(VERSATILE_PCI_MEM_BASE1 >> 28, PCI_IMAP1); + __raw_writel(VERSATILE_PCI_MEM_BASE2 >> 28, PCI_IMAP2); -void __init pci_versatile_postinit(void) -{ -} + __raw_writel(PHYS_OFFSET >> 28, PCI_SMAP0); + __raw_writel(PHYS_OFFSET >> 28, PCI_SMAP1); + __raw_writel(PHYS_OFFSET >> 28, PCI_SMAP2); + __raw_writel(1, SYS_PCICTL); +} /* * map the specified device/slot/pin to an IRQ. Different backplanes may need to modify this. @@ -326,16 +331,15 @@ static int __init versatile_map_irq(struct pci_dev *dev, u8 slot, u8 pin) int irq; int devslot = PCI_SLOT(dev->devfn); - /* slot, pin, irq - 24 1 27 - 25 1 28 untested - 26 1 29 - 27 1 30 untested - */ + /* slot, pin, irq + * 24 1 27 + * 25 1 28 + * 26 1 29 + * 27 1 30 + */ + irq = 27 + ((slot + pin - 1) & 3); - irq = 27 + ((slot + pin + 2) % 3); /* Fudged */ - - printk("map irq: slot %d, pin %d, devslot %d, irq: %d\n",slot,pin,devslot,irq); + printk("PCI map irq: slot %d, pin %d, devslot %d, irq: %d\n",slot,pin,devslot,irq); return irq; } @@ -347,7 +351,6 @@ static struct hw_pci versatile_pci __initdata = { .setup = pci_versatile_setup, .scan = pci_versatile_scan_bus, .preinit = pci_versatile_preinit, - .postinit = pci_versatile_postinit, }; static int __init versatile_pci_init(void) From d7353b25c855b3a8103647503deaa98f512bd439 Mon Sep 17 00:00:00 2001 From: Alessandro Zummo Date: Wed, 22 Feb 2006 21:12:05 +0000 Subject: [PATCH 080/102] [ARM] 3342/1: NSLU2: Protect power button init routine with machine_is_nslu2() Patch from Alessandro Zummo The power button exit routine for the Linksys NSLU2 was not protected by a machine_is_nslu2(). This patch fixes it. Signed-off-by: Rod Whitby Signed-off-by: Alessandro Zummo Signed-off-by: Russell King --- arch/arm/mach-ixp4xx/nslu2-power.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/arm/mach-ixp4xx/nslu2-power.c b/arch/arm/mach-ixp4xx/nslu2-power.c index b0ad9e901f6..d80c362bc53 100644 --- a/arch/arm/mach-ixp4xx/nslu2-power.c +++ b/arch/arm/mach-ixp4xx/nslu2-power.c @@ -77,6 +77,9 @@ static int __init nslu2_power_init(void) static void __exit nslu2_power_exit(void) { + if (!(machine_is_nslu2())) + return; + free_irq(NSLU2_RB_IRQ, NULL); free_irq(NSLU2_PB_IRQ, NULL); } From af898b8f602441a3bebe918a3b26adc92b30762e Mon Sep 17 00:00:00 2001 From: Alessandro Zummo Date: Wed, 22 Feb 2006 21:12:06 +0000 Subject: [PATCH 081/102] [ARM] 3343/1: NAS100d: Fix incorrect I2C pin assignment Patch from Alessandro Zummo The I2C pin assignment for the Iomega NAS100d board was incorrect. This patch fixes it. The correct assignment has now been tested using the new RTC class and a new driver for the RTC on the NAS100d. Signed-off-by: Rod Whitby Signed-off-by: Alessandro Zummo Signed-off-by: Russell King --- include/asm-arm/arch-ixp4xx/nas100d.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/include/asm-arm/arch-ixp4xx/nas100d.h b/include/asm-arm/arch-ixp4xx/nas100d.h index 51ac0180427..84467a5190d 100644 --- a/include/asm-arm/arch-ixp4xx/nas100d.h +++ b/include/asm-arm/arch-ixp4xx/nas100d.h @@ -19,8 +19,8 @@ #error "Do not include this directly, instead #include " #endif -#define NAS100D_SDA_PIN 6 -#define NAS100D_SCL_PIN 5 +#define NAS100D_SDA_PIN 5 +#define NAS100D_SCL_PIN 6 /* * NAS100D PCI IRQs From bc66d4496f106a8e3a936dc24c9965a2f9c13e50 Mon Sep 17 00:00:00 2001 From: Alessandro Zummo Date: Wed, 22 Feb 2006 21:12:06 +0000 Subject: [PATCH 082/102] [ARM] 3344/1: NSLU2: beeper support Patch from Alessandro Zummo This patch adds support for the beeper embedded in the NSLU2 to the machine setup code. Signed-off-by: Alessandro Zummo Signed-off-by: Rod Whitby Signed-off-by: Russell King --- arch/arm/mach-ixp4xx/nslu2-setup.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/arch/arm/mach-ixp4xx/nslu2-setup.c b/arch/arm/mach-ixp4xx/nslu2-setup.c index f260a9d34f7..55411f21d83 100644 --- a/arch/arm/mach-ixp4xx/nslu2-setup.c +++ b/arch/arm/mach-ixp4xx/nslu2-setup.c @@ -50,6 +50,12 @@ static struct platform_device nslu2_i2c_controller = { .num_resources = 0, }; +static struct platform_device nslu2_beeper = { + .name = "ixp4xx-beeper", + .id = NSLU2_GPIO_BUZZ, + .num_resources = 0, +}; + static struct resource nslu2_uart_resources[] = { { .start = IXP4XX_UART1_BASE_PHYS, @@ -97,6 +103,7 @@ static struct platform_device *nslu2_devices[] __initdata = { &nslu2_i2c_controller, &nslu2_flash, &nslu2_uart, + &nslu2_beeper, }; static void nslu2_power_off(void) From 75d2f18088ded458f5bc4014b6c4e2d9651d41d4 Mon Sep 17 00:00:00 2001 From: Uli Luckas Date: Wed, 22 Feb 2006 21:12:07 +0000 Subject: [PATCH 083/102] [ARM] 3345/1: Fix interday RTC alarms Patch from Uli Luckas This is a bugfix. The comment in arch/arm/common/rtctime.c explains it: * FIXME: for now, we just copy the alarm time because we're lazy (and * is therefore buggy - setting a 10am alarm at 8pm will not result in * the alarm triggering.) This patch adds one day to the alarm iff the alarm wrapped beyond midnight and therefore appears to be in the past. Signed-off-by: Uli Luckas Signed-off-by: Russell King --- arch/arm/common/rtctime.c | 16 ++++++++++++---- 1 file changed, 12 insertions(+), 4 deletions(-) diff --git a/arch/arm/common/rtctime.c b/arch/arm/common/rtctime.c index 48b1e19b131..e851d86c212 100644 --- a/arch/arm/common/rtctime.c +++ b/arch/arm/common/rtctime.c @@ -128,19 +128,27 @@ EXPORT_SYMBOL(rtc_tm_to_time); /* * Calculate the next alarm time given the requested alarm time mask * and the current time. - * - * FIXME: for now, we just copy the alarm time because we're lazy (and - * is therefore buggy - setting a 10am alarm at 8pm will not result in - * the alarm triggering.) */ void rtc_next_alarm_time(struct rtc_time *next, struct rtc_time *now, struct rtc_time *alrm) { + unsigned long next_time; + unsigned long now_time; + next->tm_year = now->tm_year; next->tm_mon = now->tm_mon; next->tm_mday = now->tm_mday; next->tm_hour = alrm->tm_hour; next->tm_min = alrm->tm_min; next->tm_sec = alrm->tm_sec; + + rtc_tm_to_time(now, &now_time); + rtc_tm_to_time(next, &next_time); + + if (next_time < now_time) { + /* Advance one day */ + next_time += 60 * 60 * 24; + rtc_time_to_tm(next_time, next); + } } static inline int rtc_read_time(struct rtc_ops *ops, struct rtc_time *tm) From 43cc19816b3fc5286258e6f5e43ef4ead458f9a3 Mon Sep 17 00:00:00 2001 From: Russell King Date: Wed, 22 Feb 2006 21:13:28 +0000 Subject: [PATCH 084/102] [ARM] CONFIG_CPU_MPCORE -> CONFIG_CPU_32v6K CONFIG_CPU_MPCORE has never been a configuration symbol - it should be CONFIG_CPU_32v6K. Signed-off-by: Russell King --- arch/arm/kernel/entry-armv.S | 2 +- arch/arm/mm/abort-ev6.S | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm/kernel/entry-armv.S b/arch/arm/kernel/entry-armv.S index 964cd717506..ec48d70c6d8 100644 --- a/arch/arm/kernel/entry-armv.S +++ b/arch/arm/kernel/entry-armv.S @@ -566,7 +566,7 @@ ENTRY(__switch_to) ldr r6, [r2, #TI_CPU_DOMAIN]! #endif #if __LINUX_ARM_ARCH__ >= 6 -#ifdef CONFIG_CPU_MPCORE +#ifdef CONFIG_CPU_32v6K clrex #else strex r5, r4, [ip] @ Clear exclusive monitor diff --git a/arch/arm/mm/abort-ev6.S b/arch/arm/mm/abort-ev6.S index dbd34603312..8a7f65ba14b 100644 --- a/arch/arm/mm/abort-ev6.S +++ b/arch/arm/mm/abort-ev6.S @@ -20,7 +20,7 @@ */ .align 5 ENTRY(v6_early_abort) -#ifdef CONFIG_CPU_MPCORE +#ifdef CONFIG_CPU_32v6K clrex #else strex r0, r1, [sp] @ Clear the exclusive monitor From df666b9c510fd27fd3b1afd9ddfa1eaa62ce12b3 Mon Sep 17 00:00:00 2001 From: Andrew Victor Date: Wed, 22 Feb 2006 21:23:35 +0000 Subject: [PATCH 085/102] [ARM] 3325/2: GPIO function to control multi-drive (open collector) capability Patch from Andrew Victor This patch adds the at91_set_multi_drive() function to enable/disable the multi-drive (open collector) pin capability on the AT91RM9200 processor. This is necessary to fix the UDC (USB Gadget) driver for the AT91RM9200 board as it will not allow the board reset line to be pulled low if the pullup is not driven as an open collector output as the boards are wired to the USB connector on both the DK/EK. This version of the patch updates it to 2.6.16-rc4. Orignal patch by Jeff Warren. Signed-off-by: Andrew Victor Signed-off-by: Russell King --- arch/arm/mach-at91rm9200/devices.c | 4 +++- arch/arm/mach-at91rm9200/gpio.c | 17 +++++++++++++++++ include/asm-arm/arch-at91rm9200/gpio.h | 1 + 3 files changed, 21 insertions(+), 1 deletion(-) diff --git a/arch/arm/mach-at91rm9200/devices.c b/arch/arm/mach-at91rm9200/devices.c index 8df3e524565..57eedd5beaf 100644 --- a/arch/arm/mach-at91rm9200/devices.c +++ b/arch/arm/mach-at91rm9200/devices.c @@ -100,8 +100,10 @@ void __init at91_add_device_udc(struct at91_udc_data *data) at91_set_gpio_input(data->vbus_pin, 0); at91_set_deglitch(data->vbus_pin, 1); } - if (data->pullup_pin) + if (data->pullup_pin) { at91_set_gpio_output(data->pullup_pin, 0); + at91_set_multi_drive(data->pullup_pin, 1); + } udc_data = *data; platform_device_register(&at91rm9200_udc_device); diff --git a/arch/arm/mach-at91rm9200/gpio.c b/arch/arm/mach-at91rm9200/gpio.c index 2fd2ef583e4..a9f718bf8ba 100644 --- a/arch/arm/mach-at91rm9200/gpio.c +++ b/arch/arm/mach-at91rm9200/gpio.c @@ -159,6 +159,23 @@ int __init_or_module at91_set_deglitch(unsigned pin, int is_on) } EXPORT_SYMBOL(at91_set_deglitch); +/* + * enable/disable the multi-driver; This is only valid for output and + * allows the output pin to run as an open collector output. + */ +int __init_or_module at91_set_multi_drive(unsigned pin, int is_on) +{ + void __iomem *pio = pin_to_controller(pin); + unsigned mask = pin_to_mask(pin); + + if (!pio) + return -EINVAL; + + __raw_writel(mask, pio + (is_on ? PIO_MDER : PIO_MDDR)); + return 0; +} +EXPORT_SYMBOL(at91_set_multi_drive); + /*--------------------------------------------------------------------------*/ diff --git a/include/asm-arm/arch-at91rm9200/gpio.h b/include/asm-arm/arch-at91rm9200/gpio.h index 0f0a61e2f12..6176ab2dc41 100644 --- a/include/asm-arm/arch-at91rm9200/gpio.h +++ b/include/asm-arm/arch-at91rm9200/gpio.h @@ -183,6 +183,7 @@ extern int at91_set_B_periph(unsigned pin, int use_pullup); extern int at91_set_gpio_input(unsigned pin, int use_pullup); extern int at91_set_gpio_output(unsigned pin, int value); extern int at91_set_deglitch(unsigned pin, int is_on); +extern int at91_set_multi_drive(unsigned pin, int is_on); /* callable at any time */ extern int at91_set_gpio_value(unsigned pin, int value); From 06e4479bd092eca4125e5507e7c22619a491dab3 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?M=E5rten=20Wikstr=F6m?= Date: Wed, 22 Feb 2006 22:27:23 +0000 Subject: [PATCH 086/102] [ARM] 3347/1: Bugfix for ixp4xx_set_irq_type() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Patch from Mårten Wikström This patch fixes a bug in ixp4xx_set_irq_type() which leads to GPIO being incorrectly set to both edge triggered for raising as well as falling edge interrupt types. See the previous discussion on patch 3312/1. Signed-off-by: Mårten Wikström Signed-off-by: Russell King --- arch/arm/mach-ixp4xx/common.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) diff --git a/arch/arm/mach-ixp4xx/common.c b/arch/arm/mach-ixp4xx/common.c index 4bdc9d4526c..fbadf3021b9 100644 --- a/arch/arm/mach-ixp4xx/common.c +++ b/arch/arm/mach-ixp4xx/common.c @@ -111,24 +111,30 @@ static int ixp4xx_set_irq_type(unsigned int irq, unsigned int type) if (line < 0) return -EINVAL; - if (type & IRQT_BOTHEDGE) { + switch (type){ + case IRQT_BOTHEDGE: int_style = IXP4XX_GPIO_STYLE_TRANSITIONAL; irq_type = IXP4XX_IRQ_EDGE; - } else if (type & IRQT_RISING) { + break; + case IRQT_RISING: int_style = IXP4XX_GPIO_STYLE_RISING_EDGE; irq_type = IXP4XX_IRQ_EDGE; - } else if (type & IRQT_FALLING) { + break; + case IRQT_FALLING: int_style = IXP4XX_GPIO_STYLE_FALLING_EDGE; irq_type = IXP4XX_IRQ_EDGE; - } else if (type & IRQT_HIGH) { + break; + case IRQT_HIGH: int_style = IXP4XX_GPIO_STYLE_ACTIVE_HIGH; irq_type = IXP4XX_IRQ_LEVEL; - } else if (type & IRQT_LOW) { + break; + case IRQT_LOW: int_style = IXP4XX_GPIO_STYLE_ACTIVE_LOW; irq_type = IXP4XX_IRQ_LEVEL; - } else + break; + default: return -EINVAL; - + } ixp4xx_config_irq(irq, irq_type); if (line >= 8) { /* pins 8-15 */ From a6ceda7457b2303dcb07d3c472b25d52bbdb5a29 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 22 Feb 2006 14:35:52 -0800 Subject: [PATCH 087/102] [SCSI] esp: fix eh locking esp_reset didn't get fixed when the EH locking changed. ->eh_bus_reset_handler is now called without the host lock held. Signed-off-by: Christoph Hellwig Signed-off-by: David S. Miller --- drivers/scsi/esp.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/scsi/esp.c b/drivers/scsi/esp.c index f6900538be9..87a8c3d2072 100644 --- a/drivers/scsi/esp.c +++ b/drivers/scsi/esp.c @@ -2068,14 +2068,12 @@ static int esp_reset(struct scsi_cmnd *SCptr) { struct esp *esp = (struct esp *) SCptr->device->host->hostdata; + spin_lock_irq(esp->ehost->host_lock); (void) esp_do_resetbus(esp); - spin_unlock_irq(esp->ehost->host_lock); wait_event(esp->reset_queue, (esp->resetting_bus == 0)); - spin_lock_irq(esp->ehost->host_lock); - return SUCCESS; } From 6cec2aed8686840906f6298391dc4fd04d9ba843 Mon Sep 17 00:00:00 2001 From: Steve French Date: Wed, 22 Feb 2006 17:31:52 -0600 Subject: [PATCH 088/102] [PATCH] CIFS: CIFSSMBRead was returning an invalid pointer in buf on socket error Thanks to Adrian Bunk for debugging the problem and to Shaggy for helping find the solution. Also added a fix for 64K pages we found in loosely-related testing Signed-off-by: Dave Kleikamp Signed-off-by: Steve French Signed-off-by: Linus Torvalds --- fs/cifs/cifssmb.c | 7 ++++--- fs/cifs/connect.c | 8 ++++---- 2 files changed, 8 insertions(+), 7 deletions(-) diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 217323b0c89..b41e8b37965 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -1048,13 +1048,14 @@ CIFSSMBRead(const int xid, struct cifsTconInfo *tcon, cifs_small_buf_release(iov[0].iov_base); else if(resp_buf_type == CIFS_LARGE_BUFFER) cifs_buf_release(iov[0].iov_base); - } else /* return buffer to caller to free */ /* BB FIXME how do we tell caller if it is not a large buffer */ { - *buf = iov[0].iov_base; + } else if(resp_buf_type != CIFS_NO_BUFFER) { + /* return buffer to caller to free */ + *buf = iov[0].iov_base; if(resp_buf_type == CIFS_SMALL_BUFFER) *pbuf_type = CIFS_SMALL_BUFFER; else if(resp_buf_type == CIFS_LARGE_BUFFER) *pbuf_type = CIFS_LARGE_BUFFER; - } + } /* else no valid buffer on return - leave as null */ /* Note: On -EAGAIN error only caller can retry on handle based calls since file handle passed in no longer valid */ diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index e488603fb1e..ef5ae6f93c7 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1795,10 +1795,10 @@ cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, conjunction with 52K kvec constraint on arch with 4K page size */ - if(cifs_sb->rsize < PAGE_CACHE_SIZE) { - cifs_sb->rsize = PAGE_CACHE_SIZE; - /* Windows ME does this */ - cFYI(1,("Attempt to set readsize for mount to less than one page (4096)")); + if(cifs_sb->rsize < 2048) { + cifs_sb->rsize = 2048; + /* Windows ME may prefer this */ + cFYI(1,("readsize set to minimum 2048")); } cifs_sb->mnt_uid = volume_info.linux_uid; cifs_sb->mnt_gid = volume_info.linux_gid; From 977bdf06ca8dd7ed081fab8d30249d9e6b1c24d3 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 22 Feb 2006 11:44:58 -0800 Subject: [PATCH 089/102] [PATCH] sky2: yukon-ec-u chipset initialization Add more complete setup code for Yukon EC_U chipset. Based on matching code in 8.31 code in SysKonnect vendor driver. Signed-off-by: Stephen Hemminger --- drivers/net/sky2.c | 52 ++++++++++++++++++++++++++++++++-------- drivers/net/sky2.h | 59 ++++++++++++++++++++++++++++++++++++++-------- 2 files changed, 91 insertions(+), 20 deletions(-) diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c index bfeba5b9cd7..ce135b84a54 100644 --- a/drivers/net/sky2.c +++ b/drivers/net/sky2.c @@ -232,7 +232,17 @@ static int sky2_set_power_state(struct sky2_hw *hw, pci_power_t state) if (hw->ports > 1) reg1 |= PCI_Y2_PHY2_COMA; } + + if (hw->chip_id == CHIP_ID_YUKON_EC_U) { + pci_write_config_dword(hw->pdev, PCI_DEV_REG3, 0); + pci_read_config_dword(hw->pdev, PCI_DEV_REG4, ®1); + reg1 &= P_ASPM_CONTROL_MSK; + pci_write_config_dword(hw->pdev, PCI_DEV_REG4, reg1); + pci_write_config_dword(hw->pdev, PCI_DEV_REG5, 0); + } + pci_write_config_dword(hw->pdev, PCI_DEV_REG1, reg1); + break; case PCI_D3hot: @@ -463,16 +473,31 @@ static void sky2_phy_init(struct sky2_hw *hw, unsigned port) ledover |= PHY_M_LED_MO_RX(MO_LED_OFF); } - gm_phy_write(hw, port, PHY_MARV_LED_CTRL, ledctrl); + if (hw->chip_id == CHIP_ID_YUKON_EC_U && hw->chip_rev >= 2) { + /* apply fixes in PHY AFE */ + gm_phy_write(hw, port, 22, 255); + /* increase differential signal amplitude in 10BASE-T */ + gm_phy_write(hw, port, 24, 0xaa99); + gm_phy_write(hw, port, 23, 0x2011); + + /* fix for IEEE A/B Symmetry failure in 1000BASE-T */ + gm_phy_write(hw, port, 24, 0xa204); + gm_phy_write(hw, port, 23, 0x2002); + + /* set page register to 0 */ + gm_phy_write(hw, port, 22, 0); + } else { + gm_phy_write(hw, port, PHY_MARV_LED_CTRL, ledctrl); + + if (sky2->autoneg == AUTONEG_DISABLE || sky2->speed == SPEED_100) { + /* turn on 100 Mbps LED (LED_LINK100) */ + ledover |= PHY_M_LED_MO_100(MO_LED_ON); + } + + if (ledover) + gm_phy_write(hw, port, PHY_MARV_LED_OVER, ledover); - if (sky2->autoneg == AUTONEG_DISABLE || sky2->speed == SPEED_100) { - /* turn on 100 Mbps LED (LED_LINK100) */ - ledover |= PHY_M_LED_MO_100(MO_LED_ON); } - - if (ledover) - gm_phy_write(hw, port, PHY_MARV_LED_OVER, ledover); - /* Enable phy interrupt on auto-negotiation complete (or link up) */ if (sky2->autoneg == AUTONEG_ENABLE) gm_phy_write(hw, port, PHY_MARV_INT_MASK, PHY_M_IS_AN_COMPL); @@ -953,6 +978,12 @@ static int sky2_rx_start(struct sky2_port *sky2) sky2->rx_put = sky2->rx_next = 0; sky2_qset(hw, rxq); + + if (hw->chip_id == CHIP_ID_YUKON_EC_U && hw->chip_rev >= 2) { + /* MAC Rx RAM Read is controlled by hardware */ + sky2_write32(hw, Q_ADDR(rxq, Q_F), F_M_RX_RAM_DIS); + } + sky2_prefetch_init(hw, rxq, sky2->rx_le_map, RX_LE_SIZE - 1); rx_set_checksum(sky2); @@ -1035,9 +1066,10 @@ static int sky2_up(struct net_device *dev) RB_RST_SET); sky2_qset(hw, txqaddr[port]); - if (hw->chip_id == CHIP_ID_YUKON_EC_U) - sky2_write16(hw, Q_ADDR(txqaddr[port], Q_AL), 0x1a0); + /* Set almost empty threshold */ + if (hw->chip_id == CHIP_ID_YUKON_EC_U && hw->chip_rev == 1) + sky2_write16(hw, Q_ADDR(txqaddr[port], Q_AL), 0x1a0); sky2_prefetch_init(hw, txqaddr[port], sky2->tx_le_map, TX_RING_SIZE - 1); diff --git a/drivers/net/sky2.h b/drivers/net/sky2.h index fd12c289a23..d1c71f50d8b 100644 --- a/drivers/net/sky2.h +++ b/drivers/net/sky2.h @@ -5,14 +5,22 @@ #define _SKY2_H /* PCI config registers */ -#define PCI_DEV_REG1 0x40 -#define PCI_DEV_REG2 0x44 -#define PCI_DEV_STATUS 0x7c -#define PCI_OS_PCI_X (1<<26) +enum { + PCI_DEV_REG1 = 0x40, + PCI_DEV_REG2 = 0x44, + PCI_DEV_STATUS = 0x7c, + PCI_DEV_REG3 = 0x80, + PCI_DEV_REG4 = 0x84, + PCI_DEV_REG5 = 0x88, +}; -#define PEX_LNK_STAT 0xf2 -#define PEX_UNC_ERR_STAT 0x104 -#define PEX_DEV_CTRL 0xe8 +enum { + PEX_DEV_CAP = 0xe4, + PEX_DEV_CTRL = 0xe8, + PEX_DEV_STA = 0xea, + PEX_LNK_STAT = 0xf2, + PEX_UNC_ERR_STAT= 0x104, +}; /* Yukon-2 */ enum pci_dev_reg_1 { @@ -37,6 +45,25 @@ enum pci_dev_reg_2 { PCI_USEDATA64 = 1<<0, /* Use 64Bit Data bus ext */ }; +/* PCI_OUR_REG_4 32 bit Our Register 4 (Yukon-ECU only) */ +enum pci_dev_reg_4 { + /* (Link Training & Status State Machine) */ + P_TIMER_VALUE_MSK = 0xffL<<16, /* Bit 23..16: Timer Value Mask */ + /* (Active State Power Management) */ + P_FORCE_ASPM_REQUEST = 1<<15, /* Force ASPM Request (A1 only) */ + P_ASPM_GPHY_LINK_DOWN = 1<<14, /* GPHY Link Down (A1 only) */ + P_ASPM_INT_FIFO_EMPTY = 1<<13, /* Internal FIFO Empty (A1 only) */ + P_ASPM_CLKRUN_REQUEST = 1<<12, /* CLKRUN Request (A1 only) */ + + P_ASPM_FORCE_CLKREQ_ENA = 1<<4, /* Force CLKREQ Enable (A1b only) */ + P_ASPM_CLKREQ_PAD_CTL = 1<<3, /* CLKREQ PAD Control (A1 only) */ + P_ASPM_A1_MODE_SELECT = 1<<2, /* A1 Mode Select (A1 only) */ + P_CLK_GATE_PEX_UNIT_ENA = 1<<1, /* Enable Gate PEX Unit Clock */ + P_CLK_GATE_ROOT_COR_ENA = 1<<0, /* Enable Gate Root Core Clock */ + P_ASPM_CONTROL_MSK = P_FORCE_ASPM_REQUEST | P_ASPM_GPHY_LINK_DOWN + | P_ASPM_CLKRUN_REQUEST | P_ASPM_INT_FIFO_EMPTY, +}; + #define PCI_STATUS_ERROR_BITS (PCI_STATUS_DETECTED_PARITY | \ PCI_STATUS_SIG_SYSTEM_ERROR | \ @@ -507,6 +534,16 @@ enum { }; #define Q_ADDR(reg, offs) (B8_Q_REGS + (reg) + (offs)) +/* Q_F 32 bit Flag Register */ +enum { + F_ALM_FULL = 1<<27, /* Rx FIFO: almost full */ + F_EMPTY = 1<<27, /* Tx FIFO: empty flag */ + F_FIFO_EOF = 1<<26, /* Tag (EOF Flag) bit in FIFO */ + F_WM_REACHED = 1<<25, /* Watermark reached */ + F_M_RX_RAM_DIS = 1<<24, /* MAC Rx RAM Read Port disable */ + F_FIFO_LEVEL = 0x1fL<<16, /* Bit 23..16: # of Qwords in FIFO */ + F_WATER_MARK = 0x0007ffL, /* Bit 10.. 0: Watermark */ +}; /* Queue Prefetch Unit Offsets, use Y2_QADDR() to address (Yukon-2 only)*/ enum { @@ -909,10 +946,12 @@ enum { PHY_BCOM_ID1_C0 = 0x6044, PHY_BCOM_ID1_C5 = 0x6047, - PHY_MARV_ID1_B0 = 0x0C23, /* Yukon (PHY 88E1011) */ + PHY_MARV_ID1_B0 = 0x0C23, /* Yukon (PHY 88E1011) */ PHY_MARV_ID1_B2 = 0x0C25, /* Yukon-Plus (PHY 88E1011) */ - PHY_MARV_ID1_C2 = 0x0CC2, /* Yukon-EC (PHY 88E1111) */ - PHY_MARV_ID1_Y2 = 0x0C91, /* Yukon-2 (PHY 88E1112) */ + PHY_MARV_ID1_C2 = 0x0CC2, /* Yukon-EC (PHY 88E1111) */ + PHY_MARV_ID1_Y2 = 0x0C91, /* Yukon-2 (PHY 88E1112) */ + PHY_MARV_ID1_FE = 0x0C83, /* Yukon-FE (PHY 88E3082 Rev.A1) */ + PHY_MARV_ID1_ECU= 0x0CB0, /* Yukon-ECU (PHY 88E1149 Rev.B2?) */ }; /* Advertisement register bits */ From ff81fbbe321c3a468b6225c673ca57efa501fbed Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 22 Feb 2006 11:44:59 -0800 Subject: [PATCH 090/102] [PATCH] sky2: limit coalescing values to ring size Don't allow coalescing values to be bigger than the transmit ring. Since if you set them that big, the interrupt never happens and driver livelocks. Signed-off-by: Stephen Hemminger --- drivers/net/sky2.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c index ce135b84a54..cf7fd28d677 100644 --- a/drivers/net/sky2.c +++ b/drivers/net/sky2.c @@ -2884,11 +2884,11 @@ static int sky2_set_coalesce(struct net_device *dev, (ecmd->rx_coalesce_usecs_irq < tmin || ecmd->rx_coalesce_usecs_irq > tmax)) return -EINVAL; - if (ecmd->tx_max_coalesced_frames > 0xffff) + if (ecmd->tx_max_coalesced_frames >= TX_RING_SIZE-1) return -EINVAL; - if (ecmd->rx_max_coalesced_frames > 0xff) + if (ecmd->rx_max_coalesced_frames > RX_MAX_PENDING) return -EINVAL; - if (ecmd->rx_max_coalesced_frames_irq > 0xff) + if (ecmd->rx_max_coalesced_frames_irq >RX_MAX_PENDING) return -EINVAL; if (ecmd->tx_coalesce_usecs == 0) From a8fd6266dafd564bae6758cb78c8c152e7d4115e Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 22 Feb 2006 11:45:00 -0800 Subject: [PATCH 091/102] [PATCH] sky2: poke coalescing timer to fix hang Need to restart the interrupt coalescing timer after clearing the interrupt, to avoid races with interrupt timer and processing. Patch from Carl-Daniel Halfinger Signed-off-by: Stephen Hemminger --- drivers/net/sky2.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c index cf7fd28d677..629809433cb 100644 --- a/drivers/net/sky2.c +++ b/drivers/net/sky2.c @@ -1895,6 +1895,17 @@ static int sky2_poll(struct net_device *dev0, int *budget) sky2_write32(hw, STAT_CTRL, SC_STAT_CLR_IRQ); + /* + * Kick the STAT_LEV_TIMER_CTRL timer. + * This fixes my hangs on Yukon-EC (0xb6) rev 1. + * The if clause is there to start the timer only if it has been + * configured correctly and not been disabled via ethtool. + */ + if (sky2_read8(hw, STAT_LEV_TIMER_CTRL) == TIM_START) { + sky2_write8(hw, STAT_LEV_TIMER_CTRL, TIM_STOP); + sky2_write8(hw, STAT_LEV_TIMER_CTRL, TIM_START); + } + hwidx = sky2_read16(hw, STAT_PUT_IDX); BUG_ON(hwidx >= STATUS_RING_SIZE); rmb(); From 9a6d343188f5f1e9537e700fc4139c2d905ff129 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 22 Feb 2006 11:45:01 -0800 Subject: [PATCH 092/102] [PATCH] sky2: force early transmit status Need to force a transmit coalesce timer restart after processing transmit packets. Otherwise, can get transmit status after last update and chip doesn't send the next one. Can go with the chip defaults for coalescing timers, except for Tx timer which needs to be bigger. Signed-off-by: Stephen Hemminger --- drivers/net/sky2.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c index 629809433cb..d913d3407e9 100644 --- a/drivers/net/sky2.c +++ b/drivers/net/sky2.c @@ -1988,13 +1988,12 @@ exit_loop: sky2_tx_check(hw, 0, tx_done[0]); sky2_tx_check(hw, 1, tx_done[1]); - if (likely(work_done < to_do)) { - /* need to restart TX timer */ - if (is_ec_a1(hw)) { - sky2_write8(hw, STAT_TX_TIMER_CTRL, TIM_STOP); - sky2_write8(hw, STAT_TX_TIMER_CTRL, TIM_START); - } + if (sky2_read8(hw, STAT_TX_TIMER_CTRL) == TIM_START) { + sky2_write8(hw, STAT_TX_TIMER_CTRL, TIM_STOP); + sky2_write8(hw, STAT_TX_TIMER_CTRL, TIM_START); + } + if (likely(work_done < to_do)) { netif_rx_complete(dev0); hw->intr_mask |= Y2_IS_STAT_BMU; sky2_write32(hw, B0_IMSK, hw->intr_mask); @@ -2352,8 +2351,7 @@ static int sky2_reset(struct sky2_hw *hw) sky2_write8(hw, STAT_FIFO_ISR_WM, 16); sky2_write32(hw, STAT_TX_TIMER_INI, sky2_us2clk(hw, 1000)); - sky2_write32(hw, STAT_LEV_TIMER_INI, sky2_us2clk(hw, 100)); - sky2_write32(hw, STAT_ISR_TIMER_INI, sky2_us2clk(hw, 20)); + sky2_write32(hw, STAT_ISR_TIMER_INI, sky2_us2clk(hw, 7)); } /* enable status unit */ From 56a645cc1bc16ab33b33a3e0854a46c5d2c864f3 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 22 Feb 2006 11:45:02 -0800 Subject: [PATCH 093/102] [PATCH] sky2: use device iomem to access PCI config To avoid problems with PCI config access without ACPI (or busted ACPI tables), use the device's window into PCI config space. I know this probably will upset the purists, but I would rather have users than ACPI testers. It also generates less code. Signed-off-by: Stephen Hemminger --- drivers/net/sky2.c | 81 +++++++++++++++++++--------------------------- drivers/net/sky2.h | 21 ++++++++++++ 2 files changed, 54 insertions(+), 48 deletions(-) diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c index d913d3407e9..fbbc85532c2 100644 --- a/drivers/net/sky2.c +++ b/drivers/net/sky2.c @@ -195,11 +195,11 @@ static int sky2_set_power_state(struct sky2_hw *hw, pci_power_t state) pr_debug("sky2_set_power_state %d\n", state); sky2_write8(hw, B2_TST_CTRL1, TST_CFG_WRITE_ON); - pci_read_config_word(hw->pdev, hw->pm_cap + PCI_PM_PMC, &power_control); + power_control = sky2_pci_read16(hw, hw->pm_cap + PCI_PM_PMC); vaux = (sky2_read16(hw, B0_CTST) & Y2_VAUX_AVAIL) && (power_control & PCI_PM_CAP_PME_D3cold); - pci_read_config_word(hw->pdev, hw->pm_cap + PCI_PM_CTRL, &power_control); + power_control = sky2_pci_read16(hw, hw->pm_cap + PCI_PM_CTRL); power_control |= PCI_PM_CTRL_PME_STATUS; power_control &= ~(PCI_PM_CTRL_STATE_MASK); @@ -223,7 +223,7 @@ static int sky2_set_power_state(struct sky2_hw *hw, pci_power_t state) sky2_write8(hw, B2_Y2_CLK_GATE, 0); /* Turn off phy power saving */ - pci_read_config_dword(hw->pdev, PCI_DEV_REG1, ®1); + reg1 = sky2_pci_read32(hw, PCI_DEV_REG1); reg1 &= ~(PCI_Y2_PHY1_POWD | PCI_Y2_PHY2_POWD); /* looks like this XL is back asswards .. */ @@ -234,26 +234,26 @@ static int sky2_set_power_state(struct sky2_hw *hw, pci_power_t state) } if (hw->chip_id == CHIP_ID_YUKON_EC_U) { - pci_write_config_dword(hw->pdev, PCI_DEV_REG3, 0); - pci_read_config_dword(hw->pdev, PCI_DEV_REG4, ®1); + sky2_pci_write32(hw, PCI_DEV_REG3, 0); + reg1 = sky2_pci_read32(hw, PCI_DEV_REG4); reg1 &= P_ASPM_CONTROL_MSK; - pci_write_config_dword(hw->pdev, PCI_DEV_REG4, reg1); - pci_write_config_dword(hw->pdev, PCI_DEV_REG5, 0); + sky2_pci_write32(hw, PCI_DEV_REG4, reg1); + sky2_pci_write32(hw, PCI_DEV_REG5, 0); } - pci_write_config_dword(hw->pdev, PCI_DEV_REG1, reg1); + sky2_pci_write32(hw, PCI_DEV_REG1, reg1); break; case PCI_D3hot: case PCI_D3cold: /* Turn on phy power saving */ - pci_read_config_dword(hw->pdev, PCI_DEV_REG1, ®1); + reg1 = sky2_pci_read32(hw, PCI_DEV_REG1); if (hw->chip_id == CHIP_ID_YUKON_XL && hw->chip_rev > 1) reg1 &= ~(PCI_Y2_PHY1_POWD | PCI_Y2_PHY2_POWD); else reg1 |= (PCI_Y2_PHY1_POWD | PCI_Y2_PHY2_POWD); - pci_write_config_dword(hw->pdev, PCI_DEV_REG1, reg1); + sky2_pci_write32(hw, PCI_DEV_REG1, reg1); if (hw->chip_id == CHIP_ID_YUKON_XL && hw->chip_rev > 1) sky2_write8(hw, B2_Y2_CLK_GATE, 0); @@ -275,7 +275,7 @@ static int sky2_set_power_state(struct sky2_hw *hw, pci_power_t state) ret = -1; } - pci_write_config_byte(hw->pdev, hw->pm_cap + PCI_PM_CTRL, power_control); + sky2_pci_write16(hw, hw->pm_cap + PCI_PM_CTRL, power_control); sky2_write8(hw, B2_TST_CTRL1, TST_CFG_WRITE_OFF); return ret; } @@ -2059,13 +2059,13 @@ static void sky2_hw_intr(struct sky2_hw *hw) if (status & (Y2_IS_MST_ERR | Y2_IS_IRQ_STAT)) { u16 pci_err; - pci_read_config_word(hw->pdev, PCI_STATUS, &pci_err); + pci_err = sky2_pci_read16(hw, PCI_STATUS); if (net_ratelimit()) printk(KERN_ERR PFX "%s: pci hw error (0x%x)\n", pci_name(hw->pdev), pci_err); sky2_write8(hw, B2_TST_CTRL1, TST_CFG_WRITE_ON); - pci_write_config_word(hw->pdev, PCI_STATUS, + sky2_pci_write16(hw, PCI_STATUS, pci_err | PCI_STATUS_ERROR_BITS); sky2_write8(hw, B2_TST_CTRL1, TST_CFG_WRITE_OFF); } @@ -2074,7 +2074,7 @@ static void sky2_hw_intr(struct sky2_hw *hw) /* PCI-Express uncorrectable Error occurred */ u32 pex_err; - pci_read_config_dword(hw->pdev, PEX_UNC_ERR_STAT, &pex_err); + pex_err = sky2_pci_read32(hw, PEX_UNC_ERR_STAT); if (net_ratelimit()) printk(KERN_ERR PFX "%s: pci express error (0x%x)\n", @@ -2082,7 +2082,7 @@ static void sky2_hw_intr(struct sky2_hw *hw) /* clear the interrupt */ sky2_write32(hw, B2_TST_CTRL1, TST_CFG_WRITE_ON); - pci_write_config_dword(hw->pdev, PEX_UNC_ERR_STAT, + sky2_pci_write32(hw, PEX_UNC_ERR_STAT, 0xffffffffUL); sky2_write32(hw, B2_TST_CTRL1, TST_CFG_WRITE_OFF); @@ -2212,7 +2212,7 @@ static int sky2_reset(struct sky2_hw *hw) { u16 status; u8 t8, pmd_type; - int i, err; + int i; sky2_write8(hw, B0_CTST, CS_RST_CLR); @@ -2234,25 +2234,18 @@ static int sky2_reset(struct sky2_hw *hw) sky2_write8(hw, B0_CTST, CS_RST_CLR); /* clear PCI errors, if any */ - err = pci_read_config_word(hw->pdev, PCI_STATUS, &status); - if (err) - goto pci_err; + status = sky2_pci_read16(hw, PCI_STATUS); sky2_write8(hw, B2_TST_CTRL1, TST_CFG_WRITE_ON); - err = pci_write_config_word(hw->pdev, PCI_STATUS, - status | PCI_STATUS_ERROR_BITS); - if (err) - goto pci_err; + sky2_pci_write16(hw, PCI_STATUS, status | PCI_STATUS_ERROR_BITS); + sky2_write8(hw, B0_CTST, CS_MRST_CLR); /* clear any PEX errors */ - if (pci_find_capability(hw->pdev, PCI_CAP_ID_EXP)) { - err = pci_write_config_dword(hw->pdev, PEX_UNC_ERR_STAT, - 0xffffffffUL); - if (err) - goto pci_err; - } + if (pci_find_capability(hw->pdev, PCI_CAP_ID_EXP)) + sky2_pci_write32(hw, PEX_UNC_ERR_STAT, 0xffffffffUL); + pmd_type = sky2_read8(hw, B2_PMD_TYP); hw->copper = !(pmd_type == 'L' || pmd_type == 'S'); @@ -2362,14 +2355,6 @@ static int sky2_reset(struct sky2_hw *hw) sky2_write8(hw, STAT_ISR_TIMER_CTRL, TIM_START); return 0; - -pci_err: - /* This is to catch a BIOS bug workaround where - * mmconfig table doesn't have other buses. - */ - printk(KERN_ERR PFX "%s: can't access PCI config space\n", - pci_name(hw->pdev)); - return err; } static u32 sky2_supported_modes(const struct sky2_hw *hw) @@ -3239,17 +3224,6 @@ static int __devinit sky2_probe(struct pci_dev *pdev, } } -#ifdef __BIG_ENDIAN - /* byte swap descriptors in hardware */ - { - u32 reg; - - pci_read_config_dword(pdev, PCI_DEV_REG2, ®); - reg |= PCI_REV_DESC; - pci_write_config_dword(pdev, PCI_DEV_REG2, reg); - } -#endif - err = -ENOMEM; hw = kzalloc(sizeof(*hw), GFP_KERNEL); if (!hw) { @@ -3268,6 +3242,17 @@ static int __devinit sky2_probe(struct pci_dev *pdev, } hw->pm_cap = pm_cap; +#ifdef __BIG_ENDIAN + /* byte swap descriptors in hardware */ + { + u32 reg; + + reg = sky2_pci_read32(hw, PCI_DEV_REG2); + reg |= PCI_REV_DESC; + sky2_pci_write32(hw, PCI_DEV_REG2, reg); + } +#endif + /* ring for status responses */ hw->st_le = pci_alloc_consistent(hw->pdev, STATUS_LE_BYTES, &hw->st_dma); diff --git a/drivers/net/sky2.h b/drivers/net/sky2.h index d1c71f50d8b..9e40766150b 100644 --- a/drivers/net/sky2.h +++ b/drivers/net/sky2.h @@ -1951,4 +1951,25 @@ static inline void gma_set_addr(struct sky2_hw *hw, unsigned port, unsigned reg, gma_write16(hw, port, reg+4,(u16) addr[2] | ((u16) addr[3] << 8)); gma_write16(hw, port, reg+8,(u16) addr[4] | ((u16) addr[5] << 8)); } + +/* PCI config space access */ +static inline u32 sky2_pci_read32(const struct sky2_hw *hw, unsigned reg) +{ + return sky2_read32(hw, Y2_CFG_SPC + reg); +} + +static inline u16 sky2_pci_read16(const struct sky2_hw *hw, unsigned reg) +{ + return sky2_read16(hw, Y2_CFG_SPC + reg); +} + +static inline void sky2_pci_write32(struct sky2_hw *hw, unsigned reg, u32 val) +{ + sky2_write32(hw, Y2_CFG_SPC + reg, val); +} + +static inline void sky2_pci_write16(struct sky2_hw *hw, unsigned reg, u16 val) +{ + sky2_write16(hw, Y2_CFG_SPC + reg, val); +} #endif From 791917deb63c6d8beb3f347ea0911371deff1624 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 22 Feb 2006 11:45:03 -0800 Subject: [PATCH 094/102] [PATCH] sky2: close race on IRQ mask update. Need to avoid race in updating IRQ mask. This can probably be replaced smarter use of the interrupt control registers (if/when chipset docs are available). Signed-off-by: Stephen Hemminger --- drivers/net/sky2.c | 21 +++++++++++++++------ drivers/net/sky2.h | 3 ++- 2 files changed, 17 insertions(+), 7 deletions(-) diff --git a/drivers/net/sky2.c b/drivers/net/sky2.c index fbbc85532c2..ca8160d6822 100644 --- a/drivers/net/sky2.c +++ b/drivers/net/sky2.c @@ -1079,8 +1079,10 @@ static int sky2_up(struct net_device *dev) goto err_out; /* Enable interrupts from phy/mac for port */ + spin_lock_irq(&hw->hw_lock); hw->intr_mask |= (port == 0) ? Y2_IS_PORT_1 : Y2_IS_PORT_2; sky2_write32(hw, B0_IMSK, hw->intr_mask); + spin_unlock_irq(&hw->hw_lock); return 0; err_out: @@ -1380,10 +1382,10 @@ static int sky2_down(struct net_device *dev) netif_stop_queue(dev); /* Disable port IRQ */ - local_irq_disable(); + spin_lock_irq(&hw->hw_lock); hw->intr_mask &= ~((sky2->port == 0) ? Y2_IS_IRQ_PHY1 : Y2_IS_IRQ_PHY2); sky2_write32(hw, B0_IMSK, hw->intr_mask); - local_irq_enable(); + spin_unlock_irq(&hw->hw_lock); flush_scheduled_work(); @@ -1665,10 +1667,10 @@ static void sky2_phy_task(void *arg) out: up(&sky2->phy_sema); - local_irq_disable(); + spin_lock_irq(&hw->hw_lock); hw->intr_mask |= (sky2->port == 0) ? Y2_IS_IRQ_PHY1 : Y2_IS_IRQ_PHY2; sky2_write32(hw, B0_IMSK, hw->intr_mask); - local_irq_enable(); + spin_unlock_irq(&hw->hw_lock); } @@ -1994,9 +1996,13 @@ exit_loop: } if (likely(work_done < to_do)) { - netif_rx_complete(dev0); + spin_lock_irq(&hw->hw_lock); + __netif_rx_complete(dev0); + hw->intr_mask |= Y2_IS_STAT_BMU; sky2_write32(hw, B0_IMSK, hw->intr_mask); + spin_unlock_irq(&hw->hw_lock); + return 0; } else { *budget -= work_done; @@ -2128,6 +2134,7 @@ static void sky2_phy_intr(struct sky2_hw *hw, unsigned port) hw->intr_mask &= ~(port == 0 ? Y2_IS_IRQ_PHY1 : Y2_IS_IRQ_PHY2); sky2_write32(hw, B0_IMSK, hw->intr_mask); + schedule_work(&sky2->phy_task); } @@ -2141,6 +2148,7 @@ static irqreturn_t sky2_intr(int irq, void *dev_id, struct pt_regs *regs) if (status == 0 || status == ~0) return IRQ_NONE; + spin_lock(&hw->hw_lock); if (status & Y2_IS_HW_ERR) sky2_hw_intr(hw); @@ -2169,7 +2177,7 @@ static irqreturn_t sky2_intr(int irq, void *dev_id, struct pt_regs *regs) sky2_write32(hw, B0_Y2_SP_ICR, 2); - sky2_read32(hw, B0_IMSK); + spin_unlock(&hw->hw_lock); return IRQ_HANDLED; } @@ -3241,6 +3249,7 @@ static int __devinit sky2_probe(struct pci_dev *pdev, goto err_out_free_hw; } hw->pm_cap = pm_cap; + spin_lock_init(&hw->hw_lock); #ifdef __BIG_ENDIAN /* byte swap descriptors in hardware */ diff --git a/drivers/net/sky2.h b/drivers/net/sky2.h index 9e40766150b..3edb98075e0 100644 --- a/drivers/net/sky2.h +++ b/drivers/net/sky2.h @@ -1876,8 +1876,9 @@ struct sky2_port { struct sky2_hw { void __iomem *regs; struct pci_dev *pdev; - u32 intr_mask; struct net_device *dev[2]; + spinlock_t hw_lock; + u32 intr_mask; int pm_cap; int msi; From bf637ec3ef4159da3dd156ecf6f6987d8c8c5dae Mon Sep 17 00:00:00 2001 From: Stefan Richter Date: Tue, 31 Jan 2006 00:13:06 -0500 Subject: [PATCH 095/102] sbp2: fix another deadlock after disconnection If there were commands enqueued but not completed before an SBP-2 unit was unplugged (or an attempt to reconnect failed), knodemgrd or any process which tried to remove the device would sleep uninterruptibly in blk_execute_rq(). Therefore make sure that all commands are completed when sbp2 retreats. Signed-off-by: Stefan Richter Signed-off-by: Jody McIntyre (cherry picked from 61daa34c132c5d4ed8630e2c46e9bf2f0c7b3428 commit) --- drivers/ieee1394/sbp2.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/ieee1394/sbp2.c b/drivers/ieee1394/sbp2.c index c2c776fbda0..8963dd484eb 100644 --- a/drivers/ieee1394/sbp2.c +++ b/drivers/ieee1394/sbp2.c @@ -643,9 +643,15 @@ static int sbp2_remove(struct device *dev) if (!scsi_id) return 0; - /* Trigger shutdown functions in scsi's highlevel. */ - if (scsi_id->scsi_host) + if (scsi_id->scsi_host) { + /* Get rid of enqueued commands if there is no chance to + * send them. */ + if (!sbp2util_node_is_available(scsi_id)) + sbp2scsi_complete_all_commands(scsi_id, DID_NO_CONNECT); + /* scsi_remove_device() will trigger shutdown functions of SCSI + * highlevel drivers which would deadlock if blocked. */ scsi_unblock_requests(scsi_id->scsi_host); + } sdev = scsi_id->sdev; if (sdev) { scsi_id->sdev = NULL; From 35bdddb83f62978b5fad82a14fbfd78cc3a5a60c Mon Sep 17 00:00:00 2001 From: Stefan Richter Date: Tue, 31 Jan 2006 00:13:33 -0500 Subject: [PATCH 096/102] sbp2: variable status FIFO address (fix login timeout) Let the ieee1394 core select a suitable 1394 address range for sbp2's status FIFO instead of using a fixed range. Since the core only selects addresses which are guaranteed to be out of the "physical range" as per OHCI 1.1, this patch also fixes an old bug: OHCI controllers which implement a writeable PhysicalUpperBound register included sbp2's status FIFO in the physical range. That way sbp2 was never notified of a succesful login and always failed after timeout. Affected OHCI host adapters include ALi and Fujitsu controllers. As another side effect of this patch, the status FIFO is no longer located in a range for which OHCI chips perform "posted writes". Each status write now requires a response subaction. But since large data transfers involve only few status writes, there is no measurable decrease of I/O throughput. What's more, the status FIFO is now safe from potential host bus errors. Nevertheless, posted writes could be re-enabled by extensions to the ARM features of the 1394 stack. Signed-off-by: Stefan Richter Signed-off-by: Jody McIntyre (cherry picked from b2d38cccad4ef80d6b672b8f89aae5fe2907b113 commit) --- drivers/ieee1394/sbp2.c | 64 +++++++++++++++++++++++------------------ drivers/ieee1394/sbp2.h | 64 +++++++++++++++-------------------------- 2 files changed, 59 insertions(+), 69 deletions(-) diff --git a/drivers/ieee1394/sbp2.c b/drivers/ieee1394/sbp2.c index 8963dd484eb..0672224fa10 100644 --- a/drivers/ieee1394/sbp2.c +++ b/drivers/ieee1394/sbp2.c @@ -748,11 +748,6 @@ static struct scsi_id_instance_data *sbp2_alloc_device(struct unit_directory *ud hi->host = ud->ne->host; INIT_LIST_HEAD(&hi->scsi_ids); - /* Register our sbp2 status address space... */ - hpsb_register_addrspace(&sbp2_highlevel, ud->ne->host, &sbp2_ops, - SBP2_STATUS_FIFO_ADDRESS, - SBP2_STATUS_FIFO_ADDRESS + - SBP2_STATUS_FIFO_ENTRY_TO_OFFSET(SBP2_MAX_UDS_PER_NODE+1)); #ifdef CONFIG_IEEE1394_SBP2_PHYS_DMA /* Handle data movement if physical dma is not * enabled/supportedon host controller */ @@ -765,6 +760,18 @@ static struct scsi_id_instance_data *sbp2_alloc_device(struct unit_directory *ud list_add_tail(&scsi_id->scsi_list, &hi->scsi_ids); + /* Register the status FIFO address range. We could use the same FIFO + * for targets at different nodes. However we need different FIFOs per + * target in order to support multi-unit devices. */ + scsi_id->status_fifo_addr = hpsb_allocate_and_register_addrspace( + &sbp2_highlevel, ud->ne->host, &sbp2_ops, + sizeof(struct sbp2_status_block), sizeof(quadlet_t), + ~0ULL, ~0ULL); + if (!scsi_id->status_fifo_addr) { + SBP2_ERR("failed to allocate status FIFO address range"); + goto failed_alloc; + } + /* Register our host with the SCSI stack. */ scsi_host = scsi_host_alloc(&scsi_driver_template, sizeof(unsigned long)); @@ -1003,6 +1010,10 @@ static void sbp2_remove_device(struct scsi_id_instance_data *scsi_id) SBP2_DMA_FREE("single query logins data"); } + if (scsi_id->status_fifo_addr) + hpsb_unregister_addrspace(&sbp2_highlevel, hi->host, + scsi_id->status_fifo_addr); + scsi_id->ud->device.driver_data = NULL; SBP2_DEBUG("SBP-2 device removed, SCSI ID = %d", scsi_id->ud->id); @@ -1081,11 +1092,10 @@ static int sbp2_query_logins(struct scsi_id_instance_data *scsi_id) ORB_SET_QUERY_LOGINS_RESP_LENGTH(sizeof(struct sbp2_query_logins_response)); SBP2_DEBUG("sbp2_query_logins: reserved_resp_length initialized"); - scsi_id->query_logins_orb->status_FIFO_lo = SBP2_STATUS_FIFO_ADDRESS_LO + - SBP2_STATUS_FIFO_ENTRY_TO_OFFSET(scsi_id->ud->id); - scsi_id->query_logins_orb->status_FIFO_hi = (ORB_SET_NODE_ID(hi->host->node_id) | - SBP2_STATUS_FIFO_ADDRESS_HI); - SBP2_DEBUG("sbp2_query_logins: status FIFO initialized"); + scsi_id->query_logins_orb->status_fifo_hi = + ORB_SET_STATUS_FIFO_HI(scsi_id->status_fifo_addr, hi->host->node_id); + scsi_id->query_logins_orb->status_fifo_lo = + ORB_SET_STATUS_FIFO_LO(scsi_id->status_fifo_addr); sbp2util_cpu_to_be32_buffer(scsi_id->query_logins_orb, sizeof(struct sbp2_query_logins_orb)); @@ -1190,11 +1200,10 @@ static int sbp2_login_device(struct scsi_id_instance_data *scsi_id) ORB_SET_LOGIN_RESP_LENGTH(sizeof(struct sbp2_login_response)); SBP2_DEBUG("sbp2_login_device: passwd_resp_lengths initialized"); - scsi_id->login_orb->status_FIFO_lo = SBP2_STATUS_FIFO_ADDRESS_LO + - SBP2_STATUS_FIFO_ENTRY_TO_OFFSET(scsi_id->ud->id); - scsi_id->login_orb->status_FIFO_hi = (ORB_SET_NODE_ID(hi->host->node_id) | - SBP2_STATUS_FIFO_ADDRESS_HI); - SBP2_DEBUG("sbp2_login_device: status FIFO initialized"); + scsi_id->login_orb->status_fifo_hi = + ORB_SET_STATUS_FIFO_HI(scsi_id->status_fifo_addr, hi->host->node_id); + scsi_id->login_orb->status_fifo_lo = + ORB_SET_STATUS_FIFO_LO(scsi_id->status_fifo_addr); /* * Byte swap ORB if necessary @@ -1307,10 +1316,10 @@ static int sbp2_logout_device(struct scsi_id_instance_data *scsi_id) scsi_id->logout_orb->login_ID_misc |= ORB_SET_NOTIFY(1); scsi_id->logout_orb->reserved5 = 0x0; - scsi_id->logout_orb->status_FIFO_lo = SBP2_STATUS_FIFO_ADDRESS_LO + - SBP2_STATUS_FIFO_ENTRY_TO_OFFSET(scsi_id->ud->id); - scsi_id->logout_orb->status_FIFO_hi = (ORB_SET_NODE_ID(hi->host->node_id) | - SBP2_STATUS_FIFO_ADDRESS_HI); + scsi_id->logout_orb->status_fifo_hi = + ORB_SET_STATUS_FIFO_HI(scsi_id->status_fifo_addr, hi->host->node_id); + scsi_id->logout_orb->status_fifo_lo = + ORB_SET_STATUS_FIFO_LO(scsi_id->status_fifo_addr); /* * Byte swap ORB if necessary @@ -1372,10 +1381,10 @@ static int sbp2_reconnect_device(struct scsi_id_instance_data *scsi_id) scsi_id->reconnect_orb->login_ID_misc |= ORB_SET_NOTIFY(1); scsi_id->reconnect_orb->reserved5 = 0x0; - scsi_id->reconnect_orb->status_FIFO_lo = SBP2_STATUS_FIFO_ADDRESS_LO + - SBP2_STATUS_FIFO_ENTRY_TO_OFFSET(scsi_id->ud->id); - scsi_id->reconnect_orb->status_FIFO_hi = - (ORB_SET_NODE_ID(hi->host->node_id) | SBP2_STATUS_FIFO_ADDRESS_HI); + scsi_id->reconnect_orb->status_fifo_hi = + ORB_SET_STATUS_FIFO_HI(scsi_id->status_fifo_addr, hi->host->node_id); + scsi_id->reconnect_orb->status_fifo_lo = + ORB_SET_STATUS_FIFO_LO(scsi_id->status_fifo_addr); /* * Byte swap ORB if necessary @@ -2112,7 +2121,6 @@ static int sbp2_handle_status_write(struct hpsb_host *host, int nodeid, int dest { struct sbp2scsi_host_info *hi; struct scsi_id_instance_data *scsi_id = NULL, *scsi_id_tmp; - u32 id; struct scsi_cmnd *SCpnt = NULL; u32 scsi_status = SBP2_SCSI_STATUS_GOOD; struct sbp2_command_info *command; @@ -2135,12 +2143,12 @@ static int sbp2_handle_status_write(struct hpsb_host *host, int nodeid, int dest } /* - * Find our scsi_id structure by looking at the status fifo address written to by - * the sbp2 device. + * Find our scsi_id structure by looking at the status fifo address + * written to by the sbp2 device. */ - id = SBP2_STATUS_FIFO_OFFSET_TO_ENTRY((u32)(addr - SBP2_STATUS_FIFO_ADDRESS)); list_for_each_entry(scsi_id_tmp, &hi->scsi_ids, scsi_list) { - if (scsi_id_tmp->ne->nodeid == nodeid && scsi_id_tmp->ud->id == id) { + if (scsi_id_tmp->ne->nodeid == nodeid && + scsi_id_tmp->status_fifo_addr == addr) { scsi_id = scsi_id_tmp; break; } diff --git a/drivers/ieee1394/sbp2.h b/drivers/ieee1394/sbp2.h index 900ea1d25e7..e2d357a9ea3 100644 --- a/drivers/ieee1394/sbp2.h +++ b/drivers/ieee1394/sbp2.h @@ -33,15 +33,17 @@ #define ORB_DIRECTION_NO_DATA_TRANSFER 0x2 #define ORB_SET_NULL_PTR(value) ((value & 0x1) << 31) -#define ORB_SET_NOTIFY(value) ((value & 0x1) << 31) -#define ORB_SET_RQ_FMT(value) ((value & 0x3) << 29) /* unused ? */ +#define ORB_SET_NOTIFY(value) ((value & 0x1) << 31) +#define ORB_SET_RQ_FMT(value) ((value & 0x3) << 29) /* unused ? */ #define ORB_SET_NODE_ID(value) ((value & 0xffff) << 16) -#define ORB_SET_DATA_SIZE(value) (value & 0xffff) -#define ORB_SET_PAGE_SIZE(value) ((value & 0x7) << 16) -#define ORB_SET_PAGE_TABLE_PRESENT(value) ((value & 0x1) << 19) -#define ORB_SET_MAX_PAYLOAD(value) ((value & 0xf) << 20) -#define ORB_SET_SPEED(value) ((value & 0x7) << 24) -#define ORB_SET_DIRECTION(value) ((value & 0x1) << 27) +#define ORB_SET_STATUS_FIFO_HI(value, id) (value >> 32 | ORB_SET_NODE_ID(id)) +#define ORB_SET_STATUS_FIFO_LO(value) (value & 0xffffffff) +#define ORB_SET_DATA_SIZE(value) (value & 0xffff) +#define ORB_SET_PAGE_SIZE(value) ((value & 0x7) << 16) +#define ORB_SET_PAGE_TABLE_PRESENT(value) ((value & 0x1) << 19) +#define ORB_SET_MAX_PAYLOAD(value) ((value & 0xf) << 20) +#define ORB_SET_SPEED(value) ((value & 0x7) << 24) +#define ORB_SET_DIRECTION(value) ((value & 0x1) << 27) struct sbp2_command_orb { volatile u32 next_ORB_hi; @@ -76,8 +78,8 @@ struct sbp2_login_orb { u32 login_response_lo; u32 lun_misc; u32 passwd_resp_lengths; - u32 status_FIFO_hi; - u32 status_FIFO_lo; + u32 status_fifo_hi; + u32 status_fifo_lo; }; #define RESPONSE_GET_LOGIN_ID(value) (value & 0xffff) @@ -102,8 +104,8 @@ struct sbp2_query_logins_orb { u32 query_response_lo; u32 lun_misc; u32 reserved_resp_length; - u32 status_FIFO_hi; - u32 status_FIFO_lo; + u32 status_fifo_hi; + u32 status_fifo_lo; }; #define RESPONSE_GET_MAX_LOGINS(value) (value & 0xffff) @@ -123,8 +125,8 @@ struct sbp2_reconnect_orb { u32 reserved4; u32 login_ID_misc; u32 reserved5; - u32 status_FIFO_hi; - u32 status_FIFO_lo; + u32 status_fifo_hi; + u32 status_fifo_lo; }; struct sbp2_logout_orb { @@ -134,8 +136,8 @@ struct sbp2_logout_orb { u32 reserved4; u32 login_ID_misc; u32 reserved5; - u32 status_FIFO_hi; - u32 status_FIFO_lo; + u32 status_fifo_hi; + u32 status_fifo_lo; }; #define PAGE_TABLE_SET_SEGMENT_BASE_HI(value) (value & 0xffff) @@ -195,30 +197,6 @@ struct sbp2_status_block { * Miscellaneous SBP2 related config rom defines */ -/* The status fifo address definition below is used as a base for each - * node, which a chunk seperately assigned to each unit directory in the - * node. For example, 0xfffe00000000ULL is used for the first sbp2 device - * detected on node 0, 0xfffe00000020ULL for the next sbp2 device on node - * 0, and so on. - * - * Note: We could use a single status fifo address for all sbp2 devices, - * and figure out which sbp2 device the status belongs to by looking at - * the source node id of the status write... but, using separate addresses - * for each sbp2 unit directory allows for better code and the ability to - * support multiple luns within a single 1394 node. - * - * Also note that we choose the address range below as it is a region - * specified for write posting, where the ohci controller will - * automatically send an ack_complete when the status is written by the - * sbp2 device... saving a split transaction. =) - */ -#define SBP2_STATUS_FIFO_ADDRESS 0xfffe00000000ULL -#define SBP2_STATUS_FIFO_ADDRESS_HI 0xfffe -#define SBP2_STATUS_FIFO_ADDRESS_LO 0x0 - -#define SBP2_STATUS_FIFO_ENTRY_TO_OFFSET(entry) ((entry) << 5) -#define SBP2_STATUS_FIFO_OFFSET_TO_ENTRY(offset) ((offset) >> 5) - #define SBP2_UNIT_DIRECTORY_OFFSET_KEY 0xd1 #define SBP2_CSR_OFFSET_KEY 0x54 #define SBP2_UNIT_SPEC_ID_KEY 0x12 @@ -258,7 +236,6 @@ struct sbp2_status_block { */ #define SBP2_MAX_SG_ELEMENT_LENGTH 0xf000 -#define SBP2_MAX_UDS_PER_NODE 16 /* Maximum scsi devices per node */ #define SBP2_MAX_SECTORS 255 /* Max sectors supported */ #define SBP2_MAX_CMDS 8 /* This should be safe */ @@ -337,6 +314,11 @@ struct scsi_id_instance_data { u32 sbp2_lun; u32 sbp2_firmware_revision; + /* + * Address for the device to write status blocks to + */ + u64 status_fifo_addr; + /* * Variable used for logins, reconnects, logouts, query logins */ From a80614d1adba903a1e5cb22bf14ebc640fc2ba4c Mon Sep 17 00:00:00 2001 From: Stefan Richter Date: Tue, 14 Feb 2006 22:04:19 -0500 Subject: [PATCH 097/102] sbp2: update 36byte inquiry workaround (fix compatibility regression) Since about Linux 2.6.14, sbp2's inquiry workaround did not work anymore due to changes in the SCSI layer. Update it to become effective again. Testing one of the two known affected bridges has shown that skip_ms_page_8 is required as well. Also, make force_inquiry_hack tunable via /sys/module/sbp2/parameters. Signed-off-by: Stefan Richter Signed-off-by: Jody McIntyre (cherry picked from 99496037c6744fd938ffb8ccfc8fc91762322ff8 commit) --- drivers/ieee1394/sbp2.c | 42 +++++++++++++++++++---------------------- 1 file changed, 19 insertions(+), 23 deletions(-) diff --git a/drivers/ieee1394/sbp2.c b/drivers/ieee1394/sbp2.c index 0672224fa10..eca92eb475a 100644 --- a/drivers/ieee1394/sbp2.c +++ b/drivers/ieee1394/sbp2.c @@ -137,15 +137,15 @@ MODULE_PARM_DESC(exclusive_login, "Exclusive login to sbp2 device (default = 1)" /* * SCSI inquiry hack for really badly behaved sbp2 devices. Turn this on * if your sbp2 device is not properly handling the SCSI inquiry command. - * This hack makes the inquiry look more like a typical MS Windows - * inquiry. + * This hack makes the inquiry look more like a typical MS Windows inquiry + * by enforcing 36 byte inquiry and avoiding access to mode_sense page 8. * * If force_inquiry_hack=1 is required for your device to work, * please submit the logged sbp2_firmware_revision value of this device to * the linux1394-devel mailing list. */ static int force_inquiry_hack; -module_param(force_inquiry_hack, int, 0444); +module_param(force_inquiry_hack, int, 0644); MODULE_PARM_DESC(force_inquiry_hack, "Force SCSI inquiry hack (default = 0)"); /* @@ -264,18 +264,17 @@ static struct hpsb_protocol_driver sbp2_driver = { }, }; - -/* List of device firmware's that require a forced 36 byte inquiry. */ +/* + * List of device firmwares that require the inquiry hack. + * Yields a few false positives but did not break other devices so far. + */ static u32 sbp2_broken_inquiry_list[] = { - 0x00002800, /* Stefan Richter */ + 0x00002800, /* Stefan Richter */ /* DViCO Momobay CX-1 */ 0x00000200 /* Andreas Plesch */ /* QPS Fire DVDBurner */ }; -#define NUM_BROKEN_INQUIRY_DEVS \ - (sizeof(sbp2_broken_inquiry_list)/sizeof(*sbp2_broken_inquiry_list)) - /************************************** * General utility functions **************************************/ @@ -1575,7 +1574,7 @@ static void sbp2_parse_unit_directory(struct scsi_id_instance_data *scsi_id, /* Check for a blacklisted set of devices that require us to force * a 36 byte host inquiry. This can be overriden as a module param * (to force all hosts). */ - for (i = 0; i < NUM_BROKEN_INQUIRY_DEVS; i++) { + for (i = 0; i < ARRAY_SIZE(sbp2_broken_inquiry_list); i++) { if ((firmware_revision & 0xffff00) == sbp2_broken_inquiry_list[i]) { SBP2_WARN("Node " NODE_BUS_FMT ": Using 36byte inquiry workaround", @@ -2021,18 +2020,6 @@ static int sbp2_send_command(struct scsi_id_instance_data *scsi_id, return -EIO; } - /* - * The scsi stack sends down a request_bufflen which does not match the - * length field in the scsi cdb. This causes some sbp2 devices to - * reject this inquiry command. Fix the request_bufflen. - */ - if (*cmd == INQUIRY) { - if (force_inquiry_hack || scsi_id->workarounds & SBP2_BREAKAGE_INQUIRY_HACK) - request_bufflen = cmd[4] = 0x24; - else - request_bufflen = cmd[4]; - } - /* * Now actually fill in the comamnd orb and sbp2 s/g list */ @@ -2489,7 +2476,16 @@ static void sbp2scsi_complete_command(struct scsi_id_instance_data *scsi_id, static int sbp2scsi_slave_alloc(struct scsi_device *sdev) { - ((struct scsi_id_instance_data *)sdev->host->hostdata[0])->sdev = sdev; + struct scsi_id_instance_data *scsi_id = + (struct scsi_id_instance_data *)sdev->host->hostdata[0]; + + scsi_id->sdev = sdev; + + if (force_inquiry_hack || + scsi_id->workarounds & SBP2_BREAKAGE_INQUIRY_HACK) { + sdev->inquiry_len = 36; + sdev->skip_ms_page_8 = 1; + } return 0; } From 5d06a99f543e734ceb53bbc9e550537be97f0c49 Mon Sep 17 00:00:00 2001 From: Francois Romieu Date: Thu, 23 Feb 2006 00:47:58 +0100 Subject: [PATCH 098/102] r8169: fix broken ring index handling in suspend/resume rtl8169_hw_start() requires that the descriptor ring indexes be set to zero. Let a deferred invocation of rtl8169_reset_task() handle it. Enabling a few power management bits will not hurt either. suspend/resume is issued with irq on: the spinlock do not need to save the irq flag. Signed-off-by: Francois Romieu --- drivers/net/r8169.c | 102 +++++++++++++++++++++++++------------------- 1 file changed, 59 insertions(+), 43 deletions(-) diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c index 6e1018448ee..999fd6cef77 100644 --- a/drivers/net/r8169.c +++ b/drivers/net/r8169.c @@ -287,6 +287,12 @@ enum RTL8169_register_content { TxInterFrameGapShift = 24, TxDMAShift = 8, /* DMA burst value (0-7) is shift this many bits */ + /* Config1 register p.24 */ + PMEnable = (1 << 0), /* Power Management Enable */ + + /* Config5 register p.27 */ + PMEStatus = (1 << 0), /* PME status can be reset by PCI RST# */ + /* TBICSR p.28 */ TBIReset = 0x80000000, TBILoopback = 0x40000000, @@ -1442,6 +1448,11 @@ rtl8169_init_board(struct pci_dev *pdev, struct net_device **dev_out, } tp->chipset = i; + RTL_W8(Cfg9346, Cfg9346_Unlock); + RTL_W8(Config1, RTL_R8(Config1) | PMEnable); + RTL_W8(Config5, RTL_R8(Config5) & PMEStatus); + RTL_W8(Cfg9346, Cfg9346_Lock); + *ioaddr_out = ioaddr; *dev_out = dev; out: @@ -1612,49 +1623,6 @@ rtl8169_remove_one(struct pci_dev *pdev) pci_set_drvdata(pdev, NULL); } -#ifdef CONFIG_PM - -static int rtl8169_suspend(struct pci_dev *pdev, pm_message_t state) -{ - struct net_device *dev = pci_get_drvdata(pdev); - struct rtl8169_private *tp = netdev_priv(dev); - void __iomem *ioaddr = tp->mmio_addr; - unsigned long flags; - - if (!netif_running(dev)) - return 0; - - netif_device_detach(dev); - netif_stop_queue(dev); - spin_lock_irqsave(&tp->lock, flags); - - /* Disable interrupts, stop Rx and Tx */ - RTL_W16(IntrMask, 0); - RTL_W8(ChipCmd, 0); - - /* Update the error counts. */ - tp->stats.rx_missed_errors += RTL_R32(RxMissed); - RTL_W32(RxMissed, 0); - spin_unlock_irqrestore(&tp->lock, flags); - - return 0; -} - -static int rtl8169_resume(struct pci_dev *pdev) -{ - struct net_device *dev = pci_get_drvdata(pdev); - - if (!netif_running(dev)) - return 0; - - netif_device_attach(dev); - rtl8169_hw_start(dev); - - return 0; -} - -#endif /* CONFIG_PM */ - static void rtl8169_set_rxbufsize(struct rtl8169_private *tp, struct net_device *dev) { @@ -2700,6 +2668,54 @@ static struct net_device_stats *rtl8169_get_stats(struct net_device *dev) return &tp->stats; } +#ifdef CONFIG_PM + +static int rtl8169_suspend(struct pci_dev *pdev, pm_message_t state) +{ + struct net_device *dev = pci_get_drvdata(pdev); + struct rtl8169_private *tp = netdev_priv(dev); + void __iomem *ioaddr = tp->mmio_addr; + + if (!netif_running(dev)) + goto out; + + netif_device_detach(dev); + netif_stop_queue(dev); + + spin_lock_irq(&tp->lock); + + rtl8169_asic_down(ioaddr); + + tp->stats.rx_missed_errors += RTL_R32(RxMissed); + RTL_W32(RxMissed, 0); + + spin_unlock_irq(&tp->lock); + + pci_save_state(pdev); + pci_set_power_state(pdev, pci_choose_state(pdev, state)); +out: + return 0; +} + +static int rtl8169_resume(struct pci_dev *pdev) +{ + struct net_device *dev = pci_get_drvdata(pdev); + + if (!netif_running(dev)) + goto out; + + netif_device_attach(dev); + + pci_set_power_state(pdev, PCI_D0); + pci_restore_state(pdev); + + rtl8169_schedule_work(dev, rtl8169_reset_task); +out: + return 0; +} + +#endif /* CONFIG_PM */ + static struct pci_driver rtl8169_pci_driver = { .name = MODULENAME, .id_table = rtl8169_pci_tbl, From 61a4dcc2f9b5c6861e7198b80dd73dd6e9247b7b Mon Sep 17 00:00:00 2001 From: Francois Romieu Date: Thu, 23 Feb 2006 00:55:25 +0100 Subject: [PATCH 099/102] r8169: enable wake on lan Similar to 8139cp code but more inspired/lucky. Signed-off-by: Francois Romieu --- drivers/net/r8169.c | 87 +++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 87 insertions(+) diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c index 999fd6cef77..8cc0d0bbdf5 100644 --- a/drivers/net/r8169.c +++ b/drivers/net/r8169.c @@ -290,7 +290,15 @@ enum RTL8169_register_content { /* Config1 register p.24 */ PMEnable = (1 << 0), /* Power Management Enable */ + /* Config3 register p.25 */ + MagicPacket = (1 << 5), /* Wake up when receives a Magic Packet */ + LinkUp = (1 << 4), /* Wake up when the cable connection is re-established */ + /* Config5 register p.27 */ + BWF = (1 << 6), /* Accept Broadcast wakeup frame */ + MWF = (1 << 5), /* Accept Multicast wakeup frame */ + UWF = (1 << 4), /* Accept Unicast wakeup frame */ + LanWake = (1 << 1), /* LanWake enable/disable */ PMEStatus = (1 << 0), /* PME status can be reset by PCI RST# */ /* TBICSR p.28 */ @@ -439,6 +447,7 @@ struct rtl8169_private { unsigned int (*phy_reset_pending)(void __iomem *); unsigned int (*link_ok)(void __iomem *); struct work_struct task; + unsigned wol_enabled : 1; }; MODULE_AUTHOR("Realtek and the Linux r8169 crew "); @@ -613,6 +622,80 @@ static void rtl8169_link_option(int idx, u8 *autoneg, u16 *speed, u8 *duplex) *duplex = p->duplex; } +static void rtl8169_get_wol(struct net_device *dev, struct ethtool_wolinfo *wol) +{ + struct rtl8169_private *tp = netdev_priv(dev); + void __iomem *ioaddr = tp->mmio_addr; + u8 options; + + wol->wolopts = 0; + +#define WAKE_ANY (WAKE_PHY | WAKE_MAGIC | WAKE_UCAST | WAKE_BCAST | WAKE_MCAST) + wol->supported = WAKE_ANY; + + spin_lock_irq(&tp->lock); + + options = RTL_R8(Config1); + if (!(options & PMEnable)) + goto out_unlock; + + options = RTL_R8(Config3); + if (options & LinkUp) + wol->wolopts |= WAKE_PHY; + if (options & MagicPacket) + wol->wolopts |= WAKE_MAGIC; + + options = RTL_R8(Config5); + if (options & UWF) + wol->wolopts |= WAKE_UCAST; + if (options & BWF) + wol->wolopts |= WAKE_BCAST; + if (options & MWF) + wol->wolopts |= WAKE_MCAST; + +out_unlock: + spin_unlock_irq(&tp->lock); +} + +static int rtl8169_set_wol(struct net_device *dev, struct ethtool_wolinfo *wol) +{ + struct rtl8169_private *tp = netdev_priv(dev); + void __iomem *ioaddr = tp->mmio_addr; + int i; + static struct { + u32 opt; + u16 reg; + u8 mask; + } cfg[] = { + { WAKE_ANY, Config1, PMEnable }, + { WAKE_PHY, Config3, LinkUp }, + { WAKE_MAGIC, Config3, MagicPacket }, + { WAKE_UCAST, Config5, UWF }, + { WAKE_BCAST, Config5, BWF }, + { WAKE_MCAST, Config5, MWF }, + { WAKE_ANY, Config5, LanWake } + }; + + spin_lock_irq(&tp->lock); + + RTL_W8(Cfg9346, Cfg9346_Unlock); + + for (i = 0; i < ARRAY_SIZE(cfg); i++) { + u8 options = RTL_R8(cfg[i].reg) & ~cfg[i].mask; + if (wol->wolopts & cfg[i].opt) + options |= cfg[i].mask; + RTL_W8(cfg[i].reg, options); + } + + RTL_W8(Cfg9346, Cfg9346_Lock); + + tp->wol_enabled = (wol->wolopts) ? 1 : 0; + + spin_unlock_irq(&tp->lock); + + return 0; +} + static void rtl8169_get_drvinfo(struct net_device *dev, struct ethtool_drvinfo *info) { @@ -1031,6 +1114,8 @@ static struct ethtool_ops rtl8169_ethtool_ops = { .get_tso = ethtool_op_get_tso, .set_tso = ethtool_op_set_tso, .get_regs = rtl8169_get_regs, + .get_wol = rtl8169_get_wol, + .set_wol = rtl8169_set_wol, .get_strings = rtl8169_get_strings, .get_stats_count = rtl8169_get_stats_count, .get_ethtool_stats = rtl8169_get_ethtool_stats, @@ -2692,6 +2777,7 @@ static int rtl8169_suspend(struct pci_dev *pdev, pm_message_t state) spin_unlock_irq(&tp->lock); pci_save_state(pdev); + pci_enable_wake(pdev, pci_choose_state(pdev, state), tp->wol_enabled); pci_set_power_state(pdev, pci_choose_state(pdev, state)); out: return 0; @@ -2708,6 +2794,7 @@ static int rtl8169_resume(struct pci_dev *pdev) pci_set_power_state(pdev, PCI_D0); pci_restore_state(pdev); + pci_enable_wake(pdev, PCI_D0, 0); rtl8169_schedule_work(dev, rtl8169_reset_task); out: From a9cdab869ec343ccc601484fb535813e16c25f70 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 22 Feb 2006 10:28:33 -0800 Subject: [PATCH 100/102] skge: NAPI/irq race fix Fix a race in the receive NAPI, irq handling. The interrupt clear and the start need to be separated. Otherwise there is a window between the last frame received and the NAPI done level handling. Signed-off-by: Stephen Hemminger --- drivers/net/skge.c | 17 +++++------------ 1 file changed, 5 insertions(+), 12 deletions(-) diff --git a/drivers/net/skge.c b/drivers/net/skge.c index 67fb19b8fde..869c7cfb99a 100644 --- a/drivers/net/skge.c +++ b/drivers/net/skge.c @@ -2678,8 +2678,7 @@ static int skge_poll(struct net_device *dev, int *budget) /* restart receiver */ wmb(); - skge_write8(hw, Q_ADDR(rxqaddr[skge->port], Q_CSR), - CSR_START | CSR_IRQ_CL_F); + skge_write8(hw, Q_ADDR(rxqaddr[skge->port], Q_CSR), CSR_START); *budget -= work_done; dev->quota -= work_done; @@ -2856,14 +2855,6 @@ static void skge_extirq(unsigned long data) local_irq_enable(); } -static inline void skge_wakeup(struct net_device *dev) -{ - struct skge_port *skge = netdev_priv(dev); - - prefetch(skge->rx_ring.to_clean); - netif_rx_schedule(dev); -} - static irqreturn_t skge_intr(int irq, void *dev_id, struct pt_regs *regs) { struct skge_hw *hw = dev_id; @@ -2874,13 +2865,15 @@ static irqreturn_t skge_intr(int irq, void *dev_id, struct pt_regs *regs) status &= hw->intr_mask; if (status & IS_R1_F) { + skge_write8(hw, Q_ADDR(Q_R1, Q_CSR), CSR_IRQ_CL_F); hw->intr_mask &= ~IS_R1_F; - skge_wakeup(hw->dev[0]); + netif_rx_schedule(hw->dev[0]); } if (status & IS_R2_F) { + skge_write8(hw, Q_ADDR(Q_R2, Q_CSR), CSR_IRQ_CL_F); hw->intr_mask &= ~IS_R2_F; - skge_wakeup(hw->dev[1]); + netif_rx_schedule(hw->dev[1]); } if (status & IS_XA1_F) From 0781191cf69b7635e0d3ea55c6019e789d1936fa Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 22 Feb 2006 10:28:34 -0800 Subject: [PATCH 101/102] skge: genesis phy initialzation The SysKonnect Genesis based board would fail on initialization with phy_read errors caused by not waiting for last phy write. Signed-off-by: Stephen Hemminger --- drivers/net/skge.c | 37 ++++++++++++++++++++++++++----------- 1 file changed, 26 insertions(+), 11 deletions(-) diff --git a/drivers/net/skge.c b/drivers/net/skge.c index 869c7cfb99a..af2e6782031 100644 --- a/drivers/net/skge.c +++ b/drivers/net/skge.c @@ -879,13 +879,12 @@ static int __xm_phy_read(struct skge_hw *hw, int port, u16 reg, u16 *val) int i; xm_write16(hw, port, XM_PHY_ADDR, reg | hw->phy_addr); - xm_read16(hw, port, XM_PHY_DATA); + *val = xm_read16(hw, port, XM_PHY_DATA); - /* Need to wait for external PHY */ for (i = 0; i < PHY_RETRIES; i++) { - udelay(1); if (xm_read16(hw, port, XM_MMU_CMD) & XM_MMU_PHY_RDY) goto ready; + udelay(1); } return -ETIMEDOUT; @@ -918,7 +917,12 @@ static int xm_phy_write(struct skge_hw *hw, int port, u16 reg, u16 val) ready: xm_write16(hw, port, XM_PHY_DATA, val); - return 0; + for (i = 0; i < PHY_RETRIES; i++) { + if (!(xm_read16(hw, port, XM_MMU_CMD) & XM_MMU_PHY_BUSY)) + return 0; + udelay(1); + } + return -ETIMEDOUT; } static void genesis_init(struct skge_hw *hw) @@ -1168,13 +1172,17 @@ static void genesis_mac_init(struct skge_hw *hw, int port) u32 r; const u8 zero[6] = { 0 }; - /* Clear MIB counters */ - xm_write16(hw, port, XM_STAT_CMD, - XM_SC_CLR_RXC | XM_SC_CLR_TXC); - /* Clear two times according to Errata #3 */ - xm_write16(hw, port, XM_STAT_CMD, - XM_SC_CLR_RXC | XM_SC_CLR_TXC); + for (i = 0; i < 10; i++) { + skge_write16(hw, SK_REG(port, TX_MFF_CTRL1), + MFF_SET_MAC_RST); + if (skge_read16(hw, SK_REG(port, TX_MFF_CTRL1)) & MFF_SET_MAC_RST) + goto reset_ok; + udelay(1); + } + printk(KERN_WARNING PFX "%s: genesis reset failed\n", dev->name); + + reset_ok: /* Unreset the XMAC. */ skge_write16(hw, SK_REG(port, TX_MFF_CTRL1), MFF_CLR_MAC_RST); @@ -1191,7 +1199,7 @@ static void genesis_mac_init(struct skge_hw *hw, int port) r |= GP_DIR_2|GP_IO_2; skge_write32(hw, B2_GP_IO, r); - skge_read32(hw, B2_GP_IO); + /* Enable GMII interface */ xm_write16(hw, port, XM_HW_CFG, XM_HW_GMII_MD); @@ -1205,6 +1213,13 @@ static void genesis_mac_init(struct skge_hw *hw, int port) for (i = 1; i < 16; i++) xm_outaddr(hw, port, XM_EXM(i), zero); + /* Clear MIB counters */ + xm_write16(hw, port, XM_STAT_CMD, + XM_SC_CLR_RXC | XM_SC_CLR_TXC); + /* Clear two times according to Errata #3 */ + xm_write16(hw, port, XM_STAT_CMD, + XM_SC_CLR_RXC | XM_SC_CLR_TXC); + /* configure Rx High Water Mark (XM_RX_HI_WM) */ xm_write16(hw, port, XM_RX_HI_WM, 1450); From 80dd857daca1cf541b10118991569470d62c1d38 Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Wed, 22 Feb 2006 10:28:35 -0800 Subject: [PATCH 102/102] skge: protect interrupt mask There is a race between updating the irq mask and setting it which can be triggered on SMP with a bad cable. Similar patch from Ingo Molnar and Thomas Gleixner Signed-off-by: Stephen Hemminger --- drivers/net/skge.c | 21 ++++++++++++++------- drivers/net/skge.h | 1 + 2 files changed, 15 insertions(+), 7 deletions(-) diff --git a/drivers/net/skge.c b/drivers/net/skge.c index af2e6782031..25e028b7ce4 100644 --- a/drivers/net/skge.c +++ b/drivers/net/skge.c @@ -2185,8 +2185,10 @@ static int skge_up(struct net_device *dev) skge->tx_avail = skge->tx_ring.count - 1; /* Enable IRQ from port */ + spin_lock_irq(&hw->hw_lock); hw->intr_mask |= portirqmask[port]; skge_write32(hw, B0_IMSK, hw->intr_mask); + spin_unlock_irq(&hw->hw_lock); /* Initialize MAC */ spin_lock_bh(&hw->phy_lock); @@ -2244,8 +2246,10 @@ static int skge_down(struct net_device *dev) else yukon_stop(skge); + spin_lock_irq(&hw->hw_lock); hw->intr_mask &= ~portirqmask[skge->port]; skge_write32(hw, B0_IMSK, hw->intr_mask); + spin_unlock_irq(&hw->hw_lock); /* Stop transmitter */ skge_write8(hw, Q_ADDR(txqaddr[port], Q_CSR), CSR_STOP); @@ -2701,10 +2705,11 @@ static int skge_poll(struct net_device *dev, int *budget) if (work_done >= to_do) return 1; /* not done */ - netif_rx_complete(dev); - hw->intr_mask |= portirqmask[skge->port]; - skge_write32(hw, B0_IMSK, hw->intr_mask); - skge_read32(hw, B0_IMSK); + spin_lock_irq(&hw->hw_lock); + __netif_rx_complete(dev); + hw->intr_mask |= portirqmask[skge->port]; + skge_write32(hw, B0_IMSK, hw->intr_mask); + spin_unlock_irq(&hw->hw_lock); return 0; } @@ -2864,10 +2869,10 @@ static void skge_extirq(unsigned long data) } spin_unlock(&hw->phy_lock); - local_irq_disable(); + spin_lock_irq(&hw->hw_lock); hw->intr_mask |= IS_EXT_REG; skge_write32(hw, B0_IMSK, hw->intr_mask); - local_irq_enable(); + spin_unlock_irq(&hw->hw_lock); } static irqreturn_t skge_intr(int irq, void *dev_id, struct pt_regs *regs) @@ -2878,7 +2883,7 @@ static irqreturn_t skge_intr(int irq, void *dev_id, struct pt_regs *regs) if (status == 0 || status == ~0) /* hotplug or shared irq */ return IRQ_NONE; - status &= hw->intr_mask; + spin_lock(&hw->hw_lock); if (status & IS_R1_F) { skge_write8(hw, Q_ADDR(Q_R1, Q_CSR), CSR_IRQ_CL_F); hw->intr_mask &= ~IS_R1_F; @@ -2930,6 +2935,7 @@ static irqreturn_t skge_intr(int irq, void *dev_id, struct pt_regs *regs) } skge_write32(hw, B0_IMSK, hw->intr_mask); + spin_unlock(&hw->hw_lock); return IRQ_HANDLED; } @@ -3298,6 +3304,7 @@ static int __devinit skge_probe(struct pci_dev *pdev, hw->pdev = pdev; spin_lock_init(&hw->phy_lock); + spin_lock_init(&hw->hw_lock); tasklet_init(&hw->ext_tasklet, skge_extirq, (unsigned long) hw); hw->regs = ioremap_nocache(pci_resource_start(pdev, 0), 0x4000); diff --git a/drivers/net/skge.h b/drivers/net/skge.h index 2efdacc290e..941f12a333b 100644 --- a/drivers/net/skge.h +++ b/drivers/net/skge.h @@ -2402,6 +2402,7 @@ struct skge_hw { struct tasklet_struct ext_tasklet; spinlock_t phy_lock; + spinlock_t hw_lock; }; enum {