[GIT PATCH] scsi merge for 2.6.13

[GIT PATCH] scsi merge for 2.6.13

[James Bottomley]

This should be the final round of changes before we close down into bug
fix mode.  There will be a few extra driver updates in some of the -rc
stages (there's at least aacraid and zfcp still going through the
process).

The patch is available here [URL checked this time]

master.kernel.org:/pub/scm/linux/kernel/git/jejb/scsi-for-linus-2.6.git

The short change log is:

adam radford:
  o 3ware 9000: handle use_sg != 0 for emulated commands

Alan Stern:
  o Fix module removal/device add race
  o fix callers of scsi_remove_device() who already hold the scan muted
  o add missing scan mutex to scsi_scan_target()

Andrew Vasquez:
  o lpfc: use wwn_to_u64() transport helper
  o qla2xxx: use wwn_to_u64() transport helper
  o fc_transport: Generalize WWN to u64 interger conversions

Christoph Hellwig:
  o SAS transport class

Douglas Gilbert:
  o permit READ DEFECT DATA in block/scsi_ioctl
  o sg: do not set VM_IO flag on mmap-ed pages

James Bottomley:
  o Alter the scsi_add_device() API to conform to what users expect
  o SAS transport class: fixup prototype of sas_host_setup
  o core: fix leakage of scsi_cmnd's
  o core: fix leakage of scsi_cmnd's

Mike Christie:
  o set error value when failing commands in prep_fn

Neil Brown:
  o fix possible deadlock in scsi_lib.c

Tejun Heo:
  o scsi: Error handler description document

The diffstat is:

 Documentation/scsi/00-INDEX       |    2 
 Documentation/scsi/scsi_eh.txt    |  479 ++++++++++++++++++++++
 drivers/block/scsi_ioctl.c        |    1 
 drivers/ieee1394/sbp2.c           |    8 
 drivers/scsi/3w-9xxx.c            |   30 +
 drivers/scsi/Kconfig              |    7 
 drivers/scsi/Makefile             |    1 
 drivers/scsi/lpfc/lpfc_attr.c     |   22 -
 drivers/scsi/lpfc/lpfc_hbadisc.c  |    7 
 drivers/scsi/lpfc/lpfc_hw.h       |   17 
 drivers/scsi/lpfc/lpfc_init.c     |    7 
 drivers/scsi/qla2xxx/qla_attr.c   |   18 
 drivers/scsi/qla2xxx/qla_init.c   |    4 
 drivers/scsi/scsi_lib.c           |  138 ++++--
 drivers/scsi/scsi_priv.h          |    1 
 drivers/scsi/scsi_scan.c          |   88 ++--
 drivers/scsi/scsi_sysfs.c         |   28 -
 drivers/scsi/scsi_transport_sas.c |  820 ++++++++++++++++++++++++++++++++++++++
 drivers/scsi/sg.c                 |    4 
 include/scsi/scsi_device.h        |    4 
 include/scsi/scsi_transport_fc.h  |    8 
 include/scsi/scsi_transport_sas.h |  100 ++++
 22 files changed, 1648 insertions(+), 146 deletions(-)

James

[GIT PATCH] SCSI merge for 2.6.13

[James Bottomley]

This should be the entire contents of the SCSI tree I've been saving.
It also includes Jens' send SCSI requests via bios tree that he, Mike
Christie and I have been working on.

The patch is available here:

master.kernel.org:/pub/linux/kernel/scm/jejb/scsi-for-linus-2.6.git

The short changelog is:

Adrian Bunk:
  o drivers/scsi/constants.c should include scsi_dbg.h
  o git-scsi-misc: drivers/scsi/ch.c: remove devfs stuff

Alan Stern:
  o sd: pause in sd_spinup_disk for slow USB devices
  o return success after retries in scsi_eh_tur

Andrew Morton:
  o fix C syntax problem in scsi_lib.c
  o fix warning in aic7770.c
  o fix warning in scsi_softirq
  o aic79xx: needs to select SPI_TRANSPORT_ATTRS

Andrew Vasquez:
  o qla2xxx: Update version number to 8.01.00-k
  o qla2xxx: Stop firmware execution at unintialization time
  o qla2xxx: Replace schedule_timeout()
  o qla2xxx: Remove bad call to fc_remove_host() during probe failure
  o qla2xxx: Add host attributes
  o qla2xxx: Add change_queue_depth/type() API support
  o qla2xxx: Remove redundant call to pci_unmap_sg()
  o qla2xxx: Remove RISC pause/release barriers during flash manipulation
  o qla2xxx: Correct LED scheme definition
  o qla2xxx: Simplify redundant target/device reset logic
  o qla2xxx: Correct domain/area exclusion logic
  o qla2xxx: Add FDMI support
  o qla2xxx: Export class-of-service (COS) information
  o qla2xxx: Use dma_get_required_mask() in determining the 'ideal' DMA mask

Anton Blanchard:
  o Universal Xport no attach blacklist

Christoph Hellwig:
  o unexport scsi_add_timer/scsi_delete_timer
  o switch EH thread startup to the kthread API
  o fix SCSI_IOCTL_PROBE_HOST
  o fusion: whitespace fixes
  o fusion: endianess fixes
  o fusion: update LSI headers
  o fusion: extended config header support
  o aic7xxx: remove aiclib.c
  o comment cleanup for spi_execute
  o aiclib remove dead
  o aic79xx: sane pci probing
  o aic79xx: remove some dead code
  o qla1280: endianess annotations
  o qla1280: don't use bitfields for hardware access, parameters
  o qla1280: don't use bitfields for hardware access in isp_config
  o qla1280: always load microcode
  o qla1280: remove SG_SEGMENTS
  o qla1280: use SAM_ constants
  o qla1280: misc cleanups
  o qla1280: interupt posting for irq disabling/enabling
  o qla1280: remove dead per-host flag variables
  o [PATCH] ll_rw_blk.c kerneldoc updates

Dave Boutcher:
  o ibmvscsi timeout fix
  o ibmvscsi eh locking

Dave Jones:
  o blacklist addition

Douglas Gilbert:
  o sg direct io/mmap oops, st sync

Hannes Reinecke:
  o aic79xx: fixup DT setting
  o aic79xx: DV parameter settings
  o aic79xx: update to use scsi_transport_spi
  o aic79xx: Remove busyq

James Bottomley:
  o Merge by hand (conflicts in sd.c)
  o quieten messages on scsi_execute commands
  o ibmvscsi: handle large scatter/gather lists
  o This patch fixes in st.c the bug in the signed/unsigned int comparison
  o embryonic RAID class
  o attribute container final klist fixes
  o correct attribute_container list usage
  o Merge by hand (conflicts in sr.c)
  o fix sense buffer length handling problem
  o fix 3ware raid emulated commands
  o convert ch to use scsi_execute_req
  o convert sr to scsi_execute_req
  o convert sd to scsi_execute_req (and update the scsi_execute_req API)
  o convert SPI transport class to scsi_execute
  o convert the remaining mid-layer pieces to scsi_execute_req
  o Merge HEAD from ../scsi-misc-2.6-tmp
  o add missing attribute container function prototype
  o fix transport class corner case after rework
  o use scatter lists for all block pc requests and simplify hw handlers
  o update scsi_wait_req to new format for blk_rq_map_kern()
  o use scatter lists for all block pc requests and simplify hw handlers
  o fix mismerge in ll_rw_blk.c
  o correct transport class abstraction to work outside SCSI
  o add ability to deny binding to SPI transport class
  o aic7xxx: lost multifunction flags handling
  o Make the HSG80 a REPORTLUN2 device
  o aic79xx: fix boot panic with no hardware
  o add global timeout to the scsi mid-layer
  o aacraid: correct use of cmd->timeout field
  o aic7xxx/79xx: fix another potential panic due to a non existent target
  o aic7xxx: upport all sequencer and core fixes from adaptec version 6.3.9
  o aic79xx: add hold_mcs to the transport parameters
  o add missing hold_mcs parameter to the spi transport class
  o aic79xx: fix up transport settings
  o add template for scsi_host_set_state()
  o [PATCH] update blk_execute_rq to take an at_head parameter
  o [PATCH] The blk_rq_map_user() change missed an update in scsi_ioctl.c
  o [PATCH] Add scatter-gather support for the block layer SG_IO

James Smart:
  o Add Emulex as maintainer of lpfc SCSI driver
  o lpfc driver 8.0.30 : update version to 8.0.30
  o lpfc driver 8.0.30 : convert to use of int_to_scsilun()
  o lpfc driver 8.0.30 : dev_loss and nodev timeouts
  o lpfc driver 8.0.30 : fix get_stats panic
  o lpfc driver 8.0.30 : task mgmt bit clearing
  o lpfc driver 8.0.30 : fix lip/cablepull panic
  o lpfc driver 8.0.30 : fix iocb reuse initialization

Jens Axboe:
  o [PATCH] kill 'reading' variable in sg_io(), it isn't used anymore
  o [PATCH] Cleanup blk_rq_map_* interfaces
  o [PATCH] Keep the bio end_io parts inside of bio.c for blk_rq_map_kern()

Kai Mäkisara:
  o drivers/scsi/st.c: add reference count and related fixes

Mark Haverkamp:
  o aacraid: bad BUG_ON fix
  o aacraid: Fix aacraid probe breakage (updated)
  o aacraid: adapter support update
  o aacraid: sgraw command support
  o aacraid: aif registration timeout fix
  o aacraid: remove duplicate io callback code
  o aacraid: driver shutdown method
  o aacraid: driver version update
  o aacraid: interupt mitigation

Mike Anderson:
  o host state model update: mediate host add/remove race
  o host state model update: reimplement scsi_host_cancel
  o host state model update: replace old host bitmap state

Mike Christie:
  o [PATCH] Add blk_rq_map_kern()

Olaf Hering:
  o aic byteorder fixes after recent cleanup

Pete Zaitcev:
  o sr.c: Fix getting wrong size

And the diffstat:

 Documentation/scsi/aic7xxx.txt                   |    6 
 Documentation/scsi/scsi_mid_low_api.txt          |   41 
 MAINTAINERS                                      |    7 
 drivers/base/attribute_container.c               |   86 
 drivers/base/transport_class.c                   |   19 
 drivers/block/ll_rw_blk.c                        |  192 
 drivers/block/scsi_ioctl.c                       |   60 
 drivers/cdrom/cdrom.c                            |   15 
 drivers/ide/ide-disk.c                           |    2 
 drivers/message/fusion/lsi/mpi.h                 |   19 
 drivers/message/fusion/lsi/mpi_cnfg.h            |   85 
 drivers/message/fusion/lsi/mpi_history.txt       |   67 
 drivers/message/fusion/lsi/mpi_init.h            |  203 
 drivers/message/fusion/lsi/mpi_ioc.h             |   11 
 drivers/message/fusion/lsi/mpi_targ.h            |   74 
 drivers/message/fusion/mptbase.c                 |  321 -
 drivers/message/fusion/mptbase.h                 |    5 
 drivers/message/fusion/mptctl.c                  |   14 
 drivers/message/fusion/mptscsih.c                |  140 
 drivers/message/fusion/mptspi.c                  |    6 
 drivers/scsi/3w-xxxx.c                           |   57 
 drivers/scsi/Kconfig                             |    6 
 drivers/scsi/Makefile                            |    2 
 drivers/scsi/aacraid/aachba.c                    |  327 -
 drivers/scsi/aacraid/aacraid.h                   |   55 
 drivers/scsi/aacraid/commctrl.c                  |   18 
 drivers/scsi/aacraid/comminit.c                  |    4 
 drivers/scsi/aacraid/commsup.c                   |   20 
 drivers/scsi/aacraid/linit.c                     |  106 
 drivers/scsi/aacraid/rkt.c                       |   20 
 drivers/scsi/aacraid/rx.c                        |   20 
 drivers/scsi/aacraid/sa.c                        |   22 
 drivers/scsi/advansys.c                          |    4 
 drivers/scsi/aic7xxx/Kconfig.aic79xx             |    1 
 drivers/scsi/aic7xxx/aic7770.c                   |    1 
 drivers/scsi/aic7xxx/aic79xx.h                   |    6 
 drivers/scsi/aic7xxx/aic79xx_core.c              |  104 
 drivers/scsi/aic7xxx/aic79xx_osm.c               | 5502 ++++++-----------------
 drivers/scsi/aic7xxx/aic79xx_osm.h               |  288 -
 drivers/scsi/aic7xxx/aic79xx_osm_pci.c           |   82 
 drivers/scsi/aic7xxx/aic79xx_pci.c               |   14 
 drivers/scsi/aic7xxx/aic79xx_proc.c              |   88 
 drivers/scsi/aic7xxx/aic7xxx.h                   |    4 
 drivers/scsi/aic7xxx/aic7xxx.reg                 |    4 
 drivers/scsi/aic7xxx/aic7xxx.seq                 |    5 
 drivers/scsi/aic7xxx/aic7xxx_93cx6.c             |   36 
 drivers/scsi/aic7xxx/aic7xxx_core.c              |   60 
 drivers/scsi/aic7xxx/aic7xxx_osm.c               |   97 
 drivers/scsi/aic7xxx/aic7xxx_osm.h               |    2 
 drivers/scsi/aic7xxx/aic7xxx_osm_pci.c           |   29 
 drivers/scsi/aic7xxx/aic7xxx_proc.c              |   45 
 drivers/scsi/aic7xxx/aic7xxx_reg.h_shipped       |    6 
 drivers/scsi/aic7xxx/aic7xxx_reg_print.c_shipped |    4 
 drivers/scsi/aic7xxx/aic7xxx_seq.h_shipped       |  933 +--
 drivers/scsi/aic7xxx/aiclib.c                    | 1377 -----
 drivers/scsi/aic7xxx/aiclib.h                    |  890 ---
 drivers/scsi/ch.c                                |   42 
 drivers/scsi/constants.c                         |   49 
 drivers/scsi/hosts.c                             |  113 
 drivers/scsi/ibmvscsi/ibmvscsi.c                 |  179 
 drivers/scsi/ibmvscsi/ibmvscsi.h                 |    2 
 drivers/scsi/lpfc/lpfc.h                         |    5 
 drivers/scsi/lpfc/lpfc_attr.c                    |    6 
 drivers/scsi/lpfc/lpfc_ct.c                      |    2 
 drivers/scsi/lpfc/lpfc_els.c                     |    1 
 drivers/scsi/lpfc/lpfc_hbadisc.c                 |    3 
 drivers/scsi/lpfc/lpfc_init.c                    |    5 
 drivers/scsi/lpfc/lpfc_mbox.c                    |    5 
 drivers/scsi/lpfc/lpfc_mem.c                     |    5 
 drivers/scsi/lpfc/lpfc_nportdisc.c               |    1 
 drivers/scsi/lpfc/lpfc_scsi.c                    |   24 
 drivers/scsi/lpfc/lpfc_scsi.h                    |   13 
 drivers/scsi/lpfc/lpfc_sli.c                     |    2 
 drivers/scsi/lpfc/lpfc_version.h                 |    2 
 drivers/scsi/qla1280.c                           |  359 -
 drivers/scsi/qla1280.h                           |  336 -
 drivers/scsi/qla2xxx/qla_attr.c                  |  136 
 drivers/scsi/qla2xxx/qla_dbg.h                   |    7 
 drivers/scsi/qla2xxx/qla_def.h                   |  157 
 drivers/scsi/qla2xxx/qla_gbl.h                   |   12 
 drivers/scsi/qla2xxx/qla_gs.c                    |  564 ++
 drivers/scsi/qla2xxx/qla_init.c                  |   24 
 drivers/scsi/qla2xxx/qla_iocb.c                  |    6 
 drivers/scsi/qla2xxx/qla_isr.c                   |    2 
 drivers/scsi/qla2xxx/qla_mbx.c                   |   97 
 drivers/scsi/qla2xxx/qla_os.c                    |  117 
 drivers/scsi/qla2xxx/qla_sup.c                   |   34 
 drivers/scsi/qla2xxx/qla_version.h               |    4 
 drivers/scsi/raid_class.c                        |  250 +
 drivers/scsi/scsi.c                              |   17 
 drivers/scsi/scsi_devinfo.c                      |    4 
 drivers/scsi/scsi_error.c                        |   48 
 drivers/scsi/scsi_ioctl.c                        |   64 
 drivers/scsi/scsi_lib.c                          |  294 -
 drivers/scsi/scsi_priv.h                         |    3 
 drivers/scsi/scsi_scan.c                         |  128 
 drivers/scsi/scsi_sysfs.c                        |   62 
 drivers/scsi/scsi_transport_fc.c                 |    6 
 drivers/scsi/scsi_transport_spi.c                |  168 
 drivers/scsi/sd.c                                |  187 
 drivers/scsi/sg.c                                |   11 
 drivers/scsi/sr.c                                |   75 
 drivers/scsi/sr.h                                |    1 
 drivers/scsi/sr_ioctl.c                          |   62 
 drivers/scsi/st.c                                |  156 
 drivers/scsi/st.h                                |    3 
 fs/bio.c                                         |  223 
 include/linux/attribute_container.h              |   12 
 include/linux/bio.h                              |    6 
 include/linux/blkdev.h                           |   10 
 include/linux/raid_class.h                       |   59 
 include/linux/transport_class.h                  |   11 
 include/scsi/scsi_cmnd.h                         |    8 
 include/scsi/scsi_dbg.h                          |    2 
 include/scsi/scsi_device.h                       |   22 
 include/scsi/scsi_eh.h                           |   11 
 include/scsi/scsi_host.h                         |   26 
 include/scsi/scsi_request.h                      |   16 
 include/scsi/scsi_transport_spi.h                |    6 
 119 files changed, 6830 insertions(+), 9047 deletions(-)

James

Re: [GIT PATCH] SCSI merge for 2.6.13

[Patrick Mansfield]

On Tue, Sep 06, 2005 at 07:37:32PM -0500, James Bottomley wrote:
> This should be the entire contents of the SCSI tree I've been saving.
> It also includes Jens' send SCSI requests via bios tree that he, Mike
> Christie and I have been working on.
> 
> The patch is available here:
> 
> master.kernel.org:/pub/linux/kernel/scm/jejb/scsi-for-linus-2.6.git
> 

>   o convert ch to use scsi_execute_req
>   o convert sr to scsi_execute_req
>   o convert sd to scsi_execute_req (and update the scsi_execute_req API)
>   o convert SPI transport class to scsi_execute
>   o convert the remaining mid-layer pieces to scsi_execute_req

The scsi_execute() retries argument is still not used.

How is this going to work?

For example, multiple unit attentions (power on / reset) during scanning.
We send REPORT LUN, READ CAPACITY, etc., and would not retry if we got a
unit attention.

-- Patrick Mansfield

Re: [GIT PATCH] SCSI merge for 2.6.13

[James Bottomley]

On Wed, 2005-09-07 at 10:47 -0700, Patrick Mansfield wrote:
> The scsi_execute() retries argument is still not used.

Yes, Jens was thinking about how to add this to the block layer.

> How is this going to work?
> 
> For example, multiple unit attentions (power on / reset) during scanning.
> We send REPORT LUN, READ CAPACITY, etc., and would not retry if we got a
> unit attention.

Well ... all of the places we were specifying retries mostly had three
of them, so the scsi_request_fn() actually hard codes three retries.  A
hack, but it works until the real method is in place.

James

Re: [GIT PATCH] SCSI merge for 2.6.13

[Jens Axboe]

On Wed, Sep 07 2005, James Bottomley wrote:
> On Wed, 2005-09-07 at 10:47 -0700, Patrick Mansfield wrote:
> > The scsi_execute() retries argument is still not used.
> 
> Yes, Jens was thinking about how to add this to the block layer.

Yeah we probably have to, FAILFAST just doesn't cover all the needed
cases.

-- 
Jens Axboe

Re: [GIT PATCH] SCSI merge for 2.6.13

[Linus Torvalds]

On Tue, 6 Sep 2005, James Bottomley wrote:
> 
> master.kernel.org:/pub/linux/kernel/scm/jejb/scsi-for-linus-2.6.git

No it isn't.

You meant 

  master.kernel.org:/pub/scm/linux/kernel/git/jejb/scsi-for-linus-2.6.git

I do believe,

		Linus

Re: [GIT PATCH] SCSI merge for 2.6.13

[James Bottomley]

On Wed, 2005-09-07 at 17:31 -0700, Linus Torvalds wrote:
> On Tue, 6 Sep 2005, James Bottomley wrote:
> > 
> > master.kernel.org:/pub/linux/kernel/scm/jejb/scsi-for-linus-2.6.git
> 
> No it isn't.
> 
> You meant 
> 
>   master.kernel.org:/pub/scm/linux/kernel/git/jejb/scsi-for-linus-2.6.git
> 
> I do believe,

It's a fair cop guv'nor.

James

Re: [GIT PATCH] SCSI merge for 2.6.13

[Linus Torvalds]

On Wed, 7 Sep 2005, Linus Torvalds wrote:
>
> You meant 
> 
>   master.kernel.org:/pub/scm/linux/kernel/git/jejb/scsi-for-linus-2.6.git
> 
> I do believe,

And also, this fails to compile due to the klist problem.

Quite frankly, what's the point in asking people to pull a tree that is 
known to not compile?

Now I made the mistake of pushing the thing out before I realized that the 
thing that you asked me to pull was the same thing that was known to be 
incomplete and non-working. That just makes me irritated.

		Linus

Re: [GIT PATCH] SCSI merge for 2.6.13

[Linus Torvalds]

On Wed, 7 Sep 2005, Linus Torvalds wrote:
> 
> Quite frankly, what's the point in asking people to pull a tree that is 
> known to not compile?

Btw, I see the patch that is supposed to fix it, but I'm in no position to
know whether it's even acceptable to basically double the size of the
"struct klist", for example. There may be a good reason why Greg hasn't 
been merging the klist stuff, and just assuming that they are merged not 
only screws up everybody down-stream, it's not necessarily valid in the 
first place.

In other words, I think I will have to just revert the commit that
introduces this bogus "assume a patch that wasn't merged" (commit ID
2b7d6a8cb9718fc1d9e826201b64909c44a915f4) for now.

And once more strongly complain about it getting sent to me in the first
place since it was known to not even compile.

		Linus

Re: [GIT PATCH] SCSI merge for 2.6.13

[James Bottomley]

On Wed, 2005-09-07 at 17:49 -0700, Linus Torvalds wrote:
> > Quite frankly, what's the point in asking people to pull a tree that is 
> > known to not compile?
> 
> Btw, I see the patch that is supposed to fix it, but I'm in no position to
> know whether it's even acceptable to basically double the size of the
> "struct klist", for example. There may be a good reason why Greg hasn't 
> been merging the klist stuff, and just assuming that they are merged not 
> only screws up everybody down-stream, it's not necessarily valid in the 
> first place.
> 
> In other words, I think I will have to just revert the commit that
> introduces this bogus "assume a patch that wasn't merged" (commit ID
> 2b7d6a8cb9718fc1d9e826201b64909c44a915f4) for now.
> 
> And once more strongly complain about it getting sent to me in the first
> place since it was known to not even compile.

He's been on holiday, but he did send me a sign off for that particular
patch so I could put it through the SCSI tree.  However, because Andrew
sent you the patch before I could do this, there didn't seem to be any
necessity ...

James

Re: [GIT PATCH] SCSI merge for 2.6.13

[Andrew Morton]

Linus Torvalds <torvalds@osdl.org> wrote:
>
> 
> 
> On Wed, 7 Sep 2005, Linus Torvalds wrote:
> > 
> > Quite frankly, what's the point in asking people to pull a tree that is 
> > known to not compile?

James was assuming you'd merged the klist patch whcih I'd sent beforehand.

> Btw, I see the patch that is supposed to fix it, but I'm in no position to
> know whether it's even acceptable to basically double the size of the
> "struct klist", for example. There may be a good reason why Greg hasn't 
> been merging the klist stuff,

This patch came out around the time Greg went walkabout.

I'd suggest we merge the klist patch anyway, so it doesn't hold up scsi. 
If Greg wants to fix that bug by other means later on, we can do that
anytime.

Re: [GIT PATCH] SCSI merge for 2.6.13

[James Bottomley]

On Wed, 2005-09-07 at 17:41 -0700, Linus Torvalds wrote:
> And also, this fails to compile due to the klist problem.
> 
> Quite frankly, what's the point in asking people to pull a tree that is 
> known to not compile?
> 
> Now I made the mistake of pushing the thing out before I realized that the 
> thing that you asked me to pull was the same thing that was known to be 
> incomplete and non-working. That just makes me irritated.

That would be why I told you the tree was still waiting for the klist
fix, and why I waited to request a pull until I saw Andrew send it to
you.

James