[git pull] device-mapper update for 2.6.25

[git pull] device-mapper update for 2.6.25

[Alasdair G Kergon]

Please pull from:

  master.kernel.org:/pub/scm/linux/kernel/git/agk/linux-2.6-dm.git

to get the following device-mapper update for 2.6.25:

Alasdair G Kergon (5):
      dm: mark function lists static
      dm ioctl: remove lock_kernel
      dm: targets no longer experimental
      dm mpath: add missing static
      dm crypt: move queue functions

Andrew Morton (3):
      dm snapshot: use uninitialized_var
      dm: table use uninitialized_var
      dm ioctl: use uninitialized_var

Brian Wood (2):
      dm: stripe trigger event on failure
      dm: stripe enhanced status return

Daniel Walker (1):
      dm: convert suspend_lock semaphore to mutex

Jonathan Brassow (6):
      dm log: auto load modules
      dm raid1: handle write failures
      dm raid1: handle recovery failures
      dm raid1: fix EIO after log failure
      dm raid1: handle read failures
      dm raid1: report fault status

Jun'ichi Nomura (1):
      dm: table remove unused total

Milan Broz (23):
      dm: add missing memory barrier to dm_suspend
      dm ioctl: move compat code
      dm: tidy alloc_dev labels
      dm: refactor deferred bio_list processing
      dm: tidy dm_suspend
      dm: split dm_suspend io_lock hold into two
      dm: refactor dm_suspend completion wait
      dm crypt: move convert_context inside dm_crypt_io
      dm crypt: remove unnecessary crypt_context write parm
      dm crypt: move error setting outside crypt_dec_pending
      dm crypt: tidy crypt_endio
      dm crypt: adjust io processing functions
      dm crypt: store sector mapping in dm_crypt_io
      dm crypt: abstract crypt_write_done
      dm crypt: introduce crypt_write_io_loop
      dm crypt: tidy io ref counting
      dm crypt: extract scatterlist processing
      dm crypt: add async request mempool
      dm crypt: add completion for async
      dm crypt: prepare async callback fn
      dm crypt: use async crypto
      dm: move deferred bio flushing to workqueue
      dm snapshot: combine consecutive exceptions in memory

Paul Jimenez (1):
      dm: table use list_for_each

Robert P. J. Day (1):
      dm snapshot: use rounddown_pow_of_two

Vasily Averin (1):
      dm: table remove unused variable

 drivers/md/Kconfig              |   24 +-
 drivers/md/dm-crypt.c           |  486 +++++++++++++++++++----------
 drivers/md/dm-exception-store.c |    2 +-
 drivers/md/dm-ioctl.c           |   32 ++-
 drivers/md/dm-log.c             |   51 +++-
 drivers/md/dm-mpath.c           |    2 +-
 drivers/md/dm-raid1.c           |  664 +++++++++++++++++++++++++++++++++-----
 drivers/md/dm-snap.c            |   95 ++++--
 drivers/md/dm-snap.h            |   50 +++-
 drivers/md/dm-stripe.c          |  105 ++++++-
 drivers/md/dm-table.c           |   20 +-
 drivers/md/dm.c                 |  238 +++++++++------
 fs/compat_ioctl.c               |   34 --
 include/linux/device-mapper.h   |   18 +-
 include/linux/dm-ioctl.h        |   34 +--
 15 files changed, 1372 insertions(+), 483 deletions(-)

Alasdair
-- 
agk@redhat.com

Re: [git pull] device-mapper update for 2.6.25

[Andrew Morton]

On Fri, 8 Feb 2008 02:45:32 +0000 Alasdair G Kergon <agk@redhat.com> wrote:

>       dm: move deferred bio flushing to workqueue

To my unpracticed eye this one looks fairly fundamental.  Unfortunately the
changelog:

  Add a single-thread workqueue for each mapped device and move flushing
  of the lists of pushback and deferred bios to this new workqueue.

tells us nothing of the reasons for this change.  (It should have done so!)

It _looks_ pretty unscalable.  Perhaps these deferred BIOs don't happen
very often?

Re: [git pull] device-mapper update for 2.6.25

[Alasdair G Kergon]

On Thu, Feb 07, 2008 at 07:04:47PM -0800, Andrew Morton wrote:
> On Fri, 8 Feb 2008 02:45:32 +0000 Alasdair G Kergon <agk@redhat.com> wrote:
> 
> >       dm: move deferred bio flushing to workqueue
> 
> To my unpracticed eye this one looks fairly fundamental.  Unfortunately the
> changelog:
> 
>   Add a single-thread workqueue for each mapped device and move flushing
>   of the lists of pushback and deferred bios to this new workqueue.
> 
> tells us nothing of the reasons for this change.  (It should have done so!)
> 
> It _looks_ pretty unscalable.  Perhaps these deferred BIOs don't happen
> very often?
 
Indeed - it's a rare occurrence - happens when the dm device is
suspended/resumed which usually means it was being reconfigured by something in
userspace.

It's the first patch of a series that will provide barrier support (slated for
2.6.26).

The appearance of all these workqueues in dm goes back to this old change:
  md-dm-reduce-stack-usage-with-stacked-block-devices.patch

Alasdair
-- 
agk@redhat.com