]> git.karo-electronics.de Git - karo-tx-linux.git/log
karo-tx-linux.git
13 years agothp: mremap support and TLB optimization
Andrea Arcangeli [Wed, 28 Sep 2011 00:50:14 +0000 (10:50 +1000)]
thp: mremap support and TLB optimization

This adds THP support to mremap (decreases the number of split_huge_page()
calls).

Here are also some benchmarks with a proggy like this:

===
#define _GNU_SOURCE
#include <sys/mman.h>
#include <stdlib.h>
#include <stdio.h>
#include <string.h>
#include <sys/time.h>

#define SIZE (5UL*1024*1024*1024)

int main()
{
        static struct timeval oldstamp, newstamp;
long diffsec;
char *p, *p2, *p3, *p4;
if (posix_memalign((void **)&p, 2*1024*1024, SIZE))
perror("memalign"), exit(1);
if (posix_memalign((void **)&p2, 2*1024*1024, SIZE))
perror("memalign"), exit(1);
if (posix_memalign((void **)&p3, 2*1024*1024, 4096))
perror("memalign"), exit(1);

memset(p, 0xff, SIZE);
memset(p2, 0xff, SIZE);
memset(p3, 0x77, 4096);
gettimeofday(&oldstamp, NULL);
p4 = mremap(p, SIZE, SIZE, MREMAP_FIXED|MREMAP_MAYMOVE, p3);
gettimeofday(&newstamp, NULL);
diffsec = newstamp.tv_sec - oldstamp.tv_sec;
diffsec = newstamp.tv_usec - oldstamp.tv_usec + 1000000 * diffsec;
printf("usec %ld\n", diffsec);
if (p == MAP_FAILED || p4 != p3)
//if (p == MAP_FAILED)
perror("mremap"), exit(1);
if (memcmp(p4, p2, SIZE))
printf("mremap bug\n"), exit(1);
printf("ok\n");

return 0;
}
===

THP on

 Performance counter stats for './largepage13' (3 runs):

          69195836 dTLB-loads                 ( +-   3.546% )  (scaled from 50.30%)
             60708 dTLB-load-misses           ( +-  11.776% )  (scaled from 52.62%)
         676266476 dTLB-stores                ( +-   5.654% )  (scaled from 69.54%)
             29856 dTLB-store-misses          ( +-   4.081% )  (scaled from 89.22%)
        1055848782 iTLB-loads                 ( +-   4.526% )  (scaled from 80.18%)
              8689 iTLB-load-misses           ( +-   2.987% )  (scaled from 58.20%)

        7.314454164  seconds time elapsed   ( +-   0.023% )

THP off

 Performance counter stats for './largepage13' (3 runs):

        1967379311 dTLB-loads                 ( +-   0.506% )  (scaled from 60.59%)
           9238687 dTLB-load-misses           ( +-  22.547% )  (scaled from 61.87%)
        2014239444 dTLB-stores                ( +-   0.692% )  (scaled from 60.40%)
           3312335 dTLB-store-misses          ( +-   7.304% )  (scaled from 67.60%)
        6764372065 iTLB-loads                 ( +-   0.925% )  (scaled from 79.00%)
              8202 iTLB-load-misses           ( +-   0.475% )  (scaled from 70.55%)

        9.693655243  seconds time elapsed   ( +-   0.069% )

grep thp /proc/vmstat
thp_fault_alloc 35849
thp_fault_fallback 0
thp_collapse_alloc 3
thp_collapse_alloc_failed 0
thp_split 0

thp_split 0 confirms no thp split despite plenty of hugepages allocated.

The measurement of only the mremap time (so excluding the 3 long
memset and final long 10GB memory accessing memcmp):

THP on

usec 14824
usec 14862
usec 14859

THP off

usec 256416
usec 255981
usec 255847

With an older kernel without the mremap optimizations (the below patch
optimizes the non THP version too).

THP on

usec 392107
usec 390237
usec 404124

THP off

usec 444294
usec 445237
usec 445820

I guess with a threaded program that sends more IPI on large SMP it'd
create an even larger difference.

All debug options are off except DEBUG_VM to avoid skewing the
results.

The only problem for native 2M mremap like it happens above both the
source and destination address must be 2M aligned or the hugepmd can't be
moved without a split but that is an hardware limitation.

Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Johannes Weiner <jweiner@redhat.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <>
13 years agomremap: avoid sending one IPI per page
Andrea Arcangeli [Wed, 28 Sep 2011 00:50:13 +0000 (10:50 +1000)]
mremap: avoid sending one IPI per page

This replaces ptep_clear_flush() with ptep_get_and_clear() and a single
flush_tlb_range() at the end of the loop, to avoid sending one IPI for
each page.

The mmu_notifier_invalidate_range_start/end section is enlarged
accordingly but this is not going to fundamentally change things.  It was
more by accident that the region under mremap was for the most part still
available for secondary MMUs: the primary MMU was never allowed to
reliably access that region for the duration of the mremap (modulo
trapping SIGSEGV on the old address range which sounds unpractical and
flakey).  If users wants secondary MMUs not to lose access to a large
region under mremap they should reduce the mremap size accordingly in
userland and run multiple calls.  Overall this will run faster so it's
actually going to reduce the time the region is under mremap for the
primary MMU which should provide a net benefit to apps.

For KVM this is a noop because the guest physical memory is never
mremapped, there's just no point it ever moving it while guest runs.  One
target of this optimization is JVM GC (so unrelated to the mmu notifier
logic).

Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Johannes Weiner <jweiner@redhat.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <>
13 years agomremap: check for overflow using deltas
Andrea Arcangeli [Wed, 28 Sep 2011 00:50:13 +0000 (10:50 +1000)]
mremap: check for overflow using deltas

Using "- 1" relies on the old_end to be page aligned and PAGE_SIZE > 1,
those are reasonable requirements but the check remains obscure and it
looks more like an off by one error than an overflow check.  This I feel
will improve readability.

Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Johannes Weiner <jweiner@redhat.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <>
13 years agomemblock: add NO_BOOTMEM config symbol
Sam Ravnborg [Wed, 28 Sep 2011 00:50:12 +0000 (10:50 +1000)]
memblock: add NO_BOOTMEM config symbol

With the NO_BOOTMEM symbol added architectures may now use the following
syntax to tell that they do not need bootmem:

select NO_BOOTMEM

This is much more convinient than adding a new kconfig symbol which was
otherwise required.

Adding this symbol does not conflict with the architctures that already
define their own symbol.

Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Cc: Yinghai Lu <yinghai@kernel.org>
Acked-by: Tejun Heo <tj@kernel.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Andrew Morton <>
13 years agomemblock: add memblock_start_of_DRAM()
Sam Ravnborg [Wed, 28 Sep 2011 00:50:12 +0000 (10:50 +1000)]
memblock: add memblock_start_of_DRAM()

SPARC32 require access to the start address.  Add a new helper
memblock_start_of_DRAM() to give access to the address of the first
memblock - which contains the lowest address.

The awkward name was chosen to match the already present
memblock_end_of_DRAM().

Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Yinghai Lu <yinghai@kernel.org>
Acked-by: Tejun Heo <tj@kernel.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Andrew Morton <>
13 years agovmscan: activate executable pages after first usage
Konstantin Khlebnikov [Wed, 28 Sep 2011 00:50:12 +0000 (10:50 +1000)]
vmscan: activate executable pages after first usage

Logic added in commit 8cab4754d24a0 ("vmscan: make mapped executable pages
the first class citizen") was noticeably weakened in commit
645747462435d84 ("vmscan: detect mapped file pages used only once").

Currently these pages can become "first class citizens" only after second
usage.  After this patch page_check_references() will activate they after
first usage, and executable code gets yet better chance to stay in memory.

Signed-off-by: Konstantin Khlebnikov <khlebnikov@openvz.org>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: Minchan Kim <minchan.kim@gmail.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Nick Piggin <npiggin@kernel.dk>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Shaohua Li <shaohua.li@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agovmscan: promote shared file mapped pages
Konstantin Khlebnikov [Wed, 28 Sep 2011 00:50:11 +0000 (10:50 +1000)]
vmscan: promote shared file mapped pages

Commit 645747462435 ("vmscan: detect mapped file pages used only once")
greatly decreases lifetime of single-used mapped file pages.
Unfortunately it also decreases life time of all shared mapped file pages.
Because after commit bf3f3bc5e7347 ("mm: don't mark_page_accessed
in fault path") page-fault handler does not mark page active or even
referenced.

Thus page_check_references() activates file page only if it was used twice
while it stays in inactive list, meanwhile it activates anon pages after
first access.  Inactive list can be small enough, this way reclaimer can
accidentally throw away any widely used page if it wasn't used twice in
short period.

After this patch page_check_references() also activate file mapped page at
first inactive list scan if this page is already used multiple times via
several ptes.

I found this while trying to fix degragation in rhel6 (~2.6.32) from rhel5
(~2.6.18).  There a complete mess with >100 web/mail/spam/ftp containers,
they share all their files but there a lot of anonymous pages: ~500mb
shared file mapped memory and 15-20Gb non-shared anonymous memory.  In
this situation major-pagefaults are very costly, because all containers
share the same page.  In my load kernel created a disproportionate
pressure on the file memory, compared with the anonymous, they equaled
only if I raise swappiness up to 150 =)

These patches actually wasn't helped a lot in my problem, but I saw
noticable (10-20 times) reduce in count and average time of
major-pagefault in file-mapped areas.

Actually both patches are fixes for commit v2.6.33-5448-g6457474, because
it was aimed at one scenario (singly used pages), but it breaks the logic
in other scenarios (shared and/or executable pages)

Signed-off-by: Konstantin Khlebnikov <khlebnikov@openvz.org>
Acked-by: Pekka Enberg <penberg@kernel.org>
Acked-by: Minchan Kim <minchan.kim@gmail.com>
Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Nick Piggin <npiggin@kernel.dk>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Shaohua Li <shaohua.li@intel.com>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agomm: avoid null pointer access in vm_struct via /proc/vmallocinfo
Mitsuo Hayasaka [Wed, 28 Sep 2011 00:50:11 +0000 (10:50 +1000)]
mm: avoid null pointer access in vm_struct via /proc/vmallocinfo

The /proc/vmallocinfo shows information about vmalloc allocations in
vmlist that is a linklist of vm_struct.  It, however, may access pages
field of vm_struct where a page was not allocated.  This results in a null
pointer access and leads to a kernel panic.

Why this happens: In __vmalloc_node_range() called from vmalloc(), newly
allocated vm_struct is added to vmlist at __get_vm_area_node() and then,
some fields of vm_struct such as nr_pages and pages are set at
__vmalloc_area_node().  In other words, it is added to vmlist before it is
fully initialized.  At the same time, when the /proc/vmallocinfo is read,
it accesses the pages field of vm_struct according to the nr_pages field
at show_numa_info().  Thus, a null pointer access happens.

The patch adds the newly allocated vm_struct to the vmlist *after* it is
fully initialized.  So, it can avoid accessing the pages field with
unallocated page when show_numa_info() is called.

Signed-off-by: Mitsuo Hayasaka <mitsuo.hayasaka.hu@hitachi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <>
13 years agovmscan-use-atomic-long-for-shrinker-batching-fix
Andrew Morton [Wed, 28 Sep 2011 00:50:11 +0000 (10:50 +1000)]
vmscan-use-atomic-long-for-shrinker-batching-fix

massage atomic.h inclusions

Cc: Dave Chinner <david@fromorbit.com>
Cc: Konstantin Khlebnikov <khlebnikov@openvz.org>
Signed-off-by: Andrew Morton <>
13 years agovmscan: use atomic-long for shrinker batching
Konstantin Khlebnikov [Wed, 28 Sep 2011 00:50:10 +0000 (10:50 +1000)]
vmscan: use atomic-long for shrinker batching

Use atomic-long operations instead of looping around cmpxchg().

Signed-off-by: Konstantin Khlebnikov <khlebnikov@openvz.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <>
13 years agovmscan: fix initial shrinker size handling
Konstantin Khlebnikov [Wed, 28 Sep 2011 00:50:10 +0000 (10:50 +1000)]
vmscan: fix initial shrinker size handling

A shrinker function can return -1, means that it cannot do anything
without a risk of deadlock.  For example prune_super() does this if it
cannot grab a superblock refrence, even if nr_to_scan=0.  Currently we
interpret this -1 as a ULONG_MAX size shrinker and evaluate `total_scan'
according to this.  So the next time around this shrinker can cause really
big pressure.  Let's skip such shrinkers instead.

Also make total_scan signed, otherwise the check (total_scan < 0) below
never works.

Signed-off-by: Konstantin Khlebnikov <khlebnikov@openvz.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <>
13 years agomm/debug-pagealloc.c: use memchr_inv
Akinobu Mita [Wed, 28 Sep 2011 00:50:10 +0000 (10:50 +1000)]
mm/debug-pagealloc.c: use memchr_inv

Use newly introduced memchr_inv() for page verification.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Andrew Morton <>
13 years agolib-stringc-introduce-memchr_inv-fix-kernel-doc-for-memchr_inv
Akinobu Mita [Wed, 28 Sep 2011 00:50:09 +0000 (10:50 +1000)]
lib-stringc-introduce-memchr_inv-fix-kernel-doc-for-memchr_inv

This fixes kernel-doc for memchr_inv() which is introduced by
lib-stringc-introduce-memchr_inv.patch in mmotm 2011-08-24-14-08

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Acked-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <>
13 years agolib/string.c: introduce memchr_inv()
Akinobu Mita [Wed, 28 Sep 2011 00:50:09 +0000 (10:50 +1000)]
lib/string.c: introduce memchr_inv()

memchr_inv() is mainly used to check whether the whole buffer is filled
with just a specified byte.

The function name and prototype are stolen from logfs and the
implementation is from SLUB.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Acked-by: Christoph Lameter <cl@linux-foundation.org>
Acked-by: Pekka Enberg <penberg@kernel.org>
Cc: Matt Mackall <mpm@selenic.com>
Acked-by: Joern Engel <joern@logfs.org>
Cc: Marcin Slusarz <marcin.slusarz@gmail.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: Andrew Morton <>
13 years agomm/debug-pagealloc.c: use plain __ratelimit() instead of printk_ratelimit()
Akinobu Mita [Wed, 28 Sep 2011 00:50:08 +0000 (10:50 +1000)]
mm/debug-pagealloc.c: use plain __ratelimit() instead of printk_ratelimit()

printk_ratelimit() should not be used, because it shares ratelimiting
state with all other unrelated printk_ratelimit() callsites.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Andrew Morton <>
13 years agovmscan: count pages into balanced for zone with good watermark
Shaohua Li [Wed, 28 Sep 2011 00:50:08 +0000 (10:50 +1000)]
vmscan: count pages into balanced for zone with good watermark

It's possible a zone watermark is ok when entering the balance_pgdat()
loop, while the zone is within the requested classzone_idx.  Count pages
from this zone into `balanced'.  In this way, we can skip shrinking zones
too much for high order allocation.

Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Signed-off-by: Andrew Morton <>
13 years agomm: vmscan: immediately reclaim end-of-LRU dirty pages when writeback completes
Mel Gorman [Wed, 28 Sep 2011 00:50:08 +0000 (10:50 +1000)]
mm: vmscan: immediately reclaim end-of-LRU dirty pages when writeback completes

When direct reclaim encounters a dirty page, it gets recycled around the
LRU for another cycle.  This patch marks the page PageReclaim similar to
deactivate_page() so that the page gets reclaimed almost immediately after
the page gets cleaned.  This is to avoid reclaiming clean pages that are
younger than a dirty page encountered at the end of the LRU that might
have been something like a use-once page.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Johannes Weiner <jweiner@redhat.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Minchan Kim <minchan.kim@gmail.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Alex Elder <aelder@sgi.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: Dave Hansen <dave@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <>
13 years agomm-vmscan-throttle-reclaim-if-encountering-too-many-dirty-pages-under-writeback-update
Mel Gorman [Wed, 28 Sep 2011 00:50:07 +0000 (10:50 +1000)]
mm-vmscan-throttle-reclaim-if-encountering-too-many-dirty-pages-under-writeback-update

This patch expands on a comment on how we throttle from reclaim context.
It should be merged with
mm-vmscan-throttle-reclaim-if-encountering-too-many-dirty-pages-under-writeback.patch

Signed-off-by: Mel Gorman <mgorman@suse.de>
Signed-off-by: Andrew Morton <>
13 years agomm: vmscan: throttle reclaim if encountering too many dirty pages under writeback
Mel Gorman [Wed, 28 Sep 2011 00:50:07 +0000 (10:50 +1000)]
mm: vmscan: throttle reclaim if encountering too many dirty pages under writeback

Workloads that are allocating frequently and writing files place a large
number of dirty pages on the LRU.  With use-once logic, it is possible for
them to reach the end of the LRU quickly requiring the reclaimer to scan
more to find clean pages.  Ordinarily, processes that are dirtying memory
will get throttled by dirty balancing but this is a global heuristic and
does not take into account that LRUs are maintained on a per-zone basis.
This can lead to a situation whereby reclaim is scanning heavily, skipping
over a large number of pages under writeback and recycling them around the
LRU consuming CPU.

This patch checks how many of the number of pages isolated from the LRU
were dirty and under writeback.  If a percentage of them under writeback,
the process will be throttled if a backing device or the zone is
congested.  Note that this applies whether it is anonymous or file-backed
pages that are under writeback meaning that swapping is potentially
throttled.  This is intentional due to the fact if the swap device is
congested, scanning more pages and dispatching more IO is not going to
help matters.

The percentage that must be in writeback depends on the priority.  At
default priority, all of them must be dirty.  At DEF_PRIORITY-1, 50% of
them must be, DEF_PRIORITY-2, 25% etc.  i.e.  as pressure increases the
greater the likelihood the process will get throttled to allow the flusher
threads to make some progress.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Acked-by: Johannes Weiner <jweiner@redhat.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Alex Elder <aelder@sgi.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: Dave Hansen <dave@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <>
13 years agomm: vmscan: do not writeback filesystem pages in kswapd except in high priority
Mel Gorman [Wed, 28 Sep 2011 00:50:07 +0000 (10:50 +1000)]
mm: vmscan: do not writeback filesystem pages in kswapd except in high priority

It is preferable that no dirty pages are dispatched for cleaning from the
page reclaim path.  At normal priorities, this patch prevents kswapd
writing pages.

However, page reclaim does have a requirement that pages be freed in a
particular zone.  If it is failing to make sufficient progress (reclaiming
< SWAP_CLUSTER_MAX at any priority priority), the priority is raised to
scan more pages.  A priority of DEF_PRIORITY - 3 is considered to be the
point where kswapd is getting into trouble reclaiming pages.  If this
priority is reached, kswapd will dispatch pages for writing.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Alex Elder <aelder@sgi.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: Dave Hansen <dave@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <>
13 years agoext4: warn if direct reclaim tries to writeback pages
Mel Gorman [Wed, 28 Sep 2011 00:50:06 +0000 (10:50 +1000)]
ext4: warn if direct reclaim tries to writeback pages

Direct reclaim should never writeback pages.  Warn if an attempt is made.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Minchan Kim <minchan.kim@gmail.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Alex Elder <aelder@sgi.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: Dave Hansen <dave@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <>
13 years agoxfs: warn if direct reclaim tries to writeback pages
Mel Gorman [Wed, 28 Sep 2011 00:50:06 +0000 (10:50 +1000)]
xfs: warn if direct reclaim tries to writeback pages

Direct reclaim should never writeback pages.  For now, handle the
situation and warn about it.  Ultimately, this will be a BUG_ON.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Minchan Kim <minchan.kim@gmail.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Alex Elder <aelder@sgi.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: Dave Hansen <dave@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <>
13 years agomm: vmscan: remove dead code related to lumpy reclaim waiting on pages under writeback
Mel Gorman [Wed, 28 Sep 2011 00:50:06 +0000 (10:50 +1000)]
mm: vmscan: remove dead code related to lumpy reclaim waiting on pages under writeback

Lumpy reclaim worked with two passes - the first which queued pages for IO
and the second which waited on writeback.  As direct reclaim can no longer
write pages there is some dead code.  This patch removes it but direct
reclaim will continue to wait on pages under writeback while in
synchronous reclaim mode.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Minchan Kim <minchan.kim@gmail.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Alex Elder <aelder@sgi.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: Dave Hansen <dave@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <>
13 years agomm: vmscan: do not writeback filesystem pages in direct reclaim
Mel Gorman [Wed, 28 Sep 2011 00:50:05 +0000 (10:50 +1000)]
mm: vmscan: do not writeback filesystem pages in direct reclaim

Testing from the XFS folk revealed that there is still too much I/O from
the end of the LRU in kswapd.  Previously it was considered acceptable by
VM people for a small number of pages to be written back from reclaim with
testing generally showing about 0.3% of pages reclaimed were written back
(higher if memory was low).  That writing back a small number of pages is
ok has been heavily disputed for quite some time and Dave Chinner
explained it well;

It doesn't have to be a very high number to be a problem. IO
is orders of magnitude slower than the CPU time it takes to
flush a page, so the cost of making a bad flush decision is
very high. And single page writeback from the LRU is almost
always a bad flush decision.

To complicate matters, filesystems respond very differently to requests
from reclaim according to Christoph Hellwig;

xfs tries to write it back if the requester is kswapd
ext4 ignores the request if it's a delayed allocation
btrfs ignores the request

As a result, each filesystem has different performance characteristics
when under memory pressure and there are many pages being dirtied.  In
some cases, the request is ignored entirely so the VM cannot depend on the
IO being dispatched.

The objective of this series is to reduce writing of filesystem-backed
pages from reclaim, play nicely with writeback that is already in progress
and throttle reclaim appropriately when writeback pages are encountered.
The assumption is that the flushers will always write pages faster than if
reclaim issues the IO.

A secondary goal is to avoid the problem whereby direct reclaim splices
two potentially deep call stacks together.

There is a potential new problem as reclaim has less control over how long
before a page in a particularly zone or container is cleaned and direct
reclaimers depend on kswapd or flusher threads to do the necessary work.
However, as filesystems sometimes ignore direct reclaim requests already,
it is not expected to be a serious issue.

Patch 1 disables writeback of filesystem pages from direct reclaim
entirely. Anonymous pages are still written.

Patch 2 removes dead code in lumpy reclaim as it is no longer able
to synchronously write pages. This hurts lumpy reclaim but
there is an expectation that compaction is used for hugepage
allocations these days and lumpy reclaim's days are numbered.

Patches 3-4 add warnings to XFS and ext4 if called from
direct reclaim. With patch 1, this "never happens" and is
intended to catch regressions in this logic in the future.

Patch 5 disables writeback of filesystem pages from kswapd unless
the priority is raised to the point where kswapd is considered
to be in trouble.

Patch 6 throttles reclaimers if too many dirty pages are being
encountered and the zones or backing devices are congested.

Patch 7 invalidates dirty pages found at the end of the LRU so they
are reclaimed quickly after being written back rather than
waiting for a reclaimer to find them

I consider this series to be orthogonal to the writeback work but it is
worth noting that the writeback work affects the viability of patch 8 in
particular.

I tested this on ext4 and xfs using fs_mark, a simple writeback test based
on dd and a micro benchmark that does a streaming write to a large mapping
(exercises use-once LRU logic) followed by streaming writes to a mix of
anonymous and file-backed mappings.  The command line for fs_mark when
botted with 512M looked something like

./fs_mark -d  /tmp/fsmark-2676  -D  100  -N  150  -n  150  -L  25  -t  1  -S0  -s  10485760

The number of files was adjusted depending on the amount of available
memory so that the files created was about 3xRAM.  For multiple threads,
the -d switch is specified multiple times.

The test machine is x86-64 with an older generation of AMD processor with
4 cores.  The underlying storage was 4 disks configured as RAID-0 as this
was the best configuration of storage I had available.  Swap is on a
separate disk.  Dirty ratio was tuned to 40% instead of the default of
20%.

Testing was run with and without monitors to both verify that the patches
were operating as expected and that any performance gain was real and not
due to interference from monitors.

Here is a summary of results based on testing XFS.

512M1P-xfs           Files/s  mean                 32.69 ( 0.00%)     34.44 ( 5.08%)
512M1P-xfs           Elapsed Time fsmark                    51.41     48.29
512M1P-xfs           Elapsed Time simple-wb                114.09    108.61
512M1P-xfs           Elapsed Time mmap-strm                113.46    109.34
512M1P-xfs           Kswapd efficiency fsmark                 62%       63%
512M1P-xfs           Kswapd efficiency simple-wb              56%       61%
512M1P-xfs           Kswapd efficiency mmap-strm              44%       42%
512M-xfs             Files/s  mean                 30.78 ( 0.00%)     35.94 (14.36%)
512M-xfs             Elapsed Time fsmark                    56.08     48.90
512M-xfs             Elapsed Time simple-wb                112.22     98.13
512M-xfs             Elapsed Time mmap-strm                219.15    196.67
512M-xfs             Kswapd efficiency fsmark                 54%       56%
512M-xfs             Kswapd efficiency simple-wb              54%       55%
512M-xfs             Kswapd efficiency mmap-strm              45%       44%
512M-4X-xfs          Files/s  mean                 30.31 ( 0.00%)     33.33 ( 9.06%)
512M-4X-xfs          Elapsed Time fsmark                    63.26     55.88
512M-4X-xfs          Elapsed Time simple-wb                100.90     90.25
512M-4X-xfs          Elapsed Time mmap-strm                261.73    255.38
512M-4X-xfs          Kswapd efficiency fsmark                 49%       50%
512M-4X-xfs          Kswapd efficiency simple-wb              54%       56%
512M-4X-xfs          Kswapd efficiency mmap-strm              37%       36%
512M-16X-xfs         Files/s  mean                 60.89 ( 0.00%)     65.22 ( 6.64%)
512M-16X-xfs         Elapsed Time fsmark                    67.47     58.25
512M-16X-xfs         Elapsed Time simple-wb                103.22     90.89
512M-16X-xfs         Elapsed Time mmap-strm                237.09    198.82
512M-16X-xfs         Kswapd efficiency fsmark                 45%       46%
512M-16X-xfs         Kswapd efficiency simple-wb              53%       55%
512M-16X-xfs         Kswapd efficiency mmap-strm              33%       33%

Up until 512-4X, the FSmark improvements were statistically significant.
For the 4X and 16X tests the results were within standard deviations but
just barely.  The time to completion for all tests is improved which is an
important result.  In general, kswapd efficiency is not affected by
skipping dirty pages.

1024M1P-xfs          Files/s  mean                 39.09 ( 0.00%)     41.15 ( 5.01%)
1024M1P-xfs          Elapsed Time fsmark                    84.14     80.41
1024M1P-xfs          Elapsed Time simple-wb                210.77    184.78
1024M1P-xfs          Elapsed Time mmap-strm                162.00    160.34
1024M1P-xfs          Kswapd efficiency fsmark                 69%       75%
1024M1P-xfs          Kswapd efficiency simple-wb              71%       77%
1024M1P-xfs          Kswapd efficiency mmap-strm              43%       44%
1024M-xfs            Files/s  mean                 35.45 ( 0.00%)     37.00 ( 4.19%)
1024M-xfs            Elapsed Time fsmark                    94.59     91.00
1024M-xfs            Elapsed Time simple-wb                229.84    195.08
1024M-xfs            Elapsed Time mmap-strm                405.38    440.29
1024M-xfs            Kswapd efficiency fsmark                 79%       71%
1024M-xfs            Kswapd efficiency simple-wb              74%       74%
1024M-xfs            Kswapd efficiency mmap-strm              39%       42%
1024M-4X-xfs         Files/s  mean                 32.63 ( 0.00%)     35.05 ( 6.90%)
1024M-4X-xfs         Elapsed Time fsmark                   103.33     97.74
1024M-4X-xfs         Elapsed Time simple-wb                204.48    178.57
1024M-4X-xfs         Elapsed Time mmap-strm                528.38    511.88
1024M-4X-xfs         Kswapd efficiency fsmark                 81%       70%
1024M-4X-xfs         Kswapd efficiency simple-wb              73%       72%
1024M-4X-xfs         Kswapd efficiency mmap-strm              39%       38%
1024M-16X-xfs        Files/s  mean                 42.65 ( 0.00%)     42.97 ( 0.74%)
1024M-16X-xfs        Elapsed Time fsmark                   103.11     99.11
1024M-16X-xfs        Elapsed Time simple-wb                200.83    178.24
1024M-16X-xfs        Elapsed Time mmap-strm                397.35    459.82
1024M-16X-xfs        Kswapd efficiency fsmark                 84%       69%
1024M-16X-xfs        Kswapd efficiency simple-wb              74%       73%
1024M-16X-xfs        Kswapd efficiency mmap-strm              39%       40%

All FSMark tests up to 16X had statistically significant improvements.
For the most part, tests are completing faster with the exception of the
streaming writes to a mixture of anonymous and file-backed mappings which
were slower in two cases

In the cases where the mmap-strm tests were slower, there was more
swapping due to dirty pages being skipped.  The number of additional pages
swapped is almost identical to the fewer number of pages written from
reclaim.  In other words, roughly the same number of pages were reclaimed
but swapping was slower.  As the test is a bit unrealistic and stresses
memory heavily, the small shift is acceptable.

4608M1P-xfs          Files/s  mean                 29.75 ( 0.00%)     30.96 ( 3.91%)
4608M1P-xfs          Elapsed Time fsmark                   512.01    492.15
4608M1P-xfs          Elapsed Time simple-wb                618.18    566.24
4608M1P-xfs          Elapsed Time mmap-strm                488.05    465.07
4608M1P-xfs          Kswapd efficiency fsmark                 93%       86%
4608M1P-xfs          Kswapd efficiency simple-wb              88%       84%
4608M1P-xfs          Kswapd efficiency mmap-strm              46%       45%
4608M-xfs            Files/s  mean                 27.60 ( 0.00%)     28.85 ( 4.33%)
4608M-xfs            Elapsed Time fsmark                   555.96    532.34
4608M-xfs            Elapsed Time simple-wb                659.72    571.85
4608M-xfs            Elapsed Time mmap-strm               1082.57   1146.38
4608M-xfs            Kswapd efficiency fsmark                 89%       91%
4608M-xfs            Kswapd efficiency simple-wb              88%       82%
4608M-xfs            Kswapd efficiency mmap-strm              48%       46%
4608M-4X-xfs         Files/s  mean                 26.00 ( 0.00%)     27.47 ( 5.35%)
4608M-4X-xfs         Elapsed Time fsmark                   592.91    564.00
4608M-4X-xfs         Elapsed Time simple-wb                616.65    575.07
4608M-4X-xfs         Elapsed Time mmap-strm               1773.02   1631.53
4608M-4X-xfs         Kswapd efficiency fsmark                 90%       94%
4608M-4X-xfs         Kswapd efficiency simple-wb              87%       82%
4608M-4X-xfs         Kswapd efficiency mmap-strm              43%       43%
4608M-16X-xfs        Files/s  mean                 26.07 ( 0.00%)     26.42 ( 1.32%)
4608M-16X-xfs        Elapsed Time fsmark                   602.69    585.78
4608M-16X-xfs        Elapsed Time simple-wb                606.60    573.81
4608M-16X-xfs        Elapsed Time mmap-strm               1549.75   1441.86
4608M-16X-xfs        Kswapd efficiency fsmark                 98%       98%
4608M-16X-xfs        Kswapd efficiency simple-wb              88%       82%
4608M-16X-xfs        Kswapd efficiency mmap-strm              44%       42%

Unlike the other tests, the fsmark results are not statistically
significant but the min and max times are both improved and for the most
part, tests completed faster.

There are other indications that this is an improvement as well.  For
example, in the vast majority of cases, there were fewer pages scanned by
direct reclaim implying in many cases that stalls due to direct reclaim
are reduced.  KSwapd is scanning more due to skipping dirty pages which is
unfortunate but the CPU usage is still acceptable

In an earlier set of tests, I used blktrace and in almost all cases
throughput throughout the entire test was higher.  However, I ended up
discarding those results as recording blktrace data was too heavy for my
liking.

On a laptop, I plugged in a USB stick and ran a similar tests of tests
using it as backing storage.  A desktop environment was running and for
the entire duration of the tests, firefox and gnome terminal were
launching and exiting to vaguely simulate a user.

1024M-xfs            Files/s  mean               0.41 ( 0.00%)        0.44 ( 6.82%)
1024M-xfs            Elapsed Time fsmark               2053.52   1641.03
1024M-xfs            Elapsed Time simple-wb            1229.53    768.05
1024M-xfs            Elapsed Time mmap-strm            4126.44   4597.03
1024M-xfs            Kswapd efficiency fsmark              84%       85%
1024M-xfs            Kswapd efficiency simple-wb           92%       81%
1024M-xfs            Kswapd efficiency mmap-strm           60%       51%
1024M-xfs            Avg wait ms fsmark                5404.53     4473.87
1024M-xfs            Avg wait ms simple-wb             2541.35     1453.54
1024M-xfs            Avg wait ms mmap-strm             3400.25     3852.53

The mmap-strm results were hurt because firefox launching had a tendency
to push the test out of memory.  On the postive side, firefox launched
marginally faster with the patches applied.  Time to completion for many
tests was faster but more importantly - the "Avg wait" time as measured by
iostat was far lower implying the system would be more responsive.  It was
also the case that "Avg wait ms" on the root filesystem was lower.  I
tested it manually and while the system felt slightly more responsive
while copying data to a USB stick, it was marginal enough that it could be
my imagination.

This patch: do not writeback filesystem pages in direct reclaim.

When kswapd is failing to keep zones above the min watermark, a process
will enter direct reclaim in the same manner kswapd does.  If a dirty page
is encountered during the scan, this page is written to backing storage
using mapping->writepage.

This causes two problems.  First, it can result in very deep call stacks,
particularly if the target storage or filesystem are complex.  Some
filesystems ignore write requests from direct reclaim as a result.  The
second is that a single-page flush is inefficient in terms of IO.  While
there is an expectation that the elevator will merge requests, this does
not always happen.  Quoting Christoph Hellwig;

The elevator has a relatively small window it can operate on,
and can never fix up a bad large scale writeback pattern.

This patch prevents direct reclaim writing back filesystem pages by
checking if current is kswapd.  Anonymous pages are still written to swap
as there is not the equivalent of a flusher thread for anonymous pages.
If the dirty pages cannot be written back, they are placed back on the LRU
lists.  There is now a direct dependency on dirty page balancing to
prevent too many pages in the system being dirtied which would prevent
reclaim making forward progress.

Signed-off-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Johannes Weiner <jweiner@redhat.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Rik van Riel <riel@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Alex Elder <aelder@sgi.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: Dave Hansen <dave@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <>
13 years agomm-add-comments-to-explain-mm_struct-fields-fix
Andrew Morton [Wed, 28 Sep 2011 00:50:05 +0000 (10:50 +1000)]
mm-add-comments-to-explain-mm_struct-fields-fix

add missing ;

Cc: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <>
13 years agomm: add comments to explain mm_struct fields
Christoph Lameter [Wed, 28 Sep 2011 00:50:05 +0000 (10:50 +1000)]
mm: add comments to explain mm_struct fields

Add comments to explain the page statistics field in the mm_struct.

Signed-off-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <>
13 years agomm: distinguish between mlocked and pinned pages
Christoph Lameter [Wed, 28 Sep 2011 00:50:04 +0000 (10:50 +1000)]
mm: distinguish between mlocked and pinned pages

Some kernel components pin user space memory (infiniband and perf) (by
increasing the page count) and account that memory as "mlocked".

The difference between mlocking and pinning is:

A. mlocked pages are marked with PG_mlocked and are exempt from
   swapping. Page migration may move them around though.
   They are kept on a special LRU list.

B. Pinned pages cannot be moved because something needs to
   directly access physical memory. They may not be on any
   LRU list.

I recently saw an mlockalled process where mm->locked_vm became
bigger than the virtual size of the process (!) because some
memory was accounted for twice:

Once when the page was mlocked and once when the Infiniband
layer increased the refcount because it needt to pin the RDMA
memory.

This patch introduces a separate counter for pinned pages and
accounts them seperately.

Signed-off-by: Christoph Lameter <cl@linux.com>
Cc: Mike Marciniszyn <infinipath@qlogic.com>
Cc: Roland Dreier <roland@kernel.org>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <>
13 years agomm: vmscan: drop nr_force_scan[] from get_scan_count
Johannes Weiner [Wed, 28 Sep 2011 00:50:04 +0000 (10:50 +1000)]
mm: vmscan: drop nr_force_scan[] from get_scan_count

The nr_force_scan[] tuple holds the effective scan numbers for anon and
file pages in case the situation called for a forced scan and the
regularly calculated scan numbers turned out zero.

However, the effective scan number can always be assumed to be
SWAP_CLUSTER_MAX right before the division into anon and file.  The
numerators and denominator are properly set up for all cases, be it force
scan for just file, just anon, or both, to do the right thing.

Signed-off-by: Johannes Weiner <jweiner@redhat.com>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Cc: Ying Han <yinghan@google.com>
Cc: Balbir Singh <bsingharora@gmail.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Acked-by: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <>
13 years agomm: output a list of loaded modules when we hit bad_page()
Dave Jones [Wed, 28 Sep 2011 00:50:04 +0000 (10:50 +1000)]
mm: output a list of loaded modules when we hit bad_page()

When we get a bad_page bug report, it's useful to see what modules the
user had loaded.

Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agotmpfs: add "tmpfs" to the Kconfig prompt to make it obvious.
Robert P. J. Day [Wed, 28 Sep 2011 00:50:03 +0000 (10:50 +1000)]
tmpfs: add "tmpfs" to the Kconfig prompt to make it obvious.

Add the leading word "tmpfs" to the Kconfig string to make it blindingly
obvious that this selection refers to tmpfs.

Signed-off-by: Robert P. J. Day <rpjday@crashcourse.ca>
Acked-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <>
13 years agooom: fix race while temporarily setting current's oom_score_adj
David Rientjes [Wed, 28 Sep 2011 00:50:03 +0000 (10:50 +1000)]
oom: fix race while temporarily setting current's oom_score_adj

test_set_oom_score_adj() was introduced in 72788c385604 ("oom: replace
PF_OOM_ORIGIN with toggling oom_score_adj") to temporarily elevate
current's oom_score_adj for ksm and swapoff without requiring an
additional per-process flag.

Using that function to both set oom_score_adj to OOM_SCORE_ADJ_MAX and
then reinstate the previous value is racy since it's possible that
userspace can set the value to something else itself before the old value
is reinstated.  That results in userspace setting current's oom_score_adj
to a different value and then the kernel immediately setting it back to
its previous value without notification.

To fix this, a new compare_swap_oom_score_adj() function is introduced
with the same semantics as the compare and swap CAS instruction, or
CMPXCHG on x86.  It is used to reinstate the previous value of
oom_score_adj if and only if the present value is the same as the old
value.

Signed-off-by: David Rientjes <rientjes@google.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Ying Han <yinghan@google.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <>
13 years agooom: remove oom_disable_count
David Rientjes [Wed, 28 Sep 2011 00:50:02 +0000 (10:50 +1000)]
oom: remove oom_disable_count

This removes mm->oom_disable_count entirely since it's unnecessary and
currently buggy.  The counter was intended to be per-process but it's
currently decremented in the exit path for each thread that exits, causing
it to underflow.

The count was originally intended to prevent oom killing threads that
share memory with threads that cannot be killed since it doesn't lead to
future memory freeing.  The counter could be fixed to represent all
threads sharing the same mm, but it's better to remove the count since:

 - it is possible that the OOM_DISABLE thread sharing memory with the
   victim is waiting on that thread to exit and will actually cause
   future memory freeing, and

 - there is no guarantee that a thread is disabled from oom killing just
   because another thread sharing its mm is oom disabled.

Signed-off-by: David Rientjes <rientjes@google.com>
Reported-by: Oleg Nesterov <oleg@redhat.com>
Cc: Ying Han <yinghan@google.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <>
13 years agooom: avoid killing kthreads if they assume the oom killed thread's mm
David Rientjes [Wed, 28 Sep 2011 00:50:02 +0000 (10:50 +1000)]
oom: avoid killing kthreads if they assume the oom killed thread's mm

After selecting a task to kill, the oom killer iterates all processes and
kills all other threads that share the same mm_struct in different thread
groups.  It would not otherwise be helpful to kill a thread if its memory
would not be subsequently freed.

A kernel thread, however, may assume a user thread's mm by using
use_mm().  This is only temporary and should not result in sending a
SIGKILL to that kthread.

This patch ensures that only user threads and not kthreads are sent a
SIGKILL if they share the same mm_struct as the oom killed task.

Signed-off-by: David Rientjes <rientjes@google.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <>
13 years agomm/page-writeback.c: document bdi_min_ratio
Johannes Weiner [Wed, 28 Sep 2011 00:50:02 +0000 (10:50 +1000)]
mm/page-writeback.c: document bdi_min_ratio

Looks like someone got distracted after adding the comment characters.

Signed-off-by: Johannes Weiner <jweiner@redhat.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Andrew Morton <>
13 years agomm/page-writeback.c: make determine_dirtyable_memory static again
Johannes Weiner [Wed, 28 Sep 2011 00:50:01 +0000 (10:50 +1000)]
mm/page-writeback.c: make determine_dirtyable_memory static again

The tracing ring-buffer used this function briefly, but not anymore.
Make it local to the writeback code again.

Also, move the function so that no forward declaration needs to be
reintroduced.

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Andrew Morton <>
13 years agovmscan: add block plug for page reclaim
Shaohua Li [Wed, 28 Sep 2011 00:50:01 +0000 (10:50 +1000)]
vmscan: add block plug for page reclaim

per-task block plug can reduce block queue lock contention and increase
request merge.  Currently page reclaim doesn't support it.  I originally
thought page reclaim doesn't need it, because kswapd thread count is
limited and file cache write is done at flusher mostly.

When I test a workload with heavy swap in a 4-node machine, each CPU is
doing direct page reclaim and swap.  This causes block queue lock
contention.  In my test, without below patch, the CPU utilization is about
2% ~ 7%.  With the patch, the CPU utilization is about 1% ~ 3%.  Disk
throughput isn't changed.  This should improve normal kswapd write and
file cache write too (increase request merge for example), but might not
be so obvious as I explain above.

Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Minchan Kim <minchan.kim@gmail.com>
Signed-off-by: Andrew Morton <>
13 years agoradix_tree: clean away saw_unset_tag leftovers
Hugh Dickins [Wed, 28 Sep 2011 00:50:00 +0000 (10:50 +1000)]
radix_tree: clean away saw_unset_tag leftovers

radix_tree_tag_get()'s BUG (when it sees a tag after saw_unset_tag) was
unsafe and removed in 2.6.34, but the pointless saw_unset_tag left behind.

Remove it now, and return 0 as soon as we see unset tag - we already rely
upon the root tag to be correct, returning 0 immediately if it's not set.

Signed-off-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <>
13 years agomm: migration: clean up unmap_and_move()
Minchan Kim [Wed, 28 Sep 2011 00:50:00 +0000 (10:50 +1000)]
mm: migration: clean up unmap_and_move()

unmap_and_move() is one a big messy function.  Clean it up.

Signed-off-by: Minchan Kim <minchan.kim@gmail.com>
Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agomm-zone_reclaim-make-isolate_lru_page-filter-aware-fix
Minchan Kim [Wed, 28 Sep 2011 00:50:00 +0000 (10:50 +1000)]
mm-zone_reclaim-make-isolate_lru_page-filter-aware-fix

Signed-off-by: Minchan Kim <minchan.kim@gmail.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agomm: zone_reclaim: make isolate_lru_page() filter-aware
Minchan Kim [Wed, 28 Sep 2011 00:49:59 +0000 (10:49 +1000)]
mm: zone_reclaim: make isolate_lru_page() filter-aware

In __zone_reclaim case, we don't want to shrink mapped page.  Nonetheless,
we have isolated mapped page and re-add it into LRU's head.  It's
unnecessary CPU overhead and makes LRU churning.

Of course, when we isolate the page, the page might be mapped but when we
try to migrate the page, the page would be not mapped.  So it could be
migrated.  But race is rare and although it happens, it's no big deal.

Signed-off-by: Minchan Kim <minchan.kim@gmail.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agomm-compaction-make-isolate_lru_page-filter-aware-fix
Minchan Kim [Wed, 28 Sep 2011 00:49:58 +0000 (10:49 +1000)]
mm-compaction-make-isolate_lru_page-filter-aware-fix

Signed-off-by: Minchan Kim <minchan.kim@gmail.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agomm: compaction: make isolate_lru_page() filter-aware
Minchan Kim [Wed, 28 Sep 2011 00:49:57 +0000 (10:49 +1000)]
mm: compaction: make isolate_lru_page() filter-aware

In async mode, compaction doesn't migrate dirty or writeback pages.  So,
it's meaningless to pick the page and re-add it to lru list.

Of course, when we isolate the page in compaction, the page might be dirty
or writeback but when we try to migrate the page, the page would be not
dirty, writeback.  So it could be migrated.  But it's very unlikely as
isolate and migration cycle is much faster than writeout.

So, this patch helps cpu overhead and prevent unnecessary LRU churning.

Signed-off-by: Minchan Kim <minchan.kim@gmail.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agomm-change-isolate-mode-from-define-to-bitwise-type-fix
Minchan Kim [Wed, 28 Sep 2011 00:49:57 +0000 (10:49 +1000)]
mm-change-isolate-mode-from-define-to-bitwise-type-fix

[c1e8b0ae8, mm-change-isolate-mode-from-define-to-bitwise-type]
made a mistake on the bitwise type.

This patch corrects it.

Cc: Mel Gorman <mgorman@suse.de>
Cc: Johannes Weiner <jweiner@redhat.com>
Reviewed-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Minchan Kim <minchan.kim@gmail.com>
Signed-off-by: Andrew Morton <>
13 years agomm: change isolate mode from #define to bitwise type
Minchan Kim [Wed, 28 Sep 2011 00:49:56 +0000 (10:49 +1000)]
mm: change isolate mode from #define to bitwise type

Change ISOLATE_XXX macro with bitwise isolate_mode_t type.  Normally,
macro isn't recommended as it's type-unsafe and making debugging harder as
symbol cannot be passed throught to the debugger.

Quote from Johannes
" Hmm, it would probably be cleaner to fully convert the isolation mode
into independent flags.  INACTIVE, ACTIVE, BOTH is currently a
tri-state among flags, which is a bit ugly."

This patch moves isolate mode from swap.h to mmzone.h by memcontrol.h

Signed-off-by: Minchan Kim <minchan.kim@gmail.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Rik van Riel <riel@redhat.com>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agomm: compaction: trivial clean up in acct_isolated()
Minchan Kim [Wed, 28 Sep 2011 00:49:56 +0000 (10:49 +1000)]
mm: compaction: trivial clean up in acct_isolated()

acct_isolated of compaction uses page_lru_base_type which returns only
base type of LRU list so it never returns LRU_ACTIVE_ANON or
LRU_ACTIVE_FILE.  In addtion, cc->nr_[anon|file] is used in only
acct_isolated so it doesn't have fields in conpact_control.

This patch removes fields from compact_control and makes clear function of
acct_issolated which counts the number of anon|file pages isolated.

Signed-off-by: Minchan Kim <minchan.kim@gmail.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agocross-memory-attach-v4
Christopher Yeoh [Wed, 28 Sep 2011 00:49:56 +0000 (10:49 +1000)]
cross-memory-attach-v4

> You might get some speed benefit by optimising for the small copies
> here.  Define a local on-stack array of N page*'s and point
> process_pages at that if the number of pages is <= N.  Saves a
> malloc/free and is more cache-friendly.  But only if the result is
> measurable!

I have done some benchmarking on this, and it gains about 5-7% on a
microbenchmark with 4kb size copies and about a 1% gain with a more
realistic (but modified for smaller copies) hpcc benchmark. The
performance gain disappears into the noise by about 64kb sized copies.
No measurable overhead for larger copies. So I think its worth including

Included below is the patch (based on v4) - for ease of review the first diff
is just against the latest version of CMA which has been posted here previously.
The second is the entire CMA patch.

Signed-off-by: Chris Yeoh <cyeoh@au1.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: David Howells <dhowells@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: <linux-man@vger.kernel.org>
Cc: <linux-arch@vger.kernel.org>
Signed-off-by: Andrew Morton <>
13 years agocross-memory-attach-update
Christopher Yeoh [Wed, 28 Sep 2011 00:49:55 +0000 (10:49 +1000)]
cross-memory-attach-update

- Add x86_64 specific wire up

- Change behaviour so process_vm_readv and process_vm_writev return
  the number of bytes successfully read or written even if an error
  occurs

- Add more kernel doc interface comments

- rename some internal functions (process_vm_rw_check_iovecs,
  process_vm_rw) so they make more sense.

- Add licence message

- Fix kernel-doc comment format

Still need to do benchmarking to see if the optimisation for small copies
using a local on-stack array in process_vm_rw_core is worth it.

Signed-off-by: Chris Yeoh <cyeoh@au1.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Andrew Morton <>
13 years agoCross Memory Attach
Christopher Yeoh [Wed, 28 Sep 2011 00:49:55 +0000 (10:49 +1000)]
Cross Memory Attach

The basic idea behind cross memory attach is to allow MPI programs doing
intra-node communication to do a single copy of the message rather than a
double copy of the message via shared memory.

The following patch attempts to achieve this by allowing a destination
process, given an address and size from a source process, to copy memory
directly from the source process into its own address space via a system
call.  There is also a symmetrical ability to copy from the current
process's address space into a destination process's address space.

- Use of /proc/pid/mem has been considered, but there are issues with
  using it:
  - Does not allow for specifying iovecs for both src and dest, assuming
    preadv or pwritev was implemented either the area read from or
  written to would need to be contiguous.
  - Currently mem_read allows only processes who are currently
  ptrace'ing the target and are still able to ptrace the target to read
  from the target. This check could possibly be moved to the open call,
  but its not clear exactly what race this restriction is stopping
  (reason  appears to have been lost)
  - Having to send the fd of /proc/self/mem via SCM_RIGHTS on unix
  domain socket is a bit ugly from a userspace point of view,
  especially when you may have hundreds if not (eventually) thousands
  of processes  that all need to do this with each other
  - Doesn't allow for some future use of the interface we would like to
  consider adding in the future (see below)
  - Interestingly reading from /proc/pid/mem currently actually
  involves two copies! (But this could be fixed pretty easily)

As mentioned previously use of vmsplice instead was considered, but has
problems.  Since you need the reader and writer working co-operatively if
the pipe is not drained then you block.  Which requires some wrapping to
do non blocking on the send side or polling on the receive.  In all to all
communication it requires ordering otherwise you can deadlock.  And in the
example of many MPI tasks writing to one MPI task vmsplice serialises the
copying.

There are some cases of MPI collectives where even a single copy interface
does not get us the performance gain we could.  For example in an
MPI_Reduce rather than copy the data from the source we would like to
instead use it directly in a mathops (say the reduce is doing a sum) as
this would save us doing a copy.  We don't need to keep a copy of the data
from the source.  I haven't implemented this, but I think this interface
could in the future do all this through the use of the flags - eg could
specify the math operation and type and the kernel rather than just
copying the data would apply the specified operation between the source
and destination and store it in the destination.

Although we don't have a "second user" of the interface (though I've had
some nibbles from people who may be interested in using it for intra
process messaging which is not MPI).  This interface is something which
hardware vendors are already doing for their custom drivers to implement
fast local communication.  And so in addition to this being useful for
OpenMPI it would mean the driver maintainers don't have to fix things up
when the mm changes.

There was some discussion about how much faster a true zero copy would
go. Here's a link back to the email with some testing I did on that:

http://marc.info/?l=linux-mm&m=130105930902915&w=2

There is a basic man page for the proposed interface here:

http://ozlabs.org/~cyeoh/cma/process_vm_readv.txt

This has been implemented for x86 and powerpc, other architecture should
mainly (I think) just need to add syscall numbers for the process_vm_readv
and process_vm_writev. There are 32 bit compatibility versions for
64-bit kernels.

For arch maintainers there are some simple tests to be able to quickly
verify that the syscalls are working correctly here:

http://ozlabs.org/~cyeoh/cma/cma-test-20110718.tgz

Signed-off-by: Chris Yeoh <yeohc@au1.ibm.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: David Howells <dhowells@redhat.com>
Cc: James Morris <jmorris@namei.org>
Cc: <linux-man@vger.kernel.org>
Cc: <linux-arch@vger.kernel.org>
Signed-off-by: Andrew Morton <>
13 years agodrivers/tty/serial/pch_uart.c: add console support
Alexander Stein [Wed, 28 Sep 2011 00:49:55 +0000 (10:49 +1000)]
drivers/tty/serial/pch_uart.c: add console support

Add console support to pch_uart.  To enable append e.g.
console=ttyPCH0,115200 to your kernel command line.

This is not expected work on CM-iTC boards due to their having a different
clock.

Signed-off-by: Alexander Stein <alexander.stein@systec-electronic.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: Andrew Morton <>
13 years agoslub: add taint flag outputting to debug paths
Dave Jones [Wed, 28 Sep 2011 00:49:54 +0000 (10:49 +1000)]
slub: add taint flag outputting to debug paths

When we get corruption reports, it's useful to see if the kernel was
tainted, to rule out problems we can't do anything about.

Signed-off-by: Dave Jones <davej@redhat.com>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Andrew Morton <>
13 years agoslab: add taint flag outputting to debug paths.
Dave Jones [Wed, 28 Sep 2011 00:49:54 +0000 (10:49 +1000)]
slab: add taint flag outputting to debug paths.

When we get corruption reports, it's useful to see if the kernel was
tainted, to rule out problems we can't do anything about.

Signed-off-by: Dave Jones <davej@redhat.com>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Andrew Morton <>
13 years agodrivers/block/mtip32xx/mtip32xx.c needs compat.h
Andrew Morton [Wed, 28 Sep 2011 00:49:53 +0000 (10:49 +1000)]
drivers/block/mtip32xx/mtip32xx.c needs compat.h

Cc: Asai Thambi S P <asamymuthupa@micron.com>
Cc: Sam Bradshaw <sbradshaw@micron.com>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Andrew Morton <>
13 years agoloop-cleanup-set_status-interface-checkpatch-fixes
Andrew Morton [Wed, 28 Sep 2011 00:49:52 +0000 (10:49 +1000)]
loop-cleanup-set_status-interface-checkpatch-fixes

WARNING: line over 80 characters
#120: FILE: drivers/block/loop.c:1388:
+ (struct loop_info __user *) arg);

total: 0 errors, 1 warnings, 92 lines checked

./patches/loop-cleanup-set_status-interface.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: Dmitry Monakhov <dmonakhov@openvz.org>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Andrew Morton <>
13 years agoloop: cleanup set_status interface
Dmitry Monakhov [Wed, 28 Sep 2011 00:49:52 +0000 (10:49 +1000)]
loop: cleanup set_status interface

1) Anyone who has read access to loopdev has permission to call set_status
   and may change important parameters such as lo_offset, lo_sizelimit and
   so on, which contradicts to read access pattern and definitely equals
   to write access pattern.
2) Add lo_offset over i_size check to prevent blkdev_size overflow.
   ##Testcase_bagin
   #dd if=/dev/zero of=./file bs=1k count=1
   #losetup /dev/loop0 ./file
   /* userspace_application */
   struct loop_info64 loinf;
   fd = open("/dev/loop0", O_RDONLY);
   ioctl(fd, LOOP_GET_STATUS64, &loinf);
   /* Set offset to any value which is bigger than i_size, and sizelimit
    * to nonzero value*/
   loinf.lo_offset = 4096*1024;
   loinf.lo_sizelimit = 1024;
   ioctl(fd, LOOP_SET_STATUS64, &loinf);
   /* After this loop device will have size similar to 0x7fffffffffxxxx */
   #blockdev --getsz /dev/loop0
   ##OUTPUT: 36028797018955968
   ##Testcase_end

Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Andrew Morton <>
13 years agoloop: prevent information leak after failed read
Dmitry Monakhov [Wed, 28 Sep 2011 00:49:51 +0000 (10:49 +1000)]
loop: prevent information leak after failed read

If read was not fully successful we have to fail whole bio to prevent
information leak of old pages

##Testcase_begin
dd if=/dev/zero of=./file bs=1M count=1
losetup /dev/loop0 ./file -o 4096
truncate -s 0 ./file
# OOps loop offset is now beyond i_size, so read will silently fail.
# So bio's pages would not be cleared, may which result in information leak.
hexdump -C /dev/loop0
##testcase_end

Signed-off-by: Dmitry Monakhov <dmonakhov@openvz.org>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Andrew Morton <>
13 years agodrivers/message/fusion/mptbase.c: ensure NUL-termination of MptCallbacksName elements
Ferenc Wagner [Wed, 28 Sep 2011 00:49:51 +0000 (10:49 +1000)]
drivers/message/fusion/mptbase.c: ensure NUL-termination of MptCallbacksName elements

I just stumbled upon this while pondering over
https://bugzilla.kernel.org/show_bug.cgi?id=26692 and thought this could
be made better.

[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: Ferenc Wagner <wferi@niif.hu>
Cc: Desai <kashyap.desai@lsi.com>
Signed-off-by: Andrew Morton <>
13 years agodrivers/scsi/mpt2sas/mpt2sas_base.c: fix mismatch in mpt2sas_base_hard_reset_handler...
Alexey Khoroshilov [Wed, 28 Sep 2011 00:49:50 +0000 (10:49 +1000)]
drivers/scsi/mpt2sas/mpt2sas_base.c: fix mismatch in mpt2sas_base_hard_reset_handler() mutex lock-unlock

If ioc->pci_error_recovery is set, goto out in
mpt2sas_base_hard_reset_handler() leads to unlock unheld
ioc->reset_in_progress_mutex.

Fix the issue by jumping afer mutex_unlock() call.

Found by Linux Driver Verification project (linuxtesting.org).

Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Cc: Kashyap Desai <kashyap.desai@lsi.com>
Signed-off-by: Andrew Morton <>
13 years agodrivers/scsi/sg.c: convert to kstrtoul_from_user()
Stephen Boyd [Wed, 28 Sep 2011 00:49:50 +0000 (10:49 +1000)]
drivers/scsi/sg.c: convert to kstrtoul_from_user()

Instead of open coding this function use kstrtoul_from_user() directly.

Signed-off-by: Stephen Boyd <sboyd@codeaurora.org>
Cc: Doug Gilbert <dgilbert@interlog.com>
Cc: Douglas Gilbert <dougg@torque.net>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Andrew Morton <>
13 years agodrivers/scsi/osd/osd_uld.c: use ida_simple_get() to handle id
Jonathan Cameron [Wed, 28 Sep 2011 00:49:49 +0000 (10:49 +1000)]
drivers/scsi/osd/osd_uld.c: use ida_simple_get() to handle id

This does involve additional use of the spin lock in idr.c.  Is this an
issue?

Also, some error mangling was needed to keep the interface the same.  Does
this matter or can we return -ENOSPC instead of -EBUSY?

Signed-off-by: Jonathan Cameron <jic23@cam.ac.uk>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Tejun Heo <tj@kernel.org>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Cc: Boaz Harrosh <bharrosh@panasas.com>
Cc: Benny Halevy <bhalevy@panasas.com>
Signed-off-by: Andrew Morton <>
13 years agodrivers/scsi/sd.c: use ida_simple_get() and ida_simple_remove() in place of boilerpla...
Jonathan Cameron [Wed, 28 Sep 2011 00:49:49 +0000 (10:49 +1000)]
drivers/scsi/sd.c: use ida_simple_get() and ida_simple_remove() in place of boilerplate code

Some mangling of errors was necessary to maintain current interface.

Signed-off-by: Jonathan Cameron <jic23@cam.ac.uk>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Tejun Heo <tj@kernel.org>
Cc: Guenter Roeck <guenter.roeck@ericsson.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Andrew Morton <>
13 years agodrivers/scsi/aacraid/commctrl.c: fix mem leak in aac_send_raw_srb()
Jesper Juhl [Wed, 28 Sep 2011 00:49:49 +0000 (10:49 +1000)]
drivers/scsi/aacraid/commctrl.c: fix mem leak in aac_send_raw_srb()

We leak in drivers/scsi/aacraid/commctrl.c::aac_send_raw_srb() :

We allocate memory:
        ...
                        struct user_sgmap* usg;
                        usg = kmalloc(actual_fibsize - sizeof(struct aac_srb)
                          + sizeof(struct sgmap), GFP_KERNEL);
and then neglect to free it:
        ...
                        for (i = 0; i < usg->count; i++) {
                                u64 addr;
                                void* p;
                                if (usg->sg[i].count >
                                    ((dev->adapter_info.options &
                                     AAC_OPT_NEW_COMM) ?
                                      (dev->scsi_host_ptr->max_sectors << 9) :
                                      65536)) {
                                        rcode = -EINVAL;
                                        goto cleanup;
        ... this 'goto' makes 'usg' go out of scope and leak the memory we
            allocated.
            Other exits properly kfree(usg), it's just here it is neglected.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Andrew Morton <>
13 years agodrivers/scsi/megaraid.c: fix sparse warnings
Randy Dunlap [Wed, 28 Sep 2011 00:49:48 +0000 (10:49 +1000)]
drivers/scsi/megaraid.c: fix sparse warnings

Fix sparse warnings of right shift bigger than source value size:

drivers/scsi/megaraid.c:311:65: warning: right shift by bigger than source value
drivers/scsi/megaraid.c:313:65: warning: right shift by bigger than source value
drivers/scsi/megaraid.c:317:67: warning: right shift by bigger than source value
drivers/scsi/megaraid.c:319:67: warning: right shift by bigger than source value

Patch suggestion from email by Al Viro:

"Since both are claimed to be strings, I really suspect that this >> 8 is
misspelled >> 4 and they have a character followed by pair of two-digit
packed decimals in there..."

Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Neela Syam Kolli <megaraidlinux@lsi.com>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Andrew Morton <>
13 years agoscsi: fix a header to include linux/types.h
Alexander Shishkin [Wed, 28 Sep 2011 00:49:48 +0000 (10:49 +1000)]
scsi: fix a header to include linux/types.h

For headers that get exported to userland and make use of u32 style
type names, it is advised to include linux/types.h.

This fixes a headers_check warning.

Signed-off-by: Alexander Shishkin <virtuoso@slind.org>
Cc: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: Andrew Morton <>
13 years agodrivers/firmware/dmi_scan.c: make dmi_name_in_vendors more focused
Jean Delvare [Wed, 28 Sep 2011 00:49:48 +0000 (10:49 +1000)]
drivers/firmware/dmi_scan.c: make dmi_name_in_vendors more focused

The current implementation of dmi_name_in_vendors() is an invitation to
lazy coding and false positives [1].  Searching for a string in 8 know
what you're looking for, so you should know where to look.  strstr isn't
fast, especially when it fails, so we should avoid calling it when it just
can't succeed.

Looking at the current users of the function, it seems clear to me that
they are looking for a system or board vendor name, so let's limit
dmi_name_in_vendors to these two DMI fields.  This much better matches the
function name, BTW.

[1] We currently have code looking for short names in DMI data, such
as "IBM", "ASUS" or "Acer". I let you guess what will happen the day
other vendors ship products named, for example, "SCHREIBMEISTER",
"PEGASUS" or "Acerola".

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Andrew Morton <>
13 years agoparisc, exec: remove redundant set_fs(USER_DS)
Mathias Krause [Wed, 28 Sep 2011 00:49:47 +0000 (10:49 +1000)]
parisc, exec: remove redundant set_fs(USER_DS)

The address limit is already set in flush_old_exec() so those calls to
set_fs(USER_DS) are redundant.

Signed-off-by: Mathias Krause <minipli@googlemail.com>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: Helge Deller <deller@gmx.de>
Signed-off-by: Andrew Morton <>
13 years agoocfs2: avoid unaligned access to dqc_bitmap
Akinobu Mita [Wed, 28 Sep 2011 00:49:47 +0000 (10:49 +1000)]
ocfs2: avoid unaligned access to dqc_bitmap

The dqc_bitmap field of struct ocfs2_local_disk_chunk is 32-bit aligned,
but not 64-bit aligned.  The dqc_bitmap is accessed by ocfs2_set_bit(),
ocfs2_clear_bit(), ocfs2_test_bit(), or ocfs2_find_next_zero_bit().  These
are wrapper macros for ext2_*_bit() which need to take an unsigned long
aligned address (though some architectures are able to handle unaligned
address correctly)

So some 64bit architectures may not be able to access the dqc_bitmap
correctly.

This avoids such unaligned access by using another wrapper functions for
ext2_*_bit().  The code is taken from fs/ext4/mballoc.c which also need to
handle unaligned bitmap access.

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Acked-by: Joel Becker <jlbec@evilplan.org>
Cc: Mark Fasheh <mfasheh@suse.com>
Signed-off-by: Andrew Morton <>
13 years agoext4: use proper little-endian bitops
Akinobu Mita [Wed, 28 Sep 2011 00:49:47 +0000 (10:49 +1000)]
ext4: use proper little-endian bitops

ext4_{set,clear}_bit() is defined as __test_and_{set,clear}_bit_le() for
ext4.  Only two ext4_{set,clear}_bit() calls check the return value.  The
rest of calls ignore the return value and they can be replaced with
__{set,clear}_bit_le().

This changes ext4_{set,clear}_bit() from __test_and_{set,clear}_bit_le()
to __{set,clear}_bit_le() and introduces ext4_test_and_{set,clear}_bit()
for the two places where old bit needs to be returned.

This ext4_{set,clear}_bit() change is considered safe, because if someone
uses these macros without noticing the change, new ext4_{set,clear}_bit
don't have return value and causes compiler errors where the return value
is used.

This also removes unused ext4_find_first_zero_bit().

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 <>
13 years agokernel/timer.c: use debugobjects to catch deletion of uninitialized timers
Christine Chan [Wed, 28 Sep 2011 00:49:46 +0000 (10:49 +1000)]
kernel/timer.c: use debugobjects to catch deletion of uninitialized timers

del_timer_sync() calls debug_object_assert_init() to assert that a timer
has been initialized before calling lock_timer_base().  lock_timer_base()
would spin forever on a NULL(uninit-ed) base.  The check is added to
del_timer() to prevent silent failure, even though it would not get stuck
in an infinite loop.

Signed-off-by: Christine Chan <cschan@codeaurora.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: John Stultz <john.stultz@linaro.org>
Signed-off-by: Andrew Morton <>
13 years agodebugobjects: extend debugobjects to assert that an object is initialized
Christine Chan [Wed, 28 Sep 2011 00:49:46 +0000 (10:49 +1000)]
debugobjects: extend debugobjects to assert that an object is initialized

Add new check (assert_init) to make sure objects are initialized and
tracked by debugobjects.

Signed-off-by: Christine Chan <cschan@codeaurora.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: John Stultz <john.stultz@linaro.org>
Signed-off-by: Andrew Morton <>
13 years agounicore32, exec: remove redundant set_fs(USER_DS)
Mathias Krause [Wed, 28 Sep 2011 00:49:45 +0000 (10:49 +1000)]
unicore32, exec: remove redundant set_fs(USER_DS)

The address limit is already set in flush_old_exec() so this
set_fs(USER_DS) is redundant.

Signed-off-by: Mathias Krause <minipli@googlemail.com>
Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Signed-off-by: Andrew Morton <>
13 years agoia64, exec: remove redundant set_fs(USER_DS)
Mathias Krause [Wed, 28 Sep 2011 00:49:44 +0000 (10:49 +1000)]
ia64, exec: remove redundant set_fs(USER_DS)

The address limit is already set in flush_old_exec() so this
set_fs(USER_DS) is redundant.

Signed-off-by: Mathias Krause <minipli@googlemail.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Signed-off-by: Andrew Morton <>
13 years agodrivers/gpu/vga/vgaarb.c: add missing kfree
Julia Lawall [Wed, 28 Sep 2011 00:49:44 +0000 (10:49 +1000)]
drivers/gpu/vga/vgaarb.c: add missing kfree

kbuf is a buffer that is local to this function, so all of the error paths
leaving the function should release it.

Signed-off-by: Julia Lawall <julia@diku.dk>
Cc: Dave Airlie <airlied@redhat.com>
Cc: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Andrew Morton <>
13 years agodrivers/edac/mpc85xx_edac.c: fix memory controller compatible for edac
Shaohui Xie [Wed, 28 Sep 2011 00:49:44 +0000 (10:49 +1000)]
drivers/edac/mpc85xx_edac.c: fix memory controller compatible for edac

compatible in dts has been changed, so the driver needs to be updated
accordingly.

Signed-off-by: Shaohui Xie <Shaohui.Xie@freescale.com>
Cc: Kumar Gala <galak@kernel.crashing.org>
Cc: Grant Likely <grant.likely@secretlab.ca>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Andrew Morton <>
13 years agobtrfs: don't dereference extent_mapping if NULL
Roel Kluin [Wed, 28 Sep 2011 00:49:43 +0000 (10:49 +1000)]
btrfs: don't dereference extent_mapping if NULL

Don't dereference em if it's NULL or an error pointer.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Cc: Chris Mason <chris.mason@oracle.com>
Signed-off-by: Andrew Morton <>
13 years agoaudit: always follow va_copy() with va_end()
Jesper Juhl [Wed, 28 Sep 2011 00:49:43 +0000 (10:49 +1000)]
audit: always follow va_copy() with va_end()

A call to va_copy() should always be followed by a call to va_end() in the
same function.  In kernel/autit.c::audit_log_vformat() this is not always
done.  This patch makes sure va_end() is always called.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Eric Paris <eparis@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agoarm, exec: remove redundant set_fs(USER_DS)
Mathias Krause [Wed, 28 Sep 2011 00:49:43 +0000 (10:49 +1000)]
arm, exec: remove redundant set_fs(USER_DS)

The address limit is already set in flush_old_exec() so this
set_fs(USER_DS) is redundant.

Signed-off-by: Mathias Krause <minipli@googlemail.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Signed-off-by: Andrew Morton <>
13 years agoarch/arm/mach-ux500/mbox-db5500.c: world-writable sysfs fifo file
Vasiliy Kulikov [Wed, 28 Sep 2011 00:49:42 +0000 (10:49 +1000)]
arch/arm/mach-ux500/mbox-db5500.c: world-writable sysfs fifo file

Don't allow everybody to use a modem.

Signed-off-by: Vasiliy Kulikov <segoon@openwall.com>
Cc: Srinidhi Kasagar <srinidhi.kasagar@stericsson.com>
Cc: Linus Walleij <linus.walleij@stericsson.com>
Cc: Russell King <linux@arm.linux.org.uk>
Signed-off-by: Andrew Morton <>
13 years agox86: tlb flush avoid superflous leave_mm()
Shaohua Li [Wed, 28 Sep 2011 00:49:42 +0000 (10:49 +1000)]
x86: tlb flush avoid superflous leave_mm()

If just one page VA tlb is required to be flushed and current task is in
lazy TLB state, doing leave_mm() is superfluous because it flushes the
whole TLB.  This can reduce some TLB miss.

Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <>
13 years agoarch/x86/mm/pageattr.c: quiet sparse noise; local functions should be static
H Hartley Sweeten [Wed, 28 Sep 2011 00:49:42 +0000 (10:49 +1000)]
arch/x86/mm/pageattr.c: quiet sparse noise; local functions should be static

Local functions should be marked static.  This quiets the following
sparse noise:

warning: symbol '_set_memory_array' was not declared. Should it be static?

Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: Andrew Morton <>
13 years agoarch/x86/kernel/ptrace.c: quiet sparse noise
H Hartley Sweeten [Wed, 28 Sep 2011 00:49:41 +0000 (10:49 +1000)]
arch/x86/kernel/ptrace.c: quiet sparse noise

ptrace_set_debugreg() is only used in this file and should be static.
This quiets the following sparse warning:

warning: symbol 'ptrace_set_debugreg' was not declared. Should it be static?

Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: Andrew Morton <>
13 years agoarch/x86/kernel/e820.c: quiet sparse noise about plain integer as NULL pointer
H Hartley Sweeten [Wed, 28 Sep 2011 00:49:41 +0000 (10:49 +1000)]
arch/x86/kernel/e820.c: quiet sparse noise about plain integer as NULL pointer

The last parameter to sort() is a pointer to the function used to swap
items.  This parameter should be NULL, not 0, when not used.  This quiets
the following sparse warning:

warning: Using plain integer as NULL pointer

Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: Andrew Morton <>
13 years agox86,mrst: add mapping for bma023
William Douglas [Wed, 28 Sep 2011 00:49:41 +0000 (10:49 +1000)]
x86,mrst: add mapping for bma023

There is now an upstream bma023 driver so instead of submitting ours we
use that one.  The defaults are just fine so it's a simple mapping entry.

(Thanks go to Erik Andersson for incorporating the changes we needed into his
 version)

Signed-off-by: William Douglas <william.douglas@intel.com>
Signed-off-by: Alan Cox <alan@linux.intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Andrew Morton <>
13 years agodrivers/power/intel_mid_battery.c: fix build
Andrew Morton [Wed, 28 Sep 2011 00:49:40 +0000 (10:49 +1000)]
drivers/power/intel_mid_battery.c: fix build

Seems that nobody's even trying any more.

Cc: Nithish Mahalingam <nithish.mahalingam@intel.com>
Cc: Alan Cox <alan@linux.intel.com>
Cc: Anton Vorontsov <cbouatmailru@gmail.com>
Cc: Major Lee <major_lee@wistron.com>
Signed-off-by: Andrew Morton <>
13 years agomrst: battery fixes
Major Lee [Wed, 28 Sep 2011 00:49:40 +0000 (10:49 +1000)]
mrst: battery fixes

When DCDC input line over current detecting, PMIC will change charging
current automatically.  Logging event is enough.

Signed-off-by: Major Lee <major_lee@wistron.com>
Signed-off-by: Alan Cox <alan@linux.intel.com>
Cc: Mathias Nyman <mathias.nyman@linux.intel.com>
Cc: Feng Tang <feng.tang@intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Andrew Morton <>
13 years agox86: rtc: don't register a platform RTC device for Intel MID platforms
Mathias Nyman [Wed, 28 Sep 2011 00:49:40 +0000 (10:49 +1000)]
x86: rtc: don't register a platform RTC device for Intel MID platforms

Intel MID x86 platforms have a memory mapped virtual RTC instead.  No MID
platform have the default ports (and accessing them may do weird stuff)

Signed-off-by: Mathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: Alan Cox <alan@linux.intel.com>
Cc: Feng Tang <feng.tang@intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Andrew Morton <>
13 years agovrtc: change its year offset from 1960 to 1972
Feng Tang [Wed, 28 Sep 2011 00:49:39 +0000 (10:49 +1000)]
vrtc: change its year offset from 1960 to 1972

Real world year equals the value in vrtc YEAR register plus an offset.  We
used 1960 for original developepment as the offset to make leap year
consistent, but for a device's first use, its YEAR register is 0 and the
system year will be parsed as 1960 which is not a valid UNIX time and will
cause many applications to fail mysteriously.  Devices use 1972 instead to
fix this issue.

Updated patch which adds a sanity check suggested by Mathias

Signed-off-by: Feng Tang <feng.tang@intel.com>
Signed-off-by: Alan Cox <alan@linux.intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Mathias Nyman <mathias.nyman@linux.intel.com>
Signed-off-by: Andrew Morton <>
13 years agoarch/x86/kernel/e820.c: eliminate bubble sort from sanitize_e820_map
Mike Ditto [Wed, 28 Sep 2011 00:49:39 +0000 (10:49 +1000)]
arch/x86/kernel/e820.c: eliminate bubble sort from sanitize_e820_map

Replace the bubble sort in sanitize_e820_map() with a call to the generic
kernel sort function to avoid pathological performance with large maps.

On large (thousands of entries) E820 maps, the previous code took minutes
to run; with this change it's now milliseconds.

Signed-off-by: Mike Ditto <mditto@google.com>
Cc: Stefan Assmann <sassmann@kpanic.de>
Cc: Nancy Yuen <yuenn@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Andrew Morton <>
13 years agox86: fix mmap random address range
Ludwig Nussel [Wed, 28 Sep 2011 00:49:39 +0000 (10:49 +1000)]
x86: fix mmap random address range

On x86_32 casting the unsigned int result of get_random_int() to long may
result in a negative value.  On x86_32 the range of mmap_rnd() therefore
was -255 to 255.  The 32bit mode on x86_64 used 0 to 255 as intended.

The bug was introduced by 675a081 ("x86: unify mmap_{32|64}.c") in January
2008.

Signed-off-by: Ludwig Nussel <ludwig.nussel@suse.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Harvey Harrison <harvey.harrison@gmail.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <>
13 years agohp_accel: Add axis-mapping for HP ProBook / EliteBook
Takashi Iwai [Wed, 28 Sep 2011 00:49:38 +0000 (10:49 +1000)]
hp_accel: Add axis-mapping for HP ProBook / EliteBook

Add the corrected axis-mapping for some HP laptops.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Cc: Matthew Garrett <mjg@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agohp_accel: Add a new PNP id
Takashi Iwai [Wed, 28 Sep 2011 00:49:38 +0000 (10:49 +1000)]
hp_accel: Add a new PNP id

New HP laptops assign a new PNP id "HPQ6000" for DriveGuard.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Cc: Matthew Garrett <mjg@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agoarch/x86/platform/iris/iris.c: register a platform device and a platform driver
Shérab [Wed, 28 Sep 2011 00:49:37 +0000 (10:49 +1000)]
arch/x86/platform/iris/iris.c: register a platform device and a platform driver

This makes the iris driver use the platform API, so it is properly exposed
in /sys.

[akpm@linux-foundation.org: remove commented-out code, add missing space to printk, clean up code layout]
Signed-off-by: Shérab <Sebastien.Hinderer@ens-lyon.org>
Cc: Len Brown <lenb@kernel.org>
Cc: Matthew Garrett <mjg@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agoacerhdf: add support for Aspire 1410 BIOS v1.3314
Clay Carpenter [Wed, 28 Sep 2011 00:49:37 +0000 (10:49 +1000)]
acerhdf: add support for Aspire 1410 BIOS v1.3314

Add support for Aspire 1410 BIOS v1.3314.  Fixes the following error:

acerhdf: unknown (unsupported) BIOS version Acer/Aspire 1410/v1.3314,
please report, aborting!

Signed-off-by: Clay Carpenter <claycarpenter@gmail.com>
Signed-off-by: Peter Feuerer <peter@piie.net>
Cc: Matthew Garrett <mjg@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agoreadlinkat: ensure we return ENOENT for the empty pathname for normal lookups
Andy Whitcroft [Wed, 28 Sep 2011 00:49:37 +0000 (10:49 +1000)]
readlinkat: ensure we return ENOENT for the empty pathname for normal lookups

Since the commit below which added O_PATH support to the *at() calls, the
error return for readlink/readlinkat for the empty pathname has switched
from ENOENT to EINVAL:

  commit 65cfc6722361570bfe255698d9cd4dccaf47570d
  Author: Al Viro <viro@zeniv.linux.org.uk>
  Date:   Sun Mar 13 15:56:26 2011 -0400

    readlinkat(), fchownat() and fstatat() with empty relative pathnames

This is both unexpected for userspace and makes readlink/readlinkat
inconsistant with all other interfaces; and inconsistant with our stated
return for these pathnames.

As the readlinkat call does not have a flags parameter we cannot use the
AT_EMPTY_PATH approach used in the other calls.  Therefore expose whether
the original path is infact entry via a new user_path_at_empty() path
lookup function.  Use this to determine whether to default to EINVAL or
ENOENT for failures.

BugLink: http://bugs.launchpad.net/bugs/817187
Signed-off-by: Andy Whitcroft <apw@canonical.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <>
13 years agonet/netfilter/nf_conntrack_netlink.c: fix Oops on container destroy
Alex Bligh [Wed, 28 Sep 2011 00:49:36 +0000 (10:49 +1000)]
net/netfilter/nf_conntrack_netlink.c: fix Oops on container destroy

Problem:

A repeatable Oops can be caused if a container with networking
unshared is destroyed when it has nf_conntrack entries yet to expire.

A copy of the oops follows below. A perl program generating the oops
repeatably is attached inline below.

Analysis:

The oops is called from cleanup_net when the namespace is
destroyed. conntrack iterates through outstanding events and calls
death_by_timeout on each of them, which in turn produces a call to
ctnetlink_conntrack_event. This calls nf_netlink_has_listeners, which
oopses because net->nfnl is NULL.

The perl program generates the container through fork() then
clone(NS_NEWNET). I does not explicitly set up netlink
explicitly set up netlink, but I presume it was set up else net->nfnl
would have been NULL earlier (i.e. when an earlier connection
timed out). This would thus suggest that net->nfnl is made NULL
during the destruction of the container, which I think is done by
nfnetlink_net_exit_batch.

I can see that the various subsystems are deinitialised in the opposite
order to which the relevant register_pernet_subsys calls are called,
and both nf_conntrack and nfnetlink_net_ops register their relevant
subsystems. If nfnetlink_net_ops registered later than nfconntrack,
then its exit routine would have been called first, which would cause
the oops described. I am not sure there is anything to prevent this
happening in a container environment.

Whilst there's perhaps a more complex problem revolving around ordering
of subsystem deinit, it seems to me that missing a netlink event on a
container that is dying is not a disaster. An early check for net->nfnl
being non-NULL in ctnetlink_conntrack_event appears to fix this. There
may remain a potential race condition if it becomes NULL immediately
after being checked (I am not sure any lock is held at this point or
how synchronisation for subsystem deinitialization works).

Patch:

The patch attached should apply on everything from 2.6.26 (if not before)
onwards; it appears to be a problem on all kernels. This was taken against
Ubuntu-3.0.0-11.17 which is very close to 3.0.4. I have torture-tested it
with the above perl script for 15 minutes or so; the perl script hung the
machine within 20 seconds without this patch.

Applicability:

If this is the right solution, it should be applied to all stable kernels
as well as head. Apart from the minor overhead of checking one variable
against NULL, it can never 'do the wrong thing', because if net->nfnl
is NULL, an oops will inevitably result. Therefore, checking is a reasonable
thing to do unless it can be proven than net->nfnl will never be NULL.

Check net->nfnl for NULL in ctnetlink_conntrack_event to avoid Oops on
container destroy

Signed-off-by: Alex Bligh <alex@alex.org.uk>
Cc: Patrick McHardy <kaber@trash.net>
Cc: David Miller <davem@davemloft.net>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <>
13 years agodrm: fix kconfig unmet dependency warning
Randy Dunlap [Wed, 28 Sep 2011 00:49:36 +0000 (10:49 +1000)]
drm: fix kconfig unmet dependency warning

Fix kconfig unmet dependency warning.  BACKLIGHT_CLASS_DEVICE depends on
BACKLIGHT_LCD_SUPPORT, so select the latter along with the former.

warning: (DRM_RADEON_KMS && DRM_I915 && STUB_POULSBO && FB_BACKLIGHT && PANEL_SHARP_LS037V7DW01 && PANEL_ACX565AKM && USB_APPLEDISPLAY && FB_OLPC_DCON && ASUS_LAPTOP && SONY_LAPTOP && THINKPAD_ACPI && EEEPC_LAPTOP && ACPI_ASUS && ACPI_CMPC && SAMSUNG_Q10) selects BACKLIGHT_CLASS_DEVICE which has unmet direct dependencies (HAS_IOMEM && BACKLIGHT_LCD_SUPPORT)

Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Cc: David Airlie <airlied@linux.ie>
Signed-off-by: Andrew Morton <>
13 years agodrivers/net/ethernet/i825xx/3c505.c: fix build with dynamic debug
Andrew Morton [Wed, 28 Sep 2011 00:49:36 +0000 (10:49 +1000)]
drivers/net/ethernet/i825xx/3c505.c: fix build with dynamic debug

The `#define filename' screws up the expansion of
DEFINE_DYNAMIC_DEBUG_METADATA:

drivers/net/ethernet/i825xx/3c505.c: In function 'send_pcb':
drivers/net/ethernet/i825xx/3c505.c:390: error: expected identifier before string constant
drivers/net/ethernet/i825xx/3c505.c:390: error: expected '}' before '.' token
drivers/net/ethernet/i825xx/3c505.c:436: error: expected identifier before string constant
drivers/net/ethernet/i825xx/3c505.c:435: error: expected '}' before '.' token
drivers/net/ethernet/i825xx/3c505.c: In function 'start_receive':
drivers/net/ethernet/i825xx/3c505.c:557: error: expected identifier before string constant
drivers/net/ethernet/i825xx/3c505.c:557: error: expected '}' before '.' token
drivers/net/ethernet/i825xx/3c505.c: In function 'receive_packet':
drivers/net/ethernet/i825xx/3c505.c:629: error: expected identifier before string constant

etc

Cc: Philip Blundell <philb@gnu.org>
Cc: David Miller <davem@davemloft.net>
Cc: Jason Baron <jbaron@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agoarch/x86/kernel/cpu/perf_event_amd.c needs asm/apicdef.h
Andrew Morton [Wed, 28 Sep 2011 00:49:35 +0000 (10:49 +1000)]
arch/x86/kernel/cpu/perf_event_amd.c needs asm/apicdef.h

x86_64 allnoconfig:

arch/x86/kernel/cpu/perf_event_amd.c: In function 'amd_pmu_cpu_starting':
arch/x86/kernel/cpu/perf_event_amd.c:350: error: 'BAD_APICID' undeclared (first use in this function)
arch/x86/kernel/cpu/perf_event_amd.c:350: error: (Each undeclared identifier is reported only once
arch/x86/kernel/cpu/perf_event_amd.c:350: error: for each function it appears in.)

Cc: Kevin Winchester <kjwinchester@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Ingo Molnar <mingo@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agoinclude/linux/dmar.h: forward-declare struct acpi_dmar_header
Andrew Morton [Wed, 28 Sep 2011 00:49:35 +0000 (10:49 +1000)]
include/linux/dmar.h: forward-declare struct acpi_dmar_header

x86_64 allnoconfig:

In file included from arch/x86/kernel/pci-dma.c:3:
include/linux/dmar.h:248: warning: 'struct acpi_dmar_header' declared inside parameter list
include/linux/dmar.h:248: warning: its scope is only this definition or declaration, which is probably not what you want

Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Ingo Molnar <mingo@redhat.com>
Signed-off-by: Andrew Morton <>
13 years agoMerge remote-tracking branch 'kvmtool/master'
Stephen Rothwell [Tue, 4 Oct 2011 07:33:53 +0000 (18:33 +1100)]
Merge remote-tracking branch 'kvmtool/master'

Conflicts:
include/net/9p/9p.h

13 years agoMerge remote-tracking branch 'moduleh/for-sfr'
Stephen Rothwell [Tue, 4 Oct 2011 06:41:29 +0000 (17:41 +1100)]
Merge remote-tracking branch 'moduleh/for-sfr'

Conflicts:
arch/arm/mach-bcmring/mm.c
drivers/media/dvb/frontends/dibx000_common.c
drivers/misc/altera-stapl/altera.c
drivers/net/ethernet/oki-semi/pch_gbe/pch_gbe_main.c
drivers/scsi/libfc/fc_lport.c
drivers/staging/iio/adc/ad7152.c
drivers/staging/iio/adc/ad7291.c
drivers/staging/iio/adc/ad7745.c
drivers/staging/iio/industrialio-ring.c
include/linux/dmaengine.h
sound/soc/soc-io.c