[GIT PULL] target: Updates for v3.2-rc1 (round one)

[GIT PULL] target: Updates for v3.2-rc1 (round one)

[Nicholas A. Bellinger]

Hi Linus,

The following is the first round of target updates for v3.2-rc1.  It's
available directly from kernel.org here:

  git://git.kernel.org/pub/scm/linux/kernel/git/nab/target-pending.git for-next

What can I say...?  Christoph really outdid himself this round with a
long list of cleanups culminating with the conversion to push the core
completion path out of transport_processing_thread() and into a separate
target workqueue.   There are still a number of performance improvements
in the works, most of which at this point will end up being v3.3 items.
Also included are a handful of bugfixes reported by various folks that
are marked for stable@kernel.org, and other various improvements +
removal of legacy code.

The bulk of this series has been posted for review in the last weeks
here:

[PATCH 00/37] target: Updates for v3.2-rc1 (round one)
http://marc.info/?l=linux-scsi&m=131855299419466&w=2

Unfortunately this code has only gotten minimal linux-next testing as
the target-pending.git tree did not return to kernel.org until last
week.  I've been able to verify it does not conflict with linux-next
from 20111014, and has been getting some testing in lio-core.git and I'm
content enough for an merge into v3.2-rc1.

For the second round of v3.2-rc1 changes, a new ib_srpt module will be
merged once Roland is happy with the recent review cleanups.  There is
still some outstanding non-critical breakage with last rounds conversion
to use errno.h in certain areas that still needs to be addressed, as
well as a active I/O shutdown refactoring that I’m also targeting for
v3.2-rc1.

Thank you,

--nab

Andy Shevchenko (3):
  target: simplify target_parse_naa_6h_vendor_specific()
  iscsi-target: use native hex2bin for chap_string_to_hex
  tcm_fc: remove custom hex_to_bin in ft_parse_wwn

Christoph Hellwig (39):
  tfm_fc: use transport_handle_cdb_direct
  iscsi-target: always call transport_handle_cdb_direct
  target: remove transport_generic_handle_cdb
  target: don't opencode transport_release_cmd in
    transport_release_fe_cmd
  target: remove transport_free_se_cmd
  target: simplify transport_generic_remove
  target: simplify transport_put_cmd
  target: remove transport_generic_remove
  target: push session reinstatement out of transport_generic_free_cmd
  target: cleanup iblock bio submission
  target: cleanup pscsi request submission
  target: remove unused se_subsystem_api methods
  target: Cleanup unused target_core_base.h bits
  target: Cleanup unused se_task bits
  target: make more use of the task_flags field in se_task
  target: Remove unnecessary se_task members
  target: pack struct se_task more tightly
  target: make the ->get_cdb method optional
  target: replace ->get_cdb with a target_get_task_cdb helper
  target: remove the ->transport_split_cdb callback in se_cmd
  target: make iblock_emulate_sync_cache asynchronous
  target: clean up the backend interface to caching parameters
  target: remove the transport_qf_callback se_cmd callback
  target: remove SCF_EMULATE_QUEUE_FULL
  target: do not pass the queue object to
    transport_remove_cmd_from_queue
  target: use transport_cmd_check_stop_to_fabric consistently
  target: fix list walking in transport_free_dev_tasks
  target: factor some duplicate code for stopping a task
  target: remove TF_TIMER_STOP
  target: stop task timers earlier
  target: move depth_left manipulation out of
    transport_generic_request_failure
  target: remove the TRANSPORT_REMOVE state
  target: remove TRANSPORT_DEFERRED_CMD state
  target: remove unused TRANSPORT_ states
  target: use a workqueue for I/O completions
  target: remove the task_sg_bidi field se_task and pSCSI BIDI support
  target: merge transport_new_cmd_obj into transport_generic_new_cmd
  target: remove transport_allocate_tasks
  target: re-use the command S/G list for single-task commands

Dan Carpenter (1):
  target: Make pscsi_create_virtdevice use ERR_CAST

Jesper Juhl (1):
  target: Remove unneeded version.h includes
Nicholas Bellinger (16):
  target: Re-org of core_tmr_lun_reset
  target: Prevent TRANSPORT_FREE_CMD_INTR processing in
    core_tmr_drain_cmd_list
  target: Fix transport_cmd_finish_abort queue removal bug
  target: Prevent transport_send_task_abort when CHECK_CONDITION status
  target: Remove session_reinstatement parameter from
    ->transport_wait_for_tasks
  target: Convert ->transport_wait_for_tasks usage to
    transport_generic_free_cmd
  iscsi-target: Remove SCF_SE_LUN_CMD flag abuses
  target: Merge transport_cmd_finish_abort_tmr into
    transport_cmd_finish_abort
  target: Remove legacy + unused device active I/O shutdown code
  loopback: Prevent uninitialized use of tl_tpg in
    tcm_loop_queuecommand
  target: Fix REPORT TARGET PORT GROUPS handling with small allocation
    length
  target: transport_subsystem_check_init cleanups
  target: Fix BIDI t_task_cdb handling in transport_generic_new_cmd
  target: Fix incorrect transport_sent usage
  target: Remove legacy se_task->task_timer and associated logic
  target: Fix compile warning w/ missing module.h include

Roland Dreier (2):
  target: Prevent cmd->se_queue_node double add
  target: Have core_tmr_alloc_req() take an explicit GFP_xxx flag

 drivers/target/Makefile                      |    1 -
 drivers/target/iscsi/iscsi_target.c          |   39 +-
 drivers/target/iscsi/iscsi_target_auth.c     |   34 +-
 drivers/target/iscsi/iscsi_target_core.h     |    4 +-
 drivers/target/iscsi/iscsi_target_erl2.c     |   49 +-
 drivers/target/iscsi/iscsi_target_tmr.c      |    6 +-
 drivers/target/iscsi/iscsi_target_util.c     |   20 +-
 drivers/target/iscsi/iscsi_target_util.h     |    1 +
 drivers/target/loopback/tcm_loop.c           |   18 +-
 drivers/target/target_core_alua.c            |   21 +-
 drivers/target/target_core_cdb.c             |   79 ++-
 drivers/target/target_core_configfs.c        |   16 +-
 drivers/target/target_core_device.c          |   70 +-
 drivers/target/target_core_fabric_configfs.c |    1 -
 drivers/target/target_core_file.c            |   56 +-
 drivers/target/target_core_file.h            |    4 +-
 drivers/target/target_core_iblock.c          |  274 ++----
 drivers/target/target_core_iblock.h          |    2 -
 drivers/target/target_core_pr.c              |    1 -
 drivers/target/target_core_pscsi.c           |  261 ++----
 drivers/target/target_core_pscsi.h           |    1 -
 drivers/target/target_core_rd.c              |   19 +-
 drivers/target/target_core_rd.h              |    2 -
 drivers/target/target_core_scdb.c            |  105 --
 drivers/target/target_core_scdb.h            |   10 -
 drivers/target/target_core_stat.c            |    1 -
 drivers/target/target_core_tmr.c             |  251 +++--
 drivers/target/target_core_transport.c       | 1477 ++++++++------------------
 drivers/target/target_core_ua.c              |    1 -
 drivers/target/tcm_fc/tfc_cmd.c              |   18 +-
 drivers/target/tcm_fc/tfc_conf.c             |   13 +-
 drivers/target/tcm_fc/tfc_io.c               |    1 -
 drivers/target/tcm_fc/tfc_sess.c             |    1 -
 include/target/target_core_base.h            |   71 +-
 include/target/target_core_tmr.h             |    2 +-
 include/target/target_core_transport.h       |   75 +--
 36 files changed, 937 insertions(+), 2068 deletions(-)
 delete mode 100644 drivers/target/target_core_scdb.c
 delete mode 100644 drivers/target/target_core_scdb.h

Re: [GIT PULL] target: Updates for v3.2-rc1 (round one)

[Linus Torvalds]

On Tue, Oct 25, 2011 at 11:00 AM, Nicholas A. Bellinger
<nab@linux-iscsi.org> wrote:
>
>  36 files changed, 937 insertions(+), 2068 deletions(-)

Hmm. This is not what I get. It might be a git version issue or diff
algorithm one - although that's actually pretty unusual. Do you do
something odd/special?

             Linus

Re: [GIT PULL] target: Updates for v3.2-rc1 (round one)

[Christoph Hellwig]

On Tue, Oct 25, 2011 at 11:29:39AM +0200, Linus Torvalds wrote:
> On Tue, Oct 25, 2011 at 11:00 AM, Nicholas A. Bellinger
> <nab@linux-iscsi.org> wrote:
> >
> > ?36 files changed, 937 insertions(+), 2068 deletions(-)
> 
> Hmm. This is not what I get. It might be a git version issue or diff
> algorithm one - although that's actually pretty unusual. Do you do
> something odd/special?

It's mostly me looking at the code and factoring code and removing
unneeded things from old versions.

Re: [GIT PULL] target: Updates for v3.2-rc1 (round one)

[Linus Torvalds]

On Tue, Oct 25, 2011 at 11:29 AM, Linus Torvalds
<torvalds@linux-foundation.org> wrote:
>
> Hmm. This is not what I get. It might be a git version issue or diff
> algorithm one - although that's actually pretty unusual. Do you do
> something odd/special?

Another question: you picked a very particular and odd place to start
development. It's commit 65112dccf8a ("Merge
git://git.samba.org/sfrench/cifs-2.6") which just *seems* to be some
"random Linux tree kernel of the day". I'm not seeing why you picked
that one.

This is not a deal-breaker, and *please* don't use this as an argument
for rebasing (quite the opposite), but it's basically a note to ask
you to just pick a better starting point next time. It's usually much
better to start from an actual release (preferably a *real* release,
like 3.0) than from a "random daily tree". Of course, using an -rc1
release is usually a really bad idea also, but in the rough timeframe
we're talking about (October 10 - two weeks ago - that's apparently
when you started worrying about this next merge window) you could have
at least picked something like -rc9 as the base.

So basically, I'm asking that people not rebase, but also spend a bit
of time thinking about "where should I start development".

                      Linus

Re: [GIT PULL] target: Updates for v3.2-rc1 (round one)

[Nicholas A. Bellinger]

On Tue, 2011-10-25 at 11:29 +0200, Linus Torvalds wrote:
> On Tue, Oct 25, 2011 at 11:00 AM, Nicholas A. Bellinger
> <nab@linux-iscsi.org> wrote:
> >
> >  36 files changed, 937 insertions(+), 2068 deletions(-)
> 
> Hmm. This is not what I get. It might be a git version issue or diff
> algorithm one - although that's actually pretty unusual. Do you do
> something odd/special?
> 

I don't believe so.  The diff-stat was generated using git-format-patch
from target-pending.git with git 1.5.6.5 from the following HEAD:

commit 65112dccf8a113737684366349d7f9ec373ddc47
Merge: bbf5e97... 9d1e397...
Author: Linus Torvalds <torvalds@linux-foundation.org>
Date:   Mon Oct 10 14:53:11 2011 +1200

    Merge git://git.samba.org/sfrench/cifs-2.6

After pulling into a linux-next tree from 20111014, this does appear
different slightly different for some reason..?

 30 files changed, 937 insertions(+), 2052 deletions(-)

I'm certain it's from the same commit that starts the series here: 

commit 9375b1bfd2555c8bc828d394a4419a212b46ba71
Author: Jesper Juhl <jj@chaosbits.net>
Date:   Mon Aug 1 23:29:11 2011 +0200

    target: Remove unneeded version.h includes

--nab

Re: [GIT PULL] target: Updates for v3.2-rc1 (round one)

[Linus Torvalds]

On Tue, Oct 25, 2011 at 11:42 AM, Christoph Hellwig <hch@lst.de> wrote:
> On Tue, Oct 25, 2011 at 11:29:39AM +0200, Linus Torvalds wrote:
>> On Tue, Oct 25, 2011 at 11:00 AM, Nicholas A. Bellinger
>> <nab@linux-iscsi.org> wrote:
>> >
>> > ?36 files changed, 937 insertions(+), 2068 deletions(-)
>>
>> Hmm. This is not what I get. It might be a git version issue or diff
>> algorithm one - although that's actually pretty unusual. Do you do
>> something odd/special?
>
> It's mostly me looking at the code and factoring code and removing
> unneeded things from old versions.

No, I'm saying that the diffstat doesn't match what I get when I pull.

I get

 36 files changed, 955 insertions(+), 2086 deletions(-)

which is *close*, but not identical. I'm wondering where the numbers
Nicholas has came from..

             Linus

Re: [GIT PULL] target: Updates for v3.2-rc1 (round one)

[Linus Torvalds]

On Tue, Oct 25, 2011 at 11:45 AM, Nicholas A. Bellinger
<nab@linux-iscsi.org> wrote:
>
> I don't believe so.  The diff-stat was generated using git-format-patch
> from target-pending.git with git 1.5.6.5 from the following HEAD:

It might just be that you apparently have a rather old git version.
I'm rather more current, and some of the xdiff options have changed (I
think that 1.5.6 had the old "generate minimal diff" that got disabled
because it was very slow for some cases)

I am running git version 1.7.6.4.

             Linus

Re: [GIT PULL] target: Updates for v3.2-rc1 (round one)

[Nicholas A. Bellinger]

On Tue, 2011-10-25 at 11:44 +0200, Linus Torvalds wrote:
> On Tue, Oct 25, 2011 at 11:29 AM, Linus Torvalds
> <torvalds@linux-foundation.org> wrote:
> >
> > Hmm. This is not what I get. It might be a git version issue or diff
> > algorithm one - although that's actually pretty unusual. Do you do
> > something odd/special?
> 
> Another question: you picked a very particular and odd place to start
> development. It's commit 65112dccf8a ("Merge
> git://git.samba.org/sfrench/cifs-2.6") which just *seems* to be some
> "random Linux tree kernel of the day". I'm not seeing why you picked
> that one.
> 
> This is not a deal-breaker, and *please* don't use this as an argument
> for rebasing (quite the opposite), but it's basically a note to ask
> you to just pick a better starting point next time. It's usually much
> better to start from an actual release (preferably a *real* release,
> like 3.0) than from a "random daily tree". Of course, using an -rc1
> release is usually a really bad idea also, but in the rough timeframe
> we're talking about (October 10 - two weeks ago - that's apparently
> when you started worrying about this next merge window) you could have
> at least picked something like -rc9 as the base.
> 
> So basically, I'm asking that people not rebase, but also spend a bit
> of time thinking about "where should I start development".
> 

Point understood.   I'll make sure to pick a tagged -rc release for the
target-pending.git base next time.

Thank you,

--nab

Re: [GIT PULL] target: Updates for v3.2-rc1 (round one)

[Nicholas A. Bellinger]

On Tue, 2011-10-25 at 11:48 +0200, Linus Torvalds wrote:
> On Tue, Oct 25, 2011 at 11:45 AM, Nicholas A. Bellinger
> <nab@linux-iscsi.org> wrote:
> >
> > I don't believe so.  The diff-stat was generated using git-format-patch
> > from target-pending.git with git 1.5.6.5 from the following HEAD:
> 
> It might just be that you apparently have a rather old git version.
> I'm rather more current, and some of the xdiff options have changed (I
> think that 1.5.6 had the old "generate minimal diff" that got disabled
> because it was very slow for some cases)
> 
> I am running git version 1.7.6.4.
> 

That appears to be it, as the gateway I'm pushing to target-pending.git
is running this rather old version of git.  I'm happy to verify this
again from another machine with a more modern version of git, and send a
a followup PULL request later today if you're more comfortable with
that.

Thanks,

--nab

Re: [GIT PULL] target: Updates for v3.2-rc1 (round one)

[Nicholas A. Bellinger]

On Tue, 2011-10-25 at 11:45 +0200, Linus Torvalds wrote:
> On Tue, Oct 25, 2011 at 11:42 AM, Christoph Hellwig <hch@lst.de> wrote:
> > On Tue, Oct 25, 2011 at 11:29:39AM +0200, Linus Torvalds wrote:
> >> On Tue, Oct 25, 2011 at 11:00 AM, Nicholas A. Bellinger
> >> <nab@linux-iscsi.org> wrote:
> >> >
> >> > ?36 files changed, 937 insertions(+), 2068 deletions(-)
> >>
> >> Hmm. This is not what I get. It might be a git version issue or diff
> >> algorithm one - although that's actually pretty unusual. Do you do
> >> something odd/special?
> >
> > It's mostly me looking at the code and factoring code and removing
> > unneeded things from old versions.
> 
> No, I'm saying that the diffstat doesn't match what I get when I pull.
> 
> I get
> 
>  36 files changed, 955 insertions(+), 2086 deletions(-)
> 
> which is *close*, but not identical. I'm wondering where the numbers
> Nicholas has came from..
> 

Ok, just verified with git 1.7.2.3 from target-pending.git and the
numbers match up with your pull.   Here is the full diff-stat for
reference.  My apologies for the extra noise.

--nab

 drivers/target/Makefile                      |    1 -
 drivers/target/iscsi/iscsi_target.c          |   39 +-
 drivers/target/iscsi/iscsi_target_auth.c     |   34 +-
 drivers/target/iscsi/iscsi_target_core.h     |    4 +-
 drivers/target/iscsi/iscsi_target_erl2.c     |   49 +-
 drivers/target/iscsi/iscsi_target_tmr.c      |    6 +-
 drivers/target/iscsi/iscsi_target_util.c     |   20 +-
 drivers/target/iscsi/iscsi_target_util.h     |    1 +
 drivers/target/loopback/tcm_loop.c           |   18 +-
 drivers/target/target_core_alua.c            |   21 +-
 drivers/target/target_core_cdb.c             |   79 ++-
 drivers/target/target_core_configfs.c        |   16 +-
 drivers/target/target_core_device.c          |   70 +-
 drivers/target/target_core_fabric_configfs.c |    1 -
 drivers/target/target_core_file.c            |   56 +-
 drivers/target/target_core_file.h            |    4 +-
 drivers/target/target_core_iblock.c          |  274 ++----
 drivers/target/target_core_iblock.h          |    2 -
 drivers/target/target_core_pr.c              |    1 -
 drivers/target/target_core_pscsi.c           |  261 ++----
 drivers/target/target_core_pscsi.h           |    1 -
 drivers/target/target_core_rd.c              |   19 +-
 drivers/target/target_core_rd.h              |    2 -
 drivers/target/target_core_scdb.c            |  105 --
 drivers/target/target_core_scdb.h            |   10 -
 drivers/target/target_core_stat.c            |    1 -
 drivers/target/target_core_tmr.c             |  251 +++--
 drivers/target/target_core_transport.c       | 1513 ++++++++------------------
 drivers/target/target_core_ua.c              |    1 -
 drivers/target/tcm_fc/tfc_cmd.c              |   18 +-
 drivers/target/tcm_fc/tfc_conf.c             |   13 +-
 drivers/target/tcm_fc/tfc_io.c               |    1 -
 drivers/target/tcm_fc/tfc_sess.c             |    1 -
 include/target/target_core_base.h            |   71 +-
 include/target/target_core_tmr.h             |    2 +-
 include/target/target_core_transport.h       |   75 +-
 36 files changed, 955 insertions(+), 2086 deletions(-)
 delete mode 100644 drivers/target/target_core_scdb.c
 delete mode 100644 drivers/target/target_core_scdb.h