Transparent huge pages can change page->flags (PG_compound_lock) without
taking Slab lock. Since THP can not break slab pages we can safely access
compound page without taking compound lock.
Specifically this patch fixes a race between compound_unlock() and slab
functions which perform page-flags updates. This can occur when
get_page()/put_page() is called on a page from slab.
[akpm@linux-foundation.org: tweak comment text, fix comment layout, fix label indenting]
Reported-by: Amey Bhide <abhide@nicira.com>
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Reviewed-by: Christoph Lameter <cl@linux.com>
Acked-by: Andrea Arcangeli <aarcange@redhat.com>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The transfer of ->flags causes some of the static mapping virtual
addresses to be prematurely freed (before the mapping is removed) because
VM_LAZY_FREE gets "set" if tmp->flags has VM_IOREMAP set. This might
cause subsequent vmalloc/ioremap calls to fail because it might allocate
one of the freed virtual address ranges that aren't unmapped.
va->flags has different types of flags from tmp->flags. If a region with
VM_IOREMAP set is registered with vm_area_add_early(), it will be removed
by __purge_vmap_area_lazy().
Fix vmalloc_init() to correctly initialize vmap_area for the given
vm_struct.
Also initialise va->vm. If it is not set, find_vm_area() for the early
vm regions will always fail.
Signed-off-by: KyongHo Cho <pullip.cho@samsung.com>
Cc: "Olav Haugan" <ohaugan@codeaurora.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The oom_score_adj scale ranges from -1000 to 1000 and represents the
proportion of memory available to the process at allocation time. This
means an oom_score_adj value of 300, for example, will bias a process as
though it was using an extra 30.0% of available memory and a value of
-350 will discount 35.0% of available memory from its usage.
The oom killer badness heuristic also uses this scale to report the oom
score for each eligible process in determining the "best" process to
kill. Thus, it can only differentiate each process's memory usage by
0.1% of system RAM.
On large systems, this can end up being a large amount of memory: 256MB
on 256GB systems, for example.
This can be fixed by having the badness heuristic to use the actual
memory usage in scoring threads and then normalizing it to the
oom_score_adj scale for userspace. This results in better comparison
between eligible threads for kill and no change from the userspace
perspective.
Suggested-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Tested-by: Dave Jones <davej@redhat.com>
Signed-off-by: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Sometimes we'd like to avoid swapping out anonymous memory. In
particular, avoid swapping out pages of important process or process
groups while there is a reasonable amount of pagecache on RAM so that we
can satisfy our customers' requirements.
OTOH, we can control how aggressive the kernel will swap memory pages with
/proc/sys/vm/swappiness for global and
/sys/fs/cgroup/memory/memory.swappiness for each memcg.
But with current reclaim implementation, the kernel may swap out even if
we set swappiness=0 and there is pagecache in RAM.
This patch changes the behavior with swappiness==0. If we set
swappiness==0, the kernel does not swap out completely (for global reclaim
until the amount of free pages and filebacked pages in a zone has been
reduced to something very very small (nr_free + nr_filebacked < high
watermark)).
Signed-off-by: Satoru Moriya <satoru.moriya@hds.com>
Acked-by: Minchan Kim <minchan@kernel.org>
Reviewed-by: Rik van Riel <riel@redhat.com>
Acked-by: Jerome Marchand <jmarchan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When called for anonymous (non-shared) mappings, hugetlb_reserve_pages()
does a resv_map_alloc(). It depends on code in hugetlbfs's
vm_ops->close() to release that allocation.
However, in the mmap() failure path, we do a plain unmap_region() without
the remove_vma() which actually calls vm_ops->close().
This is a decent fix. This leak could get reintroduced if new code (say,
after hugetlb_reserve_pages() in hugetlbfs_file_mmap()) decides to return
an error. But, I think it would have to unroll the reservation anyway.
Christoph's test case:
http://marc.info/?l=linux-mm&m=133728900729735
This patch applies to 3.4 and later. A version for earlier kernels is at
https://lkml.org/lkml/2012/5/22/418.
Signed-off-by: Dave Hansen <dave@linux.vnet.ibm.com>
Acked-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Reported-by: Christoph Lameter <cl@linux.com>
Tested-by: Christoph Lameter <cl@linux.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: <stable@vger.kernel.org> [2.6.32+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The objects of "struct bootmem_data_t" are linked together to form
double-linked list sequentially based on its minimal page frame number.
The current implementation implicitly supports the following cases,
which means the inserting point for current bootmem data depends on how
"list_for_each" works. That makes the code a little hard to read.
Besides, "list_for_each" and "list_entry" can be replaced with
"list_for_each_entry".
- The linked list is empty.
- There has no entry in the linked list, whose minimal page
frame number is bigger than current one.
Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit 6457474624 ("vmscan: detect mapped file pages used only once")
made mapped pages have another round in inactive list because they might
be just short lived and so we could consider them again next time. This
heuristic helps to reduce pressure on the active list with a streaming
IO worklods.
This patch fixes a regression introduced by this commit for heavy shmem
based workloads because unlike Anon pages, which are excluded from this
heuristic because they are usually long lived, shmem pages are handled
as a regular page cache.
This doesn't work quite well, unfortunately, if the workload is mostly
backed by shmem (in memory database sitting on 80% of memory) with a
streaming IO in the background (backup - up to 20% of memory). Anon
inactive list is full of (dirty) shmem pages when watermarks are hit.
Shmem pages are kept in the inactive list (they are referenced) in the
first round and it is hard to reclaim anything else so we reach lower
scanning priorities very quickly which leads to an excessive swap out.
Let's fix this by excluding all swap backed pages (they tend to be long
lived wrt. the regular page cache anyway) from used-once heuristic and
rather activate them if they are referenced.
The customer's workload is shmem backed database (80% of RAM) and they
are measuring transactions/s with an IO in the background (20%).
Transactions touch more or less random rows in the table. The
transaction rate fell by a factor of 3 (in the worst case) because of
commit 64574746. This patch restores the previous numbers.
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Minchan Kim <minchan@kernel.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Reviewed-by: Rik van Riel <riel@redhat.com>
Cc: <stable@vger.kernel.org> [2.6.34+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
String tables with names of enum items are always prone to go out of
sync with the enums themselves. Ensure during compile time that the
name table of page flags has the same size as the page flags enum.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The array pageflag_names[] does conversion from page flags into their
corresponding names so that a meaningful representation of the
corresponding page flag can be printed. This mechanism is used while
dumping page frames. However, the array missed PG_compound_lock. So
the PG_compound_lock page flag would be printed as a digital number
instead of a meaningful string.
The patch fixes that and prints "compound_lock" for the PG_compound_lock
page flag.
Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
It is better to define readahead(2) in mm/readahead.c than in
mm/filemap.c.
Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Fengguang Wu <fengguang.wu@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
It's quite easy for tmpfs to scan the radix_tree to support llseek's new
SEEK_DATA and SEEK_HOLE options: so add them while the minutiae are still
on my mind (in particular, the !PageUptodate-ness of pages fallocated but
still unwritten).
But I don't know who actually uses SEEK_DATA or SEEK_HOLE, and whether it
would be of any use to them on tmpfs. This code adds 92 lines and 752
bytes on x86_64 - is that bloat or worthwhile?
[akpm@linux-foundation.org: fix warning with CONFIG_TMPFS=n]
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Josef Bacik <josef@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Andreas Dilger <adilger@dilger.ca>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Marco Stornelli <marco.stornelli@gmail.com>
Cc: Jeff liu <jeff.liu@oracle.com>
Cc: Chris Mason <chris.mason@oracle.com>
Cc: Sunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
As it stands, a large fallocate() on tmpfs is liable to fill memory with
pages, freed on failure except when they run into swap, at which point
they become fixed into the file despite the failure. That feels quite
wrong, to be consuming resources precisely when they're in short supply.
Go the other way instead: shmem_fallocate() indicate the range it has
fallocated to shmem_writepage(), keeping count of pages it's allocating;
shmem_writepage() reactivate instead of swapping out pages fallocated by
this syscall (but happily swap out those from earlier occasions), keeping
count; shmem_fallocate() compare counts and give up once the reactivated
pages have started to coming back to writepage (approximately: some zones
would in fact recycle faster than others).
This is a little unusual, but works well: although we could consider the
failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling
added in swapfile.c and memcontrol.c, I doubt that we shall ever want to.
(If there's no swap, an over-large fallocate() on tmpfs is limited in the
same way as writing: stopped by rlimit, or by tmpfs mount size if that was
set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or
OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill
others as writing would, but stops and frees if interrupted.)
Now that everything is freed on failure, we can then skip updating ctime.
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Cong Wang <amwang@redhat.com>
Cc: Kay Sievers <kay@vrfy.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In the previous episode, we left the already-fallocated pages attached to
the file when shmem_fallocate() fails part way through.
Now try to do better, by extending the earlier optimization of !Uptodate
pages (then always under page lock) to !Uptodate pages (outside of page
lock), representing fallocated pages. And don't waste time clearing them
at the time of fallocate(), leave that until later if necessary.
Adapt shmem_truncate_range() to shmem_undo_range(), so that a failing
fallocate can recognize and remove precisely those !Uptodate allocations
which it added (and were not independently allocated by racing tasks).
But unless we start playing with swapfile.c and memcontrol.c too, once one
of our fallocated pages reaches shmem_writepage(), we do then have to
instantiate it as an ordinarily allocated page, before swapping out. This
is unsatisfactory, but improved in the next episode.
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Cong Wang <amwang@redhat.com>
Cc: Kay Sievers <kay@vrfy.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The systemd plumbers expressed a wish that tmpfs support preallocation.
Cong Wang wrote a patch, but several kernel guys expressed scepticism:
https://lkml.org/lkml/2011/11/18/137
Christoph Hellwig: What for exactly? Please explain why preallocating on
tmpfs would make any sense.
Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files
on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or
-EOPNOTSUPP] on fallocate is just ugly.
Hugh Dickins: If tmpfs is going to support
fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the
deallocation but fail the allocation. Christoph Hellwig: Agreed.
Now that we do have shmem_fallocate() for hole-punching, plumb in basic
support for preallocation mode too. It's fairly straightforward (though
quite a few details needed attention), except for when it fails part way
through. What a pity that fallocate(2) was not specified to return the
length allocated, permitting short fallocations!
As it is, when it fails part way through, we ought to free what has just
been allocated by this system call; but must be very sure not to free any
allocated earlier, or any allocated by racing accesses (not all excluded
by i_mutex).
But we cannot distinguish them: so in this patch simply leak allocations
on partial failure (they will be freed later if the file is removed).
An attractive alternative approach would have been for fallocate() not to
allocate pages at all, but note reservations by entries in the radix-tree.
But that would give less assurance, and, critically, would be hard to fit
with mem cgroups (who owns the reservations?): allocating pages lets
fallocate() behave in just the same way as write().
Based-on-patch-by: Cong Wang <amwang@redhat.com>
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Cong Wang <amwang@redhat.com>
Cc: Kay Sievers <kay@vrfy.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Remove vmtruncate_range(), and remove the truncate_range method from
struct inode_operations: only tmpfs ever supported it, and tmpfs has now
converted over to using the fallocate method of file_operations.
Update Documentation accordingly, adding (setlease and) fallocate lines.
And while we're in mm.h, remove duplicate declarations of shmem_lock() and
shmem_file_setup(): everyone is now using the ones in shmem_fs.h.
Based-on-patch-by: Cong Wang <amwang@redhat.com>
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Cong Wang <amwang@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Now tmpfs supports hole-punching via fallocate(), switch madvise_remove()
to use do_fallocate() instead of vmtruncate_range(): which extends
madvise(,,MADV_REMOVE) support from tmpfs to ext4, ocfs2 and xfs.
There is one more user of vmtruncate_range() in our tree,
staging/android's ashmem_shrink(): convert it to use do_fallocate() too
(but if its unpinned areas are already unmapped - I don't know - then it
would do better to use shmem_truncate_range() directly).
Based-on-patch-by: Cong Wang <amwang@redhat.com>
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Colin Cross <ccross@android.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Greg Kroah-Hartman <gregkh@linux-foundation.org>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Andreas Dilger <adilger@dilger.ca>
Cc: Mark Fasheh <mfasheh@suse.de>
Cc: Joel Becker <jlbec@evilplan.org>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Ben Myers <bpm@sgi.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
tmpfs has supported hole-punching since 2.6.16, via
madvise(,,MADV_REMOVE).
But nowadays fallocate(,FALLOC_FL_PUNCH_HOLE|FALLOC_FL_KEEP_SIZE,,) is
the agreed way to punch holes.
So add shmem_fallocate() to support that, and tweak shmem_truncate_range()
to support partial pages at both the beginning and end of range (never
needed for madvise, which demands rounded addr and rounds up length).
Based-on-patch-by: Cong Wang <amwang@redhat.com>
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Cong Wang <amwang@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Nick proposed years ago that tmpfs should avoid clearing its pages where
write will overwrite them with new data, as ramfs has long done. But I
messed it up and just got bad data. Tried again recently, it works
fine.
Here's time output for writing 4GiB 16 times on this Core i5 laptop:
before: real 0m21.169s user 0m0.028s sys 0m21.057s
real 0m21.382s user 0m0.016s sys 0m21.289s
real 0m21.311s user 0m0.020s sys 0m21.217s
after: real 0m18.273s user 0m0.032s sys 0m18.165s
real 0m18.354s user 0m0.020s sys 0m18.265s
real 0m18.440s user 0m0.032s sys 0m18.337s
ramfs: real 0m16.860s user 0m0.028s sys 0m16.765s
real 0m17.382s user 0m0.040s sys 0m17.273s
real 0m17.133s user 0m0.044s sys 0m17.021s
Yes, I have done perf reports, but they need more explanation than they
deserve: in summary, clear_page vanishes, its cache loading shifts into
copy_user_generic_unrolled; shmem_getpage_gfp goes down, and
surprisingly mark_page_accessed goes way up - I think because they are
respectively where the cache gets to be reloaded after being purged by
clear or copy.
Suggested-by: Nick Piggin <npiggin@gmail.com>
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Let tmpfs into the NOSEC optimization (avoiding file_remove_suid()
overhead on most common writes): set MS_NOSEC on its superblocks.
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the
backing RAM has to be below 4GB. Not a problem while the boards
supported only 4GB: but now Intel's D2700MUD boards support 8GB, and
their GMA3600 is managed by the GMA500 driver.
shmem/tmpfs has never pretended to support hardware restrictions on the
backing memory, but it might have appeared to do so before v3.1, and
even now it works fine until a page is swapped out then back in. When
read_cache_page_gfp() supplied a freshly allocated page for copy, that
compensated for whatever choice might have been made by earlier swapin
readahead; but swapoff was likely to destroy the illusion.
We'd like to continue to support GMA500, so now add a new
shmem_should_replace_page() check on the zone when about to move a page
from swapcache to filecache (in swapin and swapoff cases), with
shmem_replace_page() to allocate and substitute a suitable page (given
gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32).
This does involve a minor extension to mem_cgroup_replace_page_cache()
(the page may or may not have already been charged); and I've removed a
comment and call to mem_cgroup_uncharge_cache_page(), which in fact is
always a no-op while PageSwapCache.
Also removed optimization of an unlikely path in shmem_getpage_gfp(),
now that we need to check PageSwapCache more carefully (a racing caller
might already have made the copy). And at one point shmem_unuse_inode()
needs to use the hitherto private page_swapcount(), to guard against
racing with inode eviction.
It would make sense to extend shmem_should_replace_page(), to cover
cpuset and NUMA mempolicy restrictions too, but set that aside for now:
needs a cleanup of shmem mempolicy handling, and more testing, and ought
to handle swap faults in do_swap_page() as well as shmem.
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@infradead.org>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Stephane Marchesin <marcheu@chromium.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Dave Airlie <airlied@gmail.com>
Cc: Daniel Vetter <daniel@ffwll.ch>
Cc: Rob Clark <rob.clark@linaro.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When MIGRATE_UNMOVABLE pages are freed from MIGRATE_UNMOVABLE type
pageblock (and some MIGRATE_MOVABLE pages are left in it) waiting until an
allocation takes ownership of the block may take too long. The type of
the pageblock remains unchanged so the pageblock cannot be used as a
migration target during compaction.
Fix it by:
* Adding enum compact_mode (COMPACT_ASYNC_[MOVABLE,UNMOVABLE], and
COMPACT_SYNC) and then converting sync field in struct compact_control
to use it.
* Adding nr_pageblocks_skipped field to struct compact_control and
tracking how many destination pageblocks were of MIGRATE_UNMOVABLE type.
If COMPACT_ASYNC_MOVABLE mode compaction ran fully in
try_to_compact_pages() (COMPACT_COMPLETE) it implies that there is not a
suitable page for allocation. In this case then check how if there were
enough MIGRATE_UNMOVABLE pageblocks to try a second pass in
COMPACT_ASYNC_UNMOVABLE mode.
* Scanning the MIGRATE_UNMOVABLE pageblocks (during COMPACT_SYNC and
COMPACT_ASYNC_UNMOVABLE compaction modes) and building a count based on
finding PageBuddy pages, page_count(page) == 0 or PageLRU pages. If all
pages within the MIGRATE_UNMOVABLE pageblock are in one of those three
sets change the whole pageblock type to MIGRATE_MOVABLE.
My particular test case (on a ARM EXYNOS4 device with 512 MiB, which means
131072 standard 4KiB pages in 'Normal' zone) is to:
- allocate 120000 pages for kernel's usage
- free every second page (60000 pages) of memory just allocated
- allocate and use 60000 pages from user space
- free remaining 60000 pages of kernel memory
(now we have fragmented memory occupied mostly by user space pages)
- try to allocate 100 order-9 (2048 KiB) pages for kernel's usage
The results:
- with compaction disabled I get 11 successful allocations
- with compaction enabled - 14 successful allocations
- with this patch I'm able to get all 100 successful allocations
NOTE: If we can make kswapd aware of order-0 request during compaction, we
can enhance kswapd with changing mode to COMPACT_ASYNC_FULL
(COMPACT_ASYNC_MOVABLE + COMPACT_ASYNC_UNMOVABLE). Please see the
following thread:
http://marc.info/?l=linux-mm&m=133552069417068&w=2
[minchan@kernel.org: minor cleanups]
Cc: Mel Gorman <mgorman@suse.de>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
alloc_bootmem_section() derives allocation area constraints from the
specified sparsemem section. This is a bit specific for a generic memory
allocator like bootmem, though, so move it over to sparsemem.
As __alloc_bootmem_node_nopanic() already retries failed allocations with
relaxed area constraints, the fallback code in sparsemem.c can be removed
and the code becomes a bit more compact overall.
[akpm@linux-foundation.org: fix build]
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Pass down the node descriptor instead of the more specific bootmem node
descriptor down the call stack, like nobootmem does, when there is no good
reason for the two to be different.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
While the panicking node-specific allocation function tries to satisfy
node+goal, goal, node, anywhere, the non-panicking function still does
node+goal, goal, anywhere.
Make it simpler: define the panicking version in terms of the non-panicking
one, like the node-agnostic interface, so they always behave the same way
apart from how to deal with allocation failure.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
__alloc_bootmem_node and __alloc_bootmem_low_node documentation claims
the functions panic on allocation failure. Do it.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Yinghai Lu <yinghai@kernel.org>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
While the panicking node-specific allocation function tries to satisfy
node+goal, goal, node, anywhere, the non-panicking function still does
node+goal, goal, anywhere.
Make it simpler: define the panicking version in terms of the
non-panicking one, like the node-agnostic interface, so they always behave
the same way apart from how to deal with allocation failure.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Match the nobootmem version of __alloc_bootmem_node. Try to satisfy both
the node and the goal, then just the goal, then just the node, then
allocate anywhere before panicking.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Matching the desired goal to the right node is one thing, dropping the
goal when it can not be satisfied is another. Split this into separate
functions so that subsequent patches can use the node-finding but drop and
handle the goal fallback on their own terms.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Callsites need to provide a bootmem_data_t *, make the naming more
descriptive.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When bootmem releases an unaligned BITS_PER_LONG pages chunk of memory
to the page allocator, it checks the bitmap if there are still
unreserved pages in the chunk (set bits), but also if the offset in the
chunk indicates BITS_PER_LONG loop iterations already.
But since the consulted bitmap is only a one-word-excerpt of the full
per-node bitmap, there can not be more than BITS_PER_LONG bits set in
it. The additional offset check is unnecessary.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When bootmem releases an unaligned chunk of memory at the beginning of a
node to the page allocator, it iterates from that unaligned PFN but
checks an aligned word of the page bitmap. The checked bits do not
correspond to the PFNs and, as a result, reserved pages can be freed.
Properly shift the bitmap word so that the lowest bit corresponds to the
starting PFN before entering the freeing loop.
This bug has been around since commit 41546c1741 ("bootmem: clean up
free_all_bootmem_core") (2.6.27) without known reports.
Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Tejun Heo <tj@kernel.org>
Acked-by: David S. Miller <davem@davemloft.net>
Cc: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This has always been broken: one version takes an unsigned int and the
other version takes no arguments. This bug was hidden because one
version of set_pageblock_order() was a macro which doesn't evaluate its
argument.
Simplify it all and remove pageblock_default_order() altogether.
Reported-by: rajman mekaco <rajman.mekaco@gmail.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Minchan Kim <minchan.kim@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Print physical address info in a style consistent with the %pR style used
elsewhere in the kernel. For example:
-Zone PFN ranges:
+Zone ranges:
- DMA32 0x00000010 -> 0x00100000
+ DMA32 [mem 0x00010000-0xffffffff]
- Normal 0x00100000 -> 0x01080000
+ Normal [mem 0x100000000-0x107fffffff]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add a Kconfig option to allow people who don't want cross memory attach to
not have it included in their build.
Signed-off-by: Chris Yeoh <yeohc@au1.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm->page_table_lock is hotly contested for page fault tests and isn't
necessary to do mem_cgroup_uncharge_page() in do_huge_pmd_wp_page().
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Andrew pointed out that the is_mlocked_vma() is misnamed. A function
with name like that would expect bool return and no side-effects.
Since it is called on the fault path for new page, rename it in this
patch.
Signed-off-by: Ying Han <yinghan@google.com>
Reviewed-by: Rik van Riel <riel@redhat.com>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujtisu.com>
Reviewed-by: Minchan Kim <minchan@kernel.org>
[akpm@linux-foundation.org: s/mlock_vma_newpage/mlock_vma_newpage/, per Minchan]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The rmap walker checking page table references has historically ignored
references from VMAs that were not part of the memcg that was being
reclaimed during memcg hard limit reclaim.
When transitioning global reclaim to memcg hierarchy reclaim, I missed
that bit and now references from outside a memcg are ignored even during
global reclaim.
Reverting back to traditional behaviour - count all references during
global reclaim and only mind references of the memcg being reclaimed
during limit reclaim would be one option.
However, the more generic idea is to ignore references exactly then when
they are outside the hierarchy that is currently under reclaim; because
only then will their reclamation be of any use to help the pressure
situation. It makes no sense to ignore references from a sibling memcg
and then evict a page that will be immediately refaulted by that sibling
which contributes to the same usage of the common ancestor under
reclaim.
The solution: make the rmap walker ignore references from VMAs that are
not part of the hierarchy that is being reclaimed.
Flat limit reclaim will stay the same, hierarchical limit reclaim will
mind the references only to pages that the hierarchy owns. Global
reclaim, since it reclaims from all memcgs, will be fixed to regard all
references.
[akpm@linux-foundation.org: name the args in the declaration]
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reported-by: Konstantin Khlebnikov <khlebnikov@openvz.org>
Acked-by: Konstantin Khlebnikov<khlebnikov@openvz.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: Li Zefan <lizf@cn.fujitsu.com>
Cc: Li Zefan <lizf@cn.fujitsu.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Library functions should not grab locks when the callsites can do it,
even if the lock nests like the rcu read-side lock does.
Push the rcu_read_lock() from css_is_ancestor() to its single user,
mem_cgroup_same_or_subtree() in preparation for another user that may
already hold the rcu read-side lock.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Konstantin Khlebnikov <khlebnikov@openvz.org>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Acked-by: Li Zefan <lizf@cn.fujitsu.com>
Cc: Li Zefan <lizf@cn.fujitsu.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
s/from_nodes/from and s/to_nodes/to/. The "_nodes" is redundant - it
duplicates the argument's type.
Done in a fit of irritation over 80-col issues :(
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: KOSAKI Motohiro <mkosaki@redhat.com>
Cc: Larry Woodman <lwoodman@redhat.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
While running an application that moves tasks from one cpuset to another
I noticed that it takes much longer and moves many more pages than
expected.
The reason for this is do_migrate_pages() does its best to preserve the
relative node differential from the first node of the cpuset because the
application may have been written with that in mind. If memory was
interleaved on the nodes of the source cpuset by an application
do_migrate_pages() will try its best to maintain that interleaving on
the nodes of the destination cpuset. This means copying the memory from
all source nodes to the destination nodes even if the source and
destination nodes overlap.
This is a problem for userspace NUMA placement tools. The amount of
time spent doing extra memory moves cancels out some of the NUMA
performance improvements. Furthermore, if the number of source and
destination nodes are to maintain the previous interleaving layout
anyway.
This patch changes do_migrate_pages() to only preserve the relative
layout inside the program if the number of NUMA nodes in the source and
destination mask are the same. If the number is different, we do a much
more efficient migration by not touching memory that is in an allowed
node.
This preserves the old behaviour for programs that want it, while
allowing a userspace NUMA placement tool to use the new, faster
migration. This improves performance in our tests by up to a factor of
7.
Without this change migrating tasks from a cpuset containing nodes 0-7
to a cpuset containing nodes 3-4, we migrate from ALL the nodes even if
they are in the both the source and destination nodesets:
Migrating 7 to 4
Migrating 6 to 3
Migrating 5 to 4
Migrating 4 to 3
Migrating 1 to 4
Migrating 3 to 4
Migrating 0 to 3
Migrating 2 to 3
With this change we only migrate from nodes that are not in the
destination nodesets:
Migrating 7 to 4
Migrating 6 to 3
Migrating 5 to 4
Migrating 2 to 3
Migrating 1 to 4
Migrating 0 to 3
Yet if we move from a cpuset containing nodes 2,3,4 to a cpuset
containing 3,4,5 we still do move everything so that we preserve the
desired NUMA offsets:
Migrating 4 to 5
Migrating 3 to 4
Migrating 2 to 3
As far as performance is concerned this simple patch improves the time
it takes to move 14, 20 and 26 large tasks from a cpuset containing
nodes 0-7 to a cpuset containing nodes 1 & 3 by up to a factor of 7.
Here are the timings with and without the patch:
BEFORE PATCH -- Move times: 59, 140, 651 seconds
============
Moving 14 tasks from nodes (0-7) to nodes (1,3)
numad(8780) do_migrate_pages (mm=0xffff88081d414400
from_nodes=0xffff880818c81d28 to_nodes=0xffff880818c81ce8 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d414400 source=0x7 dest=0x3 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d414400 source=0x6 dest=0x1 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d414400 source=0x5 dest=0x3 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d414400 source=0x4 dest=0x1 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d414400 source=0x2 dest=0x1 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d414400 source=0x1 dest=0x3 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d414400 source=0x0 dest=0x1 flags=0x4)
(Above moves repeated for each of the 14 tasks...)
PID 8890 moved to node(s) 1,3 in 59.2 seconds
Moving 20 tasks from nodes (0-7) to nodes (1,4-5)
numad(8780) do_migrate_pages (mm=0xffff88081d88c700
from_nodes=0xffff880818c81d28 to_nodes=0xffff880818c81ce8 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d88c700 source=0x7 dest=0x4 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d88c700 source=0x6 dest=0x1 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d88c700 source=0x3 dest=0x1 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d88c700 source=0x2 dest=0x5 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d88c700 source=0x1 dest=0x4 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d88c700 source=0x0 dest=0x1 flags=0x4)
(Above moves repeated for each of the 20 tasks...)
PID 8962 moved to node(s) 1,4-5 in 139.88 seconds
Moving 26 tasks from nodes (0-7) to nodes (1-3,5)
numad(8780) do_migrate_pages (mm=0xffff88081d5bc740
from_nodes=0xffff880818c81d28 to_nodes=0xffff880818c81ce8 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d5bc740 source=0x7 dest=0x5 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d5bc740 source=0x6 dest=0x3 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d5bc740 source=0x5 dest=0x2 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d5bc740 source=0x3 dest=0x5 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d5bc740 source=0x2 dest=0x3 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d5bc740 source=0x1 dest=0x2 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d5bc740 source=0x0 dest=0x1 flags=0x4)
numad(8780) migrate_to_node (mm=0xffff88081d5bc740 source=0x4 dest=0x1 flags=0x4)
(Above moves repeated for each of the 26 tasks...)
PID 9058 moved to node(s) 1-3,5 in 651.45 seconds
AFTER PATCH -- Move times: 42, 56, 93 seconds
===========
Moving 14 tasks from nodes (0-7) to nodes (5,7)
numad(33209) do_migrate_pages (mm=0xffff88101d5ff140
from_nodes=0xffff88101e7b5d28 to_nodes=0xffff88101e7b5ce8 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d5ff140 source=0x6 dest=0x5 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d5ff140 source=0x4 dest=0x5 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d5ff140 source=0x3 dest=0x7 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d5ff140 source=0x2 dest=0x5 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d5ff140 source=0x1 dest=0x7 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d5ff140 source=0x0 dest=0x5 flags=0x4)
(Above moves repeated for each of the 14 tasks...)
PID 33221 moved to node(s) 5,7 in 41.67 seconds
Moving 20 tasks from nodes (0-7) to nodes (1,3,5)
numad(33209) do_migrate_pages (mm=0xffff88101d6c37c0
from_nodes=0xffff88101e7b5d28 to_nodes=0xffff88101e7b5ce8 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d6c37c0 source=0x7 dest=0x3 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d6c37c0 source=0x6 dest=0x1 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d6c37c0 source=0x4 dest=0x3 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d6c37c0 source=0x2 dest=0x5 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d6c37c0 source=0x0 dest=0x1 flags=0x4)
(Above moves repeated for each of the 20 tasks...)
PID 33289 moved to node(s) 1,3,5 in 56.3 seconds
Moving 26 tasks from nodes (0-7) to nodes (1,3,5,7)
numad(33209) do_migrate_pages (mm=0xffff88101d924400
from_nodes=0xffff88101e7b5d28 to_nodes=0xffff88101e7b5ce8 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d924400 source=0x6 dest=0x5 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d924400 source=0x4 dest=0x1 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d924400 source=0x2 dest=0x5 flags=0x4)
numad(33209) migrate_to_node (mm=0xffff88101d924400 source=0x0 dest=0x1 flags=0x4)
(Above moves repeated for each of the 26 tasks...)
PID 33372 moved to node(s) 1,3,5,7 in 92.67 seconds
[akpm@linux-foundation.org: clean up comment layout]
Signed-off-by: Larry Woodman <lwoodman@redhat.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Reviewed-by: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
On COW, a new hugepage is allocated and charged to the memcg. If the
system is oom or the charge to the memcg fails, however, the fault
handler will return VM_FAULT_OOM which results in an oom kill.
Instead, it's possible to fallback to splitting the hugepage so that the
COW results only in an order-0 page being allocated and charged to the
memcg which has a higher liklihood to succeed. This is expensive
because the hugepage must be split in the page fault handler, but it is
much better than unnecessarily oom killing a process.
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Johannes Weiner <jweiner@redhat.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Remove debug fs files and directory on failure. Since no one is using
"extfrag_debug_root" dentry outside of extfrag_debug_init(), make it
local to the function.
Signed-off-by: Sasikantha babu <sasikanth.v19@gmail.com>
Acked-by: David Rientjes <rientjes@google.com>
Acked-by: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The "if (mm)" check is not required in find_vma, as the kernel code
calls find_vma only when it is absolutely sure that the mm_struct arg to
it is non-NULL.
Remove the if(mm) check and adding the a WARN_ONCE(!mm) for now. This
will serve the purpose of mandating that the execution
context(user-mode/kernel-mode) be known before find_vma is called. Also
fixed 2 checkpatch.pl errors in the declaration of the rb_node and
vma_tmp local variables.
I was browsing through the internet and read a discussion at
https://lkml.org/lkml/2012/3/27/342 which discusses removal of the
validation check within find_vma. Since no-one responded, I decided to
send this patch with Andrew's suggestions.
[akpm@linux-foundation.org: add remove-me comment]
Signed-off-by: Rajman Mekaco <rajman.mekaco@gmail.com>
Cc: Kautuk Consul <consul.kautuk@gmail.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The advantage of kcalloc is, that will prevent integer overflows which
could result from the multiplication of number of elements and size and
it is also a bit nicer to read.
The semantic patch that makes this change is available in
https://lkml.org/lkml/2011/11/25/107
Signed-off-by: Thomas Meyer <thomas@m3y3r.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
There is little motiviation for reclaim_mode_t once RECLAIM_MODE_[A]SYNC
and lumpy reclaim have been removed. This patch gets rid of
reclaim_mode_t as well and improves the documentation about what
reclaim/compaction is and when it is triggered.
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Konstantin Khlebnikov <khlebnikov@openvz.org>
Cc: Hugh Dickins <hughd@google.com>
Cc: Ying Han <yinghan@google.com>
Cc: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch stops reclaim/compaction entering sync reclaim as this was
only intended for lumpy reclaim and an oversight. Page migration has
its own logic for stalling on writeback pages if necessary and memory
compaction is already using it.
Waiting on page writeback is bad for a number of reasons but the primary
one is that waiting on writeback to a slow device like USB can take a
considerable length of time. Page reclaim instead uses
wait_iff_congested() to throttle if too many dirty pages are being
scanned.
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Konstantin Khlebnikov <khlebnikov@openvz.org>
Cc: Hugh Dickins <hughd@google.com>
Cc: Ying Han <yinghan@google.com>
Cc: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This series removes lumpy reclaim and some stalling logic that was
unintentionally being used by memory compaction. The end result is that
stalling on dirty pages during page reclaim now depends on
wait_iff_congested().
Four kernels were compared
3.3.0 vanilla
3.4.0-rc2 vanilla
3.4.0-rc2 lumpyremove-v2 is patch one from this series
3.4.0-rc2 nosync-v2r3 is the full series
Removing lumpy reclaim saves almost 900 bytes of text whereas the full
series removes 1200 bytes.
text data bss dec hex filename
6740375 1927944 2260992 10929311 a6c49f vmlinux-3.4.0-rc2-vanilla
6739479 1927944 2260992 10928415 a6c11f vmlinux-3.4.0-rc2-lumpyremove-v2
6739159 1927944 2260992 10928095 a6bfdf vmlinux-3.4.0-rc2-nosync-v2
There are behaviour changes in the series and so tests were run with
monitoring of ftrace events. This disrupts results so the performance
results are distorted but the new behaviour should be clearer.
fs-mark running in a threaded configuration showed little of interest as
it did not push reclaim aggressively
FS-Mark Multi Threaded
3.3.0-vanilla rc2-vanilla lumpyremove-v2r3 nosync-v2r3
Files/s min 3.20 ( 0.00%) 3.20 ( 0.00%) 3.20 ( 0.00%) 3.20 ( 0.00%)
Files/s mean 3.20 ( 0.00%) 3.20 ( 0.00%) 3.20 ( 0.00%) 3.20 ( 0.00%)
Files/s stddev 0.00 ( 0.00%) 0.00 ( 0.00%) 0.00 ( 0.00%) 0.00 ( 0.00%)
Files/s max 3.20 ( 0.00%) 3.20 ( 0.00%) 3.20 ( 0.00%) 3.20 ( 0.00%)
Overhead min 508667.00 ( 0.00%) 521350.00 (-2.49%) 544292.00 (-7.00%) 547168.00 (-7.57%)
Overhead mean 551185.00 ( 0.00%) 652690.73 (-18.42%) 991208.40 (-79.83%) 570130.53 (-3.44%)
Overhead stddev 18200.69 ( 0.00%) 331958.29 (-1723.88%) 1579579.43 (-8578.68%) 9576.81 (47.38%)
Overhead max 576775.00 ( 0.00%) 1846634.00 (-220.17%) 6901055.00 (-1096.49%) 585675.00 (-1.54%)
MMTests Statistics: duration
Sys Time Running Test (seconds) 309.90 300.95 307.33 298.95
User+Sys Time Running Test (seconds) 319.32 309.67 315.69 307.51
Total Elapsed Time (seconds) 1187.85 1193.09 1191.98 1193.73
MMTests Statistics: vmstat
Page Ins 80532 82212 81420 79480
Page Outs 111434984 111456240 111437376 111582628
Swap Ins 0 0 0 0
Swap Outs 0 0 0 0
Direct pages scanned 44881 27889 27453 34843
Kswapd pages scanned 25841428 25860774 25861233 25843212
Kswapd pages reclaimed 25841393 25860741 25861199 25843179
Direct pages reclaimed 44881 27889 27453 34843
Kswapd efficiency 99% 99% 99% 99%
Kswapd velocity 21754.791 21675.460 21696.029 21649.127
Direct efficiency 100% 100% 100% 100%
Direct velocity 37.783 23.375 23.031 29.188
Percentage direct scans 0% 0% 0% 0%
ftrace showed that there was no stalling on writeback or pages submitted
for IO from reclaim context.
postmark was similar and while it was more interesting, it also did not
push reclaim heavily.
POSTMARK
3.3.0-vanilla rc2-vanilla lumpyremove-v2r3 nosync-v2r3
Transactions per second: 16.00 ( 0.00%) 20.00 (25.00%) 18.00 (12.50%) 17.00 ( 6.25%)
Data megabytes read per second: 18.80 ( 0.00%) 24.27 (29.10%) 22.26 (18.40%) 20.54 ( 9.26%)
Data megabytes written per second: 35.83 ( 0.00%) 46.25 (29.08%) 42.42 (18.39%) 39.14 ( 9.24%)
Files created alone per second: 28.00 ( 0.00%) 38.00 (35.71%) 34.00 (21.43%) 30.00 ( 7.14%)
Files create/transact per second: 8.00 ( 0.00%) 10.00 (25.00%) 9.00 (12.50%) 8.00 ( 0.00%)
Files deleted alone per second: 556.00 ( 0.00%) 1224.00 (120.14%) 3062.00 (450.72%) 6124.00 (1001.44%)
Files delete/transact per second: 8.00 ( 0.00%) 10.00 (25.00%) 9.00 (12.50%) 8.00 ( 0.00%)
MMTests Statistics: duration
Sys Time Running Test (seconds) 113.34 107.99 109.73 108.72
User+Sys Time Running Test (seconds) 145.51 139.81 143.32 143.55
Total Elapsed Time (seconds) 1159.16 899.23 980.17 1062.27
MMTests Statistics: vmstat
Page Ins 13710192 13729032 13727944 13760136
Page Outs 43071140 42987228 42733684 42931624
Swap Ins 0 0 0 0
Swap Outs 0 0 0 0
Direct pages scanned 0 0 0 0
Kswapd pages scanned 99416139937443 9939085 9929154
Kswapd pages reclaimed 9940926 9936751 9938397 9928465
Direct pages reclaimed 0 0 0 0
Kswapd efficiency 99% 99% 99% 99%
Kswapd velocity 8576.567 11051.058 10140.164 9347.109
Direct efficiency 100% 100% 100% 100%
Direct velocity 0.000 0.000 0.000 0.000
It looks like here that the full series regresses performance but as
ftrace showed no usage of wait_iff_congested() or sync reclaim I am
assuming it's a disruption due to monitoring. Other data such as memory
usage, page IO, swap IO all looked similar.
Running a benchmark with a plain DD showed nothing very interesting.
The full series stalled in wait_iff_congested() slightly less but stall
times on vanilla kernels were marginal.
Running a benchmark that hammered on file-backed mappings showed stalls
due to congestion but not in sync writebacks
MICRO
3.3.0-vanilla rc2-vanilla lumpyremove-v2r3 nosync-v2r3
MMTests Statistics: duration
Sys Time Running Test (seconds) 308.13 294.50 298.75 299.53
User+Sys Time Running Test (seconds) 330.45 316.28 318.93 320.79
Total Elapsed Time (seconds) 1814.90 1833.88 1821.14 1832.91
MMTests Statistics: vmstat
Page Ins 108712 120708 97224 110344
Page Outs 155514576 156017404 155813676 156193256
Swap Ins 0 0 0 0
Swap Outs 0 0 0 0
Direct pages scanned 2599253 1550480 2512822 2414760
Kswapd pages scanned 69742364 71150694 68839041 69692533
Kswapd pages reclaimed 34824488 34773341 34796602 34799396
Direct pages reclaimed 53693 94750 61792 75205
Kswapd efficiency 49% 48% 50% 49%
Kswapd velocity 38427.662 38797.901 37799.972 38022.889
Direct efficiency 2% 6% 2% 3%
Direct velocity 1432.174 845.464 1379.807 1317.446
Percentage direct scans 3% 2% 3% 3%
Page writes by reclaim 0 0 0 0
Page writes file 0 0 0 0
Page writes anon 0 0 0 0
Page reclaim immediate 0 0 0 1218
Page rescued immediate 0 0 0 0
Slabs scanned 15360 16384 13312 16384
Direct inode steals 0 0 0 0
Kswapd inode steals 4340 4327 1630 4323
FTrace Reclaim Statistics: congestion_wait
Direct number congest waited 0 0 0 0
Direct time congest waited 0ms 0ms 0ms 0ms
Direct full congest waited 0 0 0 0
Direct number conditional waited 900 870 754 789
Direct time conditional waited 0ms 0ms 0ms 20ms
Direct full conditional waited 0 0 0 0
KSwapd number congest waited 2106 2308 2116 1915
KSwapd time congest waited 139924ms 157832ms 125652ms 132516ms
KSwapd full congest waited 1346 1530 1202 1278
KSwapd number conditional waited 12922 16320 10943 14670
KSwapd time conditional waited 0ms 0ms 0ms 0ms
KSwapd full conditional waited 0 0 0 0
Reclaim statistics are not radically changed. The stall times in kswapd
are massive but it is clear that it is due to calls to congestion_wait()
and that is almost certainly the call in balance_pgdat(). Otherwise
stalls due to dirty pages are non-existant.
I ran a benchmark that stressed high-order allocation. This is very
artifical load but was used in the past to evaluate lumpy reclaim and
compaction. Generally I look at allocation success rates and latency
figures.
STRESS-HIGHALLOC
3.3.0-vanilla rc2-vanilla lumpyremove-v2r3 nosync-v2r3
Pass 1 81.00 ( 0.00%) 28.00 (-53.00%) 24.00 (-57.00%) 28.00 (-53.00%)
Pass 2 82.00 ( 0.00%) 39.00 (-43.00%) 38.00 (-44.00%) 43.00 (-39.00%)
while Rested 88.00 ( 0.00%) 87.00 (-1.00%) 88.00 ( 0.00%) 88.00 ( 0.00%)
MMTests Statistics: duration
Sys Time Running Test (seconds) 740.93 681.42 685.14 684.87
User+Sys Time Running Test (seconds) 2922.65 3269.52 3281.35 3279.44
Total Elapsed Time (seconds) 1161.73 1152.49 1159.55 1161.44
MMTests Statistics: vmstat
Page Ins 4486020 2807256 2855944 2876244
Page Outs 7261600 7973688 7975320 7986120
Swap Ins 31694 0 0 0
Swap Outs 98179 0 0 0
Direct pages scanned 53494 57731 34406 113015
Kswapd pages scanned 6271173 1287481 1278174 1219095
Kswapd pages reclaimed 2029240 1281025 1260708 1201583
Direct pages reclaimed 1468 14564 16649 92456
Kswapd efficiency 32% 99% 98% 98%
Kswapd velocity 5398.133 1117.130 1102.302 1049.641
Direct efficiency 2% 25% 48% 81%
Direct velocity 46.047 50.092 29.672 97.306
Percentage direct scans 0% 4% 2% 8%
Page writes by reclaim 1616049 0 0 0
Page writes file 1517870 0 0 0
Page writes anon 98179 0 0 0
Page reclaim immediate 103778 27339 9796 17831
Page rescued immediate 0 0 0 0
Slabs scanned 1096704 986112 980992 998400
Direct inode steals 223 215040 216736 247881
Kswapd inode steals 175331 61548 68444 63066
Kswapd skipped wait 21991 0 1 0
THP fault alloc 1 135 125 134
THP collapse alloc 393 311 228 236
THP splits 25 13 7 8
THP fault fallback 0 0 0 0
THP collapse fail 3 5 7 7
Compaction stalls 865 1270 1422 1518
Compaction success 370 401 353 383
Compaction failures 495 869 1069 1135
Compaction pages moved 870155 3828868 4036106 4423626
Compaction move failure 26429 23865 29742 27514
Success rates are completely hosed for 3.4-rc2 which is almost certainly
due to commit fe2c2a1066 ("vmscan: reclaim at order 0 when compaction
is enabled"). I expected this would happen for kswapd and impair
allocation success rates (https://lkml.org/lkml/2012/1/25/166) but I did
not anticipate this much a difference: 80% less scanning, 37% less
reclaim by kswapd
In comparison, reclaim/compaction is not aggressive and gives up easily
which is the intended behaviour. hugetlbfs uses __GFP_REPEAT and would
be much more aggressive about reclaim/compaction than THP allocations
are. The stress test above is allocating like neither THP or hugetlbfs
but is much closer to THP.
Mainline is now impaired in terms of high order allocation under heavy
load although I do not know to what degree as I did not test with
__GFP_REPEAT. Keep this in mind for bugs related to hugepage pool
resizing, THP allocation and high order atomic allocation failures from
network devices.
In terms of congestion throttling, I see the following for this test
FTrace Reclaim Statistics: congestion_wait
Direct number congest waited 3 0 0 0
Direct time congest waited 0ms 0ms 0ms 0ms
Direct full congest waited 0 0 0 0
Direct number conditional waited 957 512 1081 1075
Direct time conditional waited 0ms 0ms 0ms 0ms
Direct full conditional waited 0 0 0 0
KSwapd number congest waited 36 4 3 5
KSwapd time congest waited 3148ms 400ms 300ms 500ms
KSwapd full congest waited 30 4 3 5
KSwapd number conditional waited 88514 197 332 542
KSwapd time conditional waited 4980ms 0ms 0ms 0ms
KSwapd full conditional waited 49 0 0 0
The "conditional waited" times are the most interesting as this is
directly impacted by the number of dirty pages encountered during scan.
As lumpy reclaim is no longer scanning contiguous ranges, it is finding
fewer dirty pages. This brings wait times from about 5 seconds to 0.
kswapd itself is still calling congestion_wait() so it'll still stall but
it's a lot less.
In terms of the type of IO we were doing, I see this
FTrace Reclaim Statistics: mm_vmscan_writepage
Direct writes anon sync 0 0 0 0
Direct writes anon async 0 0 0 0
Direct writes file sync 0 0 0 0
Direct writes file async 0 0 0 0
Direct writes mixed sync 0 0 0 0
Direct writes mixed async 0 0 0 0
KSwapd writes anon sync 0 0 0 0
KSwapd writes anon async 91682 0 0 0
KSwapd writes file sync 0 0 0 0
KSwapd writes file async 822629 0 0 0
KSwapd writes mixed sync 0 0 0 0
KSwapd writes mixed async 0 0 0 0
In 3.2, kswapd was doing a bunch of async writes of pages but
reclaim/compaction was never reaching a point where it was doing sync
IO. This does not guarantee that reclaim/compaction was not calling
wait_on_page_writeback() but I would consider it unlikely. It indicates
that merging patches 2 and 3 to stop reclaim/compaction calling
wait_on_page_writeback() should be safe.
This patch:
Lumpy reclaim had a purpose but in the mind of some, it was to kick the
system so hard it trashed. For others the purpose was to complicate
vmscan.c. Over time it was giving softer shoes and a nicer attitude but
memory compaction needs to step up and replace it so this patch sends
lumpy reclaim to the farm.
The tracepoint format changes for isolating LRU pages with this patch
applied. Furthermore reclaim/compaction can no longer queue dirty pages
in pageout() if the underlying BDI is congested. Lumpy reclaim used
this logic and reclaim/compaction was using it in error.
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Konstantin Khlebnikov <khlebnikov@openvz.org>
Cc: Hugh Dickins <hughd@google.com>
Cc: Ying Han <yinghan@google.com>
Cc: Andy Whitcroft <apw@shadowen.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The swap token code no longer fits in with the current VM model. It
does not play well with cgroups or the better NUMA placement code in
development, since we have only one swap token globally.
It also has the potential to mess with scalability of the system, by
increasing the number of non-reclaimable pages on the active and
inactive anon LRU lists.
Last but not least, the swap token code has been broken for a year
without complaints, as reported by Konstantin Khlebnikov. This suggests
we no longer have much use for it.
The days of sub-1G memory systems with heavy use of swap are over. If
we ever need thrashing reducing code in the future, we will have to
implement something that does scale.
Signed-off-by: Rik van Riel <riel@redhat.com>
Cc: Konstantin Khlebnikov <khlebnikov@openvz.org>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Cc: Hugh Dickins <hughd@google.com>
Acked-by: Bob Picco <bpicco@meloft.net>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The transparent hugepages feature is careful to not invoke the oom
killer when a hugepage cannot be allocated.
pte_alloc_one() failing in __do_huge_pmd_anonymous_page(), however,
currently results in VM_FAULT_OOM which invokes the pagefault oom killer
to kill a memory-hogging task.
This is unnecessary since it's possible to drop the reference to the
hugepage and fallback to allocating a small page.
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>