dect
/
linux-2.6
Archived
13
0
Fork 0
This repository has been archived on 2022-02-17. You can view files and clone it, but cannot push or open issues or pull requests.
linux-2.6/drivers/scsi/u14-34f.c

1972 lines
66 KiB
C
Raw Normal View History

/*
* u14-34f.c - Low-level driver for UltraStor 14F/34F SCSI host adapters.
*
* 03 Jun 2003 Rev. 8.10 for linux-2.5.70
* + Update for new IRQ API.
* + Use "goto" when appropriate.
* + Drop u14-34f.h.
* + Update for new module_param API.
* + Module parameters can now be specified only in the
* same format as the kernel boot options.
*
* boot option old module param
* ----------- ------------------
* addr,... io_port=addr,...
* lc:[y|n] linked_comm=[1|0]
* mq:xx max_queue_depth=xx
* tm:[0|1|2] tag_mode=[0|1|2]
* et:[y|n] ext_tran=[1|0]
* of:[y|n] have_old_firmware=[1|0]
*
* A valid example using the new parameter format is:
* modprobe u14-34f "u14-34f=0x340,0x330,lc:y,tm:0,mq:4"
*
* which is equivalent to the old format:
* modprobe u14-34f io_port=0x340,0x330 linked_comm=1 tag_mode=0 \
* max_queue_depth=4
*
* With actual module code, u14-34f and u14_34f are equivalent
* as module parameter names.
*
* 12 Feb 2003 Rev. 8.04 for linux 2.5.60
* + Release irq before calling scsi_register.
*
* 12 Nov 2002 Rev. 8.02 for linux 2.5.47
* + Release driver_lock before calling scsi_register.
*
* 11 Nov 2002 Rev. 8.01 for linux 2.5.47
* + Fixed bios_param and scsicam_bios_param calling parameters.
*
* 28 Oct 2002 Rev. 8.00 for linux 2.5.44-ac4
* + Use new tcq and adjust_queue_depth api.
* + New command line option (tm:[0-2]) to choose the type of tags:
* 0 -> disable tagging ; 1 -> simple tags ; 2 -> ordered tags.
* Default is tm:0 (tagged commands disabled).
* For compatibility the "tc:" option is an alias of the "tm:"
* option; tc:n is equivalent to tm:0 and tc:y is equivalent to
* tm:1.
*
* 10 Oct 2002 Rev. 7.70 for linux 2.5.42
* + Foreport from revision 6.70.
*
* 25 Jun 2002 Rev. 6.70 for linux 2.4.19
* + Fixed endian-ness problem due to bitfields.
*
* 21 Feb 2002 Rev. 6.52 for linux 2.4.18
* + Backport from rev. 7.22 (use io_request_lock).
*
* 20 Feb 2002 Rev. 7.22 for linux 2.5.5
* + Remove any reference to virt_to_bus().
* + Fix pio hang while detecting multiple HBAs.
*
* 01 Jan 2002 Rev. 7.20 for linux 2.5.1
* + Use the dynamic DMA mapping API.
*
* 19 Dec 2001 Rev. 7.02 for linux 2.5.1
* + Use SCpnt->sc_data_direction if set.
* + Use sglist.page instead of sglist.address.
*
* 11 Dec 2001 Rev. 7.00 for linux 2.5.1
* + Use host->host_lock instead of io_request_lock.
*
* 1 May 2001 Rev. 6.05 for linux 2.4.4
* + Fix data transfer direction for opcode SEND_CUE_SHEET (0x5d)
*
* 25 Jan 2001 Rev. 6.03 for linux 2.4.0
* + "check_region" call replaced by "request_region".
*
* 22 Nov 2000 Rev. 6.02 for linux 2.4.0-test11
* + Removed old scsi error handling support.
* + The obsolete boot option flag eh:n is silently ignored.
* + Removed error messages while a disk drive is powered up at
* boot time.
* + Improved boot messages: all tagged capable device are
* indicated as "tagged".
*
* 16 Sep 1999 Rev. 5.11 for linux 2.2.12 and 2.3.18
* + Updated to the new __setup interface for boot command line options.
* + When loaded as a module, accepts the new parameter boot_options
* which value is a string with the same format of the kernel boot
* command line options. A valid example is:
* modprobe u14-34f 'boot_options="0x230,0x340,lc:y,mq:4"'
*
* 22 Jul 1999 Rev. 5.00 for linux 2.2.10 and 2.3.11
* + Removed pre-2.2 source code compatibility.
*
* 26 Jul 1998 Rev. 4.33 for linux 2.0.35 and 2.1.111
* Added command line option (et:[y|n]) to use the existing
* translation (returned by scsicam_bios_param) as disk geometry.
* The default is et:n, which uses the disk geometry jumpered
* on the board.
* The default value et:n is compatible with all previous revisions
* of this driver.
*
* 28 May 1998 Rev. 4.32 for linux 2.0.33 and 2.1.104
* Increased busy timeout from 10 msec. to 200 msec. while
* processing interrupts.
*
* 18 May 1998 Rev. 4.31 for linux 2.0.33 and 2.1.102
* Improved abort handling during the eh recovery process.
*
* 13 May 1998 Rev. 4.30 for linux 2.0.33 and 2.1.101
* The driver is now fully SMP safe, including the
* abort and reset routines.
* Added command line options (eh:[y|n]) to choose between
* new_eh_code and the old scsi code.
* If linux version >= 2.1.101 the default is eh:y, while the eh
* option is ignored for previous releases and the old scsi code
* is used.
*
* 18 Apr 1998 Rev. 4.20 for linux 2.0.33 and 2.1.97
* Reworked interrupt handler.
*
* 11 Apr 1998 rev. 4.05 for linux 2.0.33 and 2.1.95
* Major reliability improvement: when a batch with overlapping
* requests is detected, requests are queued one at a time
* eliminating any possible board or drive reordering.
*
* 10 Apr 1998 rev. 4.04 for linux 2.0.33 and 2.1.95
* Improved SMP support (if linux version >= 2.1.95).
*
* 9 Apr 1998 rev. 4.03 for linux 2.0.33 and 2.1.94
* Performance improvement: when sequential i/o is detected,
* always use direct sort instead of reverse sort.
*
* 4 Apr 1998 rev. 4.02 for linux 2.0.33 and 2.1.92
* io_port is now unsigned long.
*
* 17 Mar 1998 rev. 4.01 for linux 2.0.33 and 2.1.88
* Use new scsi error handling code (if linux version >= 2.1.88).
* Use new interrupt code.
*
* 12 Sep 1997 rev. 3.11 for linux 2.0.30 and 2.1.55
* Use of udelay inside the wait loops to avoid timeout
* problems with fast cpus.
* Removed check about useless calls to the interrupt service
* routine (reported on SMP systems only).
* At initialization time "sorted/unsorted" is displayed instead
* of "linked/unlinked" to reinforce the fact that "linking" is
* nothing but "elevator sorting" in the actual implementation.
*
* 17 May 1997 rev. 3.10 for linux 2.0.30 and 2.1.38
* Use of serial_number_at_timeout in abort and reset processing.
* Use of the __initfunc and __initdata macro in setup code.
* Minor cleanups in the list_statistics code.
*
* 24 Feb 1997 rev. 3.00 for linux 2.0.29 and 2.1.26
* When loading as a module, parameter passing is now supported
* both in 2.0 and in 2.1 style.
* Fixed data transfer direction for some SCSI opcodes.
* Immediate acknowledge to request sense commands.
* Linked commands to each disk device are now reordered by elevator
* sorting. Rare cases in which reordering of write requests could
* cause wrong results are managed.
*
* 18 Jan 1997 rev. 2.60 for linux 2.1.21 and 2.0.28
* Added command line options to enable/disable linked commands
* (lc:[y|n]), old firmware support (of:[y|n]) and to set the max
* queue depth (mq:xx). Default is "u14-34f=lc:n,of:n,mq:8".
* Improved command linking.
*
* 8 Jan 1997 rev. 2.50 for linux 2.1.20 and 2.0.27
* Added linked command support.
*
* 3 Dec 1996 rev. 2.40 for linux 2.1.14 and 2.0.27
* Added queue depth adjustment.
*
* 22 Nov 1996 rev. 2.30 for linux 2.1.12 and 2.0.26
* The list of i/o ports to be probed can be overwritten by the
* "u14-34f=port0,port1,...." boot command line option.
* Scatter/gather lists are now allocated by a number of kmalloc
* calls, in order to avoid the previous size limit of 64Kb.
*
* 16 Nov 1996 rev. 2.20 for linux 2.1.10 and 2.0.25
* Added multichannel support.
*
* 27 Sep 1996 rev. 2.12 for linux 2.1.0
* Portability cleanups (virtual/bus addressing, little/big endian
* support).
*
* 09 Jul 1996 rev. 2.11 for linux 2.0.4
* "Data over/under-run" no longer implies a redo on all targets.
* Number of internal retries is now limited.
*
* 16 Apr 1996 rev. 2.10 for linux 1.3.90
* New argument "reset_flags" to the reset routine.
*
* 21 Jul 1995 rev. 2.02 for linux 1.3.11
* Fixed Data Transfer Direction for some SCSI commands.
*
* 13 Jun 1995 rev. 2.01 for linux 1.2.10
* HAVE_OLD_UX4F_FIRMWARE should be defined for U34F boards when
* the firmware prom is not the latest one (28008-006).
*
* 11 Mar 1995 rev. 2.00 for linux 1.2.0
* Fixed a bug which prevented media change detection for removable
* disk drives.
*
* 23 Feb 1995 rev. 1.18 for linux 1.1.94
* Added a check for scsi_register returning NULL.
*
* 11 Feb 1995 rev. 1.17 for linux 1.1.91
* U14F qualified to run with 32 sglists.
* Now DEBUG_RESET is disabled by default.
*
* 9 Feb 1995 rev. 1.16 for linux 1.1.90
* Use host->wish_block instead of host->block.
*
* 8 Feb 1995 rev. 1.15 for linux 1.1.89
* Cleared target_time_out counter while performing a reset.
*
* 28 Jan 1995 rev. 1.14 for linux 1.1.86
* Added module support.
* Log and do a retry when a disk drive returns a target status
* different from zero on a recovered error.
* Auto detects if U14F boards have an old firmware revision.
* Max number of scatter/gather lists set to 16 for all boards
* (most installation run fine using 33 sglists, while other
* has problems when using more than 16).
*
* 16 Jan 1995 rev. 1.13 for linux 1.1.81
* Display a message if check_region detects a port address
* already in use.
*
* 15 Dec 1994 rev. 1.12 for linux 1.1.74
* The host->block flag is set for all the detected ISA boards.
*
* 30 Nov 1994 rev. 1.11 for linux 1.1.68
* Redo i/o on target status CHECK_CONDITION for TYPE_DISK only.
* Added optional support for using a single board at a time.
*
* 14 Nov 1994 rev. 1.10 for linux 1.1.63
*
* 28 Oct 1994 rev. 1.09 for linux 1.1.58 Final BETA release.
* 16 Jul 1994 rev. 1.00 for linux 1.1.29 Initial ALPHA release.
*
* This driver is a total replacement of the original UltraStor
* scsi driver, but it supports ONLY the 14F and 34F boards.
* It can be configured in the same kernel in which the original
* ultrastor driver is configured to allow the original U24F
* support.
*
* Multiple U14F and/or U34F host adapters are supported.
*
* Copyright (C) 1994-2003 Dario Ballabio (ballabio_dario@emc.com)
*
* Alternate email: dario.ballabio@inwind.it, dario.ballabio@tiscalinet.it
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that redistributions of source
* code retain the above copyright notice and this comment without
* modification.
*
* WARNING: if your 14/34F board has an old firmware revision (see below)
* you must change "#undef" into "#define" in the following
* statement.
*/
#undef HAVE_OLD_UX4F_FIRMWARE
/*
* The UltraStor 14F, 24F, and 34F are a family of intelligent, high
* performance SCSI-2 host adapters.
* Here is the scoop on the various models:
*
* 14F - ISA first-party DMA HA with floppy support and WD1003 emulation.
* 24F - EISA Bus Master HA with floppy support and WD1003 emulation.
* 34F - VESA Local-Bus Bus Master HA (no WD1003 emulation).
*
* This code has been tested with up to two U14F boards, using both
* firmware 28004-005/38004-004 (BIOS rev. 2.00) and the latest firmware
* 28004-006/38004-005 (BIOS rev. 2.01).
*
* The latest firmware is required in order to get reliable operations when
* clustering is enabled. ENABLE_CLUSTERING provides a performance increase
* up to 50% on sequential access.
*
* Since the struct scsi_host_template structure is shared among all 14F and 34F,
* the last setting of use_clustering is in effect for all of these boards.
*
* Here a sample configuration using two U14F boards:
*
U14F0: ISA 0x330, BIOS 0xc8000, IRQ 11, DMA 5, SG 32, MB 16, of:n, lc:y, mq:8.
U14F1: ISA 0x340, BIOS 0x00000, IRQ 10, DMA 6, SG 32, MB 16, of:n, lc:y, mq:8.
*
* The boot controller must have its BIOS enabled, while other boards can
* have their BIOS disabled, or enabled to an higher address.
* Boards are named Ux4F0, Ux4F1..., according to the port address order in
* the io_port[] array.
*
* The following facts are based on real testing results (not on
* documentation) on the above U14F board.
*
* - The U14F board should be jumpered for bus on time less or equal to 7
* microseconds, while the default is 11 microseconds. This is order to
* get acceptable performance while using floppy drive and hard disk
* together. The jumpering for 7 microseconds is: JP13 pin 15-16,
* JP14 pin 7-8 and pin 9-10.
* The reduction has a little impact on scsi performance.
*
* - If scsi bus length exceeds 3m., the scsi bus speed needs to be reduced
* from 10Mhz to 5Mhz (do this by inserting a jumper on JP13 pin 7-8).
*
* - If U14F on board firmware is older than 28004-006/38004-005,
* the U14F board is unable to provide reliable operations if the scsi
* request length exceeds 16Kbyte. When this length is exceeded the
* behavior is:
* - adapter_status equal 0x96 or 0xa3 or 0x93 or 0x94;
* - adapter_status equal 0 and target_status equal 2 on for all targets
* in the next operation following the reset.
* This sequence takes a long time (>3 seconds), so in the meantime
* the SD_TIMEOUT in sd.c could expire giving rise to scsi aborts
* (SD_TIMEOUT has been increased from 3 to 6 seconds in 1.1.31).
* Because of this I had to DISABLE_CLUSTERING and to work around the
* bus reset in the interrupt service routine, returning DID_BUS_BUSY
* so that the operations are retried without complains from the scsi.c
* code.
* Any reset of the scsi bus is going to kill tape operations, since
* no retry is allowed for tapes. Bus resets are more likely when the
* scsi bus is under heavy load.
* Requests using scatter/gather have a maximum length of 16 x 1024 bytes
* when DISABLE_CLUSTERING is in effect, but unscattered requests could be
* larger than 16Kbyte.
*
* The new firmware has fixed all the above problems.
*
* For U34F boards the latest bios prom is 38008-002 (BIOS rev. 2.01),
* the latest firmware prom is 28008-006. Older firmware 28008-005 has
* problems when using more than 16 scatter/gather lists.
*
* The list of i/o ports to be probed can be totally replaced by the
* boot command line option: "u14-34f=port0,port1,port2,...", where the
* port0, port1... arguments are ISA/VESA addresses to be probed.
* For example using "u14-34f=0x230,0x340", the driver probes only the two
* addresses 0x230 and 0x340 in this order; "u14-34f=0" totally disables
* this driver.
*
* After the optional list of detection probes, other possible command line
* options are:
*
* et:y use disk geometry returned by scsicam_bios_param;
* et:n use disk geometry jumpered on the board;
* lc:y enables linked commands;
* lc:n disables linked commands;
* tm:0 disables tagged commands (same as tc:n);
* tm:1 use simple queue tags (same as tc:y);
* tm:2 use ordered queue tags (same as tc:2);
* of:y enables old firmware support;
* of:n disables old firmware support;
* mq:xx set the max queue depth to the value xx (2 <= xx <= 8).
*
* The default value is: "u14-34f=lc:n,of:n,mq:8,tm:0,et:n".
* An example using the list of detection probes could be:
* "u14-34f=0x230,0x340,lc:y,tm:2,of:n,mq:4,et:n".
*
* When loading as a module, parameters can be specified as well.
* The above example would be (use 1 in place of y and 0 in place of n):
*
* modprobe u14-34f io_port=0x230,0x340 linked_comm=1 have_old_firmware=0 \
* max_queue_depth=4 ext_tran=0 tag_mode=2
*
* ----------------------------------------------------------------------------
* In this implementation, linked commands are designed to work with any DISK
* or CD-ROM, since this linking has only the intent of clustering (time-wise)
* and reordering by elevator sorting commands directed to each device,
* without any relation with the actual SCSI protocol between the controller
* and the device.
* If Q is the queue depth reported at boot time for each device (also named
* cmds/lun) and Q > 2, whenever there is already an active command to the
* device all other commands to the same device (up to Q-1) are kept waiting
* in the elevator sorting queue. When the active command completes, the
* commands in this queue are sorted by sector address. The sort is chosen
* between increasing or decreasing by minimizing the seek distance between
* the sector of the commands just completed and the sector of the first
* command in the list to be sorted.
* Trivial math assures that the unsorted average seek distance when doing
* random seeks over S sectors is S/3.
* When (Q-1) requests are uniformly distributed over S sectors, the average
* distance between two adjacent requests is S/((Q-1) + 1), so the sorted
* average seek distance for (Q-1) random requests over S sectors is S/Q.
* The elevator sorting hence divides the seek distance by a factor Q/3.
* The above pure geometric remarks are valid in all cases and the
* driver effectively reduces the seek distance by the predicted factor
* when there are Q concurrent read i/o operations on the device, but this
* does not necessarily results in a noticeable performance improvement:
* your mileage may vary....
*
* Note: command reordering inside a batch of queued commands could cause
* wrong results only if there is at least one write request and the
* intersection (sector-wise) of all requests is not empty.
* When the driver detects a batch including overlapping requests
* (a really rare event) strict serial (pid) order is enforced.
* ----------------------------------------------------------------------------
*
* The boards are named Ux4F0, Ux4F1,... according to the detection order.
*
* In order to support multiple ISA boards in a reliable way,
* the driver sets host->wish_block = TRUE for all ISA boards.
*/
#include <linux/string.h>
#include <linux/kernel.h>
#include <linux/ioport.h>
#include <linux/delay.h>
#include <asm/io.h>
#include <asm/byteorder.h>
#include <linux/proc_fs.h>
#include <linux/blkdev.h>
#include <linux/interrupt.h>
#include <linux/stat.h>
#include <linux/pci.h>
#include <linux/init.h>
#include <linux/ctype.h>
#include <linux/spinlock.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/slab.h>
#include <asm/dma.h>
#include <asm/irq.h>
#include <scsi/scsi.h>
#include <scsi/scsi_cmnd.h>
#include <scsi/scsi_device.h>
#include <scsi/scsi_host.h>
#include <scsi/scsi_tcq.h>
#include <scsi/scsicam.h>
static int u14_34f_detect(struct scsi_host_template *);
static int u14_34f_release(struct Scsi_Host *);
static int u14_34f_queuecommand(struct Scsi_Host *, struct scsi_cmnd *);
static int u14_34f_eh_abort(struct scsi_cmnd *);
static int u14_34f_eh_host_reset(struct scsi_cmnd *);
static int u14_34f_bios_param(struct scsi_device *, struct block_device *,
sector_t, int *);
static int u14_34f_slave_configure(struct scsi_device *);
static struct scsi_host_template driver_template = {
.name = "UltraStor 14F/34F rev. 8.10.00 ",
.detect = u14_34f_detect,
.release = u14_34f_release,
.queuecommand = u14_34f_queuecommand,
.eh_abort_handler = u14_34f_eh_abort,
.eh_host_reset_handler = u14_34f_eh_host_reset,
.bios_param = u14_34f_bios_param,
.slave_configure = u14_34f_slave_configure,
.this_id = 7,
.unchecked_isa_dma = 1,
.use_clustering = ENABLE_CLUSTERING,
};
#if !defined(__BIG_ENDIAN_BITFIELD) && !defined(__LITTLE_ENDIAN_BITFIELD)
#error "Adjust your <asm/byteorder.h> defines"
#endif
/* Values for the PRODUCT_ID ports for the 14/34F */
#define PRODUCT_ID1 0x56
#define PRODUCT_ID2 0x40 /* NOTE: Only upper nibble is used */
/* Subversion values */
#define ISA 0
#define ESA 1
#define OP_HOST_ADAPTER 0x1
#define OP_SCSI 0x2
#define OP_RESET 0x4
#define DTD_SCSI 0x0
#define DTD_IN 0x1
#define DTD_OUT 0x2
#define DTD_NONE 0x3
#define HA_CMD_INQUIRY 0x1
#define HA_CMD_SELF_DIAG 0x2
#define HA_CMD_READ_BUFF 0x3
#define HA_CMD_WRITE_BUFF 0x4
#undef DEBUG_LINKED_COMMANDS
#undef DEBUG_DETECT
#undef DEBUG_INTERRUPT
#undef DEBUG_RESET
#undef DEBUG_GENERATE_ERRORS
#undef DEBUG_GENERATE_ABORTS
#undef DEBUG_GEOMETRY
#define MAX_ISA 3
#define MAX_VESA 1
#define MAX_EISA 0
#define MAX_PCI 0
#define MAX_BOARDS (MAX_ISA + MAX_VESA + MAX_EISA + MAX_PCI)
#define MAX_CHANNEL 1
#define MAX_LUN 8
#define MAX_TARGET 8
#define MAX_MAILBOXES 16
#define MAX_SGLIST 32
#define MAX_SAFE_SGLIST 16
#define MAX_INTERNAL_RETRIES 64
#define MAX_CMD_PER_LUN 2
#define MAX_TAGGED_CMD_PER_LUN (MAX_MAILBOXES - MAX_CMD_PER_LUN)
#define SKIP ULONG_MAX
#define FALSE 0
#define TRUE 1
#define FREE 0
#define IN_USE 1
#define LOCKED 2
#define IN_RESET 3
#define IGNORE 4
#define READY 5
#define ABORTING 6
#define NO_DMA 0xff
#define MAXLOOP 10000
#define TAG_DISABLED 0
#define TAG_SIMPLE 1
#define TAG_ORDERED 2
#define REG_LCL_MASK 0
#define REG_LCL_INTR 1
#define REG_SYS_MASK 2
#define REG_SYS_INTR 3
#define REG_PRODUCT_ID1 4
#define REG_PRODUCT_ID2 5
#define REG_CONFIG1 6
#define REG_CONFIG2 7
#define REG_OGM 8
#define REG_ICM 12
#define REGION_SIZE 13UL
#define BSY_ASSERTED 0x01
#define IRQ_ASSERTED 0x01
#define CMD_RESET 0xc0
#define CMD_OGM_INTR 0x01
#define CMD_CLR_INTR 0x01
#define CMD_ENA_INTR 0x81
#define ASOK 0x00
#define ASST 0x91
#define YESNO(a) ((a) ? 'y' : 'n')
#define TLDEV(type) ((type) == TYPE_DISK || (type) == TYPE_ROM)
#define PACKED __attribute__((packed))
struct sg_list {
unsigned int address; /* Segment Address */
unsigned int num_bytes; /* Segment Length */
};
/* MailBox SCSI Command Packet */
struct mscp {
#if defined(__BIG_ENDIAN_BITFIELD)
unsigned char sg:1, ca:1, dcn:1, xdir:2, opcode:3;
unsigned char lun: 3, channel:2, target:3;
#else
unsigned char opcode: 3, /* type of command */
xdir: 2, /* data transfer direction */
dcn: 1, /* disable disconnect */
ca: 1, /* use cache (if available) */
sg: 1; /* scatter/gather operation */
unsigned char target: 3, /* SCSI target id */
channel: 2, /* SCSI channel number */
lun: 3; /* SCSI logical unit number */
#endif
unsigned int data_address PACKED; /* transfer data pointer */
unsigned int data_len PACKED; /* length in bytes */
unsigned int link_address PACKED; /* for linking command chains */
unsigned char clink_id; /* identifies command in chain */
unsigned char use_sg; /* (if sg is set) 8 bytes per list */
unsigned char sense_len;
unsigned char cdb_len; /* 6, 10, or 12 */
unsigned char cdb[12]; /* SCSI Command Descriptor Block */
unsigned char adapter_status; /* non-zero indicates HA error */
unsigned char target_status; /* non-zero indicates target error */
unsigned int sense_addr PACKED;
/* Additional fields begin here. */
struct scsi_cmnd *SCpnt;
unsigned int cpp_index; /* cp index */
/* All the cp structure is zero filled by queuecommand except the
following CP_TAIL_SIZE bytes, initialized by detect */
dma_addr_t cp_dma_addr; /* dma handle for this cp structure */
struct sg_list *sglist; /* pointer to the allocated SG list */
};
#define CP_TAIL_SIZE (sizeof(struct sglist *) + sizeof(dma_addr_t))
struct hostdata {
struct mscp cp[MAX_MAILBOXES]; /* Mailboxes for this board */
unsigned int cp_stat[MAX_MAILBOXES]; /* FREE, IN_USE, LOCKED, IN_RESET */
unsigned int last_cp_used; /* Index of last mailbox used */
unsigned int iocount; /* Total i/o done for this board */
int board_number; /* Number of this board */
char board_name[16]; /* Name of this board */
int in_reset; /* True if board is doing a reset */
int target_to[MAX_TARGET][MAX_CHANNEL]; /* N. of timeout errors on target */
int target_redo[MAX_TARGET][MAX_CHANNEL]; /* If TRUE redo i/o on target */
unsigned int retries; /* Number of internal retries */
unsigned long last_retried_pid; /* Pid of last retried command */
unsigned char subversion; /* Bus type, either ISA or ESA */
struct pci_dev *pdev; /* Always NULL */
unsigned char heads;
unsigned char sectors;
char board_id[256]; /* data from INQUIRY on this board */
};
static struct Scsi_Host *sh[MAX_BOARDS + 1];
static const char *driver_name = "Ux4F";
static char sha[MAX_BOARDS];
static DEFINE_SPINLOCK(driver_lock);
/* Initialize num_boards so that ihdlr can work while detect is in progress */
static unsigned int num_boards = MAX_BOARDS;
static unsigned long io_port[] = {
/* Space for MAX_INT_PARAM ports usable while loading as a module */
SKIP, SKIP, SKIP, SKIP, SKIP, SKIP, SKIP, SKIP,
SKIP, SKIP,
/* Possible ISA/VESA ports */
0x330, 0x340, 0x230, 0x240, 0x210, 0x130, 0x140,
/* End of list */
0x0
};
#define HD(board) ((struct hostdata *) &sh[board]->hostdata)
#define BN(board) (HD(board)->board_name)
/* Device is Little Endian */
#define H2DEV(x) cpu_to_le32(x)
#define DEV2H(x) le32_to_cpu(x)
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 13:55:46 +00:00
static irqreturn_t do_interrupt_handler(int, void *);
static void flush_dev(struct scsi_device *, unsigned long, unsigned int, unsigned int);
static int do_trace = FALSE;
static int setup_done = FALSE;
static int link_statistics;
static int ext_tran = FALSE;
#if defined(HAVE_OLD_UX4F_FIRMWARE)
static int have_old_firmware = TRUE;
#else
static int have_old_firmware = FALSE;
#endif
#if defined(CONFIG_SCSI_U14_34F_TAGGED_QUEUE)
static int tag_mode = TAG_SIMPLE;
#else
static int tag_mode = TAG_DISABLED;
#endif
#if defined(CONFIG_SCSI_U14_34F_LINKED_COMMANDS)
static int linked_comm = TRUE;
#else
static int linked_comm = FALSE;
#endif
#if defined(CONFIG_SCSI_U14_34F_MAX_TAGS)
static int max_queue_depth = CONFIG_SCSI_U14_34F_MAX_TAGS;
#else
static int max_queue_depth = MAX_CMD_PER_LUN;
#endif
#define MAX_INT_PARAM 10
#define MAX_BOOT_OPTIONS_SIZE 256
static char boot_options[MAX_BOOT_OPTIONS_SIZE];
#if defined(MODULE)
#include <linux/module.h>
#include <linux/moduleparam.h>
module_param_string(u14_34f, boot_options, MAX_BOOT_OPTIONS_SIZE, 0);
MODULE_PARM_DESC(u14_34f, " equivalent to the \"u14-34f=...\" kernel boot " \
"option." \
" Example: modprobe u14-34f \"u14_34f=0x340,0x330,lc:y,tm:0,mq:4\"");
MODULE_AUTHOR("Dario Ballabio");
MODULE_LICENSE("GPL");
MODULE_DESCRIPTION("UltraStor 14F/34F SCSI Driver");
#endif
static int u14_34f_slave_configure(struct scsi_device *dev) {
int j, tqd, utqd;
char *tag_suffix, *link_suffix;
struct Scsi_Host *host = dev->host;
j = ((struct hostdata *) host->hostdata)->board_number;
utqd = MAX_CMD_PER_LUN;
tqd = max_queue_depth;
if (TLDEV(dev->type) && dev->tagged_supported)
if (tag_mode == TAG_SIMPLE) {
scsi_adjust_queue_depth(dev, MSG_SIMPLE_TAG, tqd);
tag_suffix = ", simple tags";
}
else if (tag_mode == TAG_ORDERED) {
scsi_adjust_queue_depth(dev, MSG_ORDERED_TAG, tqd);
tag_suffix = ", ordered tags";
}
else {
scsi_adjust_queue_depth(dev, 0, tqd);
tag_suffix = ", no tags";
}
else if (TLDEV(dev->type) && linked_comm) {
scsi_adjust_queue_depth(dev, 0, tqd);
tag_suffix = ", untagged";
}
else {
scsi_adjust_queue_depth(dev, 0, utqd);
tag_suffix = "";
}
if (TLDEV(dev->type) && linked_comm && dev->queue_depth > 2)
link_suffix = ", sorted";
else if (TLDEV(dev->type))
link_suffix = ", unsorted";
else
link_suffix = "";
sdev_printk(KERN_INFO, dev, "cmds/lun %d%s%s.\n",
dev->queue_depth, link_suffix, tag_suffix);
return FALSE;
}
static int wait_on_busy(unsigned long iobase, unsigned int loop) {
while (inb(iobase + REG_LCL_INTR) & BSY_ASSERTED) {
udelay(1L);
if (--loop == 0) return TRUE;
}
return FALSE;
}
static int board_inquiry(unsigned int j) {
struct mscp *cpp;
dma_addr_t id_dma_addr;
unsigned int limit = 0;
unsigned long time;
id_dma_addr = pci_map_single(HD(j)->pdev, HD(j)->board_id,
sizeof(HD(j)->board_id), PCI_DMA_BIDIRECTIONAL);
cpp = &HD(j)->cp[0];
cpp->cp_dma_addr = pci_map_single(HD(j)->pdev, cpp, sizeof(struct mscp),
PCI_DMA_BIDIRECTIONAL);
memset(cpp, 0, sizeof(struct mscp) - CP_TAIL_SIZE);
cpp->opcode = OP_HOST_ADAPTER;
cpp->xdir = DTD_IN;
cpp->data_address = H2DEV(id_dma_addr);
cpp->data_len = H2DEV(sizeof(HD(j)->board_id));
cpp->cdb_len = 6;
cpp->cdb[0] = HA_CMD_INQUIRY;
if (wait_on_busy(sh[j]->io_port, MAXLOOP)) {
printk("%s: board_inquiry, adapter busy.\n", BN(j));
return TRUE;
}
HD(j)->cp_stat[0] = IGNORE;
/* Clear the interrupt indication */
outb(CMD_CLR_INTR, sh[j]->io_port + REG_SYS_INTR);
/* Store pointer in OGM address bytes */
outl(H2DEV(cpp->cp_dma_addr), sh[j]->io_port + REG_OGM);
/* Issue OGM interrupt */
outb(CMD_OGM_INTR, sh[j]->io_port + REG_LCL_INTR);
spin_unlock_irq(&driver_lock);
time = jiffies;
while ((jiffies - time) < HZ && limit++ < 20000) udelay(100L);
spin_lock_irq(&driver_lock);
if (cpp->adapter_status || HD(j)->cp_stat[0] != FREE) {
HD(j)->cp_stat[0] = FREE;
printk("%s: board_inquiry, err 0x%x.\n", BN(j), cpp->adapter_status);
return TRUE;
}
pci_unmap_single(HD(j)->pdev, cpp->cp_dma_addr, sizeof(struct mscp),
PCI_DMA_BIDIRECTIONAL);
pci_unmap_single(HD(j)->pdev, id_dma_addr, sizeof(HD(j)->board_id),
PCI_DMA_BIDIRECTIONAL);
return FALSE;
}
static int port_detect \
(unsigned long port_base, unsigned int j, struct scsi_host_template *tpnt) {
unsigned char irq, dma_channel, subversion, i;
unsigned char in_byte;
char *bus_type, dma_name[16];
/* Allowed BIOS base addresses (NULL indicates reserved) */
unsigned long bios_segment_table[8] = {
0,
0xc4000, 0xc8000, 0xcc000, 0xd0000,
0xd4000, 0xd8000, 0xdc000
};
/* Allowed IRQs */
unsigned char interrupt_table[4] = { 15, 14, 11, 10 };
/* Allowed DMA channels for ISA (0 indicates reserved) */
unsigned char dma_channel_table[4] = { 5, 6, 7, 0 };
/* Head/sector mappings */
struct {
unsigned char heads;
unsigned char sectors;
} mapping_table[4] = {
{ 16, 63 }, { 64, 32 }, { 64, 63 }, { 64, 32 }
};
struct config_1 {
#if defined(__BIG_ENDIAN_BITFIELD)
unsigned char dma_channel: 2, interrupt:2,
removable_disks_as_fixed:1, bios_segment: 3;
#else
unsigned char bios_segment: 3, removable_disks_as_fixed: 1,
interrupt: 2, dma_channel: 2;
#endif
} config_1;
struct config_2 {
#if defined(__BIG_ENDIAN_BITFIELD)
unsigned char tfr_port: 2, bios_drive_number: 1,
mapping_mode: 2, ha_scsi_id: 3;
#else
unsigned char ha_scsi_id: 3, mapping_mode: 2,
bios_drive_number: 1, tfr_port: 2;
#endif
} config_2;
char name[16];
sprintf(name, "%s%d", driver_name, j);
if (!request_region(port_base, REGION_SIZE, driver_name)) {
#if defined(DEBUG_DETECT)
printk("%s: address 0x%03lx in use, skipping probe.\n", name, port_base);
#endif
goto fail;
}
spin_lock_irq(&driver_lock);
if (inb(port_base + REG_PRODUCT_ID1) != PRODUCT_ID1) goto freelock;
in_byte = inb(port_base + REG_PRODUCT_ID2);
if ((in_byte & 0xf0) != PRODUCT_ID2) goto freelock;
*(char *)&config_1 = inb(port_base + REG_CONFIG1);
*(char *)&config_2 = inb(port_base + REG_CONFIG2);
irq = interrupt_table[config_1.interrupt];
dma_channel = dma_channel_table[config_1.dma_channel];
subversion = (in_byte & 0x0f);
/* Board detected, allocate its IRQ */
if (request_irq(irq, do_interrupt_handler,
IRQF_DISABLED | ((subversion == ESA) ? IRQF_SHARED : 0),
driver_name, (void *) &sha[j])) {
printk("%s: unable to allocate IRQ %u, detaching.\n", name, irq);
goto freelock;
}
if (subversion == ISA && request_dma(dma_channel, driver_name)) {
printk("%s: unable to allocate DMA channel %u, detaching.\n",
name, dma_channel);
goto freeirq;
}
if (have_old_firmware) tpnt->use_clustering = DISABLE_CLUSTERING;
spin_unlock_irq(&driver_lock);
sh[j] = scsi_register(tpnt, sizeof(struct hostdata));
spin_lock_irq(&driver_lock);
if (sh[j] == NULL) {
printk("%s: unable to register host, detaching.\n", name);
goto freedma;
}
sh[j]->io_port = port_base;
sh[j]->unique_id = port_base;
sh[j]->n_io_port = REGION_SIZE;
sh[j]->base = bios_segment_table[config_1.bios_segment];
sh[j]->irq = irq;
sh[j]->sg_tablesize = MAX_SGLIST;
sh[j]->this_id = config_2.ha_scsi_id;
sh[j]->can_queue = MAX_MAILBOXES;
sh[j]->cmd_per_lun = MAX_CMD_PER_LUN;
#if defined(DEBUG_DETECT)
{
unsigned char sys_mask, lcl_mask;
sys_mask = inb(sh[j]->io_port + REG_SYS_MASK);
lcl_mask = inb(sh[j]->io_port + REG_LCL_MASK);
printk("SYS_MASK 0x%x, LCL_MASK 0x%x.\n", sys_mask, lcl_mask);
}
#endif
/* Probably a bogus host scsi id, set it to the dummy value */
if (sh[j]->this_id == 0) sh[j]->this_id = -1;
/* If BIOS is disabled, force enable interrupts */
if (sh[j]->base == 0) outb(CMD_ENA_INTR, sh[j]->io_port + REG_SYS_MASK);
memset(HD(j), 0, sizeof(struct hostdata));
HD(j)->heads = mapping_table[config_2.mapping_mode].heads;
HD(j)->sectors = mapping_table[config_2.mapping_mode].sectors;
HD(j)->subversion = subversion;
HD(j)->pdev = NULL;
HD(j)->board_number = j;
if (have_old_firmware) sh[j]->sg_tablesize = MAX_SAFE_SGLIST;
if (HD(j)->subversion == ESA) {
sh[j]->unchecked_isa_dma = FALSE;
sh[j]->dma_channel = NO_DMA;
sprintf(BN(j), "U34F%d", j);
bus_type = "VESA";
}
else {
unsigned long flags;
sh[j]->unchecked_isa_dma = TRUE;
flags=claim_dma_lock();
disable_dma(dma_channel);
clear_dma_ff(dma_channel);
set_dma_mode(dma_channel, DMA_MODE_CASCADE);
enable_dma(dma_channel);
release_dma_lock(flags);
sh[j]->dma_channel = dma_channel;
sprintf(BN(j), "U14F%d", j);
bus_type = "ISA";
}
sh[j]->max_channel = MAX_CHANNEL - 1;
sh[j]->max_id = MAX_TARGET;
sh[j]->max_lun = MAX_LUN;
if (HD(j)->subversion == ISA && !board_inquiry(j)) {
HD(j)->board_id[40] = 0;
if (strcmp(&HD(j)->board_id[32], "06000600")) {
printk("%s: %s.\n", BN(j), &HD(j)->board_id[8]);
printk("%s: firmware %s is outdated, FW PROM should be 28004-006.\n",
BN(j), &HD(j)->board_id[32]);
sh[j]->hostt->use_clustering = DISABLE_CLUSTERING;
sh[j]->sg_tablesize = MAX_SAFE_SGLIST;
}
}
if (dma_channel == NO_DMA) sprintf(dma_name, "%s", "BMST");
else sprintf(dma_name, "DMA %u", dma_channel);
spin_unlock_irq(&driver_lock);
for (i = 0; i < sh[j]->can_queue; i++)
HD(j)->cp[i].cp_dma_addr = pci_map_single(HD(j)->pdev,
&HD(j)->cp[i], sizeof(struct mscp), PCI_DMA_BIDIRECTIONAL);
for (i = 0; i < sh[j]->can_queue; i++)
if (! ((&HD(j)->cp[i])->sglist = kmalloc(
sh[j]->sg_tablesize * sizeof(struct sg_list),
(sh[j]->unchecked_isa_dma ? GFP_DMA : 0) | GFP_ATOMIC))) {
printk("%s: kmalloc SGlist failed, mbox %d, detaching.\n", BN(j), i);
goto release;
}
if (max_queue_depth > MAX_TAGGED_CMD_PER_LUN)
max_queue_depth = MAX_TAGGED_CMD_PER_LUN;
if (max_queue_depth < MAX_CMD_PER_LUN) max_queue_depth = MAX_CMD_PER_LUN;
if (tag_mode != TAG_DISABLED && tag_mode != TAG_SIMPLE)
tag_mode = TAG_ORDERED;
if (j == 0) {
printk("UltraStor 14F/34F: Copyright (C) 1994-2003 Dario Ballabio.\n");
printk("%s config options -> of:%c, tm:%d, lc:%c, mq:%d, et:%c.\n",
driver_name, YESNO(have_old_firmware), tag_mode,
YESNO(linked_comm), max_queue_depth, YESNO(ext_tran));
}
printk("%s: %s 0x%03lx, BIOS 0x%05x, IRQ %u, %s, SG %d, MB %d.\n",
BN(j), bus_type, (unsigned long)sh[j]->io_port, (int)sh[j]->base,
sh[j]->irq, dma_name, sh[j]->sg_tablesize, sh[j]->can_queue);
if (sh[j]->max_id > 8 || sh[j]->max_lun > 8)
printk("%s: wide SCSI support enabled, max_id %u, max_lun %u.\n",
BN(j), sh[j]->max_id, sh[j]->max_lun);
for (i = 0; i <= sh[j]->max_channel; i++)
printk("%s: SCSI channel %u enabled, host target ID %d.\n",
BN(j), i, sh[j]->this_id);
return TRUE;
freedma:
if (subversion == ISA) free_dma(dma_channel);
freeirq:
free_irq(irq, &sha[j]);
freelock:
spin_unlock_irq(&driver_lock);
release_region(port_base, REGION_SIZE);
fail:
return FALSE;
release:
u14_34f_release(sh[j]);
return FALSE;
}
static void internal_setup(char *str, int *ints) {
int i, argc = ints[0];
char *cur = str, *pc;
if (argc > 0) {
if (argc > MAX_INT_PARAM) argc = MAX_INT_PARAM;
for (i = 0; i < argc; i++) io_port[i] = ints[i + 1];
io_port[i] = 0;
setup_done = TRUE;
}
while (cur && (pc = strchr(cur, ':'))) {
int val = 0, c = *++pc;
if (c == 'n' || c == 'N') val = FALSE;
else if (c == 'y' || c == 'Y') val = TRUE;
else val = (int) simple_strtoul(pc, NULL, 0);
if (!strncmp(cur, "lc:", 3)) linked_comm = val;
else if (!strncmp(cur, "of:", 3)) have_old_firmware = val;
else if (!strncmp(cur, "tm:", 3)) tag_mode = val;
else if (!strncmp(cur, "tc:", 3)) tag_mode = val;
else if (!strncmp(cur, "mq:", 3)) max_queue_depth = val;
else if (!strncmp(cur, "ls:", 3)) link_statistics = val;
else if (!strncmp(cur, "et:", 3)) ext_tran = val;
if ((cur = strchr(cur, ','))) ++cur;
}
return;
}
static int option_setup(char *str) {
int ints[MAX_INT_PARAM];
char *cur = str;
int i = 1;
while (cur && isdigit(*cur) && i < MAX_INT_PARAM) {
ints[i++] = simple_strtoul(cur, NULL, 0);
if ((cur = strchr(cur, ',')) != NULL) cur++;
}
ints[0] = i - 1;
internal_setup(cur, ints);
return 1;
}
static int u14_34f_detect(struct scsi_host_template *tpnt) {
unsigned int j = 0, k;
tpnt->proc_name = "u14-34f";
if(strlen(boot_options)) option_setup(boot_options);
#if defined(MODULE)
/* io_port could have been modified when loading as a module */
if(io_port[0] != SKIP) {
setup_done = TRUE;
io_port[MAX_INT_PARAM] = 0;
}
#endif
for (k = 0; k < MAX_BOARDS + 1; k++) sh[k] = NULL;
for (k = 0; io_port[k]; k++) {
if (io_port[k] == SKIP) continue;
if (j < MAX_BOARDS && port_detect(io_port[k], j, tpnt)) j++;
}
num_boards = j;
return j;
}
static void map_dma(unsigned int i, unsigned int j) {
unsigned int data_len = 0;
unsigned int k, pci_dir;
int count;
struct scatterlist *sg;
struct mscp *cpp;
struct scsi_cmnd *SCpnt;
cpp = &HD(j)->cp[i]; SCpnt = cpp->SCpnt;
pci_dir = SCpnt->sc_data_direction;
if (SCpnt->sense_buffer)
cpp->sense_addr = H2DEV(pci_map_single(HD(j)->pdev, SCpnt->sense_buffer,
SCSI_SENSE_BUFFERSIZE, PCI_DMA_FROMDEVICE));
cpp->sense_len = SCSI_SENSE_BUFFERSIZE;
if (scsi_bufflen(SCpnt)) {
count = scsi_dma_map(SCpnt);
BUG_ON(count < 0);
scsi_for_each_sg(SCpnt, sg, count, k) {
cpp->sglist[k].address = H2DEV(sg_dma_address(sg));
cpp->sglist[k].num_bytes = H2DEV(sg_dma_len(sg));
data_len += sg->length;
}
cpp->sg = TRUE;
cpp->use_sg = scsi_sg_count(SCpnt);
cpp->data_address =
H2DEV(pci_map_single(HD(j)->pdev, cpp->sglist,
cpp->use_sg * sizeof(struct sg_list),
pci_dir));
cpp->data_len = H2DEV(data_len);
} else {
pci_dir = PCI_DMA_BIDIRECTIONAL;
cpp->data_len = H2DEV(scsi_bufflen(SCpnt));
}
}
static void unmap_dma(unsigned int i, unsigned int j) {
unsigned int pci_dir;
struct mscp *cpp;
struct scsi_cmnd *SCpnt;
cpp = &HD(j)->cp[i]; SCpnt = cpp->SCpnt;
pci_dir = SCpnt->sc_data_direction;
if (DEV2H(cpp->sense_addr))
pci_unmap_single(HD(j)->pdev, DEV2H(cpp->sense_addr),
DEV2H(cpp->sense_len), PCI_DMA_FROMDEVICE);
scsi_dma_unmap(SCpnt);
if (!DEV2H(cpp->data_len)) pci_dir = PCI_DMA_BIDIRECTIONAL;
if (DEV2H(cpp->data_address))
pci_unmap_single(HD(j)->pdev, DEV2H(cpp->data_address),
DEV2H(cpp->data_len), pci_dir);
}
static void sync_dma(unsigned int i, unsigned int j) {
unsigned int pci_dir;
struct mscp *cpp;
struct scsi_cmnd *SCpnt;
cpp = &HD(j)->cp[i]; SCpnt = cpp->SCpnt;
pci_dir = SCpnt->sc_data_direction;
if (DEV2H(cpp->sense_addr))
pci_dma_sync_single_for_cpu(HD(j)->pdev, DEV2H(cpp->sense_addr),
DEV2H(cpp->sense_len), PCI_DMA_FROMDEVICE);
if (scsi_sg_count(SCpnt))
pci_dma_sync_sg_for_cpu(HD(j)->pdev, scsi_sglist(SCpnt),
scsi_sg_count(SCpnt), pci_dir);
if (!DEV2H(cpp->data_len)) pci_dir = PCI_DMA_BIDIRECTIONAL;
if (DEV2H(cpp->data_address))
pci_dma_sync_single_for_cpu(HD(j)->pdev, DEV2H(cpp->data_address),
DEV2H(cpp->data_len), pci_dir);
}
static void scsi_to_dev_dir(unsigned int i, unsigned int j) {
unsigned int k;
static const unsigned char data_out_cmds[] = {
0x0a, 0x2a, 0x15, 0x55, 0x04, 0x07, 0x18, 0x1d, 0x24, 0x2e,
0x30, 0x31, 0x32, 0x38, 0x39, 0x3a, 0x3b, 0x3d, 0x3f, 0x40,
0x41, 0x4c, 0xaa, 0xae, 0xb0, 0xb1, 0xb2, 0xb6, 0xea, 0x1b, 0x5d
};
static const unsigned char data_none_cmds[] = {
0x01, 0x0b, 0x10, 0x11, 0x13, 0x16, 0x17, 0x19, 0x2b, 0x1e,
0x2c, 0xac, 0x2f, 0xaf, 0x33, 0xb3, 0x35, 0x36, 0x45, 0x47,
0x48, 0x49, 0xa9, 0x4b, 0xa5, 0xa6, 0xb5, 0x00
};
struct mscp *cpp;
struct scsi_cmnd *SCpnt;
cpp = &HD(j)->cp[i]; SCpnt = cpp->SCpnt;
if (SCpnt->sc_data_direction == DMA_FROM_DEVICE) {
cpp->xdir = DTD_IN;
return;
}
else if (SCpnt->sc_data_direction == DMA_TO_DEVICE) {
cpp->xdir = DTD_OUT;
return;
}
else if (SCpnt->sc_data_direction == DMA_NONE) {
cpp->xdir = DTD_NONE;
return;
}
if (SCpnt->sc_data_direction != DMA_BIDIRECTIONAL)
panic("%s: qcomm, invalid SCpnt->sc_data_direction.\n", BN(j));
cpp->xdir = DTD_IN;
for (k = 0; k < ARRAY_SIZE(data_out_cmds); k++)
if (SCpnt->cmnd[0] == data_out_cmds[k]) {
cpp->xdir = DTD_OUT;
break;
}
if (cpp->xdir == DTD_IN)
for (k = 0; k < ARRAY_SIZE(data_none_cmds); k++)
if (SCpnt->cmnd[0] == data_none_cmds[k]) {
cpp->xdir = DTD_NONE;
break;
}
}
static int u14_34f_queuecommand_lck(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *)) {
unsigned int i, j, k;
struct mscp *cpp;
/* j is the board number */
j = ((struct hostdata *) SCpnt->device->host->hostdata)->board_number;
if (SCpnt->host_scribble)
panic("%s: qcomm, SCpnt %p already active.\n",
BN(j), SCpnt);
/* i is the mailbox number, look for the first free mailbox
starting from last_cp_used */
i = HD(j)->last_cp_used + 1;
for (k = 0; k < sh[j]->can_queue; k++, i++) {
if (i >= sh[j]->can_queue) i = 0;
if (HD(j)->cp_stat[i] == FREE) {
HD(j)->last_cp_used = i;
break;
}
}
if (k == sh[j]->can_queue) {
printk("%s: qcomm, no free mailbox.\n", BN(j));
return 1;
}
/* Set pointer to control packet structure */
cpp = &HD(j)->cp[i];
memset(cpp, 0, sizeof(struct mscp) - CP_TAIL_SIZE);
SCpnt->scsi_done = done;
cpp->cpp_index = i;
SCpnt->host_scribble = (unsigned char *) &cpp->cpp_index;
if (do_trace) printk("%s: qcomm, mbox %d, target %d.%d:%d.\n",
BN(j), i, SCpnt->device->channel, SCpnt->device->id,
SCpnt->device->lun);
cpp->opcode = OP_SCSI;
cpp->channel = SCpnt->device->channel;
cpp->target = SCpnt->device->id;
cpp->lun = SCpnt->device->lun;
cpp->SCpnt = SCpnt;
cpp->cdb_len = SCpnt->cmd_len;
memcpy(cpp->cdb, SCpnt->cmnd, SCpnt->cmd_len);
/* Use data transfer direction SCpnt->sc_data_direction */
scsi_to_dev_dir(i, j);
/* Map DMA buffers and SG list */
map_dma(i, j);
if (linked_comm && SCpnt->device->queue_depth > 2
&& TLDEV(SCpnt->device->type)) {
HD(j)->cp_stat[i] = READY;
block: convert to pos and nr_sectors accessors With recent cleanups, there is no place where low level driver directly manipulates request fields. This means that the 'hard' request fields always equal the !hard fields. Convert all rq->sectors, nr_sectors and current_nr_sectors references to accessors. While at it, drop superflous blk_rq_pos() < 0 test in swim.c. [ Impact: use pos and nr_sectors accessors ] Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Geert Uytterhoeven <Geert.Uytterhoeven@sonycom.com> Tested-by: Grant Likely <grant.likely@secretlab.ca> Acked-by: Grant Likely <grant.likely@secretlab.ca> Tested-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Mike Miller <mike.miller@hp.com> Cc: James Bottomley <James.Bottomley@HansenPartnership.com> Cc: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> Cc: Borislav Petkov <petkovbb@googlemail.com> Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com> Cc: Eric Moore <Eric.Moore@lsi.com> Cc: Alan Stern <stern@rowland.harvard.edu> Cc: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Cc: Pete Zaitcev <zaitcev@redhat.com> Cc: Stephen Rothwell <sfr@canb.auug.org.au> Cc: Paul Clements <paul.clements@steeleye.com> Cc: Tim Waugh <tim@cyberelk.net> Cc: Jeff Garzik <jgarzik@pobox.com> Cc: Jeremy Fitzhardinge <jeremy@xensource.com> Cc: Alex Dubov <oakad@yahoo.com> Cc: David Woodhouse <dwmw2@infradead.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Dario Ballabio <ballabio_dario@emc.com> Cc: David S. Miller <davem@davemloft.net> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: unsik Kim <donari75@gmail.com> Cc: Laurent Vivier <Laurent@lvivier.info> Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2009-05-07 13:24:39 +00:00
flush_dev(SCpnt->device, blk_rq_pos(SCpnt->request), j, FALSE);
return 0;
}
if (wait_on_busy(sh[j]->io_port, MAXLOOP)) {
unmap_dma(i, j);
SCpnt->host_scribble = NULL;
scmd_printk(KERN_INFO, SCpnt,
"qcomm, adapter busy.\n");
return 1;
}
/* Store pointer in OGM address bytes */
outl(H2DEV(cpp->cp_dma_addr), sh[j]->io_port + REG_OGM);
/* Issue OGM interrupt */
outb(CMD_OGM_INTR, sh[j]->io_port + REG_LCL_INTR);
HD(j)->cp_stat[i] = IN_USE;
return 0;
}
static DEF_SCSI_QCMD(u14_34f_queuecommand)
static int u14_34f_eh_abort(struct scsi_cmnd *SCarg) {
unsigned int i, j;
j = ((struct hostdata *) SCarg->device->host->hostdata)->board_number;
if (SCarg->host_scribble == NULL) {
scmd_printk(KERN_INFO, SCarg, "abort, command inactive.\n");
return SUCCESS;
}
i = *(unsigned int *)SCarg->host_scribble;
scmd_printk(KERN_INFO, SCarg, "abort, mbox %d.\n", i);
if (i >= sh[j]->can_queue)
panic("%s: abort, invalid SCarg->host_scribble.\n", BN(j));
if (wait_on_busy(sh[j]->io_port, MAXLOOP)) {
printk("%s: abort, timeout error.\n", BN(j));
return FAILED;
}
if (HD(j)->cp_stat[i] == FREE) {
printk("%s: abort, mbox %d is free.\n", BN(j), i);
return SUCCESS;
}
if (HD(j)->cp_stat[i] == IN_USE) {
printk("%s: abort, mbox %d is in use.\n", BN(j), i);
if (SCarg != HD(j)->cp[i].SCpnt)
panic("%s: abort, mbox %d, SCarg %p, cp SCpnt %p.\n",
BN(j), i, SCarg, HD(j)->cp[i].SCpnt);
if (inb(sh[j]->io_port + REG_SYS_INTR) & IRQ_ASSERTED)
printk("%s: abort, mbox %d, interrupt pending.\n", BN(j), i);
return FAILED;
}
if (HD(j)->cp_stat[i] == IN_RESET) {
printk("%s: abort, mbox %d is in reset.\n", BN(j), i);
return FAILED;
}
if (HD(j)->cp_stat[i] == LOCKED) {
printk("%s: abort, mbox %d is locked.\n", BN(j), i);
return SUCCESS;
}
if (HD(j)->cp_stat[i] == READY || HD(j)->cp_stat[i] == ABORTING) {
unmap_dma(i, j);
SCarg->result = DID_ABORT << 16;
SCarg->host_scribble = NULL;
HD(j)->cp_stat[i] = FREE;
printk("%s, abort, mbox %d ready, DID_ABORT, done.\n", BN(j), i);
SCarg->scsi_done(SCarg);
return SUCCESS;
}
panic("%s: abort, mbox %d, invalid cp_stat.\n", BN(j), i);
}
static int u14_34f_eh_host_reset(struct scsi_cmnd *SCarg) {
unsigned int i, j, k, c, limit = 0;
unsigned long time;
int arg_done = FALSE;
struct scsi_cmnd *SCpnt;
j = ((struct hostdata *) SCarg->device->host->hostdata)->board_number;
scmd_printk(KERN_INFO, SCarg, "reset, enter.\n");
spin_lock_irq(sh[j]->host_lock);
if (SCarg->host_scribble == NULL)
printk("%s: reset, inactive.\n", BN(j));
if (HD(j)->in_reset) {
printk("%s: reset, exit, already in reset.\n", BN(j));
spin_unlock_irq(sh[j]->host_lock);
return FAILED;
}
if (wait_on_busy(sh[j]->io_port, MAXLOOP)) {
printk("%s: reset, exit, timeout error.\n", BN(j));
spin_unlock_irq(sh[j]->host_lock);
return FAILED;
}
HD(j)->retries = 0;
for (c = 0; c <= sh[j]->max_channel; c++)
for (k = 0; k < sh[j]->max_id; k++) {
HD(j)->target_redo[k][c] = TRUE;
HD(j)->target_to[k][c] = 0;
}
for (i = 0; i < sh[j]->can_queue; i++) {
if (HD(j)->cp_stat[i] == FREE) continue;
if (HD(j)->cp_stat[i] == LOCKED) {
HD(j)->cp_stat[i] = FREE;
printk("%s: reset, locked mbox %d forced free.\n", BN(j), i);
continue;
}
if (!(SCpnt = HD(j)->cp[i].SCpnt))
panic("%s: reset, mbox %d, SCpnt == NULL.\n", BN(j), i);
if (HD(j)->cp_stat[i] == READY || HD(j)->cp_stat[i] == ABORTING) {
HD(j)->cp_stat[i] = ABORTING;
printk("%s: reset, mbox %d aborting.\n", BN(j), i);
}
else {
HD(j)->cp_stat[i] = IN_RESET;
printk("%s: reset, mbox %d in reset.\n", BN(j), i);
}
if (SCpnt->host_scribble == NULL)
panic("%s: reset, mbox %d, garbled SCpnt.\n", BN(j), i);
if (*(unsigned int *)SCpnt->host_scribble != i)
panic("%s: reset, mbox %d, index mismatch.\n", BN(j), i);
if (SCpnt->scsi_done == NULL)
panic("%s: reset, mbox %d, SCpnt->scsi_done == NULL.\n", BN(j), i);
if (SCpnt == SCarg) arg_done = TRUE;
}
if (wait_on_busy(sh[j]->io_port, MAXLOOP)) {
printk("%s: reset, cannot reset, timeout error.\n", BN(j));
spin_unlock_irq(sh[j]->host_lock);
return FAILED;
}
outb(CMD_RESET, sh[j]->io_port + REG_LCL_INTR);
printk("%s: reset, board reset done, enabling interrupts.\n", BN(j));
#if defined(DEBUG_RESET)
do_trace = TRUE;
#endif
HD(j)->in_reset = TRUE;
spin_unlock_irq(sh[j]->host_lock);
time = jiffies;
while ((jiffies - time) < (10 * HZ) && limit++ < 200000) udelay(100L);
spin_lock_irq(sh[j]->host_lock);
printk("%s: reset, interrupts disabled, loops %d.\n", BN(j), limit);
for (i = 0; i < sh[j]->can_queue; i++) {
if (HD(j)->cp_stat[i] == IN_RESET) {
SCpnt = HD(j)->cp[i].SCpnt;
unmap_dma(i, j);
SCpnt->result = DID_RESET << 16;
SCpnt->host_scribble = NULL;
/* This mailbox is still waiting for its interrupt */
HD(j)->cp_stat[i] = LOCKED;
printk("%s, reset, mbox %d locked, DID_RESET, done.\n", BN(j), i);
}
else if (HD(j)->cp_stat[i] == ABORTING) {
SCpnt = HD(j)->cp[i].SCpnt;
unmap_dma(i, j);
SCpnt->result = DID_RESET << 16;
SCpnt->host_scribble = NULL;
/* This mailbox was never queued to the adapter */
HD(j)->cp_stat[i] = FREE;
printk("%s, reset, mbox %d aborting, DID_RESET, done.\n", BN(j), i);
}
else
/* Any other mailbox has already been set free by interrupt */
continue;
SCpnt->scsi_done(SCpnt);
}
HD(j)->in_reset = FALSE;
do_trace = FALSE;
if (arg_done) printk("%s: reset, exit, done.\n", BN(j));
else printk("%s: reset, exit.\n", BN(j));
spin_unlock_irq(sh[j]->host_lock);
return SUCCESS;
}
static int u14_34f_bios_param(struct scsi_device *disk,
struct block_device *bdev, sector_t capacity, int *dkinfo) {
unsigned int j = 0;
unsigned int size = capacity;
dkinfo[0] = HD(j)->heads;
dkinfo[1] = HD(j)->sectors;
dkinfo[2] = size / (HD(j)->heads * HD(j)->sectors);
if (ext_tran && (scsicam_bios_param(bdev, capacity, dkinfo) < 0)) {
dkinfo[0] = 255;
dkinfo[1] = 63;
dkinfo[2] = size / (dkinfo[0] * dkinfo[1]);
}
#if defined (DEBUG_GEOMETRY)
printk ("%s: bios_param, head=%d, sec=%d, cyl=%d.\n", driver_name,
dkinfo[0], dkinfo[1], dkinfo[2]);
#endif
return FALSE;
}
static void sort(unsigned long sk[], unsigned int da[], unsigned int n,
unsigned int rev) {
unsigned int i, j, k, y;
unsigned long x;
for (i = 0; i < n - 1; i++) {
k = i;
for (j = k + 1; j < n; j++)
if (rev) {
if (sk[j] > sk[k]) k = j;
}
else {
if (sk[j] < sk[k]) k = j;
}
if (k != i) {
x = sk[k]; sk[k] = sk[i]; sk[i] = x;
y = da[k]; da[k] = da[i]; da[i] = y;
}
}
return;
}
static int reorder(unsigned int j, unsigned long cursec,
unsigned int ihdlr, unsigned int il[], unsigned int n_ready) {
struct scsi_cmnd *SCpnt;
struct mscp *cpp;
unsigned int k, n;
unsigned int rev = FALSE, s = TRUE, r = TRUE;
unsigned int input_only = TRUE, overlap = FALSE;
unsigned long sl[n_ready], pl[n_ready], ll[n_ready];
unsigned long maxsec = 0, minsec = ULONG_MAX, seek = 0, iseek = 0;
unsigned long ioseek = 0;
static unsigned int flushcount = 0, batchcount = 0, sortcount = 0;
static unsigned int readycount = 0, ovlcount = 0, inputcount = 0;
static unsigned int readysorted = 0, revcount = 0;
static unsigned long seeksorted = 0, seeknosort = 0;
if (link_statistics && !(++flushcount % link_statistics))
printk("fc %d bc %d ic %d oc %d rc %d rs %d sc %d re %d"\
" av %ldK as %ldK.\n", flushcount, batchcount, inputcount,
ovlcount, readycount, readysorted, sortcount, revcount,
seeknosort / (readycount + 1),
seeksorted / (readycount + 1));
if (n_ready <= 1) return FALSE;
for (n = 0; n < n_ready; n++) {
k = il[n]; cpp = &HD(j)->cp[k]; SCpnt = cpp->SCpnt;
if (!(cpp->xdir == DTD_IN)) input_only = FALSE;
block: convert to pos and nr_sectors accessors With recent cleanups, there is no place where low level driver directly manipulates request fields. This means that the 'hard' request fields always equal the !hard fields. Convert all rq->sectors, nr_sectors and current_nr_sectors references to accessors. While at it, drop superflous blk_rq_pos() < 0 test in swim.c. [ Impact: use pos and nr_sectors accessors ] Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Geert Uytterhoeven <Geert.Uytterhoeven@sonycom.com> Tested-by: Grant Likely <grant.likely@secretlab.ca> Acked-by: Grant Likely <grant.likely@secretlab.ca> Tested-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Mike Miller <mike.miller@hp.com> Cc: James Bottomley <James.Bottomley@HansenPartnership.com> Cc: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> Cc: Borislav Petkov <petkovbb@googlemail.com> Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com> Cc: Eric Moore <Eric.Moore@lsi.com> Cc: Alan Stern <stern@rowland.harvard.edu> Cc: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Cc: Pete Zaitcev <zaitcev@redhat.com> Cc: Stephen Rothwell <sfr@canb.auug.org.au> Cc: Paul Clements <paul.clements@steeleye.com> Cc: Tim Waugh <tim@cyberelk.net> Cc: Jeff Garzik <jgarzik@pobox.com> Cc: Jeremy Fitzhardinge <jeremy@xensource.com> Cc: Alex Dubov <oakad@yahoo.com> Cc: David Woodhouse <dwmw2@infradead.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Dario Ballabio <ballabio_dario@emc.com> Cc: David S. Miller <davem@davemloft.net> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: unsik Kim <donari75@gmail.com> Cc: Laurent Vivier <Laurent@lvivier.info> Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2009-05-07 13:24:39 +00:00
if (blk_rq_pos(SCpnt->request) < minsec)
minsec = blk_rq_pos(SCpnt->request);
if (blk_rq_pos(SCpnt->request) > maxsec)
maxsec = blk_rq_pos(SCpnt->request);
block: convert to pos and nr_sectors accessors With recent cleanups, there is no place where low level driver directly manipulates request fields. This means that the 'hard' request fields always equal the !hard fields. Convert all rq->sectors, nr_sectors and current_nr_sectors references to accessors. While at it, drop superflous blk_rq_pos() < 0 test in swim.c. [ Impact: use pos and nr_sectors accessors ] Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Geert Uytterhoeven <Geert.Uytterhoeven@sonycom.com> Tested-by: Grant Likely <grant.likely@secretlab.ca> Acked-by: Grant Likely <grant.likely@secretlab.ca> Tested-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Mike Miller <mike.miller@hp.com> Cc: James Bottomley <James.Bottomley@HansenPartnership.com> Cc: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> Cc: Borislav Petkov <petkovbb@googlemail.com> Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com> Cc: Eric Moore <Eric.Moore@lsi.com> Cc: Alan Stern <stern@rowland.harvard.edu> Cc: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Cc: Pete Zaitcev <zaitcev@redhat.com> Cc: Stephen Rothwell <sfr@canb.auug.org.au> Cc: Paul Clements <paul.clements@steeleye.com> Cc: Tim Waugh <tim@cyberelk.net> Cc: Jeff Garzik <jgarzik@pobox.com> Cc: Jeremy Fitzhardinge <jeremy@xensource.com> Cc: Alex Dubov <oakad@yahoo.com> Cc: David Woodhouse <dwmw2@infradead.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Dario Ballabio <ballabio_dario@emc.com> Cc: David S. Miller <davem@davemloft.net> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: unsik Kim <donari75@gmail.com> Cc: Laurent Vivier <Laurent@lvivier.info> Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2009-05-07 13:24:39 +00:00
sl[n] = blk_rq_pos(SCpnt->request);
ioseek += blk_rq_sectors(SCpnt->request);
if (!n) continue;
if (sl[n] < sl[n - 1]) s = FALSE;
if (sl[n] > sl[n - 1]) r = FALSE;
if (link_statistics) {
if (sl[n] > sl[n - 1])
seek += sl[n] - sl[n - 1];
else
seek += sl[n - 1] - sl[n];
}
}
if (link_statistics) {
if (cursec > sl[0]) seek += cursec - sl[0]; else seek += sl[0] - cursec;
}
if (cursec > ((maxsec + minsec) / 2)) rev = TRUE;
if (ioseek > ((maxsec - minsec) / 2)) rev = FALSE;
if (!((rev && r) || (!rev && s))) sort(sl, il, n_ready, rev);
if (!input_only) for (n = 0; n < n_ready; n++) {
k = il[n]; cpp = &HD(j)->cp[k]; SCpnt = cpp->SCpnt;
block: convert to pos and nr_sectors accessors With recent cleanups, there is no place where low level driver directly manipulates request fields. This means that the 'hard' request fields always equal the !hard fields. Convert all rq->sectors, nr_sectors and current_nr_sectors references to accessors. While at it, drop superflous blk_rq_pos() < 0 test in swim.c. [ Impact: use pos and nr_sectors accessors ] Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Geert Uytterhoeven <Geert.Uytterhoeven@sonycom.com> Tested-by: Grant Likely <grant.likely@secretlab.ca> Acked-by: Grant Likely <grant.likely@secretlab.ca> Tested-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Mike Miller <mike.miller@hp.com> Cc: James Bottomley <James.Bottomley@HansenPartnership.com> Cc: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> Cc: Borislav Petkov <petkovbb@googlemail.com> Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com> Cc: Eric Moore <Eric.Moore@lsi.com> Cc: Alan Stern <stern@rowland.harvard.edu> Cc: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Cc: Pete Zaitcev <zaitcev@redhat.com> Cc: Stephen Rothwell <sfr@canb.auug.org.au> Cc: Paul Clements <paul.clements@steeleye.com> Cc: Tim Waugh <tim@cyberelk.net> Cc: Jeff Garzik <jgarzik@pobox.com> Cc: Jeremy Fitzhardinge <jeremy@xensource.com> Cc: Alex Dubov <oakad@yahoo.com> Cc: David Woodhouse <dwmw2@infradead.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Dario Ballabio <ballabio_dario@emc.com> Cc: David S. Miller <davem@davemloft.net> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: unsik Kim <donari75@gmail.com> Cc: Laurent Vivier <Laurent@lvivier.info> Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2009-05-07 13:24:39 +00:00
ll[n] = blk_rq_sectors(SCpnt->request); pl[n] = SCpnt->serial_number;
if (!n) continue;
if ((sl[n] == sl[n - 1]) || (!rev && ((sl[n - 1] + ll[n - 1]) > sl[n]))
|| (rev && ((sl[n] + ll[n]) > sl[n - 1]))) overlap = TRUE;
}
if (overlap) sort(pl, il, n_ready, FALSE);
if (link_statistics) {
if (cursec > sl[0]) iseek = cursec - sl[0]; else iseek = sl[0] - cursec;
batchcount++; readycount += n_ready; seeknosort += seek / 1024;
if (input_only) inputcount++;
if (overlap) { ovlcount++; seeksorted += iseek / 1024; }
else seeksorted += (iseek + maxsec - minsec) / 1024;
if (rev && !r) { revcount++; readysorted += n_ready; }
if (!rev && !s) { sortcount++; readysorted += n_ready; }
}
#if defined(DEBUG_LINKED_COMMANDS)
if (link_statistics && (overlap || !(flushcount % link_statistics)))
for (n = 0; n < n_ready; n++) {
k = il[n]; cpp = &HD(j)->cp[k]; SCpnt = cpp->SCpnt;
printk("%s %d.%d:%d mb %d fc %d nr %d sec %ld ns %u"\
" cur %ld s:%c r:%c rev:%c in:%c ov:%c xd %d.\n",
(ihdlr ? "ihdlr" : "qcomm"), SCpnt->channel, SCpnt->target,
SCpnt->lun, k, flushcount, n_ready,
block: convert to pos and nr_sectors accessors With recent cleanups, there is no place where low level driver directly manipulates request fields. This means that the 'hard' request fields always equal the !hard fields. Convert all rq->sectors, nr_sectors and current_nr_sectors references to accessors. While at it, drop superflous blk_rq_pos() < 0 test in swim.c. [ Impact: use pos and nr_sectors accessors ] Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Geert Uytterhoeven <Geert.Uytterhoeven@sonycom.com> Tested-by: Grant Likely <grant.likely@secretlab.ca> Acked-by: Grant Likely <grant.likely@secretlab.ca> Tested-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Mike Miller <mike.miller@hp.com> Cc: James Bottomley <James.Bottomley@HansenPartnership.com> Cc: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> Cc: Borislav Petkov <petkovbb@googlemail.com> Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com> Cc: Eric Moore <Eric.Moore@lsi.com> Cc: Alan Stern <stern@rowland.harvard.edu> Cc: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Cc: Pete Zaitcev <zaitcev@redhat.com> Cc: Stephen Rothwell <sfr@canb.auug.org.au> Cc: Paul Clements <paul.clements@steeleye.com> Cc: Tim Waugh <tim@cyberelk.net> Cc: Jeff Garzik <jgarzik@pobox.com> Cc: Jeremy Fitzhardinge <jeremy@xensource.com> Cc: Alex Dubov <oakad@yahoo.com> Cc: David Woodhouse <dwmw2@infradead.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Dario Ballabio <ballabio_dario@emc.com> Cc: David S. Miller <davem@davemloft.net> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: unsik Kim <donari75@gmail.com> Cc: Laurent Vivier <Laurent@lvivier.info> Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2009-05-07 13:24:39 +00:00
blk_rq_pos(SCpnt->request), blk_rq_sectors(SCpnt->request),
cursec, YESNO(s), YESNO(r), YESNO(rev), YESNO(input_only),
YESNO(overlap), cpp->xdir);
}
#endif
return overlap;
}
static void flush_dev(struct scsi_device *dev, unsigned long cursec, unsigned int j,
unsigned int ihdlr) {
struct scsi_cmnd *SCpnt;
struct mscp *cpp;
unsigned int k, n, n_ready = 0, il[MAX_MAILBOXES];
for (k = 0; k < sh[j]->can_queue; k++) {
if (HD(j)->cp_stat[k] != READY && HD(j)->cp_stat[k] != IN_USE) continue;
cpp = &HD(j)->cp[k]; SCpnt = cpp->SCpnt;
if (SCpnt->device != dev) continue;
if (HD(j)->cp_stat[k] == IN_USE) return;
il[n_ready++] = k;
}
if (reorder(j, cursec, ihdlr, il, n_ready)) n_ready = 1;
for (n = 0; n < n_ready; n++) {
k = il[n]; cpp = &HD(j)->cp[k]; SCpnt = cpp->SCpnt;
if (wait_on_busy(sh[j]->io_port, MAXLOOP)) {
scmd_printk(KERN_INFO, SCpnt,
"%s, mbox %d, adapter"
" busy, will abort.\n", (ihdlr ? "ihdlr" : "qcomm"),
k);
HD(j)->cp_stat[k] = ABORTING;
continue;
}
outl(H2DEV(cpp->cp_dma_addr), sh[j]->io_port + REG_OGM);
outb(CMD_OGM_INTR, sh[j]->io_port + REG_LCL_INTR);
HD(j)->cp_stat[k] = IN_USE;
}
}
static irqreturn_t ihdlr(unsigned int j)
{
struct scsi_cmnd *SCpnt;
unsigned int i, k, c, status, tstatus, reg, ret;
struct mscp *spp, *cpp;
int irq = sh[j]->irq;
/* Check if this board need to be serviced */
if (!((reg = inb(sh[j]->io_port + REG_SYS_INTR)) & IRQ_ASSERTED)) goto none;
HD(j)->iocount++;
if (do_trace) printk("%s: ihdlr, enter, irq %d, count %d.\n", BN(j), irq,
HD(j)->iocount);
/* Check if this board is still busy */
if (wait_on_busy(sh[j]->io_port, 20 * MAXLOOP)) {
outb(CMD_CLR_INTR, sh[j]->io_port + REG_SYS_INTR);
printk("%s: ihdlr, busy timeout error, irq %d, reg 0x%x, count %d.\n",
BN(j), irq, reg, HD(j)->iocount);
goto none;
}
ret = inl(sh[j]->io_port + REG_ICM);
/* Clear interrupt pending flag */
outb(CMD_CLR_INTR, sh[j]->io_port + REG_SYS_INTR);
/* Find the mailbox to be serviced on this board */
for (i = 0; i < sh[j]->can_queue; i++)
if (H2DEV(HD(j)->cp[i].cp_dma_addr) == ret) break;
if (i >= sh[j]->can_queue)
panic("%s: ihdlr, invalid mscp bus address %p, cp0 %p.\n", BN(j),
(void *)ret, (void *)H2DEV(HD(j)->cp[0].cp_dma_addr));
cpp = &(HD(j)->cp[i]);
spp = cpp;
#if defined(DEBUG_GENERATE_ABORTS)
if ((HD(j)->iocount > 500) && ((HD(j)->iocount % 500) < 3)) goto handled;
#endif
if (HD(j)->cp_stat[i] == IGNORE) {
HD(j)->cp_stat[i] = FREE;
goto handled;
}
else if (HD(j)->cp_stat[i] == LOCKED) {
HD(j)->cp_stat[i] = FREE;
printk("%s: ihdlr, mbox %d unlocked, count %d.\n", BN(j), i,
HD(j)->iocount);
goto handled;
}
else if (HD(j)->cp_stat[i] == FREE) {
printk("%s: ihdlr, mbox %d is free, count %d.\n", BN(j), i,
HD(j)->iocount);
goto handled;
}
else if (HD(j)->cp_stat[i] == IN_RESET)
printk("%s: ihdlr, mbox %d is in reset.\n", BN(j), i);
else if (HD(j)->cp_stat[i] != IN_USE)
panic("%s: ihdlr, mbox %d, invalid cp_stat: %d.\n",
BN(j), i, HD(j)->cp_stat[i]);
HD(j)->cp_stat[i] = FREE;
SCpnt = cpp->SCpnt;
if (SCpnt == NULL) panic("%s: ihdlr, mbox %d, SCpnt == NULL.\n", BN(j), i);
if (SCpnt->host_scribble == NULL)
panic("%s: ihdlr, mbox %d, SCpnt %p garbled.\n", BN(j), i,
SCpnt);
if (*(unsigned int *)SCpnt->host_scribble != i)
panic("%s: ihdlr, mbox %d, index mismatch %d.\n",
BN(j), i, *(unsigned int *)SCpnt->host_scribble);
sync_dma(i, j);
if (linked_comm && SCpnt->device->queue_depth > 2
&& TLDEV(SCpnt->device->type))
block: convert to pos and nr_sectors accessors With recent cleanups, there is no place where low level driver directly manipulates request fields. This means that the 'hard' request fields always equal the !hard fields. Convert all rq->sectors, nr_sectors and current_nr_sectors references to accessors. While at it, drop superflous blk_rq_pos() < 0 test in swim.c. [ Impact: use pos and nr_sectors accessors ] Signed-off-by: Tejun Heo <tj@kernel.org> Acked-by: Geert Uytterhoeven <Geert.Uytterhoeven@sonycom.com> Tested-by: Grant Likely <grant.likely@secretlab.ca> Acked-by: Grant Likely <grant.likely@secretlab.ca> Tested-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Adrian McMenamin <adrian@mcmen.demon.co.uk> Acked-by: Mike Miller <mike.miller@hp.com> Cc: James Bottomley <James.Bottomley@HansenPartnership.com> Cc: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com> Cc: Borislav Petkov <petkovbb@googlemail.com> Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com> Cc: Eric Moore <Eric.Moore@lsi.com> Cc: Alan Stern <stern@rowland.harvard.edu> Cc: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Cc: Pete Zaitcev <zaitcev@redhat.com> Cc: Stephen Rothwell <sfr@canb.auug.org.au> Cc: Paul Clements <paul.clements@steeleye.com> Cc: Tim Waugh <tim@cyberelk.net> Cc: Jeff Garzik <jgarzik@pobox.com> Cc: Jeremy Fitzhardinge <jeremy@xensource.com> Cc: Alex Dubov <oakad@yahoo.com> Cc: David Woodhouse <dwmw2@infradead.org> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Dario Ballabio <ballabio_dario@emc.com> Cc: David S. Miller <davem@davemloft.net> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: unsik Kim <donari75@gmail.com> Cc: Laurent Vivier <Laurent@lvivier.info> Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2009-05-07 13:24:39 +00:00
flush_dev(SCpnt->device, blk_rq_pos(SCpnt->request), j, TRUE);
tstatus = status_byte(spp->target_status);
#if defined(DEBUG_GENERATE_ERRORS)
if ((HD(j)->iocount > 500) && ((HD(j)->iocount % 200) < 2))
spp->adapter_status = 0x01;
#endif
switch (spp->adapter_status) {
case ASOK: /* status OK */
/* Forces a reset if a disk drive keeps returning BUSY */
if (tstatus == BUSY && SCpnt->device->type != TYPE_TAPE)
status = DID_ERROR << 16;
/* If there was a bus reset, redo operation on each target */
else if (tstatus != GOOD && SCpnt->device->type == TYPE_DISK
&& HD(j)->target_redo[scmd_id(SCpnt)][scmd_channel(SCpnt)])
status = DID_BUS_BUSY << 16;
/* Works around a flaw in scsi.c */
else if (tstatus == CHECK_CONDITION
&& SCpnt->device->type == TYPE_DISK
&& (SCpnt->sense_buffer[2] & 0xf) == RECOVERED_ERROR)
status = DID_BUS_BUSY << 16;
else
status = DID_OK << 16;
if (tstatus == GOOD)
HD(j)->target_redo[scmd_id(SCpnt)][scmd_channel(SCpnt)] = FALSE;
if (spp->target_status && SCpnt->device->type == TYPE_DISK &&
(!(tstatus == CHECK_CONDITION && HD(j)->iocount <= 1000 &&
(SCpnt->sense_buffer[2] & 0xf) == NOT_READY)))
scmd_printk(KERN_INFO, SCpnt,
"ihdlr, target_status 0x%x, sense key 0x%x.\n",
spp->target_status,
SCpnt->sense_buffer[2]);
HD(j)->target_to[scmd_id(SCpnt)][scmd_channel(SCpnt)] = 0;
if (HD(j)->last_retried_pid == SCpnt->serial_number) HD(j)->retries = 0;
break;
case ASST: /* Selection Time Out */
if (HD(j)->target_to[scmd_id(SCpnt)][scmd_channel(SCpnt)] > 1)
status = DID_ERROR << 16;
else {
status = DID_TIME_OUT << 16;
HD(j)->target_to[scmd_id(SCpnt)][scmd_channel(SCpnt)]++;
}
break;
/* Perform a limited number of internal retries */
case 0x93: /* Unexpected bus free */
case 0x94: /* Target bus phase sequence failure */
case 0x96: /* Illegal SCSI command */
case 0xa3: /* SCSI bus reset error */
for (c = 0; c <= sh[j]->max_channel; c++)
for (k = 0; k < sh[j]->max_id; k++)
HD(j)->target_redo[k][c] = TRUE;
case 0x92: /* Data over/under-run */
if (SCpnt->device->type != TYPE_TAPE
&& HD(j)->retries < MAX_INTERNAL_RETRIES) {
#if defined(DID_SOFT_ERROR)
status = DID_SOFT_ERROR << 16;
#else
status = DID_BUS_BUSY << 16;
#endif
HD(j)->retries++;
HD(j)->last_retried_pid = SCpnt->serial_number;
}
else
status = DID_ERROR << 16;
break;
case 0x01: /* Invalid command */
case 0x02: /* Invalid parameters */
case 0x03: /* Invalid data list */
case 0x84: /* SCSI bus abort error */
case 0x9b: /* Auto request sense error */
case 0x9f: /* Unexpected command complete message error */
case 0xff: /* Invalid parameter in the S/G list */
default:
status = DID_ERROR << 16;
break;
}
SCpnt->result = status | spp->target_status;
#if defined(DEBUG_INTERRUPT)
if (SCpnt->result || do_trace)
#else
if ((spp->adapter_status != ASOK && HD(j)->iocount > 1000) ||
(spp->adapter_status != ASOK &&
spp->adapter_status != ASST && HD(j)->iocount <= 1000) ||
do_trace || msg_byte(spp->target_status))
#endif
scmd_printk(KERN_INFO, SCpnt, "ihdlr, mbox %2d, err 0x%x:%x,"\
" reg 0x%x, count %d.\n",
i, spp->adapter_status, spp->target_status,
reg, HD(j)->iocount);
unmap_dma(i, j);
/* Set the command state to inactive */
SCpnt->host_scribble = NULL;
SCpnt->scsi_done(SCpnt);
if (do_trace) printk("%s: ihdlr, exit, irq %d, count %d.\n", BN(j), irq,
HD(j)->iocount);
handled:
return IRQ_HANDLED;
none:
return IRQ_NONE;
}
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 13:55:46 +00:00
static irqreturn_t do_interrupt_handler(int irq, void *shap) {
unsigned int j;
unsigned long spin_flags;
irqreturn_t ret;
/* Check if the interrupt must be processed by this handler */
if ((j = (unsigned int)((char *)shap - sha)) >= num_boards) return IRQ_NONE;
spin_lock_irqsave(sh[j]->host_lock, spin_flags);
ret = ihdlr(j);
spin_unlock_irqrestore(sh[j]->host_lock, spin_flags);
return ret;
}
static int u14_34f_release(struct Scsi_Host *shpnt) {
unsigned int i, j;
for (j = 0; sh[j] != NULL && sh[j] != shpnt; j++);
if (sh[j] == NULL)
panic("%s: release, invalid Scsi_Host pointer.\n", driver_name);
for (i = 0; i < sh[j]->can_queue; i++)
kfree((&HD(j)->cp[i])->sglist);
for (i = 0; i < sh[j]->can_queue; i++)
pci_unmap_single(HD(j)->pdev, HD(j)->cp[i].cp_dma_addr,
sizeof(struct mscp), PCI_DMA_BIDIRECTIONAL);
free_irq(sh[j]->irq, &sha[j]);
if (sh[j]->dma_channel != NO_DMA)
free_dma(sh[j]->dma_channel);
release_region(sh[j]->io_port, sh[j]->n_io_port);
scsi_unregister(sh[j]);
return FALSE;
}
#include "scsi_module.c"
#ifndef MODULE
__setup("u14-34f=", option_setup);
#endif /* end MODULE */