]> git.karo-electronics.de Git - karo-tx-linux.git/log
karo-tx-linux.git
12 years agorbtree: adjust root color in rb_insert_color() only when necessary
Michel Lespinasse [Sat, 21 Jul 2012 00:54:56 +0000 (10:54 +1000)]
rbtree: adjust root color in rb_insert_color() only when necessary

The root node of an rbtree must always be black.  However,
rb_insert_color() only needs to maintain this invariant when it has been
broken - that is, when it exits the loop due to the current (red) node
being the root.  In all other cases (exiting after tree rotations, or
exiting due to an existing black parent) the invariant is already
satisfied, so there is no need to adjust the root node color.

Signed-off-by: Michel Lespinasse <walken@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: David Woodhouse <David.Woodhouse@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Daniel Santos <daniel.santos@pobox.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agorbtree: break out of rb_insert_color loop after tree rotation
Michel Lespinasse [Sat, 21 Jul 2012 00:54:56 +0000 (10:54 +1000)]
rbtree: break out of rb_insert_color loop after tree rotation

It is a well known property of rbtrees that insertion never requires more
than two tree rotations.  In our implementation, after one loop iteration
identified one or two necessary tree rotations, we would iterate and look
for more.  However at that point the node's parent would always be black,
which would cause us to exit the loop.

We can make the code flow more obvious by just adding a break statement
after the tree rotations, where we know we are done.  Additionally, in the
cases where two tree rotations are necessary, we don't have to update the
'node' pointer as it wouldn't be used until the next loop iteration, which
we now avoid due to this break statement.

Signed-off-by: Michel Lespinasse <walken@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: David Woodhouse <David.Woodhouse@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Daniel Santos <daniel.santos@pobox.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agorbtree: performance and correctness test
Michel Lespinasse [Sat, 21 Jul 2012 00:54:56 +0000 (10:54 +1000)]
rbtree: performance and correctness test

This small module helps measure the performance of rbtree insert and
erase.

Additionally, we run a few correctness tests to check that the rbtrees
have all desired properties:

- contains the right number of nodes in the order desired,
- never two consecutive red nodes on any path,
- all paths to leaf nodes have the same number of black nodes,
- root node is black

Signed-off-by: Michel Lespinasse <walken@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: David Woodhouse <David.Woodhouse@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Daniel Santos <daniel.santos@pobox.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agorbtree: fix jffs2 build issue due to renamed __rb_parent_color field
David Woodhouse [Sat, 21 Jul 2012 00:54:55 +0000 (10:54 +1000)]
rbtree: fix jffs2 build issue due to renamed __rb_parent_color field

... and clean up the comments to better explain why it's acceptable to
do it this way instead of using rb_erase() "properly".

Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Cc: Michel Lespinasse <walken@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agorbtree: move some implementation details from rbtree.h to rbtree.c
Michel Lespinasse [Sat, 21 Jul 2012 00:54:55 +0000 (10:54 +1000)]
rbtree: move some implementation details from rbtree.h to rbtree.c

rbtree users must use the documented APIs to manipulate the tree
structure.  Low-level helpers to manipulate node colors and parenthood are
not part of that API, so move them to lib/rbtree.c

Signed-off-by: Michel Lespinasse <walken@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: David Woodhouse <David.Woodhouse@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Daniel Santos <daniel.santos@pobox.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agorbtree: fix incorrect rbtree node insertion in fs/proc/proc_sysctl.c
Michel Lespinasse [Sat, 21 Jul 2012 00:54:55 +0000 (10:54 +1000)]
rbtree: fix incorrect rbtree node insertion in fs/proc/proc_sysctl.c

The recently added code to use rbtrees in sysctl did not follow the proper
rbtree interface on insertion - it was calling rb_link_node() which
inserts a new node into the binary tree, but missed the call to
rb_insert_color() which properly balances the rbtree and establishes all
expected rbtree invariants.

I found out about this only because faulty commit also used
rb_init_node(), which I am removing within this patchset.  But I think
it's an easy mistake to make, and it makes me wonder if we should change
the rbtree API so that insertions would be done with a single rb_insert()
call (even if its implementation could still inline the rb_link_node()
part and call a private __rb_insert_color function to do the rebalancing).

Signed-off-by: Michel Lespinasse <walken@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: David Woodhouse <David.Woodhouse@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Daniel Santos <daniel.santos@pobox.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agorbtree: empty nodes have no color
Michel Lespinasse [Sat, 21 Jul 2012 00:54:54 +0000 (10:54 +1000)]
rbtree: empty nodes have no color

Empty nodes have no color.  We can make use of this property to simplify
the code emitted by the RB_EMPTY_NODE and RB_CLEAR_NODE macros.  Also, we
can get rid of the rb_init_node function which had been introduced by
88d19cf37952 ("timers: Add rb_init_node() to allow for stack allocated rb
nodes") to avoid some issue with the empty node's color not being
initialized.

I'm not sure what the RB_EMPTY_NODE checks in rb_prev() / rb_next() are
doing there, though.  axboe introduced them in 10fd48f2376d ("rbtree:
fixed reversed RB_EMPTY_NODE and rb_next/prev").  The way I see it, the
'empty node' abstraction is only used by rbtree users to flag nodes that
they haven't inserted in any rbtree, so asking the predecessor or
successor of such nodes doesn't make any sense.

One final rb_init_node() caller was recently added in sysctl code to
implement faster sysctl name lookups.  This code doesn't make use of
RB_EMPTY_NODE at all, and from what I could see it only called
rb_init_node() under the mistaken assumption that such initialization was
required before node insertion.

Signed-off-by: Michel Lespinasse <walken@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: David Woodhouse <David.Woodhouse@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Daniel Santos <daniel.santos@pobox.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: John Stultz <john.stultz@linaro.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agorbtree: reference Documentation/rbtree.txt for usage instructions
Michel Lespinasse [Sat, 21 Jul 2012 00:54:54 +0000 (10:54 +1000)]
rbtree: reference Documentation/rbtree.txt for usage instructions

I recently started looking at the rbtree code (with an eye towards
improving the augmented rbtree support, but I haven't gotten there yet).
I noticed a lot of possible speed improvements, which I am now proposing
in this patch set.

Patches 1-4 are preparatory: remove internal functions from rbtree.h so
that users won't be tempted to use them instead of the documented APIs,
clean up some incorrect usages I've noticed (in particular, with the
recently added fs/proc/proc_sysctl.c rbtree usage), reference the
documentation so that people have one less excuse to miss it, etc.

Patch 5 is a small module I wrote to check the rbtree performance.  It
creates 100 nodes with random keys and repeatedly inserts and erases them
from an rbtree.  Additionally, it has code to check for rbtree invariants
after each insert or erase operation.

Patches 6-12 is where the rbtree optimizations are done, and they touch
only that one file, lib/rbtree.c .  I am getting good results out of these
- in my small benchmark doing rbtree insertion (including search) and
erase, I'm seeing a 30% runtime reduction on Sandybridge E5, which is more
than I initially thought would be possible.  (the results aren't as
impressive on my two other test hosts though, AMD barcelona and Intel
Westmere, where I am seeing 14% runtime reduction only).  The code size -
both source (ommiting comments) and compiled - is also shorter after these
changes.  However, I do admit that the updated code is more arduous to
read - one big reason for that is the removal of the tree rotation
helpers, which added some overhead but also made it easier to reason about
things locally.  Overall, I believe this is an acceptable compromise,
given that this code doesn't get modified very often, and that I have good
tests for it.

Upon Peter's suggestion, I added comments showing the rtree configuration
before every rotation.  I think they help; however it's still best to have
a copy of the cormen/leiserson/rivest book when digging into this code.

This patch: reference Documentation/rbtree.txt for usage instructions

include/linux/rbtree.h included some basic usage instructions, while
Documentation/rbtree.txt had some more complete and easier to follow
instructions.  Replacing the former with a reference to the latter.

Signed-off-by: Michel Lespinasse <walken@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: David Woodhouse <David.Woodhouse@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Daniel Santos <daniel.santos@pobox.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoipc/mqueue: remove unnecessary rb_init_node() calls
Michel Lespinasse [Sat, 21 Jul 2012 00:54:54 +0000 (10:54 +1000)]
ipc/mqueue: remove unnecessary rb_init_node() calls

d6629859 ("ipc/mqueue: improve performance of send/recv") and ce2d52cc
("ipc/mqueue: add rbtree node caching support") introduced an rbtree of
message priorities, and usage of rb_init_node() to initialize the
corresponding nodes.  As it turns out, rb_init_node() is unnecessary here,
as the nodes are fully initialized on insertion by rb_link_node() and the
code doesn't access nodes that aren't inserted on the rbtree.

Removing the rb_init_node() calls as I removed that function during
rbtree API cleanups (the only other use of it was in a place that similarly
didn't require it).

Signed-off-by: Michel Lespinasse <walken@google.com>
Acked-by: Doug Ledford <dledford@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoext4: use memweight()
Akinobu Mita [Sat, 21 Jul 2012 00:54:54 +0000 (10:54 +1000)]
ext4: use memweight()

Convert ext4_count_free() to use memweight() instead of table lookup based
counting clear bits implementation.  This change only affects the code
segments enabled by EXT4FS_DEBUG.

Note that this memweight() call can't be replaced with a single
bitmap_weight() call, although the pointer to the memory area is aligned
to long-word boundary.  Because the size of the memory area may not be a
multiple of BITS_PER_LONG, then it returns wrong value on big-endian
architecture.

This also includes the following change.

- Remove unnecessary map == NULL check in ext4_count_free() which
  always takes non-null pointer as the memory area.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoext3: use memweight()
Akinobu Mita [Sat, 21 Jul 2012 00:54:53 +0000 (10:54 +1000)]
ext3: use memweight()

Convert ext3_count_free() to use memweight() instead of table lookup based
counting clear bits implementation.  This change only affects the code
segments enabled by EXT3FS_DEBUG.

Note that this memweight() call can't be replaced with a single
bitmap_weight() call, although the pointer to the memory area is aligned
to long-word boundary.  Because the size of the memory area may not be a
multiple of BITS_PER_LONG, then it returns wrong value on big-endian
architecture.

This also includes the following changes.

- Remove unnecessary map == NULL check in ext3_count_free() which
  always takes non-null pointer as the memory area.

- Fix printk format warning that only reveals with EXT3FS_DEBUG.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Acked-by: Jan Kara <jack@suse.cz>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoext2: use memweight()
Akinobu Mita [Sat, 21 Jul 2012 00:54:53 +0000 (10:54 +1000)]
ext2: use memweight()

Convert ext2_count_free() to use memweight() instead of table lookup based
counting clear bits implementation.  This change only affects the code
segments enabled by EXT2FS_DEBUG.

Note that this memweight() call can't be replaced with a single
bitmap_weight() call, although the pointer to the memory area is aligned
to long-word boundary.  Because the size of the memory area may not be a
multiple of BITS_PER_LONG, then it returns wrong value on big-endian
architecture.

This also includes the following changes.

- Remove unnecessary map == NULL check in ext2_count_free() which
  always takes non-null pointer as the memory area.

- Fix printk format warning that only reveals with EXT2FS_DEBUG.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Acked-by: Jan Kara <jack@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoocfs2: use memweight()
Akinobu Mita [Sat, 21 Jul 2012 00:54:53 +0000 (10:54 +1000)]
ocfs2: use memweight()

Use memweight to count the total number of bits set in memory area.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agovideo/uvc: use memweight()
Akinobu Mita [Sat, 21 Jul 2012 00:54:52 +0000 (10:54 +1000)]
video/uvc: use memweight()

Use memweight() to count the total number of bits set in memory area.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Acked-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoaffs: use memweight()
Akinobu Mita [Sat, 21 Jul 2012 00:54:52 +0000 (10:54 +1000)]
affs: use memweight()

Use memweight() to count the total number of bits set in memory area.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agodm: use memweight()
Akinobu Mita [Sat, 21 Jul 2012 00:54:52 +0000 (10:54 +1000)]
dm: use memweight()

Use memweight() to count the total number of bits set in memory area.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Cc: Alasdair Kergon <agk@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoqnx4fs: use memweight()
Akinobu Mita [Sat, 21 Jul 2012 00:54:51 +0000 (10:54 +1000)]
qnx4fs: use memweight()

Use memweight() to count the total number of bits clear in memory area.

Note that this memweight() call can't be replaced with a single
bitmap_weight() call, although the pointer to the memory area is aligned
to long-word boundary.  Because the size of the memory area may not be a
multiple of BITS_PER_LONG, then it returns wrong value on big-endian
architecture.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Acked-by: Anders Larsen <al@alarsen.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agostring: fix build error caused by memweight() introduction
Akinobu Mita [Sat, 21 Jul 2012 00:54:51 +0000 (10:54 +1000)]
string: fix build error caused by memweight() introduction

Tony Luck reports a build error on the ia64 sim_defconfig:

 LD      arch/ia64/hp/sim/boot/bootloader
lib/lib.a(string.o): In function `bitmap_weight':
.../linux-next/include/linux/bitmap.h:280: undefined reference to `__bitmap_weight'

It fails because it pulls in lib/lib.a(string.o) to get some
innocuous function like strcpy() ... but it also gets
given memweight() which relies on __bitmap_weight()
which it doesn't have, because it doesn't include lib/built-in.o
(which is where bitmap.o, the definer of __bitmap_weight(), has
been linked).

This build error is introduced by the patch string-introduce-memweight.patch
in -mm tree.  Fix it by creating own file lib/memweight.c.

Reported-by: Tony Luck <tony.luck@gmail.com>
Suggested-by: Tony Luck <tony.luck@gmail.com>
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agostring-introduce-memweight-fix
Andrew Morton [Sat, 21 Jul 2012 00:54:51 +0000 (10:54 +1000)]
string-introduce-memweight-fix

rename `w' to `ret'

Cc: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agostring: introduce memweight()
Akinobu Mita [Sat, 21 Jul 2012 00:54:50 +0000 (10:54 +1000)]
string: introduce memweight()

memweight() is the function that counts the total number of bits set in
memory area.  Unlike bitmap_weight(), memweight() takes pointer and size
in bytes to specify a memory area which does not need to be aligned to
long-word boundary.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Cc: Anders Larsen <al@alarsen.net>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Cc: Mark Fasheh <mfasheh@suse.com>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Andreas Dilger <adilger.kernel@dilger.ca>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Matthew Wilcox <matthew@wil.cx>
Cc: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobacklight: move lp855x header into platform_data directory
Kim, Milo [Sat, 21 Jul 2012 00:54:50 +0000 (10:54 +1000)]
backlight: move lp855x header into platform_data directory

The lp855x header is used only in the platform side, so it can be moved
into platform_data directory

Signed-off-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Cc: Bryan Wu <bryan.wu@canonical.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobacklight: move register definitions from header to source
Kim, Milo [Sat, 21 Jul 2012 00:54:50 +0000 (10:54 +1000)]
backlight: move register definitions from header to source

ROM boundary definitions are no need to be exported because these are used
= only in lp855x driver internally.

And few code cosmetic changes

Signed-off-by: Milo(Woogyom) Kim <milo.kim@ti.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Cc: Bryan Wu <bryan.wu@canonical.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobacklight: l4f00242t03: export and use devm_gpio_request_one()
Jingoo Han [Sat, 21 Jul 2012 00:54:49 +0000 (10:54 +1000)]
backlight: l4f00242t03: export and use devm_gpio_request_one()

The devm_ functions allocate memory that is released when a driver
detaches.  This patch uses devm_gpio_request_one() for these functions.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Cc: Alberto Panizzo <alberto@amarulasolutions.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Cc: Grant Likely <grant.likely@secretlab.ca>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobacklight: corgi_lcd: use devm_gpio_request()
Jingoo Han [Sat, 21 Jul 2012 00:54:49 +0000 (10:54 +1000)]
backlight: corgi_lcd: use devm_gpio_request()

The devm_ functions allocate memory that is released when a driver
detaches.  This patch uses devm_gpio_request() for these functions.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Cc: Eric Miao <eric.y.miao@gmail.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobacklight: lms283gf05: use devm_gpio_request()
Jingoo Han [Sat, 21 Jul 2012 00:54:49 +0000 (10:54 +1000)]
backlight: lms283gf05: use devm_gpio_request()

The devm_ functions allocate memory that is released when a driver
detaches.  This patch uses devm_gpio_request() for these functions.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Acked-by: Marek Vasut <marek.vasut@gmail.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobacklight: tosa_bl: use devm_gpio_request()
Jingoo Han [Sat, 21 Jul 2012 00:54:49 +0000 (10:54 +1000)]
backlight: tosa_bl: use devm_gpio_request()

The devm_ functions allocate memory that is released when a driver
detaches.  This patch uses devm_gpio_request() for these functions.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Cc: Dmitry Baryshkov <dbaryshkov@gmail.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobacklight: tosa_lcd: use devm_gpio_request()
Jingoo Han [Sat, 21 Jul 2012 00:54:48 +0000 (10:54 +1000)]
backlight: tosa_lcd: use devm_gpio_request()

The devm_ functions allocate memory that is released when a driver
detaches.  This patch uses devm_gpio_request() for these functions.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Cc: Dmitry Baryshkov <dbaryshkov@gmail.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobacklight: ot200_bl: use devm_gpio_request()
Jingoo Han [Sat, 21 Jul 2012 00:54:48 +0000 (10:54 +1000)]
backlight: ot200_bl: use devm_gpio_request()

The devm_ functions allocate memory that is released when a driver
detaches.  This patch uses devm_gpio_request() for these functions.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Acked-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobacklight: atmel-pwm-bl: use devm_gpio_request()
Jingoo Han [Sat, 21 Jul 2012 00:54:48 +0000 (10:54 +1000)]
backlight: atmel-pwm-bl: use devm_gpio_request()

The devm_ functions allocate memory that is released when a driver
detaches.  This patch uses devm_gpio_request() for these functions.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Cc: Hans-Christian Egtvedt <egtvedt@samfundet.no>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agodrivers/video/backlight/lm3533_bl.c: use devm_ functions
Jingoo Han [Sat, 21 Jul 2012 00:54:47 +0000 (10:54 +1000)]
drivers/video/backlight/lm3533_bl.c: use devm_ functions

The devm_ functions allocate memory that is released when a driver
detaches.  This patch uses devm_kzalloc of these functions.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Acked-by: Johan Hovold <jhovold@gmail.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agodrivers/video/backlight/ot200_bl.c: use devm_ functions
Jingoo Han [Sat, 21 Jul 2012 00:54:47 +0000 (10:54 +1000)]
drivers/video/backlight/ot200_bl.c: use devm_ functions

The devm_ functions allocate memory that is released when a driver
detaches. This patch uses devm_kzalloc of these functions

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Cc: Christian Gmeiner <christian.gmeiner@gmail.com>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agodrivers/video/backlight/atmel-pwm-bl.c: use devm_ functions
Jingoo Han [Sat, 21 Jul 2012 00:54:47 +0000 (10:54 +1000)]
drivers/video/backlight/atmel-pwm-bl.c: use devm_ functions

The devm_ functions allocate memory that is released when a driver
detaches.  This patch uses devm_kzalloc of these functions.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Acked-by: Hans-Christian Egtvedt <egtvedt@samfundet.no>
Cc: Richard Purdie <rpurdie@rpsys.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoMAINTAINERS: update EXYNOS DP DRIVER F: patterns
Jingoo Han [Sat, 21 Jul 2012 00:54:46 +0000 (10:54 +1000)]
MAINTAINERS: update EXYNOS DP DRIVER F: patterns

Add patterns for Exynos DP header to MAINTAINERS file.

Signed-off-by: Jingoo Han <jg1.han@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agolib/vsprintf.c: kptr_restrict: fix pK-error in SysRq show-all-timers(Q)
Dan Rosenberg [Sat, 21 Jul 2012 00:54:46 +0000 (10:54 +1000)]
lib/vsprintf.c: kptr_restrict: fix pK-error in SysRq show-all-timers(Q)

When using ALT+SysRq+Q all the pointers are replaced with "pK-error" like
this:

[23153.208033]   .base:               pK-error

with echo h > /proc/sysrq-trigger it works:

[23107.776363]   .base:       ffff88023e60d540

The intent behind this behavior was to return "pK-error" in cases where
the %pK format specifier was used in interrupt context, because the
CAP_SYSLOG check wouldn't be meaningful.  Clearly this should only apply
when kptr_restrict is actually enabled though.

Reported-by: Stevie Trujillo <stevie.trujillo@gmail.com>
Signed-off-by: Dan Rosenberg <dan.j.rosenberg@gmail.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agolib/vsprintf.c: remind people to update Documentation/printk-formats.txt when adding...
Andrew Morton [Sat, 21 Jul 2012 00:54:46 +0000 (10:54 +1000)]
lib/vsprintf.c: remind people to update Documentation/printk-formats.txt when adding printk formats

Cc: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoDocumentation/printk-formats.txt: add description for %pMR
Andrei Emeltchenko [Sat, 21 Jul 2012 00:54:45 +0000 (10:54 +1000)]
Documentation/printk-formats.txt: add description for %pMR

Add information about new specifier %pMR for Bluetooth addresses.

Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
Cc: Joe Perches <joe@perches.com>
Cc: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agovsprintf: add %pMR for Bluetooth MAC address
Andrei Emeltchenko [Sat, 21 Jul 2012 00:54:45 +0000 (10:54 +1000)]
vsprintf: add %pMR for Bluetooth MAC address

Bluetooth uses mostly LE byte order which is reversed for visual
interpretation.  Currently in Bluetooth in use unsafe batostr function.

This is a slightly modified version of Joe's patch (sent Sat, Dec 4,
2010).

Signed-off-by: Andrei Emeltchenko <andrei.emeltchenko@intel.com>
Cc: Joe Perches <joe@perches.com>
Cc: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoprintk: remove the now unnecessary "C" annotation for KERN_CONT
Joe Perches [Sat, 21 Jul 2012 00:54:45 +0000 (10:54 +1000)]
printk: remove the now unnecessary "C" annotation for KERN_CONT

Now that all KERN_<LEVEL> uses are prefixed with ASCII SOH, there is no
need for a KERN_CONT.  Keep it backward compatible by adding #define
KERN_CONT ""

Reduces kernel image size a thousand bytes.

Signed-off-by: Joe Perches <joe@perches.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoprintk: only look for prefix levels in kernel messages
Joe Perches [Sat, 21 Jul 2012 00:54:44 +0000 (10:54 +1000)]
printk: only look for prefix levels in kernel messages

vprintk_emit() prefix parsing should only be done for internal kernel
messages.  This allows existing behavior to be kept in all cases.

Signed-off-by: Joe Perches <joe@perches.com>
Cc: Kay Sievers <kay@vrfy.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoprintk: convert the format for KERN_<LEVEL> to a 2 byte pattern
Joe Perches [Sat, 21 Jul 2012 00:54:44 +0000 (10:54 +1000)]
printk: convert the format for KERN_<LEVEL> to a 2 byte pattern

Instead of "<.>", use an ASCII SOH for the KERN_<LEVEL> prefix initiator.

This saves 1 byte per printk, thousands of bytes in a normal kernel.

No output changes are produced as vprintk_emit converts these uses to
"<.>".

Signed-off-by: Joe Perches <joe@perches.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agosound: use printk_get_level and printk_skip_level
Joe Perches [Sat, 21 Jul 2012 00:54:44 +0000 (10:54 +1000)]
sound: use printk_get_level and printk_skip_level

Make the output logging routine independent of the KERN_<LEVEL> style.

Signed-off-by: Joe Perches <joe@perches.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Cc: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobtrfs-use-printk_get_level-and-printk_skip_level-add-__printf-fix-fallout-checkpatch...
Andrew Morton [Sat, 21 Jul 2012 00:54:44 +0000 (10:54 +1000)]
btrfs-use-printk_get_level-and-printk_skip_level-add-__printf-fix-fallout-checkpatch-fixes

Cc: Chris Mason <chris.mason@oracle.com>
WARNING: static const char * array should probably be static const char * const
#95: FILE: fs/btrfs/super.c:170:
+static const char *logtypes[] = {

total: 0 errors, 1 warnings, 115 lines checked

./patches/btrfs-use-printk_get_level-and-printk_skip_level-add-__printf-fix-fallout.patch has style problems, please review.

If any of these errors are false positives, please report
them to the maintainer, see CHECKPATCH in MAINTAINERS.

Please run checkpatch prior to sending patches

Cc: Chris Mason <chris.mason@oracle.com>
Cc: Joe Perches <joe@perches.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobtrfs-use-printk_get_level-and-printk_skip_level-add-__printf-fix-fallout-fix
Andrew Morton [Sat, 21 Jul 2012 00:54:43 +0000 (10:54 +1000)]
btrfs-use-printk_get_level-and-printk_skip_level-add-__printf-fix-fallout-fix

whitespace tweak.  checkpatch bug ;)

Cc: Joe Perches <joe@perches.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agobtrfs: use printk_get_level and printk_skip_level, add __printf, fix fallout
Joe Perches [Sat, 21 Jul 2012 00:54:43 +0000 (10:54 +1000)]
btrfs: use printk_get_level and printk_skip_level, add __printf, fix fallout

Use the generic printk_get_level() to search a message for a kern_level.
Add __printf to verify format and arguments.  Fix a few messages that had
mismatches in format and arguments.  Add #ifdef CONFIG_PRINTK blocks to
shrink the object size a bit when not using printk.

Signed-off-by: Joe Perches <joe@perches.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Cc: Chris Mason <chris.mason@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoarch: remove direct definitions of KERN_<LEVEL> uses
Joe Perches [Sat, 21 Jul 2012 00:54:43 +0000 (10:54 +1000)]
arch: remove direct definitions of KERN_<LEVEL> uses

Add #include <linux/kern_levels.h> so that the #define KERN_<LEVEL> macros
don't have to be duplicated.

Signed-off-by: Joe Perches <joe@perches.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Kay Sievers <kay@vrfy.org>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoprintk: add kern_levels.h to make KERN_<LEVEL> available for asm use
Joe Perches [Sat, 21 Jul 2012 00:54:42 +0000 (10:54 +1000)]
printk: add kern_levels.h to make KERN_<LEVEL> available for asm use

Separate the printk.h file into 2 pieces so the definitions can be used in
asm files.

Signed-off-by: Joe Perches <joe@perches.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoprintk-add-generic-functions-to-find-kern_level-headers-fix
Andrew Morton [Sat, 21 Jul 2012 00:54:42 +0000 (10:54 +1000)]
printk-add-generic-functions-to-find-kern_level-headers-fix

fix build error and warning

kernel/printk.c:1357: error: label at end of compound statement
kernel/printk.c:1360: warning: assignment discards qualifiers from pointer target type

Cc: Joe Perches <joe@perches.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Cc: Wu Fengguang <wfg@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoprintk: add generic functions to find KERN_<LEVEL> headers
Joe Perches [Sat, 21 Jul 2012 00:54:42 +0000 (10:54 +1000)]
printk: add generic functions to find KERN_<LEVEL> headers

The current form of a KERN_<LEVEL> is "<.>".

Add printk_get_level and printk_skip_level functions to handle these
formats.

These functions centralize tests of KERN_<LEVEL> so a future modification
can change the KERN_<LEVEL> style and shorten the number of bytes consumed
by these headers.

Signed-off-by: Joe Perches <joe@perches.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agokmsg: /dev/kmsg - properly return possible copy_from_user() failure
Kay Sievers [Sat, 21 Jul 2012 00:54:41 +0000 (10:54 +1000)]
kmsg: /dev/kmsg - properly return possible copy_from_user() failure

Reported-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Kay Sievers <kay@vrfy.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agokernel/sys.c: avoid argv_free(NULL)
Andrew Morton [Sat, 21 Jul 2012 00:54:41 +0000 (10:54 +1000)]
kernel/sys.c: avoid argv_free(NULL)

If argv_split() failed, the code will end up calling argv_free(NULL).  Fix
it up and clean things up a bit.

Addresses Coverity report 703573.

Cc: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: Kees Cook <keescook@chromium.org>
Cc: Serge Hallyn <serge.hallyn@canonical.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: WANG Cong <xiyou.wangcong@gmail.com>
Cc: Alan Cox <alan@linux.intel.com>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoNMI watchdog: fix for lockup detector breakage on resume
Sameer Nanda [Sat, 21 Jul 2012 00:54:41 +0000 (10:54 +1000)]
NMI watchdog: fix for lockup detector breakage on resume

On the suspend/resume path the boot CPU does not go though an
offline->online transition.  This breaks the NMI detector post-resume
since it depends on PMU state that is lost when the system gets suspended.

Fix this by forcing a CPU offline->online transition for the lockup
detector on the boot CPU during resume.

To provide more context, we enable NMI watchdog on Chrome OS.  We have
seen several reports of systems freezing up completely which indicated
that the NMI watchdog was not firing for some reason.

Debugging further, we found a simple way of repro'ing system freezes --
issuing the command 'tasket 1 sh -c "echo nmilockup > /proc/breakme"'
after the system has been suspended/resumed one or more times.

With this patch in place, the system freeze result in panics, as expected.
 These panics provide a nice stack trace for us to debug the actual issue
causing the freeze.

[akpm@linux-foundation.org: fiddle with code comment]
[akpm@linux-foundation.org: make lockup_detector_bootcpu_resume() conditional on CONFIG_SUSPEND]
[akpm@linux-foundation.org: fix section errors]
Signed-off-by: Sameer Nanda <snanda@chromium.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Mandeep Singh Baines <msb@chromium.org>
Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Cc: Anshuman Khandual <khandual@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agopanic: fix a possible deadlock in panic()
Vikram Mulukutla [Sat, 21 Jul 2012 00:54:40 +0000 (10:54 +1000)]
panic: fix a possible deadlock in panic()

panic_lock is meant to ensure that panic processing takes place only on
one cpu; if any of the other cpus encounter a panic, they will spin
waiting to be shut down.

However, this causes a regression in this scenario:

1. Cpu 0 encounters a panic and acquires the panic_lock
   and proceeds with the panic processing.
2. There is an interrupt on cpu 0 that also encounters
   an error condition and invokes panic.
3. This second invocation fails to acquire the panic_lock
   and enters the infinite while loop in panic_smp_self_stop.

Thus all panic processing is stopped, and the cpu is stuck for eternity in
the while(1) inside panic_smp_self_stop.

To address this, disable local interrupts with local_irq_disable before
acquiring the panic_lock.  This will prevent interrupt handlers from
executing during the panic processing, thus avoiding this particular
problem.

Signed-off-by: Vikram Mulukutla <markivx@codeaurora.org>
Reviewed-by: Stephen Boyd <sboyd@codeaurora.org>
Cc: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoclk: validate pointer in __clk_disable()
Fengguang Wu [Sat, 21 Jul 2012 00:54:40 +0000 (10:54 +1000)]
clk: validate pointer in __clk_disable()

clk_get() returns -ENOENT on error and some careless caller might
dereference it without error checking:

In mxc_rnga_remove():

        struct clk *clk = clk_get(&pdev->dev, "rng");

// ...

        clk_disable(clk);

Since it's insane to audit the lots of existing and future clk users,
let's add a check in the callee to avoid kernel panic and warn about
any buggy user.

Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Viresh Kumar <viresh.kumar@st.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoarch/arm/mach-netx/fb.c: reuse dummy clk routines for CONFIG_HAVE_CLK=n
Viresh Kumar [Sat, 21 Jul 2012 00:54:40 +0000 (10:54 +1000)]
arch/arm/mach-netx/fb.c: reuse dummy clk routines for CONFIG_HAVE_CLK=n

mach-netx had its own implementation of clk routines like, clk_get{put},
clk_enable{disable}, etc.  And with introduction of following patchset:

https://lkml.org/lkml/2012/4/24/154

we get compilation error for multiple definition of these routines.

Sascha had following suggestion to deal with it:

http://www.spinics.net/lists/arm-kernel/msg179369.html

So, remove this code completely.

Signed-off-by: Viresh Kumar <viresh.kumar2@arm.com>
Reported-by: Paul Gortmaker <paul.gortmaker@gmail.com>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agousb/host/r8a66597: remove conditional compilation of clk code
Viresh Kumar [Sat, 21 Jul 2012 00:54:40 +0000 (10:54 +1000)]
usb/host/r8a66597: remove conditional compilation of clk code

With addition of dummy clk_*() calls for non CONFIG_HAVE_CLK cases in
clk.h, there is no need to have clk code enclosed in #ifdef
CONFIG_HAVE_CLK, #endif macros.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agogadget/r8a66597: remove conditional compilation of clk code
Viresh Kumar [Sat, 21 Jul 2012 00:54:39 +0000 (10:54 +1000)]
gadget/r8a66597: remove conditional compilation of clk code

With addition of dummy clk_*() calls for non CONFIG_HAVE_CLK cases in
clk.h, there is no need to have clk code enclosed in #ifdef
CONFIG_HAVE_CLK, #endif macros.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agogadget/m66592: remove conditional compilation of clk code
Viresh Kumar [Sat, 21 Jul 2012 00:54:39 +0000 (10:54 +1000)]
gadget/m66592: remove conditional compilation of clk code

With addition of dummy clk_*() calls for non CONFIG_HAVE_CLK cases in
clk.h, there is no need to have clk code enclosed in #ifdef
CONFIG_HAVE_CLK, #endif macros.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agonet/stmmac: remove conditional compilation of clk code
Viresh Kumar [Sat, 21 Jul 2012 00:54:39 +0000 (10:54 +1000)]
net/stmmac: remove conditional compilation of clk code

With addition of dummy clk_*() calls for non CONFIG_HAVE_CLK cases in
clk.h, there is no need to have clk code enclosed in #ifdef
CONFIG_HAVE_CLK, #endif macros.

This also fixes error paths of probe(), as a goto is required in this patch.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Cc: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agonet/c_can: remove conditional compilation of clk code
Viresh Kumar [Sat, 21 Jul 2012 00:54:38 +0000 (10:54 +1000)]
net/c_can: remove conditional compilation of clk code

With addition of dummy clk_*() calls for non CONFIG_HAVE_CLK cases in clk.h,
there is no need to have clk code enclosed in #ifdef CONFIG_HAVE_CLK, #endif
macros.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Bhupesh Sharma <bhupesh.sharma@st.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoata/pata_arasan: remove conditional compilation of clk code
Viresh Kumar [Sat, 21 Jul 2012 00:54:38 +0000 (10:54 +1000)]
ata/pata_arasan: remove conditional compilation of clk code

With addition of dummy clk_*() calls for non CONFIG_HAVE_CLK cases in
clk.h, there is no need to have clk code enclosed in #ifdef
CONFIG_HAVE_CLK, #endif macros.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Cc: Jeff Garzik <jgarzik@redhat.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agousb/musb: remove conditional compilation of clk code
Viresh Kumar [Sat, 21 Jul 2012 00:54:38 +0000 (10:54 +1000)]
usb/musb: remove conditional compilation of clk code

With addition of dummy clk_*() calls for non CONFIG_HAVE_CLK cases in
clk.h, there is no need to have clk code enclosed in #ifdef
CONFIG_HAVE_CLK, #endif macros.

musb also has these dummy macros defined locally.  Remove them as they
aren't required anymore.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agousb/marvell: remove conditional compilation of clk code
Viresh Kumar [Sat, 21 Jul 2012 00:54:37 +0000 (10:54 +1000)]
usb/marvell: remove conditional compilation of clk code

With addition of dummy clk_*() calls for non CONFIG_HAVE_CLK cases in
clk.h, there is no need to have clk code enclosed in #ifdef
CONFIG_HAVE_CLK, #endif macros.

Marvell usb also has these dummy macros defined locally.  Remove them as
they aren't required anymore.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoi2c/i2c-pxa: remove conditional compilation of clk code
Viresh Kumar [Sat, 21 Jul 2012 00:54:37 +0000 (10:54 +1000)]
i2c/i2c-pxa: remove conditional compilation of clk code

With addition of dummy clk_*() calls for non CONFIG_HAVE_CLK cases in
clk.h, there is no need to have clk code enclosed in #ifdef
CONFIG_HAVE_CLK, #endif macros.

pxa i2c also has these dummy macros defined locally.  Remove them as they
aren't required anymore.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Acked-by: Wolfram Sang <w.sang@pengutronix.de>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoclk: remove redundant depends on from drivers/Kconfig
Viresh Kumar [Sat, 21 Jul 2012 00:54:37 +0000 (10:54 +1000)]
clk: remove redundant depends on from drivers/Kconfig

menu "Common Clock Framework" has "depends on COMMON_CLK" and so configs
defined within menu don't require these "depends on COMMON_CLK again".

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Cc: Wolfram Sang <w.sang@pengutronix.de>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Jeff Garzik <jgarzik@redhat.com>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Bhupesh Sharma <bhupesh.sharma@st.com>
Cc: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoclk: add non CONFIG_HAVE_CLK routines
Viresh Kumar [Sat, 21 Jul 2012 00:54:36 +0000 (10:54 +1000)]
clk: add non CONFIG_HAVE_CLK routines

Many drivers are shared between architectures that may or may not have
HAVE_CLK selected for them.  To remove compilation errors for them we
enclose clk_*() calls in these drivers within #ifdef CONFIG_HAVE_CLK,
#endif.

This patch removes the need of these CONFIG_HAVE_CLK statements, by
introducing dummy routines when HAVE_CLK is not selected by platforms.
So, definition of these routines will always be available.  These calls
will return error for platforms that don't select HAVE_CLK.

Signed-off-by: Viresh Kumar <viresh.kumar@st.com>
Cc: Wolfram Sang <w.sang@pengutronix.de>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Jeff Garzik <jgarzik@redhat.com>
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Bhupesh Sharma <bhupesh.sharma@st.com>
Cc: Giuseppe Cavallaro <peppe.cavallaro@st.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Mike Turquette <mturquette@linaro.org>
Cc: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Cc: viresh kumar <viresh.linux@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoavr32-mm-faultc-port-oom-changes-to-do_page_fault-fix
Andrew Morton [Sat, 21 Jul 2012 00:54:36 +0000 (10:54 +1000)]
avr32-mm-faultc-port-oom-changes-to-do_page_fault-fix

fix comment layout

Cc: Hans-Christian Egtvedt <egtvedt@samfundet.no>
Cc: Havard Skinnemoen <hskinnemoen@gmail.com>
Cc: Kautuk Consul <consul.kautuk@gmail.com>
Cc: Mohd. Faris <mohdfarisq2010@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoavr32/mm/fault.c: port OOM changes to do_page_fault
Kautuk Consul [Sat, 21 Jul 2012 00:54:36 +0000 (10:54 +1000)]
avr32/mm/fault.c: port OOM changes to do_page_fault

Commits d065bd810b6de ("mm: retry page fault when blocking on disk
transfer") and 37b23e0525d3 ("x86,mm: make pagefault killable") introduced
changes into the x86 pagefault handler for making the page fault handler
retryable as well as killable.

These changes reduce the mmap_sem hold time, which is crucial during OOM
killer invocation.

Port these changes to AVR32.

Signed-off-by: Mohd. Faris <mohdfarisq2010@gmail.com>
Signed-off-by: Kautuk Consul <consul.kautuk@gmail.com>
Acked-by: Havard Skinnemoen <hskinnemoen@gmail.com>
Cc: Hans-Christian Egtvedt <egtvedt@samfundet.no>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoalpha: remove mysterious if zero-ed out includes
Paul Bolle [Sat, 21 Jul 2012 00:54:36 +0000 (10:54 +1000)]
alpha: remove mysterious if zero-ed out includes

There's a small group of odd looking includes in smc37c669.c.  These
includes appear to be if zero-ed out ever since they were added to the
tree (in v2.1.89).  Their purpose is unclear to me.  Perhaps they were
used in someones build system.  Whatever their purpose was, nothing else
uses something comparable.  This entire if zero-ed out block might as well
be removed.

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agofrv: kill used but uninitialized variable
Geert Uytterhoeven [Sat, 21 Jul 2012 00:54:35 +0000 (10:54 +1000)]
frv: kill used but uninitialized variable

Commit 6afe1a1fe8ff83f6a ("PM: Remove legacy PM") removed the
initialization of retval, causing:

arch/frv/kernel/pm.c: In function 'sysctl_pm_do_suspend':
arch/frv/kernel/pm.c:165:5: warning: 'retval' may be used uninitialized in this function [-Wuninitialized]

Remove the variable completely to fix this, and convert to a proper
switch (...) { ... } construct to improve readability.

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agotmpfs: interleave the starting node of /dev/shmem
Nathan Zimmer [Sat, 21 Jul 2012 00:54:35 +0000 (10:54 +1000)]
tmpfs: interleave the starting node of /dev/shmem

The tmpfs superblock grants an offset for each inode as they are created.
Each inode then uses that offset to provide a preferred first node for its
interleave in the newly provided shmem_interleave.

Signed-off-by: Nathan Zimmer <nzimmer@sgi.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Nick Piggin <npiggin@gmail.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoshmem: provide vm_ops when also providing a mem policy
Nathan Zimmer [Sat, 21 Jul 2012 00:54:35 +0000 (10:54 +1000)]
shmem: provide vm_ops when also providing a mem policy

When tmpfs has the memory policy interleaved it always starts allocating
for each file at node 0.  When there are many small files the lower nodes
fill up disproportionately.

This patchset spreads out node usage by starting files at nodes other then
0.  The tmpfs superblock grants an offset for each inode as they are
created.  Each then uses that offset to proved a prefered first node for
its interleave in the shmem_interleave.

This patch:

Update shmem_get_policy() to use the vma_policy if provided.  This is to
allows us to safely provide shmem_vm_ops to the vma when the vm_file has
not been setup which is the case on the pseudo vmas.

Signed-off-by: Nathan Zimmer <nzimmer@sgi.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Nick Piggin <npiggin@gmail.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomemcg-add-mem_cgroup_from_css-helper-fix
Andrew Morton [Sat, 21 Jul 2012 00:54:34 +0000 (10:54 +1000)]
memcg-add-mem_cgroup_from_css-helper-fix

that patch was very broken

Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomemcg: add mem_cgroup_from_css() helper
Wanpeng Li [Sat, 21 Jul 2012 00:54:34 +0000 (10:54 +1000)]
memcg: add mem_cgroup_from_css() helper

Add a mem_cgroup_from_css() helper to replace open-coded invokations of
container_of().  To clarify the code and to add a little more type safety.

Signed-off-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Cc: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomemcg: further prevent OOM with too many dirty pages
Hugh Dickins [Sat, 21 Jul 2012 00:54:34 +0000 (10:54 +1000)]
memcg: further prevent OOM with too many dirty pages

The may_enter_fs test turns out to be too restrictive: though I saw no
problem with it when testing on 3.5-rc6, it very soon OOMed when I tested
on 3.5-rc6-mm1.  I don't know what the difference there is, perhaps I just
slightly changed the way I started off the testing: dd if=/dev/zero
of=/mnt/temp bs=1M count=1024; rm -f /mnt/temp; sync repeatedly, in 20M
memory.limit_in_bytes cgroup to ext4 on USB stick.

ext4 (and gfs2 and xfs) turn out to allocate new pages for writing with
AOP_FLAG_NOFS: that seems a little worrying, and it's unclear to me why
the transaction needs to be started even before allocating pagecache
memory.  But it may not be worth worrying about these days: if direct
reclaim avoids FS writeback, does __GFP_FS now mean anything?

Anyway, we insisted on the may_enter_fs test to avoid hangs with the loop
device; but since that also masks off __GFP_IO, we can test for __GFP_IO
directly, ignoring may_enter_fs and __GFP_FS.

But even so, the test still OOMs sometimes: when originally testing on
3.5-rc6, it OOMed about one time in five or ten; when testing just now on
3.5-rc6-mm1, it OOMed on the first iteration.

This residual problem comes from an accumulation of pages under ordinary
writeback, not marked PageReclaim, so rightly not causing the memcg check
to wait on their writeback: these too can prevent shrink_page_list() from
freeing any pages, so many times that memcg reclaim fails and OOMs.

Deal with these in the same way as direct reclaim now deals with dirty FS
pages: mark them PageReclaim.  It is appropriate to rotate these to tail
of list when writepage completes, but more importantly, the PageReclaim
flag makes memcg reclaim wait on them if encountered again.  Increment
NR_VMSCAN_IMMEDIATE?  That's arguable: I chose not.

Setting PageReclaim here may occasionally race with end_page_writeback()
clearing it: lru_deactivate_fn() already faced the same race, and
correctly concluded that the window is small and the issue non-critical.

With these changes, the test runs indefinitely without OOMing on ext4,
ext3 and ext2: I'll move on to test with other filesystems later.

Trivia: invert conditions for a clearer block without an else, and goto
keep_locked to do the unlock_page.

Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujtisu.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Ying Han <yinghan@google.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomemcg: prevent OOM with too many dirty pages
Michal Hocko [Sat, 21 Jul 2012 00:54:33 +0000 (10:54 +1000)]
memcg: prevent OOM with too many dirty pages

The current implementation of dirty pages throttling is not memcg aware
which makes it easy to have memcg LRUs full of dirty pages.  Without
throttling, these LRUs can be scanned faster than the rate of writeback,
leading to memcg OOM conditions when the hard limit is small.

This patch fixes the problem by throttling the allocating process
(possibly a writer) during the hard limit reclaim by waiting on
PageReclaim pages.  We are waiting only for PageReclaim pages because
those are the pages that made one full round over LRU and that means that
the writeback is much slower than scanning.

The solution is far from being ideal - long term solution is memcg aware
dirty throttling - but it is meant to be a band aid until we have a real
fix.  We are seeing this happening during nightly backups which are placed
into containers to prevent from eviction of the real working set.

The change affects only memcg reclaim and only when we encounter
PageReclaim pages which is a signal that the reclaim doesn't catch up on
with the writers so somebody should be throttled.  This could be
potentially unfair because it could be somebody else from the group who
gets throttled on behalf of the writer but as writers need to allocate as
well and they allocate in higher rate the probability that only innocent
processes would be penalized is not that high.

I have tested this change by a simple dd copying /dev/zero to tmpfs or
ext3 running under small memcg (1G copy under 5M, 60M, 300M and 2G
containers) and dd got killed by OOM killer every time.  With the patch I
could run the dd with the same size under 5M controller without any OOM.
The issue is more visible with slower devices for output.

* With the patch
================
* tmpfs size=2G
---------------
$ vim cgroup_cache_oom_test.sh
$ ./cgroup_cache_oom_test.sh 5M
using Limit 5M for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 30.4049 s, 34.5 MB/s
$ ./cgroup_cache_oom_test.sh 60M
using Limit 60M for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 31.4561 s, 33.3 MB/s
$ ./cgroup_cache_oom_test.sh 300M
using Limit 300M for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 20.4618 s, 51.2 MB/s
$ ./cgroup_cache_oom_test.sh 2G
using Limit 2G for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 1.42172 s, 738 MB/s

* ext3
------
$ ./cgroup_cache_oom_test.sh 5M
using Limit 5M for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 27.9547 s, 37.5 MB/s
$ ./cgroup_cache_oom_test.sh 60M
using Limit 60M for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 30.3221 s, 34.6 MB/s
$ ./cgroup_cache_oom_test.sh 300M
using Limit 300M for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 24.5764 s, 42.7 MB/s
$ ./cgroup_cache_oom_test.sh 2G
using Limit 2G for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 3.35828 s, 312 MB/s

* Without the patch
===================
* tmpfs size=2G
---------------
$ ./cgroup_cache_oom_test.sh 5M
using Limit 5M for group
./cgroup_cache_oom_test.sh: line 46:  4668 Killed                  dd if=/dev/zero of=$OUT/zero bs=1M count=$count
$ ./cgroup_cache_oom_test.sh 60M
using Limit 60M for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 25.4989 s, 41.1 MB/s
$ ./cgroup_cache_oom_test.sh 300M
using Limit 300M for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 24.3928 s, 43.0 MB/s
$ ./cgroup_cache_oom_test.sh 2G
using Limit 2G for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 1.49797 s, 700 MB/s

* ext3
------
$ ./cgroup_cache_oom_test.sh 5M
using Limit 5M for group
./cgroup_cache_oom_test.sh: line 46:  4689 Killed                  dd if=/dev/zero of=$OUT/zero bs=1M count=$count
$ ./cgroup_cache_oom_test.sh 60M
using Limit 60M for group
./cgroup_cache_oom_test.sh: line 46:  4692 Killed                  dd if=/dev/zero of=$OUT/zero bs=1M count=$count
$ ./cgroup_cache_oom_test.sh 300M
using Limit 300M for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 20.248 s, 51.8 MB/s
$ ./cgroup_cache_oom_test.sh 2G
using Limit 2G for group
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 2.85201 s, 368 MB/s

[akpm@linux-foundation.org: tweak changelog, reordered the test to optimize for CONFIG_CGROUP_MEM_RES_CTLR=n]
[hughd@google.com: fix deadlock with loop driver]
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujtisu.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Ying Han <yinghan@google.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Hugh Dickins <hughd@google.com>
Reviewed-by: Mel Gorman <mgorman@suse.de>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: mmu_notifier: fix freed page still mapped in secondary MMU
Xiao Guangrong [Sat, 21 Jul 2012 00:54:33 +0000 (10:54 +1000)]
mm: mmu_notifier: fix freed page still mapped in secondary MMU

mmu_notifier_release() is called when the process is exiting.  It will
delete all the mmu notifiers.  But at this time the page belonging to the
process is still present in page tables and is present on the LRU list, so
this race will happen:

      CPU 0                 CPU 1
mmu_notifier_release:    try_to_unmap:
   hlist_del_init_rcu(&mn->hlist);
                            ptep_clear_flush_notify:
                                  mmu nofifler not found
                            free page  !!!!!!
                            /*
                             * At the point, the page has been
                             * freed, but it is still mapped in
                             * the secondary MMU.
                             */

  mn->ops->release(mn, mm);

Then the box is not stable and sometimes we can get this bug:

[  738.075923] BUG: Bad page state in process migrate-perf  pfn:03bec
[  738.075931] page:ffffea00000efb00 count:0 mapcount:0 mapping:          (null) index:0x8076
[  738.075936] page flags: 0x20000000000014(referenced|dirty)

The same issue is present in mmu_notifier_unregister().

We can call ->release before deleting the notifier to ensure the page has
been unmapped from the secondary MMU before it is freed.

Signed-off-by: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Cc: Avi Kivity <avi@redhat.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: memcg: only check anon swapin page charges for swap cache
Johannes Weiner [Sat, 21 Jul 2012 00:54:33 +0000 (10:54 +1000)]
mm: memcg: only check anon swapin page charges for swap cache

shmem knows for sure that the page is in swap cache when attempting to
charge a page, because the cache charge entry function has a check for it.
Only anon pages may be removed from swap cache already when trying to
charge their swapin.

Adjust the comment, though: '4969c11 mm: fix swapin race condition' added
a stable PageSwapCache check under the page lock in the do_swap_page()
before calling the memory controller, so it's unuse_pte()'s pte_same()
that may fail.

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: David Rientjes <rientjes@google.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Wanpeng Li <liwp.linux@gmail.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: memcg: only check swap cache pages for repeated charging
Johannes Weiner [Sat, 21 Jul 2012 00:54:32 +0000 (10:54 +1000)]
mm: memcg: only check swap cache pages for repeated charging

Only anon and shmem pages in the swap cache are attempted to be charged
multiple times, from every swap pte fault or from shmem_unuse().  No other
pages require checking PageCgroupUsed().

Charging pages in the swap cache is also serialized by the page lock, and
since both the try_charge and commit_charge are called under the same page
lock section, the PageCgroupUsed() check might as well happen before the
counter charging, let alone reclaim.

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: David Rientjes <rientjes@google.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Wanpeng Li <liwp.linux@gmail.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: memcg: split swapin charge function into private and public part
Johannes Weiner [Sat, 21 Jul 2012 00:54:32 +0000 (10:54 +1000)]
mm: memcg: split swapin charge function into private and public part

When shmem is charged upon swapin, it does not need to check twice whether
the memory controller is enabled.

Also, shmem pages do not have to be checked for everything that regular
anon pages have to be checked for, so let shmem use the internal version
directly and allow future patches to move around checks that are only
required when swapping in anon pages.

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: David Rientjes <rientjes@google.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Wanpeng Li <liwp.linux@gmail.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: memcg: move swapin charge functions above callsites
Johannes Weiner [Sat, 21 Jul 2012 00:54:31 +0000 (10:54 +1000)]
mm: memcg: move swapin charge functions above callsites

Charging cache pages may require swapin in the shmem case.  Save the
forward declaration and just move the swapin functions above the cache
charging functions.

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: David Rientjes <rientjes@google.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Wanpeng Li <liwp.linux@gmail.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: swapfile: clean up unuse_pte race handling
Johannes Weiner [Sat, 21 Jul 2012 00:54:30 +0000 (10:54 +1000)]
mm: swapfile: clean up unuse_pte race handling

The conditional mem_cgroup_cancel_charge_swapin() is a leftover from when
the function would continue to reestablish the page even after
mem_cgroup_try_charge_swapin() failed.  After 85d9fc8 "memcg: fix refcnt
handling at swapoff", the condition is always true when this code is
reached.

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: Hugh Dickins <hughd@google.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Wanpeng Li <liwp.linux@gmail.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoswapfile: avoid dereferencing bd_disk during swap_entry_free for network storage
Mel Gorman [Sat, 21 Jul 2012 00:54:29 +0000 (10:54 +1000)]
swapfile: avoid dereferencing bd_disk during swap_entry_free for network storage

Commit b3a27d ("swap: Add swap slot free callback to
block_device_operations") dereferences p->bdev->bd_disk but this is a NULL
dereference if using swap-over-NFS.  This patch checks SWP_BLKDEV on the
swap_info_struct before dereferencing.

With reference to this callback, Christoph Hellwig stated "Please just
remove the callback entirely.  It has no user outside the staging tree and
was added clearly against the rules for that staging tree".  This would
also be my preference but there was not an obvious way of keeping zram in
staging/ happy.

Signed-off-by: Xiaotian Feng <dfeng@redhat.com>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Paris <eparis@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agonfs: prevent page allocator recursions with swap over NFS.
Mel Gorman [Sat, 21 Jul 2012 00:54:29 +0000 (10:54 +1000)]
nfs: prevent page allocator recursions with swap over NFS.

GFP_NOFS is _more_ permissive than GFP_NOIO in that it will initiate IO,
just not of any filesystem data.

The problem is that previously NOFS was correct because that avoids
recursion into the NFS code.  With swap-over-NFS, it is no longer correct
as swap IO can lead to this recursion.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Paris <eparis@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Neil Brown <neilb@suse.de>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Xiaotian Feng <dfeng@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agonfs: enable swap on NFS
Mel Gorman [Sat, 21 Jul 2012 00:54:29 +0000 (10:54 +1000)]
nfs: enable swap on NFS

Implement the new swapfile a_ops for NFS and hook up ->direct_IO.  This
will set the NFS socket to SOCK_MEMALLOC and run socket reconnect under
PF_MEMALLOC as well as reset SOCK_MEMALLOC before engaging the protocol
->connect() method.

PF_MEMALLOC should allow the allocation of struct socket and related
objects and the early (re)setting of SOCK_MEMALLOC should allow us to
receive the packets required for the TCP connection buildup.

[jlayton@redhat.com: Restore PF_MEMALLOC task flags in all cases]
[dfeng@redhat.com: Fix handling of multiple swap files]
[a.p.zijlstra@chello.nl: Original patch]
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Paris <eparis@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Neil Brown <neilb@suse.de>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Xiaotian Feng <dfeng@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agonfs: disable data cache revalidation for swapfiles
Mel Gorman [Sat, 21 Jul 2012 00:54:28 +0000 (10:54 +1000)]
nfs: disable data cache revalidation for swapfiles

The VM does not like PG_private set on PG_swapcache pages.  As suggested
by Trond in http://lkml.org/lkml/2006/8/25/348, this patch disables NFS
data cache revalidation on swap files.  as it does not make sense to have
other clients change the file while it is being used as swap.  This avoids
setting PG_private on swap pages, since there ought to be no further races
with invalidate_inode_pages2() to deal with.

Since we cannot set PG_private we cannot use page->private which is
already used by PG_swapcache pages to store the nfs_page.  Thus augment
the new nfs_page_find_request logic.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Paris <eparis@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Neil Brown <neilb@suse.de>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Xiaotian Feng <dfeng@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agonfs: teach the NFS client how to treat PG_swapcache pages
Mel Gorman [Sat, 21 Jul 2012 00:54:28 +0000 (10:54 +1000)]
nfs: teach the NFS client how to treat PG_swapcache pages

Replace all relevant occurences of page->index and page->mapping in the
NFS client with the new page_file_index() and page_file_mapping()
functions.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Paris <eparis@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Neil Brown <neilb@suse.de>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Xiaotian Feng <dfeng@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: add support for direct_IO to highmem pages
Mel Gorman [Sat, 21 Jul 2012 00:54:28 +0000 (10:54 +1000)]
mm: add support for direct_IO to highmem pages

The patch "mm: add support for a filesystem to activate swap files and use
direct_IO for writing swap pages" added support for using direct_IO to
write swap pages but it is insufficient for highmem pages.

To support highmem pages, this patch kmaps() the page before calling the
direct_IO() handler.  As direct_IO deals with virtual addresses an
additional helper is necessary for get_kernel_pages() to lookup the struct
page for a kmap virtual address.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Paris <eparis@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Xiaotian Feng <dfeng@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: add get_kernel_page[s] for pinning of kernel addresses for I/O
Mel Gorman [Sat, 21 Jul 2012 00:54:27 +0000 (10:54 +1000)]
mm: add get_kernel_page[s] for pinning of kernel addresses for I/O

This patch adds two new APIs get_kernel_pages() and get_kernel_page() that
may be used to pin a vector of kernel addresses for IO.  The initial user
is expected to be NFS for allowing pages to be written to swap using
aops->direct_IO().  Strictly speaking, swap-over-NFS only needs to pin one
page for IO but it makes sense to express the API in terms of a vector and
add a helper for pinning single pages.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Rik van Riel <riel@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Paris <eparis@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Xiaotian Feng <dfeng@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: swap: implement generic handler for swap_activate
Mel Gorman [Sat, 21 Jul 2012 00:54:27 +0000 (10:54 +1000)]
mm: swap: implement generic handler for swap_activate

The version of swap_activate introduced is sufficient for swap-over-NFS
but would not provide enough information to implement a generic handler.
This patch shuffles things slightly to ensure the same information is
available for aops->swap_activate() as is available to the core.

No functionality change.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Paris <eparis@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Xiaotian Feng <dfeng@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: add support for a filesystem to activate swap files and use direct_IO for writing...
Mel Gorman [Sat, 21 Jul 2012 00:54:27 +0000 (10:54 +1000)]
mm: add support for a filesystem to activate swap files and use direct_IO for writing swap pages

Currently swapfiles are managed entirely by the core VM by using ->bmap to
allocate space and write to the blocks directly.  This effectively ensures
that the underlying blocks are allocated and avoids the need for the swap
subsystem to locate what physical blocks store offsets within a file.

If the swap subsystem is to use the filesystem information to locate the
blocks, it is critical that information such as block groups, block
bitmaps and the block descriptor table that map the swap file were
resident in memory.  This patch adds address_space_operations that the VM
can call when activating or deactivating swap backed by a file.

  int swap_activate(struct file *);
  int swap_deactivate(struct file *);

The ->swap_activate() method is used to communicate to the file that the
VM relies on it, and the address_space should take adequate measures such
as reserving space in the underlying device, reserving memory for mempools
and pinning information such as the block descriptor table in memory.  The
->swap_deactivate() method is called on sys_swapoff() if ->swap_activate()
returned success.

After a successful swapfile ->swap_activate, the swapfile is marked
SWP_FILE and swapper_space.a_ops will proxy to
sis->swap_file->f_mappings->a_ops using ->direct_io to write swapcache
pages and ->readpage to read.

It is perfectly possible that direct_IO be used to read the swap pages but
it is an unnecessary complication.  Similarly, it is possible that
->writepage be used instead of direct_io to write the pages but filesystem
developers have stated that calling writepage from the VM is undesirable
for a variety of reasons and using direct_IO opens up the possibility of
writing back batches of swap pages in the future.

[a.p.zijlstra@chello.nl: Original patch]
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Paris <eparis@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Xiaotian Feng <dfeng@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: methods for teaching filesystems about PG_swapcache pages
Mel Gorman [Sat, 21 Jul 2012 00:54:26 +0000 (10:54 +1000)]
mm: methods for teaching filesystems about PG_swapcache pages

In order to teach filesystems to handle swap cache pages, three new page
functions are introduced:

  pgoff_t page_file_index(struct page *);
  loff_t page_file_offset(struct page *);
  struct address_space *page_file_mapping(struct page *);

page_file_index() - gives the offset of this page in the file in
PAGE_CACHE_SIZE blocks.  Like page->index is for mapped pages, this
function also gives the correct index for PG_swapcache pages.

page_file_offset() - uses page_file_index(), so that it will give the
expected result, even for PG_swapcache pages.

page_file_mapping() - gives the mapping backing the actual page; that is
for swap cache pages it will give swap_file->f_mapping.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Rik van Riel <riel@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Paris <eparis@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Neil Brown <neilb@suse.de>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Xiaotian Feng <dfeng@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agoselinux: tag avc cache alloc as non-critical
Mel Gorman [Sat, 21 Jul 2012 00:54:26 +0000 (10:54 +1000)]
selinux: tag avc cache alloc as non-critical

Failing to allocate a cache entry will only harm performance not
correctness.  Do not consume valuable reserve pages for something like
that.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Eric Paris <eparis@redhat.com>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: David S. Miller <davem@davemloft.net>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Neil Brown <neilb@suse.de>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Xiaotian Feng <dfeng@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agonetvm: prevent a stream-specific deadlock
Mel Gorman [Sat, 21 Jul 2012 00:54:26 +0000 (10:54 +1000)]
netvm: prevent a stream-specific deadlock

This patch series is based on top of "Swap-over-NBD without deadlocking
v15" as it depends on the same reservation of PF_MEMALLOC reserves logic.

When a user or administrator requires swap for their application, they
create a swap partition and file, format it with mkswap and activate it
with swapon.  In diskless systems this is not an option so if swap if
required then swapping over the network is considered.  The two likely
scenarios are when blade servers are used as part of a cluster where the
form factor or maintenance costs do not allow the use of disks and thin
clients.

The Linux Terminal Server Project recommends the use of the Network Block
Device (NBD) for swap but this is not always an option.  There is no
guarantee that the network attached storage (NAS) device is running Linux
or supports NBD.  However, it is likely that it supports NFS so there are
users that want support for swapping over NFS despite any performance
concern.  Some distributions currently carry patches that support swapping
over NFS but it would be preferable to support it in the mainline kernel.

Patch 1 avoids a stream-specific deadlock that potentially affects TCP.

Patch 2 is a small modification to SELinux to avoid using PFMEMALLOC
reserves.

Patch 3 adds three helpers for filesystems to handle swap cache pages.
For example, page_file_mapping() returns page->mapping for
file-backed pages and the address_space of the underlying
swap file for swap cache pages.

Patch 4 adds two address_space_operations to allow a filesystem
to pin all metadata relevant to a swapfile in memory. Upon
successful activation, the swapfile is marked SWP_FILE and
the address space operation ->direct_IO is used for writing
and ->readpage for reading in swap pages.

Patch 5 notes that patch 3 is bolting
filesystem-specific-swapfile-support onto the side and that
the default handlers have different information to what
is available to the filesystem. This patch refactors the
code so that there are generic handlers for each of the new
address_space operations.

Patch 6 adds an API to allow a vector of kernel addresses to be
translated to struct pages and pinned for IO.

Patch 7 adds support for using highmem pages for swap by kmapping
the pages before calling the direct_IO handler.

Patch 8 updates NFS to use the helpers from patch 3 where necessary.

Patch 9 avoids setting PF_private on PG_swapcache pages within NFS.

Patch 10 implements the new swapfile-related address_space operations
for NFS and teaches the direct IO handler how to manage
kernel addresses.

Patch 11 prevents page allocator recursions in NFS by using GFP_NOIO
where appropriate.

Patch 12 fixes a NULL pointer dereference that occurs when using
swap-over-NFS.

With the patches applied, it is possible to mount a swapfile that is on an
NFS filesystem.  Swap performance is not great with a swap stress test
taking roughly twice as long to complete than if the swap device was
backed by NBD.

This patch: netvm: prevent a stream-specific deadlock

It could happen that all !SOCK_MEMALLOC sockets have buffered so much data
that we're over the global rmem limit.  This will prevent SOCK_MEMALLOC
buffers from receiving data, which will prevent userspace from running,
which is needed to reduce the buffered data.

Fix this by exempting the SOCK_MEMALLOC sockets from the rmem limit.  Once
this change it applied, it is important that sockets that set
SOCK_MEMALLOC do not clear the flag until the socket is being torn down.
If this happens, a warning is generated and the tokens reclaimed to avoid
accounting errors until the bug is fixed.

[davem@davemloft.net: Warning about clearing SOCK_MEMALLOC]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: David S. Miller <davem@davemloft.net>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: Neil Brown <neilb@suse.de>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Mel Gorman <mgorman@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: account for the number of times direct reclaimers get throttled
Mel Gorman [Sat, 21 Jul 2012 00:54:25 +0000 (10:54 +1000)]
mm: account for the number of times direct reclaimers get throttled

Under significant pressure when writing back to network-backed storage,
direct reclaimers may get throttled.  This is expected to be a short-lived
event and the processes get woken up again but processes do get stalled.
This patch counts how many times such stalling occurs.  It's up to the
administrator whether to reduce these stalls by increasing
min_free_kbytes.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: David Miller <davem@davemloft.net>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: throttle direct reclaimers if PF_MEMALLOC reserves are low and swap is backed...
Mel Gorman [Sat, 21 Jul 2012 00:54:25 +0000 (10:54 +1000)]
mm: throttle direct reclaimers if PF_MEMALLOC reserves are low and swap is backed by network storage

If swap is backed by network storage such as NBD, there is a risk that a
large number of reclaimers can hang the system by consuming all
PF_MEMALLOC reserves.  To avoid these hangs, the administrator must tune
min_free_kbytes in advance which is a bit fragile.

This patch throttles direct reclaimers if half the PF_MEMALLOC reserves
are in use.  If the system is routinely getting throttled the system
administrator can increase min_free_kbytes so degradation is smoother but
the system will keep running.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: David Miller <davem@davemloft.net>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agonbd: set SOCK_MEMALLOC for access to PFMEMALLOC reserves
Mel Gorman [Sat, 21 Jul 2012 00:54:25 +0000 (10:54 +1000)]
nbd: set SOCK_MEMALLOC for access to PFMEMALLOC reserves

Set SOCK_MEMALLOC on the NBD socket to allow access to PFMEMALLOC reserves
so pages backed by NBD, particularly if swap related, can be cleaned to
prevent the machine being deadlocked.  It is still possible that the
PFMEMALLOC reserves get depleted resulting in deadlock but this can be
resolved by the administrator by increasing min_free_kbytes.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: David Miller <davem@davemloft.net>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agomm: micro-optimise slab to avoid a function call
Mel Gorman [Sat, 21 Jul 2012 00:54:24 +0000 (10:54 +1000)]
mm: micro-optimise slab to avoid a function call

Getting and putting objects in SLAB currently requires a function call but
the bulk of the work is related to PFMEMALLOC reserves which are only
consumed when network-backed storage is critical.  Use an inline function
to determine if the function call is required.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: David Miller <davem@davemloft.net>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agonetvm: set PF_MEMALLOC as appropriate during SKB processing
Mel Gorman [Sat, 21 Jul 2012 00:54:24 +0000 (10:54 +1000)]
netvm: set PF_MEMALLOC as appropriate during SKB processing

In order to make sure pfmemalloc packets receive all memory needed to
proceed, ensure processing of pfmemalloc SKBs happens under PF_MEMALLOC.
This is limited to a subset of protocols that are expected to be used for
writing to swap.  Taps are not allowed to use PF_MEMALLOC as these are
expected to communicate with userspace processes which could be paged out.

[a.p.zijlstra@chello.nl: Ideas taken from various patches]
[jslaby@suse.cz: Lock imbalance fix]
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agonetvm-propagate-page-pfmemalloc-from-skb_alloc_page-to-skb-fix
Andrew Morton [Sat, 21 Jul 2012 00:54:24 +0000 (10:54 +1000)]
netvm-propagate-page-pfmemalloc-from-skb_alloc_page-to-skb-fix

help!

Cc: Mel Gorman <mgorman@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
12 years agonetvm: propagate page->pfmemalloc from skb_alloc_page to skb
Mel Gorman [Sat, 21 Jul 2012 00:54:24 +0000 (10:54 +1000)]
netvm: propagate page->pfmemalloc from skb_alloc_page to skb

The skb->pfmemalloc flag gets set to true iff during the slab allocation
of data in __alloc_skb that the the PFMEMALLOC reserves were used.  If
page splitting is used, it is possible that pages will be allocated from
the PFMEMALLOC reserve without propagating this information to the skb.
This patch propagates page->pfmemalloc from pages allocated for fragments
to the skb.

It works by reintroducing and expanding the skb_alloc_page() API to take
an skb.  If the page was allocated from pfmemalloc reserves, it is
automatically copied.  If the driver allocates the page before the skb, it
should call skb_propagate_pfmemalloc() after the skb is allocated to
ensure the flag is copied properly.

Failure to do so is not critical.  The resulting driver may perform slower
if it is used for swap-over-NBD or swap-over-NFS but it should not result
in failure.

[davem@davemloft.net: API rename and consistency]
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Neil Brown <neilb@suse.de>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Eric B Munson <emunson@mgebm.net>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Sebastian Andrzej Siewior <sebastian@breakpoint.cc>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>