GitHub/exynos8895/android_kernel_samsung_universal8895.git
14 years agoblkfront: Clean up vbd release
Daniel Stodden [Sat, 7 Aug 2010 16:33:17 +0000 (18:33 +0200)]
blkfront: Clean up vbd release

 * Current blkfront_closing is rather a xlvbd_release_gendisk.
   Renamed in preparation of later patches (need the name again).

 * Removed the misleading comment -- this only applied to the backend
   switch handler, and the queue is already flushed btw.

 * Break out the xenbus call, callers know better when to switch
   frontend state.

Signed-off-by: Daniel Stodden <daniel.stodden@citrix.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblkfront: Fix gendisk leak
Daniel Stodden [Fri, 30 Apr 2010 22:01:17 +0000 (22:01 +0000)]
blkfront: Fix gendisk leak

Signed-off-by: Daniel Stodden <daniel.stodden@citrix.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
14 years agoblkfront: Fix backtrace in del_gendisk
Daniel Stodden [Fri, 30 Apr 2010 22:01:16 +0000 (22:01 +0000)]
blkfront: Fix backtrace in del_gendisk

The call to del_gendisk follows an non-refcounted gd->queue
pointer. We release the last ref in blk_cleanup_queue. Fixed by
reordering releases accordingly.

Signed-off-by: Daniel Stodden <daniel.stodden@citrix.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
14 years agoxenbus: Make xenbus_switch_state transactional
Daniel Stodden [Fri, 30 Apr 2010 22:01:15 +0000 (22:01 +0000)]
xenbus: Make xenbus_switch_state transactional

According to the comments, this was how it's been done years ago, but
apparently took an xbt pointer from elsewhere back then. The code was
removed because of consistency issues: cancellation wont't roll back
the saved xbdev->state.

Still, unsolicited writes to the state field remain an issue,
especially if device shutdown takes thread synchronization, and subtle
races cause accidental recreation of the device node.

Fixed by reintroducing the transaction. An internal one is sufficient,
so the xbdev->state value remains consistent.

Also fixes the original hack to prevent infinite recursion. Instead of
bailing out on the first attempt to switch to Closing, checks call
depth now.

Signed-off-by: Daniel Stodden <daniel.stodden@citrix.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
14 years agoxen/blkfront: revalidate after setting capacity
K. Y. Srinivasan [Thu, 18 Mar 2010 22:00:54 +0000 (15:00 -0700)]
xen/blkfront: revalidate after setting capacity

Signed-off-by: K. Y. Srinivasan <ksrinivasan@novell.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
14 years agoxen/blkfront: avoid compiler warning from missing cases
Jeremy Fitzhardinge [Thu, 11 Mar 2010 23:10:40 +0000 (15:10 -0800)]
xen/blkfront: avoid compiler warning from missing cases

Fix:
drivers/block/xen-blkfront.c: In function ‘blkfront_connect’:
drivers/block/xen-blkfront.c:933: warning: enumeration value ‘BLKIF_STATE_DISCONNECTED’ not handled in switch

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
14 years agoxen/front: Propagate changed size of VBDs
K. Y. Srinivasan [Thu, 11 Mar 2010 21:42:26 +0000 (13:42 -0800)]
xen/front: Propagate changed size of VBDs

Support dynamic resizing of virtual block devices. This patch supports
both file backed block devices as well as physical devices that can be
dynamically resized on the host side.

Signed-off-by: K. Y. Srinivasan <ksrinivasan@novell.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
14 years agoblkfront: don't access freed struct xenbus_device
Jan Beulich [Sat, 7 Aug 2010 16:31:12 +0000 (18:31 +0200)]
blkfront: don't access freed struct xenbus_device

Unfortunately commit "blkfront: fixes for 'xm block-detach ... --force'"
still wasn't quite right - there was a reference to freed memory left
from blkfront_closing().

Signed-off-by: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblkfront: fixes for 'xm block-detach ... --force'
Jan Beulich [Sat, 7 Aug 2010 16:28:55 +0000 (18:28 +0200)]
blkfront: fixes for 'xm block-detach ... --force'

Prevent prematurely freeing 'struct blkfront_info' instances (when the
xenbus data structures are gone, but the Linux ones are still needed).

Prevent adding a disk with the same (major, minor) [and hence the same
name and sysfs entries, which leads to oopses] when the previous
instance wasn't fully de-allocated yet.

This still doesn't address all issues resulting from forced detach:
I/O submitted after the detach still blocks forever, likely preventing
subsequent un-mounting from completing. It's not clear to me (not
knowing much about the block layer) how this can be avoided.

Signed-off-by: Jan Beulich <jbeulich@novell.com>
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoxen: use less generic names in blkfront driver.
Ian Campbell [Fri, 4 Dec 2009 15:33:54 +0000 (15:33 +0000)]
xen: use less generic names in blkfront driver.

All Xen frontend drivers have a couple of identically named functions which
makes figuring out which device went wrong from a stacktrace harder than it
needs to be. Rename them to something specificto the device type.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Cc: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
14 years agowriteback.h: needs linux/device.h
Randy Dunlap [Mon, 19 Jul 2010 23:49:17 +0000 (16:49 -0700)]
writeback.h: needs linux/device.h

include/trace/events/writeback.h uses dev_name(), so it needs to
include linux/device.h.

include/trace/events/writeback.h:12: error: implicit declaration of function 'dev_name'

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: fix problem with sending down discard that isn't of correct granularity
Jens Axboe [Thu, 15 Jul 2010 16:49:31 +0000 (10:49 -0600)]
block: fix problem with sending down discard that isn't of correct granularity

If the queue doesn't have a limit set, or it just set UINT_MAX like
we default to, we coud be sending down a discard request that isn't
of the correct granularity if the block size is > 512b.

Fix this by adjusting max_discard_sectors down to the proper
alignment.

Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblkdev: check for valid request queue before issuing flush
Dave Chinner [Tue, 13 Jul 2010 07:50:50 +0000 (17:50 +1000)]
blkdev: check for valid request queue before issuing flush

Issuing a blkdev_issue_flush() on an unconfigured loop device causes a panic as
q->make_request_fn is not configured. This can occur when trying to mount the
unconfigured loop device as an XFS filesystem. There are no guards that catch
the bio before the request function is called because we don't add a payload to
the bio. Instead, manually check this case as soon as we have a pointer to the
queue to flush.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: fix for block tracing build error
Stephen Rothwell [Fri, 9 Jul 2010 04:24:38 +0000 (14:24 +1000)]
block: fix for block tracing build error

block/compat_ioctl.c: In function 'compat_blkdev_ioctl':
block/compat_ioctl.c:754: error: 'BLKTRACESETUP32' undeclared (first use in this function)

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoscsi/i2o: restore ioctl changes
Arnd Bergmann [Thu, 8 Jul 2010 12:57:03 +0000 (14:57 +0200)]
scsi/i2o: restore ioctl changes

This restores the changes from "scsi/i2o_block: cleanup ioctl
handling", which accidentally got reverted.

Origignal changelog:
      This fixes the ioctl function of the i2o_block driver, which
      has multiple problems:

      * The BLKI2OSRSTRAT and BLKI2OSWSTRAT commands always return
        -ENOTTY on success, where they should return 0.
      * Support for 32 bit compat is missing
      * The driver should use the .ioctl function and because
        .locked_ioctl is going away.

      The use of the big kernel lock remains for now, but gets
      made explictit in the ioctl function.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoscsi/sd: remove big kernel lock
Arnd Bergmann [Wed, 7 Jul 2010 14:51:29 +0000 (16:51 +0200)]
scsi/sd: remove big kernel lock

Every user of the BKL in the sd driver is the
result of the pushdown from the block layer
into the open/close/ioctl functions.

The only place that used to rely on the BKL is
the sdkp->openers variable, which gets converted
into an atomic_t.

Nothing else seems to rely on the BKL, since the
functions do not touch global data without holding
another lock, and the open/close functions are
still protected from concurrent execution using
the bdev->bd_mutex.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: linux-scsi@vger.kernel.org
Cc: "James E.J. Bottomley" <James.Bottomley@suse.de>
Acked-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: remove BKL from partition ioctls
Arnd Bergmann [Wed, 7 Jul 2010 14:51:28 +0000 (16:51 +0200)]
block: remove BKL from partition ioctls

The blkpg_ioctl and blkdev_reread_part access fields of
the bdev and gendisk structures, yet they always do so
under the protection of bdev->bd_mutex, which seems
sufficient.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
cked-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: remove BKL from BLKROSET and BLKFLSBUF
Arnd Bergmann [Wed, 7 Jul 2010 14:51:27 +0000 (16:51 +0200)]
block: remove BKL from BLKROSET and BLKFLSBUF

We only call the functions set_device_ro(),
invalidate_bdev(), sync_filesystem() and sync_blockdev()
while holding the BKL in these commands. All
of these are also done in other code paths without
the BKL, which leads me to the conclusion that
the BKL is not needed here either.

The reason we hold it here is that it was originally
pushed down into the ioctl function from vfs_ioctl.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: push BKL into blktrace ioctls
Arnd Bergmann [Wed, 7 Jul 2010 14:51:26 +0000 (16:51 +0200)]
block: push BKL into blktrace ioctls

The blktrace driver currently needs the BKL, but
we should not need to take that in the block layer,
so just push it down into the driver itself.

It is quite likely that the BKL is not actually
required in blktrace code and could be removed
in a follow-on patch.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: push down BKL into .open and .release
Arnd Bergmann [Sat, 7 Aug 2010 16:25:34 +0000 (18:25 +0200)]
block: push down BKL into .open and .release

The open and release block_device_operations are currently
called with the BKL held. In order to change that, we must
first make sure that all drivers that currently rely
on this have no regressions.

This blindly pushes the BKL into all .open and .release
operations for all block drivers to prepare for the
next step. The drivers can subsequently replace the BKL
with their own locks or remove it completely when it can
be shown that it is not needed.

The functions blkdev_get and blkdev_put are the only
remaining users of the big kernel lock in the block
layer, besides a few uses in the ioctl code, none
of which need to serialize with blkdev_{get,put}.

Most of these two functions is also under the protection
of bdev->bd_mutex, including the actual calls to
->open and ->release, and the common code does not
access any global data structures that need the BKL.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: push down BKL into .locked_ioctl
Arnd Bergmann [Thu, 8 Jul 2010 08:18:46 +0000 (10:18 +0200)]
block: push down BKL into .locked_ioctl

As a preparation for the removal of the big kernel
lock in the block layer, this removes the BKL
from the common ioctl handling code, moving it
into every single driver still using it.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoscsi/i2o_block: cleanup ioctl handling
Arnd Bergmann [Wed, 7 Jul 2010 14:51:23 +0000 (16:51 +0200)]
scsi/i2o_block: cleanup ioctl handling

This fixes the ioctl function of the i2o_block driver, which
has multiple problems:

* The BLKI2OSRSTRAT and BLKI2OSWSTRAT commands always return
  -ENOTTY on success, where they should return 0.
* Support for 32 bit compat is missing
* The driver should use the .ioctl function and because
  .locked_ioctl is going away.

The use of the big kernel lock remains for now, but gets
made explictit in the ioctl function.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoscsi: fix discard page leak
FUJITA Tomonori [Thu, 8 Jul 2010 08:16:17 +0000 (10:16 +0200)]
scsi: fix discard page leak

We leak a page allocated for discard on some error conditions
(e.g. scsi_prep_state_check returns BLKPREP_DEFER in
scsi_setup_blk_pc_cmnd).

We unprep on requests that weren't prepped in the error path of
scsi_init_io. It makes the error path to clean up scsi commands messy.

Let's strictly apply the rule that we can't unprep on a request that
wasn't prepped.

Calling just scsi_put_command() in the error path of scsi_init_io() is
enough. We don't set REQ_DONTPREP yet.

scsi_setup_discard_cmnd can safely free a page on the error case with
the above rule.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agowriteback: Add tracing to write_cache_pages
Dave Chinner [Wed, 7 Jul 2010 03:24:08 +0000 (13:24 +1000)]
writeback: Add tracing to write_cache_pages

Add a trace event to the ->writepage loop in write_cache_pages to give
visibility into how the ->writepage call is changing variables within the
writeback control structure. Of most interest is how wbc->nr_to_write changes
from call to call, especially with filesystems that write multiple pages
in ->writepage.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agowriteback: Add tracing to balance_dirty_pages
Dave Chinner [Wed, 7 Jul 2010 03:24:07 +0000 (13:24 +1000)]
writeback: Add tracing to balance_dirty_pages

Tracing high level background writeback events is good, but it doesn't
give the entire picture. Add visibility into write throttling to catch IO
dispatched by foreground throttling of processing dirtying lots of pages.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agowriteback: Initial tracing support
Dave Chinner [Wed, 7 Jul 2010 03:24:06 +0000 (13:24 +1000)]
writeback: Initial tracing support

Trace queue/sched/exec parts of the writeback loop. This provides
insight into when and why flusher threads are scheduled to run. e.g
a sync invocation leaves traces like:

     sync-[...]: writeback_queue: bdi 8:0: sb_dev 8:1 nr_pages=7712 sync_mode=0 kupdate=0 range_cyclic=0 background=0
flush-8:0-[...]: writeback_exec: bdi 8:0: sb_dev 8:1 nr_pages=7712 sync_mode=0 kupdate=0 range_cyclic=0 background=0

This also lays the foundation for adding more writeback tracing to
provide deeper insight into the whole writeback path.

The original tracing code is from Jens Axboe, though this version is
a rewrite as a result of the code being traced changing
significantly.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: remove unused REQ_TYPE_LINUX_BLOCK
FUJITA Tomonori [Tue, 6 Jul 2010 07:03:18 +0000 (09:03 +0200)]
block: remove unused REQ_TYPE_LINUX_BLOCK

Nobody uses REQ_TYPE_LINUX_BLOCK (and its REQ_LB_OP_*).

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Acked-by: Jeff Garzik <jgarzik@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoscsi: need to reset unprep_rq_fn in sd_remove
FUJITA Tomonori [Sat, 3 Jul 2010 14:07:04 +0000 (08:07 -0600)]
scsi: need to reset unprep_rq_fn in sd_remove

This is for block's for-2.6.36.

We need to reset q->unprep_rq_fn in sd_remove. Otherwise we hit kernel
oops if we access to a scsi disk device via sg after removing scsi
disk module.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: remove q->prepare_flush_fn completely
FUJITA Tomonori [Sat, 3 Jul 2010 08:45:40 +0000 (17:45 +0900)]
block: remove q->prepare_flush_fn completely

This removes q->prepare_flush_fn completely (changes the
blk_queue_ordered API).

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoide: stop using q->prepare_flush_fn
FUJITA Tomonori [Sat, 3 Jul 2010 08:45:39 +0000 (17:45 +0900)]
ide: stop using q->prepare_flush_fn

use REQ_FLUSH flag instead.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Acked-by: David S. Miller <davem@davemloft.net>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agovirtio_blk: stop using q->prepare_flush_fn
FUJITA Tomonori [Sat, 3 Jul 2010 08:45:38 +0000 (17:45 +0900)]
virtio_blk: stop using q->prepare_flush_fn

use REQ_FLUSH flag instead.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agodm: stop using q->prepare_flush_fn
FUJITA Tomonori [Sat, 3 Jul 2010 08:45:37 +0000 (17:45 +0900)]
dm: stop using q->prepare_flush_fn

use REQ_FLUSH flag instead.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: Alasdair G Kergon <agk@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agops3disk: stop using q->prepare_flush_fn
FUJITA Tomonori [Sat, 3 Jul 2010 08:45:36 +0000 (17:45 +0900)]
ps3disk: stop using q->prepare_flush_fn

REQ_FLUSH flag enables us to kill ps3disk_prepare_flush().

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoosdblk: stop using q->prepare_flush_fn
FUJITA Tomonori [Sat, 3 Jul 2010 08:45:35 +0000 (17:45 +0900)]
osdblk: stop using q->prepare_flush_fn

use REQ_FLUSH flag instead.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoscsi: stop using q->prepare_flush_fn
FUJITA Tomonori [Sat, 3 Jul 2010 08:45:34 +0000 (17:45 +0900)]
scsi: stop using q->prepare_flush_fn

scsi-ml builds flush requests via q->prepare_flush_fn(), however,
builds discard requests via q->prep_rq_fn.

Using two different mechnisms for the similar requests (building
commands in SCSI ULD) doesn't make sense.

Handing both via q->prep_rq_fn makes the code design simpler.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: James Bottomley <James.Bottomley@suse.de>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: permit PREFLUSH and POSTFLUSH without prepare_flush_fn
FUJITA Tomonori [Sat, 3 Jul 2010 08:45:33 +0000 (17:45 +0900)]
block: permit PREFLUSH and POSTFLUSH without prepare_flush_fn

This is preparation for removing q->prepare_flush_fn.

Temporarily, blk_queue_ordered() permits QUEUE_ORDERED_DO_PREFLUSH and
QUEUE_ORDERED_DO_POSTFLUSH without prepare_flush_fn.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: introduce REQ_FLUSH flag
FUJITA Tomonori [Sat, 3 Jul 2010 08:45:32 +0000 (17:45 +0900)]
block: introduce REQ_FLUSH flag

SCSI-ml needs a way to mark a request as flush request in
q->prepare_flush_fn because it needs to identify them later (e.g. in
q->request_fn or prep_rq_fn).

queue_flush sets REQ_HARDBARRIER in rq->cmd_flags however the block
layer also sends normal REQ_TYPE_FS requests with REQ_HARDBARRIER. So
SCSI-ml can't use REQ_HARDBARRIER to identify flush requests.

We could change the block layer to clear REQ_HARDBARRIER bit before
sending non flush requests to the lower layers. However, intorudcing
the new flag looks cleaner (surely easier).

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Cc: James Bottomley <James.Bottomley@suse.de>
Cc: David S. Miller <davem@davemloft.net>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Alasdair G Kergon <agk@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoscsi: remove unused free discard page in sd_done
FUJITA Tomonori [Thu, 1 Jul 2010 10:49:19 +0000 (19:49 +0900)]
scsi: remove unused free discard page in sd_done

- sd_done isn't called for pc request so we never call the code.
- we use sd_unprep to free discard page now.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoscsi: add sd_unprep_fn to free discard page
FUJITA Tomonori [Thu, 1 Jul 2010 10:49:18 +0000 (19:49 +0900)]
scsi: add sd_unprep_fn to free discard page

This fixes discard page leak by using q->unprep_rq_fn facility.

q->unprep_rq_fn is called when all the data buffer (req->bio and
scsi_data_buffer) in the request is freed.

sd_unprep() uses rq->buffer to free discard page allocated in
sd_prepare_discard().

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: implement an unprep function corresponding directly to prep
James Bottomley [Thu, 1 Jul 2010 10:49:17 +0000 (19:49 +0900)]
block: implement an unprep function corresponding directly to prep

Reviewed-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agodrivers/cdrom: use pr_<level>
Joe Perches [Thu, 1 Jul 2010 06:24:32 +0000 (08:24 +0200)]
drivers/cdrom: use pr_<level>

- add pr_fmt.

- convert printks to pr_<level>

- add if (0) and printf argument checking to cdinfo

- coalesce consecutive printks to single pr_

- fix a typo "back ground" to "background"

- convert printks without level to pr_info

- remove VIOCD_ prefixes and use pr_fmt/pr_<level>

- add a missing newline to an OS/400 message

Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Folded in tab indentation fix from Andrew.

Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: fixup missing conversion from BIO_RW_DISCARD to REQ_DISCARD
Jens Axboe [Tue, 29 Jun 2010 11:33:38 +0000 (13:33 +0200)]
block: fixup missing conversion from BIO_RW_DISCARD to REQ_DISCARD

Didn't cause a merge conflict, so fixed this one up manually
post merge.

Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock/xd.c: fix brace typo
Randy Dunlap [Tue, 22 Jun 2010 16:03:43 +0000 (09:03 -0700)]
block/xd.c: fix brace typo

Fix extra brace typo that is causing build errors.

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agogcc-4.6: fs: fix unused but set warnings
Andi Kleen [Mon, 21 Jun 2010 09:02:48 +0000 (11:02 +0200)]
gcc-4.6: fs: fix unused but set warnings

No real bugs I believe, just some dead code, and some
shut up code.

Signed-off-by: Andi Kleen <ak@linux.intel.com>
Cc: Eric Paris <eparis@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agogcc-4.6: block: fix unused but set variables in blk-merge
Andi Kleen [Mon, 21 Jun 2010 09:02:47 +0000 (11:02 +0200)]
gcc-4.6: block: fix unused but set variables in blk-merge

Just some dead code.

Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: don't allocate a payload for discard request
Christoph Hellwig [Fri, 18 Jun 2010 14:59:42 +0000 (16:59 +0200)]
block: don't allocate a payload for discard request

Allocating a fixed payload for discard requests always was a horrible hack,
and it's not coming to byte us when adding support for discard in DM/MD.

So change the code to leave the allocation of a payload to the lowlevel
driver.  Unfortunately that means we'll need another hack, which allows
us to update the various block layer length fields indicating that we
have a payload.  Instead of hiding this in sd.c, which we already partially
do for UNMAP support add a documented helper in the core block layer for it.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agowriteback: merge bdi_writeback_task and bdi_start_fn
Christoph Hellwig [Sat, 19 Jun 2010 21:08:22 +0000 (23:08 +0200)]
writeback: merge bdi_writeback_task and bdi_start_fn

Move all code for the writeback thread into fs/fs-writeback.c instead of
splitting it over two functions in two files.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agowriteback: remove wb_list
Christoph Hellwig [Sat, 19 Jun 2010 21:08:06 +0000 (23:08 +0200)]
writeback: remove wb_list

The wb_list member of struct backing_device_info always has exactly one
element.  Just use the direct bdi->wb pointer instead and simplify some
code.

Also remove bdi_task_init which is now trivial to prepare for the next
patch.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: fix some more cmd_type cleanup fallout
Christoph Hellwig [Sat, 19 Jun 2010 15:26:47 +0000 (17:26 +0200)]
block: fix some more cmd_type cleanup fallout

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agovirtio_blk: add default case to cmd type switch
Jens Axboe [Fri, 18 Jun 2010 10:10:18 +0000 (12:10 +0200)]
virtio_blk: add default case to cmd type switch

On compilation, gcc correctly detects that we do not handle
all types:

In function ‘blk_done’:
warning: enumeration value ‘REQ_TYPE_FS’ not handled in switch
warning: enumeration value ‘REQ_TYPE_SENSE’ not handled in switch
warning: enumeration value ‘REQ_TYPE_PM_SUSPEND’ not handled in switch
warning: enumeration value ‘REQ_TYPE_PM_RESUME’ not handled in switch
warning: enumeration value ‘REQ_TYPE_PM_SHUTDOWN’ not handled in switch
warning: enumeration value ‘REQ_TYPE_LINUX_BLOCK’ not handled in switch
warning: enumeration value ‘REQ_TYPE_ATA_TASKFILE’ not handled in switch
warning: enumeration value ‘REQ_TYPE_ATA_PC’ not handled in switch

which is a bit pointless since this is at the end of the request
processessing. Add a default case that just breaks out.

Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: unify flags for struct bio and struct request
Christoph Hellwig [Sat, 7 Aug 2010 16:20:39 +0000 (18:20 +0200)]
block: unify flags for struct bio and struct request

Remove the current bio flags and reuse the request flags for the bio, too.
This allows to more easily trace the type of I/O from the filesystem
down to the block driver.  There were two flags in the bio that were
missing in the requests:  BIO_RW_UNPLUG and BIO_RW_AHEAD.  Also I've
renamed two request flags that had a superflous RW in them.

Note that the flags are in bio.h despite having the REQ_ name - as
blkdev.h includes bio.h that is the only way to go for now.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: remove wrappers for request type/flags
Christoph Hellwig [Sat, 7 Aug 2010 16:17:56 +0000 (18:17 +0200)]
block: remove wrappers for request type/flags

Remove all the trivial wrappers for the cmd_type and cmd_flags fields in
struct requests.  This allows much easier grepping for different request
types instead of unwinding through macros.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoremove needless ISA_DMA_THRESHOLD
FUJITA Tomonori [Mon, 31 May 2010 06:59:04 +0000 (15:59 +0900)]
remove needless ISA_DMA_THRESHOLD

Architectures don't need to define ISA_DMA_THRESHOLD anymore.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Acked-by: James Bottomley <James.Bottomley@suse.de>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: kill ISA_DMA_THRESHOLD usage
FUJITA Tomonori [Mon, 31 May 2010 06:59:03 +0000 (15:59 +0900)]
block: kill ISA_DMA_THRESHOLD usage

block uses ISA_DMA_THRESHOLD for BLK_BOUNCE_ISA. Only SCSI uses
ISA_DMA_THRESHOLD for ancient drivers with non-zero
unchecked_isa_dma. Nowadays drivers (and subsystems) use dma_mask
properly instead of ISA_DMA_THRESHOLD.

Documentation/scsi/scsi_mid_low_api.txt says:

unchecked_isa_dma - 1=>only use bottom 16 MB of ram (ISA DMA addressing
                   restriction), 0=>can use full 32 bit (or better) DMA
                   address space

So block simply uses DMA_BIT_MASK(24) for BLK_BOUNCE_ISA for SCSI.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Acked-by: James Bottomley <James.Bottomley@suse.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoaha1532: remove ISA_DMA_THRESHOLD usage
FUJITA Tomonori [Thu, 17 Jun 2010 12:58:21 +0000 (14:58 +0200)]
aha1532: remove ISA_DMA_THRESHOLD usage

We can safely remove ISA_DMA_THRESHOLD usage in aha1542. aha1542 uses
ISA_DMA_THRESHOLD to see if:

- the buffers in scatter/list are below 16MB.
- scsi_host is below 16MB.

Both checkings were added in the ancient times but aren't necessary
nowadays since we properly bounce the buffers and allocate scsi_host
below 16MB with non-zero unchecked_isa_dma.

Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp>
Acked-by: James Bottomley <James.Bottomley@suse.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: BARRIER request should imply SYNC
Christoph Hellwig [Thu, 17 Jun 2010 06:54:16 +0000 (08:54 +0200)]
block: BARRIER request should imply SYNC

A barrier request should by defintion have priority in get_request
and let the queue be unplugged immediately as it's blocking all forward
progress due to the queue draining.

Most filesystems already get this implicitly by the way how submit_bh
treats the buffer_ordered flag, and gfs2 sets it explicitly.  But btrfs
and XFS are still forgetting to set the flag, as is blkdev_issue_flush
and some places in DM/MD.

For XFS on metadata heavy workloads this gives a consistent speedup
in the 2-3% range.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agofloppy: use warning macros
Stephen Hemminger [Tue, 15 Jun 2010 11:21:11 +0000 (13:21 +0200)]
floppy: use warning macros

Convert assertions to use WARN().  There are several error checks in the
code for things that should never happen.  Convert them to standard
warnings so kerneloops.org will see them.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agofloppy: use wait_event_interruptible
Stephen Hemminger [Tue, 15 Jun 2010 11:21:11 +0000 (13:21 +0200)]
floppy: use wait_event_interruptible

Convert wait loops to use wait_event_ macros.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agofloppy: fix signed/unsigned warnings
Stephen Hemminger [Tue, 15 Jun 2010 11:21:11 +0000 (13:21 +0200)]
floppy: fix signed/unsigned warnings

Ioctl cmd value is unsigned, so change normalize_ioctl

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agofloppy: cmos attribute should be static
Stephen Hemminger [Tue, 15 Jun 2010 11:21:11 +0000 (13:21 +0200)]
floppy: cmos attribute should be static

As reported by sparse, cmos attribute is local.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agofloppy: use atomic type for usage_count
Stephen Hemminger [Tue, 15 Jun 2010 11:21:11 +0000 (13:21 +0200)]
floppy: use atomic type for usage_count

The usage_count was being protected by a lock which was only there to
create an atomic counter.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agofloppy: silence warning during disk test
Stephen Hemminger [Tue, 15 Jun 2010 11:21:11 +0000 (13:21 +0200)]
floppy: silence warning during disk test

The first thing the floppy does is read block 0 to test geometry and to
test for disk presence.  If disk is not present this causes a console
warning message about failed I/O.  Set flag to silence.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agofloppy: remove unnecessary inlines
Stephen Hemminger [Tue, 15 Jun 2010 11:21:11 +0000 (13:21 +0200)]
floppy: remove unnecessary inlines

These routines are all big enough that is better to let the compiler
decide to inline or not.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agofloppy: initialize debug jiffies offset
Stephen Hemminger [Tue, 15 Jun 2010 11:21:11 +0000 (13:21 +0200)]
floppy: initialize debug jiffies offset

Set debug jiffies offset at initialization.  Avoids wierd values showing
up if debugging enabled.

Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agocciss: change pad value from 32 to 0
Mike Miller [Fri, 11 Jun 2010 11:13:36 +0000 (13:13 +0200)]
cciss: change pad value from 32 to 0

Change the command padding on 32-bit systems to 0 since setting it to 32
has the identical effect.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agocciss: remove errant debug code
Mike Miller [Fri, 11 Jun 2010 11:13:36 +0000 (13:13 +0200)]
cciss: remove errant debug code

Remove a debug statement left behind by accident Ths debug statement got
left behind.  It was commented out after use but not deleted.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agocciss: move next_command function from ifdef
Mike Miller [Fri, 11 Jun 2010 11:13:35 +0000 (13:13 +0200)]
cciss: move next_command function from ifdef

The definition of next_command also ended up in wrong place It ended up
inside an "#ifdef CONFIG_PROCFS".  Already caught by Randy Dunlap and a
couple others.  Tried to put it somewhere that made sense.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agocciss: fix call to put_controller_in_performant_mode
Mike Miller [Fri, 11 Jun 2010 11:13:35 +0000 (13:13 +0200)]
cciss: fix call to put_controller_in_performant_mode

call to put_controller_in_performant_mode was in the wrong place
The call inadvertently ended up in an error path.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agocciss: make sure we request the performant mode irq
Mike Miller [Fri, 11 Jun 2010 11:13:14 +0000 (13:13 +0200)]
cciss: make sure we request the performant mode irq

Make sure we register the performant mode interrupt Another blunder.
Seemed to work because the call to put_controller_into_performant_mode was
never called.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: add helpers for the trivial queue flag sysfs show/store entries
Jens Axboe [Sat, 7 Aug 2010 16:13:50 +0000 (18:13 +0200)]
block: add helpers for the trivial queue flag sysfs show/store entries

The code for nonrot, random, and io stats are completely identical.

Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoblock: add sysfs knob for turning off disk entropy contributions
Jens Axboe [Wed, 9 Jun 2010 08:42:09 +0000 (10:42 +0200)]
block: add sysfs knob for turning off disk entropy contributions

There are two reasons for doing this:

- On SSD disks, the completion times aren't as random as they
  are for rotational drives. So it's questionable whether they
  should contribute to the random pool in the first place.

- Calling add_disk_randomness() has a lot of overhead.

This adds /sys/block/<dev>/queue/add_random that will allow you to
switch off on a per-device basis. The default setting is on, so there
should be no functional changes from this patch.

Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agocciss: new controller support and bump driver version
Mike Miller [Wed, 2 Jun 2010 19:58:09 +0000 (12:58 -0700)]
cciss: new controller support and bump driver version

Add support for new controllers due out next year.  HP must continue to
support new controllers in older distros.  All vendors require support be
upstream.  These controllers support only 16 commands in simple mode but
can support up to 1024 in performant mode.  See patch 5/6/ We have no
marketing names yet.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agocciss: add performant mode support for Stars/Sirius
Mike Miller [Wed, 2 Jun 2010 19:58:06 +0000 (12:58 -0700)]
cciss: add performant mode support for Stars/Sirius

Add a mode of controller operation called Performant Mode.  Even though
cciss has been deprecated in favor of hpsa there are new controllers due
out next year that HP must support in older vendor distros.  Vendors
require all fixes/features be upstream.  These new controllers support
only 16 commands in simple mode but support up to 1024 in performant mode.
This requires us to add this support at this late date.

The performant mode transport minimizes host PCI accesses by performinf
many completions per read.  PCI writes are posted so the host can write
then immediately get off the bus not waiting for the writwe to complete to
the target.  In the context of performant mode the host read out to a
controller pulls all posted writes into host memory ensuring the reply
queue is coherent.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agocciss: make interrupt access methods return type bool
Mike Miller [Wed, 2 Jun 2010 19:58:04 +0000 (12:58 -0700)]
cciss: make interrupt access methods return type bool

Change the return type of our interrupt access routines to bool from
unsigned long.  It makes more sense that way.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agocciss: check for msi in interrupt_not_for_us
Mike Miller [Wed, 2 Jun 2010 19:58:02 +0000 (12:58 -0700)]
cciss: check for msi in interrupt_not_for_us

Check to see if h->msi[x]_vector is set.  We need this for a following
patch.  Without this check we process one interrupt then stop because in
msi[x] mode the interrupt pending bit is not set.  Not sure why we didn't
encounter this before.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agocciss: clean up interrupt handler
Mike Miller [Wed, 2 Jun 2010 19:58:00 +0000 (12:58 -0700)]
cciss: clean up interrupt handler

Simplify the interrupt handler code to more closely match hpsa and to
hopefully make it easier to follow.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agocciss: enqueue and submit io
Mike Miller [Wed, 2 Jun 2010 19:57:58 +0000 (12:57 -0700)]
cciss: enqueue and submit io

Clean up some code where we subit our io.  The same 5 lines appeared
several times.  Also helps for a following patch.

Signed-off-by: Mike Miller <mike.miller@hp.com>
Cc: Stephen M. Cameron <scameron@beardog.cce.hp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
14 years agoLinux 2.6.35
Linus Torvalds [Sun, 1 Aug 2010 22:11:14 +0000 (15:11 -0700)]
Linux 2.6.35

14 years agoNFS: Fix a typo in include/linux/nfs_fs.h
Trond Myklebust [Sun, 1 Aug 2010 17:40:40 +0000 (13:40 -0400)]
NFS: Fix a typo in include/linux/nfs_fs.h

nfs_commit_inode() needs to be defined irrespectively of whether or not
we are supporting NFSv3 and NFSv4.

Allow the compiler to optimise away code in the NFSv2-only case by
converting it into an inlined stub function.

Reported-and-tested-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoMerge master.kernel.org:/home/rmk/linux-2.6-arm
Linus Torvalds [Sat, 31 Jul 2010 02:02:51 +0000 (19:02 -0700)]
Merge master.kernel.org:/home/rmk/linux-2.6-arm

* master.kernel.org:/home/rmk/linux-2.6-arm:
  cyber2000fb: fix console in truecolor modes
  cyber2000fb: fix machine hang on module load
  SA1111: Eliminate use after free
  ARM: Fix Versatile/Realview/VExpress MMC card detection sense
  ARM: 6279/1: highmem: fix SMP preemption bug in kmap_high_l1_vipt
  ARM: Add barriers to io{read,write}{8,16,32} accessors as well
  ARM: 6273/1: Add barriers to the I/O accessors if ARM_DMA_MEM_BUFFERABLE
  ARM: 6272/1: Convert L2x0 to use the IO relaxed operations
  ARM: 6271/1: Introduce *_relaxed() I/O accessors
  ARM: 6275/1: ux500: don't use writeb() in uncompress.h
  ARM: 6270/1: clean files in arch/arm/boot/compressed/
  ARM: Fix csum_partial_copy_from_user()

14 years agoMerge branch 'bugfixes' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6
Linus Torvalds [Sat, 31 Jul 2010 02:02:21 +0000 (19:02 -0700)]
Merge branch 'bugfixes' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6

* 'bugfixes' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6:
  NFS: Ensure that writepage respects the nonblock flag
  NFS: kswapd must not block in nfs_release_page
  nfs: include space for the NUL in root path

14 years agoMerge branch 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied...
Linus Torvalds [Sat, 31 Jul 2010 02:01:11 +0000 (19:01 -0700)]
Merge branch 'drm-fixes' of git://git./linux/kernel/git/airlied/drm-2.6

* 'drm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6:
  drm/edid: Fix the HDTV hack sync adjustment
  drm/radeon/kms: fix radeon mid power profile reporting

14 years agomm: fix ia64 crash when gcore reads gate area
Hugh Dickins [Fri, 30 Jul 2010 17:58:26 +0000 (10:58 -0700)]
mm: fix ia64 crash when gcore reads gate area

Debian's ia64 autobuilders have been seeing kernel freeze or reboot
when running the gdb testsuite (Debian bug 588574): dannf bisected to
2.6.32 62eede62dafb4a6633eae7ffbeb34c60dba5e7b1 "mm: ZERO_PAGE without
PTE_SPECIAL"; and reproduced it with gdb's gcore on a simple target.

I'd missed updating the gate_vma handling in __get_user_pages(): that
happens to use vm_normal_page() (nowadays failing on the zero page),
yet reported success even when it failed to get a page - boom when
access_process_vm() tried to copy that to its intermediate buffer.

Fix this, resisting cleanups: in particular, leave it for now reporting
success when not asked to get any pages - very probably safe to change,
but let's not risk it without testing exposure.

Why did ia64 crash with 16kB pages, but succeed with 64kB pages?
Because setup_gate() pads each 64kB of its gate area with zero pages.

Reported-by: Andreas Barth <aba@not.so.argh.org>
Bisected-by: dann frazier <dannf@debian.org>
Signed-off-by: Hugh Dickins <hughd@google.com>
Tested-by: dann frazier <dannf@dannf.org>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoCIFS: Remove __exit mark from cifs_exit_dns_resolver()
David Howells [Fri, 30 Jul 2010 14:25:19 +0000 (15:25 +0100)]
CIFS: Remove __exit mark from cifs_exit_dns_resolver()

Remove the __exit mark from cifs_exit_dns_resolver() as it's called by the
module init routine in case of error, and so may have been discarded during
linkage.

Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agocyber2000fb: fix console in truecolor modes
Ondrej Zary [Thu, 29 Jul 2010 20:40:54 +0000 (22:40 +0200)]
cyber2000fb: fix console in truecolor modes

Return value was not set to 0 in setcolreg() with truecolor modes. This causes
fb_set_cmap() to abort after first color, resulting in blank palette - and
blank console in 24bpp and 32bpp modes.

Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
14 years agocyber2000fb: fix machine hang on module load
Ondrej Zary [Thu, 29 Jul 2010 20:32:20 +0000 (22:32 +0200)]
cyber2000fb: fix machine hang on module load

I was testing two CyberPro 2000 based PCI cards on x86 and the machine always
hanged completely when the cyber2000fb module was loaded. It seems that the
card hangs when some registers are accessed too quickly after writing RAMDAC
control register. With this patch, both card work.

Add delay after RAMDAC control register write to prevent hangs on module load.

Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
14 years agoSA1111: Eliminate use after free
Julia Lawall [Fri, 30 Jul 2010 15:17:28 +0000 (17:17 +0200)]
SA1111: Eliminate use after free

__sa1111_remove always frees its argument, so the subsequent reference to
sachip->saved_state represents a use after free.  __sa1111_remove does not
appear to use the saved_state field, so the patch simply frees it first.

A simplified version of the semantic patch that finds this problem is as
follows: (http://coccinelle.lip6.fr/)

// <smpl>
@@
expression E,E2;
@@

__sa1111_remove(E)
...
(
  E = E2
|
* E
)
// </smpl>

Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
14 years agoARM: Fix Versatile/Realview/VExpress MMC card detection sense
Russell King [Thu, 29 Jul 2010 14:58:59 +0000 (15:58 +0100)]
ARM: Fix Versatile/Realview/VExpress MMC card detection sense

The MMC card detection sense has become really confused with negations
at various levels, leading to some platforms not detecting inserted
cards.  Fix this by converting everything to positive logic throughout,
thereby getting rid of these negations.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
14 years agoARM: 6279/1: highmem: fix SMP preemption bug in kmap_high_l1_vipt
Gary King [Thu, 29 Jul 2010 16:37:20 +0000 (17:37 +0100)]
ARM: 6279/1: highmem: fix SMP preemption bug in kmap_high_l1_vipt

smp_processor_id() must not be called from a preemptible context (this
is checked by CONFIG_DEBUG_PREEMPT).  kmap_high_l1_vipt() was doing so.
This lead to a problem where the wrong per_cpu kmap_high_l1_vipt_depth
could be incremented, causing a BUG_ON(*depth <= 0); in
kunmap_high_l1_vipt().

The solution is to move the call to smp_processor_id() after the call
to preempt_disable().

Originally by: Andrew Howe <ahowe@nvidia.com>

Signed-off-by: Gary King <gking@nvidia.com>
Acked-by: Nicolas Pitre <nico.as.pitre@linaro.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
14 years agoNFS: Ensure that writepage respects the nonblock flag
Trond Myklebust [Fri, 30 Jul 2010 19:31:57 +0000 (15:31 -0400)]
NFS: Ensure that writepage respects the nonblock flag

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
14 years agoNFS: kswapd must not block in nfs_release_page
Trond Myklebust [Fri, 30 Jul 2010 19:31:54 +0000 (15:31 -0400)]
NFS: kswapd must not block in nfs_release_page

See https://bugzilla.kernel.org/show_bug.cgi?id=16056

If other processes are blocked waiting for kswapd to free up some memory so
that they can make progress, then we cannot allow kswapd to block on those
processes.

Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: stable@kernel.org
14 years agonfs: include space for the NUL in root path
Dan Carpenter [Tue, 13 Jul 2010 11:34:59 +0000 (13:34 +0200)]
nfs: include space for the NUL in root path

In root_nfs_name() it does the following:

        if (strlen(buf) + strlen(cp) > NFS_MAXPATHLEN) {
                printk(KERN_ERR "Root-NFS: Pathname for remote directory too long.\n");
                return -1;
        }
        sprintf(nfs_export_path, buf, cp);

In the original code if (strlen(buf) + strlen(cp) == NFS_MAXPATHLEN)
then the sprintf() would lead to an overflow.  Generally the rest of the
code assumes that the path can have NFS_MAXPATHLEN (1024) characters and
a NUL terminator so the fix is to add space to the nfs_export_path[]
buffer.

Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
14 years agoMerge branch 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6
Linus Torvalds [Thu, 29 Jul 2010 22:23:28 +0000 (15:23 -0700)]
Merge branch 'for-linus' of git://git390.marist.edu/linux-2.6

* 'for-linus' of git://git390.marist.edu/pub/scm/linux-2.6:
  [S390] etr: fix clock synchronization race
  [S390] Fix IRQ tracing in case of PER

14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/wim/linux-2.6-watchdog
Linus Torvalds [Thu, 29 Jul 2010 22:21:44 +0000 (15:21 -0700)]
Merge git://git./linux/kernel/git/wim/linux-2.6-watchdog

* git://git.kernel.org/pub/scm/linux/kernel/git/wim/linux-2.6-watchdog:
  watchdog: update MAINTAINERS entry

14 years agoMerge branch 'fix/hda' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
Linus Torvalds [Thu, 29 Jul 2010 22:21:07 +0000 (15:21 -0700)]
Merge branch 'fix/hda' of git://git./linux/kernel/git/tiwai/sound-2.6

* 'fix/hda' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
  ALSA: hda - Add a PC-beep workaround for ASUS P5-V
  ALSA: hda - Assume PC-beep as default for Realtek
  ALSA: hda - Don't register beep input device when no beep is available
  ALSA: hda - Fix pin-detection of Nvidia HDMI

14 years agoCRED: Fix __task_cred()'s lockdep check and banner comment
David Howells [Thu, 29 Jul 2010 11:45:55 +0000 (12:45 +0100)]
CRED: Fix __task_cred()'s lockdep check and banner comment

Fix __task_cred()'s lockdep check by removing the following validation
condition:

lockdep_tasklist_lock_is_held()

as commit_creds() does not take the tasklist_lock, and nor do most of the
functions that call it, so this check is pointless and it can prevent
detection of the RCU lock not being held if the tasklist_lock is held.

Instead, add the following validation condition:

task->exit_state >= 0

to permit the access if the target task is dead and therefore unable to change
its own credentials.

Fix __task_cred()'s comment to:

 (1) discard the bit that says that the caller must prevent the target task
     from being deleted.  That shouldn't need saying.

 (2) Add a comment indicating the result of __task_cred() should not be passed
     directly to get_cred(), but rather than get_task_cred() should be used
     instead.

Also put a note into the documentation to enforce this point there too.

Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoCRED: Fix get_task_cred() and task_state() to not resurrect dead credentials
David Howells [Thu, 29 Jul 2010 11:45:49 +0000 (12:45 +0100)]
CRED: Fix get_task_cred() and task_state() to not resurrect dead credentials

It's possible for get_task_cred() as it currently stands to 'corrupt' a set of
credentials by incrementing their usage count after their replacement by the
task being accessed.

What happens is that get_task_cred() can race with commit_creds():

TASK_1 TASK_2 RCU_CLEANER
-->get_task_cred(TASK_2)
rcu_read_lock()
__cred = __task_cred(TASK_2)
-->commit_creds()
old_cred = TASK_2->real_cred
TASK_2->real_cred = ...
put_cred(old_cred)
  call_rcu(old_cred)
[__cred->usage == 0]
get_cred(__cred)
[__cred->usage == 1]
rcu_read_unlock()
-->put_cred_rcu()
[__cred->usage == 1]
panic()

However, since a tasks credentials are generally not changed very often, we can
reasonably make use of a loop involving reading the creds pointer and using
atomic_inc_not_zero() to attempt to increment it if it hasn't already hit zero.

If successful, we can safely return the credentials in the knowledge that, even
if the task we're accessing has released them, they haven't gone to the RCU
cleanup code.

We then change task_state() in procfs to use get_task_cred() rather than
calling get_cred() on the result of __task_cred(), as that suffers from the
same problem.

Without this change, a BUG_ON in __put_cred() or in put_cred_rcu() can be
tripped when it is noticed that the usage count is not zero as it ought to be,
for example:

kernel BUG at kernel/cred.c:168!
invalid opcode: 0000 [#1] SMP
last sysfs file: /sys/kernel/mm/ksm/run
CPU 0
Pid: 2436, comm: master Not tainted 2.6.33.3-85.fc13.x86_64 #1 0HR330/OptiPlex
745
RIP: 0010:[<ffffffff81069881>]  [<ffffffff81069881>] __put_cred+0xc/0x45
RSP: 0018:ffff88019e7e9eb8  EFLAGS: 00010202
RAX: 0000000000000001 RBX: ffff880161514480 RCX: 00000000ffffffff
RDX: 00000000ffffffff RSI: ffff880140c690c0 RDI: ffff880140c690c0
RBP: ffff88019e7e9eb8 R08: 00000000000000d0 R09: 0000000000000000
R10: 0000000000000001 R11: 0000000000000040 R12: ffff880140c690c0
R13: ffff88019e77aea0 R14: 00007fff336b0a5c R15: 0000000000000001
FS:  00007f12f50d97c0(0000) GS:ffff880007400000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f8f461bc000 CR3: 00000001b26ce000 CR4: 00000000000006f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process master (pid: 2436, threadinfo ffff88019e7e8000, task ffff88019e77aea0)
Stack:
 ffff88019e7e9ec8 ffffffff810698cd ffff88019e7e9ef8 ffffffff81069b45
<0> ffff880161514180 ffff880161514480 ffff880161514180 0000000000000000
<0> ffff88019e7e9f28 ffffffff8106aace 0000000000000001 0000000000000246
Call Trace:
 [<ffffffff810698cd>] put_cred+0x13/0x15
 [<ffffffff81069b45>] commit_creds+0x16b/0x175
 [<ffffffff8106aace>] set_current_groups+0x47/0x4e
 [<ffffffff8106ac89>] sys_setgroups+0xf6/0x105
 [<ffffffff81009b02>] system_call_fastpath+0x16/0x1b
Code: 48 8d 71 ff e8 7e 4e 15 00 85 c0 78 0b 8b 75 ec 48 89 df e8 ef 4a 15 00
48 83 c4 18 5b c9 c3 55 8b 07 8b 07 48 89 e5 85 c0 74 04 <0f> 0b eb fe 65 48 8b
04 25 00 cc 00 00 48 3b b8 58 04 00 00 75
RIP  [<ffffffff81069881>] __put_cred+0xc/0x45
 RSP <ffff88019e7e9eb8>
---[ end trace df391256a100ebdd ]---

Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Jiri Olsa <jolsa@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agowatchdog: update MAINTAINERS entry
Wim Van Sebroeck [Thu, 29 Jul 2010 18:02:51 +0000 (18:02 +0000)]
watchdog: update MAINTAINERS entry

Add Mailing-list and website to watchdog MAINTAINERS entry.

Signed-off-by: Wim Van Sebroeck <wim@iguana.be>
14 years agoALSA: hda - Add a PC-beep workaround for ASUS P5-V
Takashi Iwai [Thu, 29 Jul 2010 13:30:02 +0000 (15:30 +0200)]
ALSA: hda - Add a PC-beep workaround for ASUS P5-V

ASUS P5-V provides a SSID that unexpectedly matches with the value
compilant with Realtek's specification.  Thus the driver interprets
it badly, resulting in non-working PC beep.

This patch adds a white-list for such a case; a white-list of known
devices with working PC beep.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
14 years agoARM: Add barriers to io{read,write}{8,16,32} accessors as well
Russell King [Thu, 29 Jul 2010 10:38:05 +0000 (11:38 +0100)]
ARM: Add barriers to io{read,write}{8,16,32} accessors as well

The ioread/iowrite accessors also need barriers as they're used in
place of readl/writel et.al. in portable drivers.  Create __iormb()
and __iowmb() which are conditionally defined to be barriers dependent
on ARM_DMA_MEM_BUFFERABLE, and always use these macros in the accessors.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>