linux-toradex.git/block, branch v2.6.31-rc5

block: make the end_io functions be non-GPL exports

2009-07-28T20:11:24+00:00

Prior to the change for more sane end_io functions, we exported
the helpers with the normal EXPORT_SYMBOL(). That got changed
to _GPL() for the new interface. Revert that particular change,
on the basis that this is basic functionality and doesn't dip
into internal structures. If these exports can't be non-GPL,
then we may as well make EXPORT_SYMBOL() imply GPL for
everything.

Signed-off-by: Jens Axboe

block: fix improper kobject release in blk_integrity_unregister

2009-07-28T07:11:14+00:00

blk_integrity_unregister should use kobject_put to release the kobject,
otherwise after bi is freed, memory of bi->kobj->name is leaked.

Signed-off-by: Xiaotian Feng 
Signed-off-by: Jens Axboe

block: always assign default lock to queues

2009-07-28T07:07:29+00:00

Move the assignment of a default lock below blk_init_queue() to
blk_queue_make_request(), so we also get to set the default lock
for ->make_request_fn() based drivers. This is important since the
queue flag locking requires a lock to be in place.

Signed-off-by: Jens Axboe

block: sysfs fix mismatched queue_var_{store,show} in 64bit kernel

2009-07-17T07:54:15+00:00

In blk-sysfs.c, queue_var_store uses unsigned long to store data,
but queue_var_show uses unsigned int to show data.  This causes,

	# echo 70000000000 > /sys/block//queue/read_ahead_kb
	# cat /sys/block//queue/read_ahead_kb => get wrong value

Fix it by using unsigned long.

While at it, convert queue_rq_affinity_show() such that it uses bool
variable instead of explicit != 0 testing.

Signed-off-by: Xiaotian Feng 
Signed-off-by: Tejun Heo

block: fix failfast merge testing in elv_rq_merge_ok()

2009-07-17T05:50:43+00:00

Commit ab0fd1debe730ec9998678a0c53caefbd121ed10 tries to prevent merge
of requests with different failfast settings.  In elv_rq_merge_ok(),
it compares new bio's failfast flags against the merge target
request's.  However, the flag testing accessors for bio and blk don't
return boolean but the tested bit value directly and FAILFAST on bio
and blk don't match, so directly comparing them with == results in
false negative unnecessary preventing merge of readahead requests.

This patch convert the results to boolean by negating them before
comparison.

Signed-off-by: Tejun Heo 
Cc: Jens Axboe 
Cc: Boaz Harrosh 
Cc: FUJITA Tomonori 
Cc: James Bottomley 
Cc: Jeff Garzik

cfq-iosched: reset oom_cfqq in cfq_set_request()

2009-07-10T18:31:54+00:00

In case memory is scarce, we now default to oom_cfqq. Once memory is
available again, we should allocate a new cfqq and stop using oom_cfqq for
a particular io context.

Once a new request comes in, check if we are using oom_cfqq, and if yes,
try to allocate a new cfqq.

Tested the patch by forcing the use of oom_cfqq and upon next request thread
realized that it was using oom_cfqq and it allocated a new cfqq.

Signed-off-by: Vivek Goyal 
Signed-off-by: Jens Axboe

block: call blk_scsi_ioctl_init()

2009-07-10T18:31:53+00:00

Currently, blk_scsi_ioctl_init() is not called since it lacks
an initcall marking. This causes the command table to be
unitialized, hence somce commands are block when they should
not have been.

This fixes a regression introduced by commit
018e0446890661504783f92388ecce7138c1566d

Signed-off-by: FUJITA Tomonori 
Signed-off-by: Jens Axboe

block: don't merge requests of different failfast settings

2009-07-03T19:06:45+00:00

Block layer used to merge requests and bios with different failfast
settings.  This caused regular IOs to fail prematurely when they were
merged into failfast requests for readahead.

Niel Lambrechts could trigger the problem semi-reliably on ext4 when
resuming from STR.  ext4 uses readahead when reading inodes and
combined with the deterministic extra SATA PHY exception cycle during
resume on the specific configuration, non-readahead inode read would
fail causing ext4 errors.  Please read the following thread for
details.

  http://lkml.org/lkml/2009/5/23/21

This patch makes block layer reject merging if the failfast settings
don't match.  This is correct but likely to lower IO performance by
preventing regular IOs from mingling into surrounding readahead
requests.  Changes to allow such mixed merges and handle errors
correctly will be added later.

Signed-off-by: Tejun Heo 
Reported-by: Niel Lambrechts 
Cc: Theodore Tso 
Signed-off-by: Jens Axboe

cfq-iosched: remove redundant check for NULL cfqq in cfq_set_request()

2009-07-01T10:41:14+00:00

With the changes for falling back to an oom_cfqq, we never fail
to find/allocate a queue in cfq_get_queue(). So remove the check.

Signed-off-by: Shan Wei 
Signed-off-by: Jens Axboe

blocK: Restore barrier support for md and probably other virtual devices.

2009-07-01T08:56:26+00:00

The next_ordered flag is only meaningful for devices that use __make_request.
So move the test against next_ordered out of generic code and in to
__make_request

Since this test was added, barriers have not worked on md or any
devices that don't use __make_request and so don't bother to set
next_ordered.  (dm explicitly sets something other than
QUEUE_ORDERED_NONE since
  commit 99360b4c18f7675b50d283301d46d755affe75fd
but notes in the comments that it is otherwise meaningless).

Cc: Ken Milmore 
Cc: stable@kernel.org
Signed-off-by: NeilBrown 
Signed-off-by: Jens Axboe