Andrea Arcangeli [Thu, 13 Jan 2011 23:46:52 +0000 (15:46 -0800)]
thp: transparent hugepage core
Lately I've been working to make KVM use hugepages transparently without
the usual restrictions of hugetlbfs. Some of the restrictions I'd like to
see removed:
1) hugepages have to be swappable or the guest physical memory remains
locked in RAM and can't be paged out to swap
2) if a hugepage allocation fails, regular pages should be allocated
instead and mixed in the same vma without any failure and without
userland noticing
3) if some task quits and more hugepages become available in the
buddy, guest physical memory backed by regular pages should be
relocated on hugepages automatically in regions under
madvise(MADV_HUGEPAGE) (ideally event driven by waking up the
kernel deamon if the order=HPAGE_PMD_SHIFT-PAGE_SHIFT list becomes
not null)
4) avoidance of reservation and maximization of use of hugepages whenever
possible. Reservation (needed to avoid runtime fatal faliures) may be ok for
1 machine with 1 database with 1 database cache with 1 database cache size
known at boot time. It's definitely not feasible with a virtualization
hypervisor usage like RHEV-H that runs an unknown number of virtual machines
with an unknown size of each virtual machine with an unknown amount of
pagecache that could be potentially useful in the host for guest not using
O_DIRECT (aka cache=off).
hugepages in the virtualization hypervisor (and also in the guest!) are
much more important than in a regular host not using virtualization,
becasue with NPT/EPT they decrease the tlb-miss cacheline accesses from 24
to 19 in case only the hypervisor uses transparent hugepages, and they
decrease the tlb-miss cacheline accesses from 19 to 15 in case both the
linux hypervisor and the linux guest both uses this patch (though the
guest will limit the addition speedup to anonymous regions only for
now...). Even more important is that the tlb miss handler is much slower
on a NPT/EPT guest than for a regular shadow paging or no-virtualization
scenario. So maximizing the amount of virtual memory cached by the TLB
pays off significantly more with NPT/EPT than without (even if there would
be no significant speedup in the tlb-miss runtime).
The first (and more tedious) part of this work requires allowing the VM to
handle anonymous hugepages mixed with regular pages transparently on
regular anonymous vmas. This is what this patch tries to achieve in the
least intrusive possible way. We want hugepages and hugetlb to be used in
a way so that all applications can benefit without changes (as usual we
leverage the KVM virtualization design: by improving the Linux VM at
large, KVM gets the performance boost too).
The most important design choice is: always fallback to 4k allocation if
the hugepage allocation fails! This is the _very_ opposite of some large
pagecache patches that failed with -EIO back then if a 64k (or similar)
allocation failed...
Second important decision (to reduce the impact of the feature on the
existing pagetable handling code) is that at any time we can split an
hugepage into 512 regular pages and it has to be done with an operation
that can't fail. This way the reliability of the swapping isn't decreased
(no need to allocate memory when we are short on memory to swap) and it's
trivial to plug a split_huge_page* one-liner where needed without
polluting the VM. Over time we can teach mprotect, mremap and friends to
handle pmd_trans_huge natively without calling split_huge_page*. The fact
it can't fail isn't just for swap: if split_huge_page would return -ENOMEM
(instead of the current void) we'd need to rollback the mprotect from the
middle of it (ideally including undoing the split_vma) which would be a
big change and in the very wrong direction (it'd likely be simpler not to
call split_huge_page at all and to teach mprotect and friends to handle
hugepages instead of rolling them back from the middle). In short the
very value of split_huge_page is that it can't fail.
The collapsing and madvise(MADV_HUGEPAGE) part will remain separated and
incremental and it'll just be an "harmless" addition later if this initial
part is agreed upon. It also should be noted that locking-wise replacing
regular pages with hugepages is going to be very easy if compared to what
I'm doing below in split_huge_page, as it will only happen when
page_count(page) matches page_mapcount(page) if we can take the PG_lock
and mmap_sem in write mode. collapse_huge_page will be a "best effort"
that (unlike split_huge_page) can fail at the minimal sign of trouble and
we can try again later. collapse_huge_page will be similar to how KSM
works and the madvise(MADV_HUGEPAGE) will work similar to
madvise(MADV_MERGEABLE).
The default I like is that transparent hugepages are used at page fault
time. This can be changed with
/sys/kernel/mm/transparent_hugepage/enabled. The control knob can be set
to three values "always", "madvise", "never" which mean respectively that
hugepages are always used, or only inside madvise(MADV_HUGEPAGE) regions,
or never used. /sys/kernel/mm/transparent_hugepage/defrag instead
controls if the hugepage allocation should defrag memory aggressively
"always", only inside "madvise" regions, or "never".
The pmd_trans_splitting/pmd_trans_huge locking is very solid. The
put_page (from get_user_page users that can't use mmu notifier like
O_DIRECT) that runs against a __split_huge_page_refcount instead was a
pain to serialize in a way that would result always in a coherent page
count for both tail and head. I think my locking solution with a
compound_lock taken only after the page_first is valid and is still a
PageHead should be safe but it surely needs review from SMP race point of
view. In short there is no current existing way to serialize the O_DIRECT
final put_page against split_huge_page_refcount so I had to invent a new
one (O_DIRECT loses knowledge on the mapping status by the time gup_fast
returns so...). And I didn't want to impact all gup/gup_fast users for
now, maybe if we change the gup interface substantially we can avoid this
locking, I admit I didn't think too much about it because changing the gup
unpinning interface would be invasive.
If we ignored O_DIRECT we could stick to the existing compound refcounting
code, by simply adding a get_user_pages_fast_flags(foll_flags) where KVM
(and any other mmu notifier user) would call it without FOLL_GET (and if
FOLL_GET isn't set we'd just BUG_ON if nobody registered itself in the
current task mmu notifier list yet). But O_DIRECT is fundamental for
decent performance of virtualized I/O on fast storage so we can't avoid it
to solve the race of put_page against split_huge_page_refcount to achieve
a complete hugepage feature for KVM.
Swap and oom works fine (well just like with regular pages ;). MMU
notifier is handled transparently too, with the exception of the young bit
on the pmd, that didn't have a range check but I think KVM will be fine
because the whole point of hugepages is that EPT/NPT will also use a huge
pmd when they notice gup returns pages with PageCompound set, so they
won't care of a range and there's just the pmd young bit to check in that
case.
NOTE: in some cases if the L2 cache is small, this may slowdown and waste
memory during COWs because 4M of memory are accessed in a single fault
instead of 8k (the payoff is that after COW the program can run faster).
So we might want to switch the copy_huge_page (and clear_huge_page too) to
not temporal stores. I also extensively researched ways to avoid this
cache trashing with a full prefault logic that would cow in 8k/16k/32k/64k
up to 1M (I can send those patches that fully implemented prefault) but I
concluded they're not worth it and they add an huge additional complexity
and they remove all tlb benefits until the full hugepage has been faulted
in, to save a little bit of memory and some cache during app startup, but
they still don't improve substantially the cache-trashing during startup
if the prefault happens in >4k chunks. One reason is that those 4k pte
entries copied are still mapped on a perfectly cache-colored hugepage, so
the trashing is the worst one can generate in those copies (cow of 4k page
copies aren't so well colored so they trashes less, but again this results
in software running faster after the page fault). Those prefault patches
allowed things like a pte where post-cow pages were local 4k regular anon
pages and the not-yet-cowed pte entries were pointing in the middle of
some hugepage mapped read-only. If it doesn't payoff substantially with
todays hardware it will payoff even less in the future with larger l2
caches, and the prefault logic would blot the VM a lot. If one is
emebdded transparent_hugepage can be disabled during boot with sysfs or
with the boot commandline parameter transparent_hugepage=0 (or
transparent_hugepage=2 to restrict hugepages inside madvise regions) that
will ensure not a single hugepage is allocated at boot time. It is simple
enough to just disable transparent hugepage globally and let transparent
hugepages be allocated selectively by applications in the MADV_HUGEPAGE
region (both at page fault time, and if enabled with the
collapse_huge_page too through the kernel daemon).
This patch supports only hugepages mapped in the pmd, archs that have
smaller hugepages will not fit in this patch alone. Also some archs like
power have certain tlb limits that prevents mixing different page size in
the same regions so they will not fit in this framework that requires
"graceful fallback" to basic PAGE_SIZE in case of physical memory
fragmentation. hugetlbfs remains a perfect fit for those because its
software limits happen to match the hardware limits. hugetlbfs also
remains a perfect fit for hugepage sizes like 1GByte that cannot be hoped
to be found not fragmented after a certain system uptime and that would be
very expensive to defragment with relocation, so requiring reservation.
hugetlbfs is the "reservation way", the point of transparent hugepages is
not to have any reservation at all and maximizing the use of cache and
hugepages at all times automatically.
Some performance result:
vmx andrea # LD_PRELOAD=/usr/lib64/libhugetlbfs.so HUGETLB_MORECORE=yes HUGETLB_PATH=/mnt/huge/ ./largep
ages3
memset page fault 1566023
memset tlb miss 453854
memset second tlb miss 453321
random access tlb miss 41635
random access second tlb miss 41658
vmx andrea # LD_PRELOAD=/usr/lib64/libhugetlbfs.so HUGETLB_MORECORE=yes HUGETLB_PATH=/mnt/huge/ ./largepages3
memset page fault 1566471
memset tlb miss 453375
memset second tlb miss 453320
random access tlb miss 41636
random access second tlb miss 41637
vmx andrea # ./largepages3
memset page fault 1566642
memset tlb miss 453417
memset second tlb miss 453313
random access tlb miss 41630
random access second tlb miss 41647
vmx andrea # ./largepages3
memset page fault 1566872
memset tlb miss 453418
memset second tlb miss 453315
random access tlb miss 41618
random access second tlb miss 41659
vmx andrea # echo 0 > /proc/sys/vm/transparent_hugepage
vmx andrea # ./largepages3
memset page fault 2182476
memset tlb miss 460305
memset second tlb miss 460179
random access tlb miss 44483
random access second tlb miss 44186
vmx andrea # ./largepages3
memset page fault 2182791
memset tlb miss 460742
memset second tlb miss 459962
random access tlb miss 43981
random access second tlb miss 43988
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:49 +0000 (15:46 -0800)]
thp: don't alloc harder for gfp nomemalloc even if nowait
Not worth throwing away the precious reserved free memory pool for
allocations that can fail gracefully (either through mempool or because
they're transhuge allocations later falling back to 4k allocations).
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Rik van Riel <riel@redhat.com> Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.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>
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:49 +0000 (15:46 -0800)]
thp: _GFP_NO_KSWAPD
Transparent hugepage allocations must be allowed not to invoke kswapd or
any other kind of indirect reclaim (especially when the defrag sysfs is
control disabled). It's unacceptable to swap out anonymous pages
(potentially anonymous transparent hugepages) in order to create new
transparent hugepages. This is true for the MADV_HUGEPAGE areas too
(swapping out a kvm virtual machine and so having it suffer an unbearable
slowdown, so another one with guest physical memory marked MADV_HUGEPAGE
can run 30% faster if it is running memory intensive workloads, makes no
sense). If a transparent hugepage allocation fails the slowdown is minor
and there is total fallback, so kswapd should never be asked to swapout
memory to allow the high order allocation to succeed.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Rik van Riel <riel@redhat.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>
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:47 +0000 (15:46 -0800)]
thp: split_huge_page paging
Paging logic that splits the page before it is unmapped and added to swap
to ensure backwards compatibility with the legacy swap code. Eventually
swap should natively pageout the hugepages to increase performance and
decrease seeking and fragmentation of swap space. swapoff can just skip
over huge pmd as they cannot be part of swap yet. In add_to_swap be
careful to split the page only if we got a valid swap entry so we don't
split hugepages with a full swap.
In theory we could split pages before isolating them during the lru scan,
but for khugepaged to be safe, I'm relying on either mmap_sem write mode,
or PG_lock taken, so split_huge_page has to run either with mmap_sem
read/write mode or PG_lock taken. Calling it from isolate_lru_page would
make locking more complicated, in addition to that split_huge_page would
deadlock if called by __isolate_lru_page because it has to take the lru
lock to add the tail pages.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Mel Gorman <mel@csn.ul.ie> Acked-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>
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:46 +0000 (15:46 -0800)]
thp: split_huge_page_mm/vma
split_huge_page_pmd compat code. Each one of those would need to be
expanded to hundred of lines of complex code without a fully reliable
split_huge_page_pmd design.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mel@csn.ul.ie> Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:45 +0000 (15:46 -0800)]
thp: add pmd_huge_pte to mm_struct
This increase the size of the mm struct a bit but it is needed to
preallocate one pte for each hugepage so that split_huge_page will not
require a fail path. Guarantee of success is a fundamental property of
split_huge_page to avoid decrasing swapping reliability and to avoid
adding -ENOMEM fail paths that would otherwise force the hugepage-unaware
VM code to learn rolling back in the middle of its pte mangling operations
(if something we need it to learn handling pmd_trans_huge natively rather
being capable of rollback). When split_huge_page runs a pte is needed to
succeed the split, to map the newly splitted regular pages with a regular
pte. This way all existing VM code remains backwards compatible by just
adding a split_huge_page* one liner. The memory waste of those
preallocated ptes is negligible and so it is worth it.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Rik van Riel <riel@redhat.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>
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:43 +0000 (15:46 -0800)]
thp: pte alloc trans splitting
pte alloc routines must wait for split_huge_page if the pmd is not present
and not null (i.e. pmd_trans_splitting). The additional branches are
optimized away at compile time by pmd_trans_splitting if the config option
is off. However we must pass the vma down in order to know the anon_vma
lock to wait for.
[akpm@linux-foundation.org: coding-style fixes] Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Rik van Riel <riel@redhat.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>
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:41 +0000 (15:46 -0800)]
thp: add pmd mangling functions to x86
Add needed pmd mangling functions with symmetry with their pte
counterparts. pmdp_splitting_flush() is the only new addition on the pmd_
methods and it's needed to serialize the VM against split_huge_page. It
simply atomically sets the splitting bit in a similar way
pmdp_clear_flush_young atomically clears the accessed bit.
pmdp_splitting_flush() also has to flush the tlb to make it effective
against gup_fast, but it wouldn't really require to flush the tlb too.
Just the tlb flush is the simplest operation we can invoke to serialize
pmdp_splitting_flush() against gup_fast.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-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>
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:40 +0000 (15:46 -0800)]
thp: special pmd_trans_* functions
These returns 0 at compile time when the config option is disabled, to
allow gcc to eliminate the transparent hugepage function calls at compile
time without additional #ifdefs (only the export of those functions have
to be visible to gcc but they won't be required at link time and
huge_memory.o can be not built at all).
_PAGE_BIT_UNUSED1 is never used for pmd, only on pte.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Rik van Riel <riel@redhat.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>
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:36 +0000 (15:46 -0800)]
thp: add pmd paravirt ops
Paravirt ops pmd_update/pmd_update_defer/pmd_set_at. Not all might be
necessary (vmware needs pmd_update, Xen needs set_pmd_at, nobody needs
pmd_update_defer), but this is to keep full simmetry with pte paravirt
ops, which looks cleaner and simpler from a common code POV.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Rik van Riel <riel@redhat.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>
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:34 +0000 (15:46 -0800)]
thp: clear compound mapping
Clear compound mapping for anonymous compound pages like it already
happens for regular anonymous pages. But crash if mapping is set for any
tail page, also the PageAnon check is meaningless for tail pages. This
check only makes sense for the head page, for tail page it can only hide
bugs and we definitely don't want to hide bugs.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Rik van Riel <riel@redhat.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>
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:34 +0000 (15:46 -0800)]
thp: update futex compound knowledge
Futex code is smarter than most other gup_fast O_DIRECT code and knows
about the compound internals. However now doing a put_page(head_page)
will not release the pin on the tail page taken by gup-fast, leading to
all sort of refcounting bugchecks. Getting a stable head_page is a little
tricky.
page_head = page is there because if this is not a tail page it's also the
page_head. Only in case this is a tail page, compound_head is called,
otherwise it's guaranteed unnecessary. And if it's a tail page
compound_head has to run atomically inside irq disabled section
__get_user_pages_fast before returning. Otherwise ->first_page won't be a
stable pointer.
Disableing irq before __get_user_page_fast and releasing irq after running
compound_head is needed because if __get_user_page_fast returns == 1, it
means the huge pmd is established and cannot go away from under us.
pmdp_splitting_flush_notify in __split_huge_page_splitting will have to
wait for local_irq_enable before the IPI delivery can return. This means
__split_huge_page_refcount can't be running from under us, and in turn
when we run compound_head(page) we're not reading a dangling pointer from
tailpage->first_page. Then after we get to stable head page, we are
always safe to call compound_lock and after taking the compound lock on
head page we can finally re-check if the page returned by gup-fast is
still a tail page. in which case we're set and we didn't need to split
the hugepage in order to take a futex on it.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Mel Gorman <mel@csn.ul.ie> Acked-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>
Andrea Arcangeli [Thu, 13 Jan 2011 23:46:32 +0000 (15:46 -0800)]
thp: alter compound get_page/put_page
Alter compound get_page/put_page to keep references on subpages too, in
order to allow __split_huge_page_refcount to split an hugepage even while
subpages have been pinned by one of the get_user_pages() variants.
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com> Acked-by: Rik van Riel <riel@redhat.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>
I think determine_dirtyable_memory() is a rather costly function since it
need many atomic reads for gathering zone/global page state. But when we
use vm_dirty_bytes && dirty_background_bytes, we don't need that costly
calculation.
This patch eliminates such unnecessary overhead.
NOTE : newly added if condition might add overhead in normal path.
But it should be _really_ small because anyway we need the
access both vm_dirty_bytes and dirty_background_bytes so it is
likely to hit the cache.
[akpm@linux-foundation.org: fix used-uninitialised warning] Signed-off-by: Minchan Kim <minchan.kim@gmail.com> Cc: Wu Fengguang <fengguang.wu@intel.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm: set correct numa_zonelist_order string when configured on the kernel command line
When numa_zonelist_order parameter is set to "node" or "zone" on the
command line it's still showing as "default" in sysctl. That's because
early_param parsing function changes only user_zonelist_order variable.
Fix this by copying user-provided string to numa_zonelist_order if it was
successfully parsed.
Signed-off-by: Volodymyr G Lukiianyk <volodymyrgl@gmail.com> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Mel Gorman [Thu, 13 Jan 2011 23:46:26 +0000 (15:46 -0800)]
mm: kswapd: use the classzone idx that kswapd was using for sleeping_prematurely()
When kswapd is woken up for a high-order allocation, it takes account of
the highest usable zone by the caller (the classzone idx). During
allocation, this index is used to select the lowmem_reserve[] that should
be applied to the watermark calculation in zone_watermark_ok().
When balancing a node, kswapd considers the highest unbalanced zone to be
the classzone index. This will always be at least be the callers
classzone_idx and can be higher. However, sleeping_prematurely() always
considers the lowest zone (e.g. ZONE_DMA) to be the classzone index.
This means that sleeping_prematurely() can consider a zone to be balanced
that is unusable by the allocation request that originally woke kswapd.
This patch changes sleeping_prematurely() to use a classzone_idx matching
the value it used in balance_pgdat().
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Reviewed-by: Minchan Kim <minchan.kim@gmail.com> Reviewed-by: Eric B Munson <emunson@mgebm.net> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Simon Kirby <sim@hostway.ca> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Dave Hansen <dave@linux.vnet.ibm.com> Cc: Johannes Weiner <hannes@cmpxchg.org> 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>
Mel Gorman [Thu, 13 Jan 2011 23:46:24 +0000 (15:46 -0800)]
mm: kswapd: treat zone->all_unreclaimable in sleeping_prematurely similar to balance_pgdat()
After DEF_PRIORITY, balance_pgdat() considers all_unreclaimable zones to
be balanced but sleeping_prematurely does not. This can force kswapd to
stay awake longer than it should. This patch fixes it.
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Reviewed-by: Eric B Munson <emunson@mgebm.net> Cc: Minchan Kim <minchan.kim@gmail.com> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Simon Kirby <sim@hostway.ca> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Dave Hansen <dave@linux.vnet.ibm.com> Cc: Johannes Weiner <hannes@cmpxchg.org> 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>
Mel Gorman [Thu, 13 Jan 2011 23:46:23 +0000 (15:46 -0800)]
mm: kswapd: reset kswapd_max_order and classzone_idx after reading
When kswapd wakes up, it reads its order and classzone from pgdat and
calls balance_pgdat. While its awake, it potentially reclaimes at a high
order and a low classzone index. This might have been a once-off that was
not required by subsequent callers. However, because the pgdat values
were not reset, they remain artifically high while balance_pgdat() is
running and potentially kswapd enters a second unnecessary reclaim cycle.
Reset the pgdat order and classzone index after reading.
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Reviewed-by: Minchan Kim <minchan.kim@gmail.com> Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Reviewed-by: Eric B Munson <emunson@mgebm.net> Cc: Simon Kirby <sim@hostway.ca> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Dave Hansen <dave@linux.vnet.ibm.com> Cc: Johannes Weiner <hannes@cmpxchg.org> 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>
Mel Gorman [Thu, 13 Jan 2011 23:46:22 +0000 (15:46 -0800)]
mm: kswapd: use the order that kswapd was reclaiming at for sleeping_prematurely()
Before kswapd goes to sleep, it uses sleeping_prematurely() to check if
there was a race pushing a zone below its watermark. If the race
happened, it stays awake. However, balance_pgdat() can decide to reclaim
at order-0 if it decides that high-order reclaim is not working as
expected. This information is not passed back to sleeping_prematurely().
The impact is that kswapd remains awake reclaiming pages long after it
should have gone to sleep. This patch passes the adjusted order to
sleeping_prematurely and uses the same logic as balance_pgdat to decide if
it's ok to go to sleep.
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Reviewed-by: Minchan Kim <minchan.kim@gmail.com> Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Reviewed-by: Eric B Munson <emunson@mgebm.net> Cc: Simon Kirby <sim@hostway.ca> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Dave Hansen <dave@linux.vnet.ibm.com> Cc: Johannes Weiner <hannes@cmpxchg.org> 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>
Mel Gorman [Thu, 13 Jan 2011 23:46:21 +0000 (15:46 -0800)]
mm: kswapd: keep kswapd awake for high-order allocations until a percentage of the node is balanced
When reclaiming for high-orders, kswapd is responsible for balancing a
node but it should not reclaim excessively. It avoids excessive reclaim
by considering if any zone in a node is balanced then the node is
balanced. In the cases where there are imbalanced zone sizes (e.g.
ZONE_DMA with both ZONE_DMA32 and ZONE_NORMAL), kswapd can go to sleep
prematurely as just one small zone was balanced.
This alters the sleep logic of kswapd slightly. It counts the number of
pages that make up the balanced zones. If the total number of balanced
pages is more than a quarter of the zone, kswapd will go back to sleep.
This should keep a node balanced without reclaiming an excessive number of
pages.
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Reviewed-by: Minchan Kim <minchan.kim@gmail.com> Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Reviewed-by: Eric B Munson <emunson@mgebm.net> Cc: Simon Kirby <sim@hostway.ca> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Dave Hansen <dave@linux.vnet.ibm.com> Cc: Johannes Weiner <hannes@cmpxchg.org> 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>
Mel Gorman [Thu, 13 Jan 2011 23:46:20 +0000 (15:46 -0800)]
mm: kswapd: stop high-order balancing when any suitable zone is balanced
Simon Kirby reported the following problem
We're seeing cases on a number of servers where cache never fully
grows to use all available memory. Sometimes we see servers with 4 GB
of memory that never seem to have less than 1.5 GB free, even with a
constantly-active VM. In some cases, these servers also swap out while
this happens, even though they are constantly reading the working set
into memory. We have been seeing this happening for a long time; I
don't think it's anything recent, and it still happens on 2.6.36.
After some debugging work by Simon, Dave Hansen and others, the prevaling
theory became that kswapd is reclaiming order-3 pages requested by SLUB
too aggressive about it.
There are two apparent problems here. On the target machine, there is a
small Normal zone in comparison to DMA32. As kswapd tries to balance all
zones, it would continually try reclaiming for Normal even though DMA32
was balanced enough for callers. The second problem is that
sleeping_prematurely() does not use the same logic as balance_pgdat() when
deciding whether to sleep or not. This keeps kswapd artifically awake.
A number of tests were run and the figures from previous postings will
look very different for a few reasons. One, the old figures were forcing
my network card to use GFP_ATOMIC in attempt to replicate Simon's problem.
Second, I previous specified slub_min_order=3 again in an attempt to
reproduce Simon's problem. In this posting, I'm depending on Simon to say
whether his problem is fixed or not and these figures are to show the
impact to the ordinary cases. Finally, the "vmscan" figures are taken
from /proc/vmstat instead of the tracepoints. There is less information
but recording is less disruptive.
The first test of relevance was postmark with a process running in the
background reading a large amount of anonymous memory in blocks. The
objective was to vaguely simulate what was happening on Simon's machine
and it's memory intensive enough to have kswapd awake.
POSTMARK
traceonly kanyzone
Transactions per second: 156.00 ( 0.00%) 153.00 (-1.96%)
Data megabytes read per second: 21.51 ( 0.00%) 21.52 ( 0.05%)
Data megabytes written per second: 29.28 ( 0.00%) 29.11 (-0.58%)
Files created alone per second: 250.00 ( 0.00%) 416.00 (39.90%)
Files create/transact per second: 79.00 ( 0.00%) 76.00 (-3.95%)
Files deleted alone per second: 520.00 ( 0.00%) 420.00 (-23.81%)
Files delete/transact per second: 79.00 ( 0.00%) 76.00 (-3.95%)
MMTests Statistics: duration
User/Sys Time Running Test (seconds) 16.58 17.4
Total Elapsed Time (seconds) 218.48 222.47
Total pages scanned 326631 322221
Total pages reclaimed 312632 309968
%age total pages scanned/reclaimed 95.71% 96.20%
%age total pages scanned/written 3.02% 1.97%
proc vmstat: Faults
Major Faults 300 254
Minor Faults 645183 660284
Page ins 493588 486704
Page outs 49600884986704
Swap ins 1230 661
Swap outs 9869 6355
Performance is mildly affected because kswapd is no longer doing as much
work and the background memory consumer process is getting in the way.
Note that kswapd scanned and reclaimed fewer pages as it's less aggressive
and overall fewer pages were scanned and reclaimed. Swap in/out is
particularly reduced again reflecting kswapd throwing out fewer pages.
The slight performance impact is unfortunate here but it looks like a
direct result of kswapd being less aggressive. As the bug report is about
too many pages being freed by kswapd, it may have to be accepted for now.
The second test is a streaming IO benchmark that was previously used by
Johannes to show regressions in page reclaim.
MICRO
traceonly kanyzone
User/Sys Time Running Test (seconds) 29.29 28.87
Total Elapsed Time (seconds) 492.18 488.79
Total pages scanned 1773483617698943
Total pages reclaimed 86526168648834
%age total pages scanned/reclaimed 48.79% 48.87%
%age total pages scanned/written 0.00% 0.01%
proc vmstat: Faults
Major Faults 165 221
Minor Faults 96557859656506
Page ins 3880 7228
Page outs 3769294037480076
Swap ins 0 69
Swap outs 19 15
Again fewer pages are scanned and reclaimed as expected and this time the
test completed faster. Note that kswapd is hitting its watermarks faster
(low and high wmark quickly) which I expect is due to kswapd reclaiming
fewer pages.
I also ran fs-mark, iozone and sysbench but there is nothing interesting
to report in the figures. Performance is not significantly changed and
the reclaim statistics look reasonable.
Tgis patch:
When the allocator enters its slow path, kswapd is woken up to balance the
node. It continues working until all zones within the node are balanced.
For order-0 allocations, this makes perfect sense but for higher orders it
can have unintended side-effects. If the zone sizes are imbalanced,
kswapd may reclaim heavily within a smaller zone discarding an excessive
number of pages. The user-visible behaviour is that kswapd is awake and
reclaiming even though plenty of pages are free from a suitable zone.
This patch alters the "balance" logic for high-order reclaim allowing
kswapd to stop if any suitable zone becomes balanced to reduce the number
of pages it reclaims from other zones. kswapd still tries to ensure that
order-0 watermarks for all zones are met before sleeping.
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Reviewed-by: Minchan Kim <minchan.kim@gmail.com> Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Reviewed-by: Eric B Munson <emunson@mgebm.net> Cc: Simon Kirby <sim@hostway.ca> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Dave Hansen <dave@linux.vnet.ibm.com> Cc: Johannes Weiner <hannes@cmpxchg.org> 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>
Steven Rostedt [Thu, 13 Jan 2011 23:46:18 +0000 (15:46 -0800)]
mm: remove likely() from grab_cache_page_write_begin()
Running the annotated branch profiler on a box doing average work
(firefox, evolution, xchat, distcc farm), the likely() used in
grab_cache_page_write_begin() was incorrect most of the time:
correct incorrect % Function File Line
------- --------- - -------- ---- ---- 192426271332401 97 grab_cache_page_write_begin filemap.c 2206
Adding a trace_printk() and running the function tracer limited to
just this function I can see:
Steven Rostedt [Thu, 13 Jan 2011 23:46:17 +0000 (15:46 -0800)]
mm: remove unlikely() from page_mapping()
page_mapping() has a unlikely that the mapping has PAGE_MAPPING_ANON set.
But running the annotated branch profiler on a normal desktop system doing
vairous tasks (xchat, evolution, firefox, distcc), it is not really that
unlikely that the mapping here will have the PAGE_MAPPING_ANON flag set:
The page_mapping() is a static inline, which is why it shows up multiple
times.
The unlikely in page_mapping() was correct a total of 1909540379 times and
incorrect 1270533123 times, with a 39% being incorrect. With this much of
an error, it's best to simply remove the unlikely and have the compiler
and branch prediction figure this out.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Rik van Riel <riel@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Steven Rostedt [Thu, 13 Jan 2011 23:46:16 +0000 (15:46 -0800)]
mm: remove likely() from mapping_unevictable()
The mapping_unevictable() has a likely() around the mapping parameter.
This mapping parameter comes from page_mapping() which has an unlikely()
that the page will be set as PAGE_MAPPING_ANON, and if so, it will return
NULL. One would think that this unlikely() means that the mapping
returned by page_mapping() would not be NULL, but where page_mapping() is
used just above mapping_unevictable(), that unlikely() is incorrect most
of the time. This means that the "likely(mapping)" in
mapping_unevictable() is incorrect most of the time.
Running the annotated branch profiler on my main box which runs firefox,
evolution, xchat and is part of my distcc farm, I had this:
The page_mapping() is a static inline, which is why it shows up multiple
times. The mapping_unevictable() is also a static inline but seems to be
used only once in my setup.
The unlikely in page_mapping() was correct a total of 1909540379 times and
incorrect 1270533123 times, with a 39% being incorrect. Perhaps this is
enough to remove the unlikely from page_mapping() as well.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org> Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Acked-by: Nick Piggin <npiggin@kernel.dk> Acked-by: Rik van Riel <riel@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mlock: do not hold mmap_sem for extended periods of time
__get_user_pages gets a new 'nonblocking' parameter to signal that the
caller is prepared to re-acquire mmap_sem and retry the operation if
needed. This is used to split off long operations if they are going to
block on a disk transfer, or when we detect contention on the mmap_sem.
[akpm@linux-foundation.org: remove ref to rwsem_is_contended()] Signed-off-by: Michel Lespinasse <walken@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Nick Piggin <npiggin@kernel.dk> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: David Howells <dhowells@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mm: move VM_LOCKED check to __mlock_vma_pages_range()
Use a single code path for faulting in pages during mlock.
The reason to have it in this patch series is that I did not want to
update both code paths in a later change that releases mmap_sem when
blocking on disk.
Signed-off-by: Michel Lespinasse <walken@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Nick Piggin <npiggin@kernel.dk> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: David Howells <dhowells@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Move the code to mlock pages from __mlock_vma_pages_range() to
follow_page().
This allows __mlock_vma_pages_range() to not have to break down work into
16-page batches.
An additional motivation for doing this within the present patch series is
that it'll make it easier for a later chagne to drop mmap_sem when
blocking on disk (we'd like to be able to resume at the page that was read
from disk instead of at the start of a 16-page batch).
Signed-off-by: Michel Lespinasse <walken@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Nick Piggin <npiggin@kernel.dk> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: David Howells <dhowells@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mlock: only hold mmap_sem in shared mode when faulting in pages
Currently mlock() holds mmap_sem in exclusive mode while the pages get
faulted in. In the case of a large mlock, this can potentially take a
very long time, during which various commands such as 'ps auxw' will
block. This makes sysadmins unhappy:
real 14m36.232s
user 0m0.003s
sys 0m0.015s
(output from 'time ps auxw' while a 20GB file was being mlocked without
being previously preloaded into page cache)
I propose that mlock() could release mmap_sem after the VM_LOCKED bits
have been set in all appropriate VMAs. Then a second pass could be done
to actually mlock the pages, in small batches, releasing mmap_sem when we
block on disk access or when we detect some contention.
This patch:
Before this change, mlock() holds mmap_sem in exclusive mode while the
pages get faulted in. In the case of a large mlock, this can potentially
take a very long time. Various things will block while mmap_sem is held,
including 'ps auxw'. This can make sysadmins angry.
I propose that mlock() could release mmap_sem after the VM_LOCKED bits
have been set in all appropriate VMAs. Then a second pass could be done
to actually mlock the pages with mmap_sem held for reads only. We need to
recheck the vma flags after we re-acquire mmap_sem, but this is easy.
In the case where a vma has been munlocked before mlock completes, pages
that were already marked as PageMlocked() are handled by the munlock()
call, and mlock() is careful to not mark new page batches as PageMlocked()
after the munlock() call has cleared the VM_LOCKED vma flags. So, the end
result will be identical to what'd happen if munlock() had executed after
the mlock() call.
In a later change, I will allow the second pass to release mmap_sem when
blocking on disk accesses or when it is otherwise contended, so that it
won't be held for long periods of time even in shared mode.
Signed-off-by: Michel Lespinasse <walken@google.com> Tested-by: Valdis Kletnieks <Valdis.Kletnieks@vt.edu> Cc: Hugh Dickins <hughd@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Nick Piggin <npiggin@kernel.dk> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: David Howells <dhowells@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
mlock: avoid dirtying pages and triggering writeback
When faulting in pages for mlock(), we want to break COW for anonymous or
file pages within VM_WRITABLE, non-VM_SHARED vmas. However, there is no
need to write-fault into VM_SHARED vmas since shared file pages can be
mlocked first and dirtied later, when/if they actually get written to.
Skipping the write fault is desirable, as we don't want to unnecessarily
cause these pages to be dirtied and queued for writeback.
Signed-off-by: Michel Lespinasse <walken@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Kosaki Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Theodore Tso <tytso@google.com> Cc: Michael Rubin <mrubin@google.com> Cc: Suleiman Souhlal <suleiman@google.com> Cc: Dave Chinner <david@fromorbit.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>
Reorganize the code so that dirty pages are handled closer to the place
that makes them dirty (handling write fault into shared, writable VMAs).
No behavior changes.
Signed-off-by: Michel Lespinasse <walken@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Kosaki Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Theodore Tso <tytso@google.com> Cc: Michael Rubin <mrubin@google.com> Cc: Suleiman Souhlal <suleiman@google.com> Cc: Dave Chinner <david@fromorbit.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>
mlocking a shared, writable vma currently causes the corresponding pages
to be marked as dirty and queued for writeback. This seems rather
unnecessary given that the pages are not being actually modified during
mlock. It is understood that for non-shared mappings (file or anon) we
want to use a write fault in order to break COW, but there is just no such
need for shared mappings.
The first two patches in this series do not introduce any behavior change.
The intent there is to make it obvious that dirtying file pages is only
done in the (writable, shared) case. I think this clarifies the code, but
I wouldn't mind dropping these two patches if there is no consensus about
them.
The last patch is where we actually avoid dirtying shared mappings during
mlock. Note that as a side effect of this, we won't call page_mkwrite()
for the mappings that define it, and won't be pre-allocating data blocks
at the FS level if the mapped file was sparsely allocated. My
understanding is that mlock does not need to provide such guarantee, as
evidenced by the fact that it never did for the filesystems that don't
define page_mkwrite() - including some common ones like ext3. However, I
would like to gather feedback on this from filesystem people as a
precaution. If this turns out to be a showstopper, maybe block
preallocation can be added back on using a different interface.
Large shared mlocks are getting significantly (>2x) faster in my tests, as
the disk can be fully used for reading the file instead of having to share
between this and writeback.
This patch:
Reorganize the code to remove the 'reuse' flag. No behavior changes.
Signed-off-by: Michel Lespinasse <walken@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Rik van Riel <riel@redhat.com> Cc: Kosaki Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Theodore Tso <tytso@google.com> Cc: Michael Rubin <mrubin@google.com> Cc: Suleiman Souhlal <suleiman@google.com> Cc: Dave Chinner <david@fromorbit.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>
Rik van Riel [Thu, 13 Jan 2011 23:46:06 +0000 (15:46 -0800)]
mm: clear PageError bit in msync & fsync
Temporary IO failures, eg. due to loss of both multipath paths, can
permanently leave the PageError bit set on a page, resulting in msync or
fsync returning -EIO over and over again, even if IO is now getting to the
disk correctly.
We already clear the AS_ENOSPC and AS_IO bits in mapping->flags in the
filemap_fdatawait_range function. Also clearing the PageError bit on the
page allows subsequent msync or fsync calls on this file to return without
an error, if the subsequent IO succeeds.
Unfortunately data written out in the msync or fsync call that returned
-EIO can still get lost, because the page dirty bit appears to not get
restored on IO error. However, the alternative could be potentially all
of memory filling up with uncleanable dirty pages, hanging the system, so
there is no nice choice here...
Signed-off-by: Rik van Riel <riel@redhat.com> Acked-by: Valerie Aurora <vaurora@redhat.com> Acked-by: Jeff Layton <jlayton@redhat.com> Cc: Theodore Ts'o <tytso@mit.edu> Acked-by: Jan Kara <jack@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
oom: allow a non-CAP_SYS_RESOURCE proces to oom_score_adj down
We'd like to be able to oom_score_adj a process up/down as it
enters/leaves the foreground. Currently, it is not possible to oom_adj
down without CAP_SYS_RESOURCE. This patch allows a task to decrease its
oom_score_adj back to the value that a CAP_SYS_RESOURCE thread set it to
or its inherited value at fork. Assuming the thread that has forked it
has oom_score_adj of 0, each process could decrease it back from 0 upon
activation unless a CAP_SYS_RESOURCE thread elevated it to something
higher.
Alternative considered:
* a setuid binary
* a daemon with CAP_SYS_RESOURCE
Since you don't wan't all processes to be able to reduce their oom_adj, a
setuid or daemon implementation would be complex. The alternatives also
have much higher overhead.
This patch updated from original patch based on feedback from David
Rientjes.
Signed-off-by: Mandeep Singh Baines <msb@chromium.org> Acked-by: David Rientjes <rientjes@google.com> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Rik van Riel <riel@redhat.com> Cc: Ying Han <yinghan@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
David Rientjes [Thu, 13 Jan 2011 23:46:02 +0000 (15:46 -0800)]
mm: unify module_alloc code for vmalloc
Four architectures (arm, mips, sparc, x86) use __vmalloc_area() for
module_init(). Much of the code is duplicated and can be generalized in a
globally accessible function, __vmalloc_node_range().
__vmalloc_node() now calls into __vmalloc_node_range() with a range of
[VMALLOC_START, VMALLOC_END) for functionally equivalent behavior.
Each architecture may then use __vmalloc_node_range() directly to remove
the duplication of code.
Signed-off-by: David Rientjes <rientjes@google.com> Cc: Christoph Lameter <cl@linux.com> Cc: Russell King <linux@arm.linux.org.uk> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: "David S. Miller" <davem@davemloft.net> Cc: Ingo Molnar <mingo@redhat.com> 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>
Mel Gorman [Thu, 13 Jan 2011 23:46:00 +0000 (15:46 -0800)]
mm: vmscan: rename lumpy_mode to reclaim_mode
With compaction being used instead of lumpy reclaim, the name lumpy_mode
and associated variables is a bit misleading. Rename lumpy_mode to
reclaim_mode which is a better fit. There is no functional change.
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Rik van Riel <riel@redhat.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: Andy Whitcroft <apw@shadowen.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Mel Gorman [Thu, 13 Jan 2011 23:45:59 +0000 (15:45 -0800)]
mm: compaction: perform a faster migration scan when migrating asynchronously
try_to_compact_pages() is initially called to only migrate pages
asychronously and kswapd always compacts asynchronously. Both are being
optimistic so it is important to complete the work as quickly as possible
to minimise stalls.
This patch alters the scanner when asynchronous to only consider
MIGRATE_MOVABLE pageblocks as migration candidates. This reduces stalls
when allocating huge pages while not impairing allocation success rates as
a full scan will be performed if necessary after direct reclaim.
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Rik van Riel <riel@redhat.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: Andy Whitcroft <apw@shadowen.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Mel Gorman [Thu, 13 Jan 2011 23:45:58 +0000 (15:45 -0800)]
mm: migration: cleanup migrate_pages API by matching types for offlining and sync
With the introduction of the boolean sync parameter, the API looks a
little inconsistent as offlining is still an int. Convert offlining to a
bool for the sake of being tidy.
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Rik van Riel <riel@redhat.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: Andy Whitcroft <apw@shadowen.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Mel Gorman [Thu, 13 Jan 2011 23:45:57 +0000 (15:45 -0800)]
mm: migration: allow migration to operate asynchronously and avoid synchronous compaction in the faster path
Migration synchronously waits for writeback if the initial passes fails.
Callers of memory compaction do not necessarily want this behaviour if the
caller is latency sensitive or expects that synchronous migration is not
going to have a significantly better success rate.
This patch adds a sync parameter to migrate_pages() allowing the caller to
indicate if wait_on_page_writeback() is allowed within migration or not.
For reclaim/compaction, try_to_compact_pages() is first called
asynchronously, direct reclaim runs and then try_to_compact_pages() is
called synchronously as there is a greater expectation that it'll succeed.
[akpm@linux-foundation.org: build/merge fix] Signed-off-by: Mel Gorman <mel@csn.ul.ie> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Rik van Riel <riel@redhat.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: Andy Whitcroft <apw@shadowen.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Mel Gorman [Thu, 13 Jan 2011 23:45:56 +0000 (15:45 -0800)]
mm: vmscan: reclaim order-0 and use compaction instead of lumpy reclaim
Lumpy reclaim is disruptive. It reclaims a large number of pages and
ignores the age of the pages it reclaims. This can incur significant
stalls and potentially increase the number of major faults.
Compaction has reached the point where it is considered reasonably stable
(meaning it has passed a lot of testing) and is a potential candidate for
displacing lumpy reclaim. This patch introduces an alternative to lumpy
reclaim whe compaction is available called reclaim/compaction. The basic
operation is very simple - instead of selecting a contiguous range of
pages to reclaim, a number of order-0 pages are reclaimed and then
compaction is later by either kswapd (compact_zone_order()) or direct
compaction (__alloc_pages_direct_compact()).
[akpm@linux-foundation.org: fix build]
[akpm@linux-foundation.org: use conventional task_struct naming] Signed-off-by: Mel Gorman <mel@csn.ul.ie> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Rik van Riel <riel@redhat.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: Andy Whitcroft <apw@shadowen.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Mel Gorman [Thu, 13 Jan 2011 23:45:55 +0000 (15:45 -0800)]
mm: vmscan: convert lumpy_mode into a bitmask
Currently lumpy_mode is an enum and determines if lumpy reclaim is off,
syncronous or asyncronous. In preparation for using compaction instead of
lumpy reclaim, this patch converts the flags into a bitmap.
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Rik van Riel <riel@redhat.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Andy Whitcroft <apw@shadowen.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Mel Gorman [Thu, 13 Jan 2011 23:45:54 +0000 (15:45 -0800)]
mm: compaction: add trace events for memory compaction activity
In preparation for a patches promoting the use of memory compaction over
lumpy reclaim, this patch adds trace points for memory compaction
activity. Using them, we can monitor the scanning activity of the
migration and free page scanners as well as the number and success rates
of pages passed to page migration.
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Rik van Riel <riel@redhat.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Andy Whitcroft <apw@shadowen.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Bob Liu [Thu, 13 Jan 2011 23:45:49 +0000 (15:45 -0800)]
mm/page-writeback.c: fix __set_page_dirty_no_writeback() return value
__set_page_dirty_no_writeback() should return true if it actually
transitioned the page from a clean to dirty state although it seems nobody
uses its return value at present.
Signed-off-by: Bob Liu <lliubbo@gmail.com> Acked-by: Wu Fengguang <fengguang.wu@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When wb_writeback() is called in WB_SYNC_ALL mode, work->nr_to_write is
usually set to LONG_MAX. The logic in wb_writeback() then calls
__writeback_inodes_sb() with nr_to_write == MAX_WRITEBACK_PAGES and we
easily end up with non-positive nr_to_write after the function returns, if
the inode has more than MAX_WRITEBACK_PAGES dirty pages at the moment.
When nr_to_write is <= 0 wb_writeback() decides we need another round of
writeback but this is wrong in some cases! For example when a single
large file is continuously dirtied, we would never finish syncing it
because each pass would be able to write MAX_WRITEBACK_PAGES and inode
dirty timestamp never gets updated (as inode is never completely clean).
Thus __writeback_inodes_sb() would write the redirtied inode again and
again.
Fix the issue by setting nr_to_write to LONG_MAX in WB_SYNC_ALL mode. We
do not need nr_to_write in WB_SYNC_ALL mode anyway since
write_cache_pages() does livelock avoidance using page tagging in
WB_SYNC_ALL mode.
This makes wb_writeback() call __writeback_inodes_sb() only once on
WB_SYNC_ALL. The latter function won't livelock because it works on
- a finite set of files by doing queue_io() once at the beginning
- a finite set of pages by PAGECACHE_TAG_TOWRITE page tagging
After this patch, program from http://lkml.org/lkml/2010/10/24/154 is no
longer able to stall sync forever.
[fengguang.wu@intel.com: fix locking comment] Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Wu Fengguang <fengguang.wu@intel.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Jan Engelhardt <jengelh@medozas.de> Cc: Jens Axboe <axboe@kernel.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jan Kara [Thu, 13 Jan 2011 23:45:47 +0000 (15:45 -0800)]
writeback: stop background/kupdate works from livelocking other works
Background writeback is easily livelockable in a loop in wb_writeback() by
a process continuously re-dirtying pages (or continuously appending to a
file). This is in fact intended as the target of background writeback is
to write dirty pages it can find as long as we are over
dirty_background_threshold.
But the above behavior gets inconvenient at times because no other work
queued in the flusher thread's queue gets processed. In particular, since
e.g. sync(1) relies on flusher thread to do all the IO for it, sync(1)
can hang forever waiting for flusher thread to do the work.
Generally, when a flusher thread has some work queued, someone submitted
the work to achieve a goal more specific than what background writeback
does. Moreover by working on the specific work, we also reduce amount of
dirty pages which is exactly the target of background writeout. So it
makes sense to give specific work a priority over a generic page cleaning.
Thus we interrupt background writeback if there is some other work to do.
We return to the background writeback after completing all the queued
work.
This may delay the writeback of expired inodes for a while, however the
expired inodes will eventually be flushed to disk as long as the other
works won't livelock.
[fengguang.wu@intel.com: update comment] Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Wu Fengguang <fengguang.wu@intel.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Jan Engelhardt <jengelh@medozas.de> Cc: Jens Axboe <axboe@kernel.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Jan Kara [Thu, 13 Jan 2011 23:45:44 +0000 (15:45 -0800)]
writeback: integrated background writeback work
Check whether background writeback is needed after finishing each work.
When bdi flusher thread finishes doing some work check whether any kind of
background writeback needs to be done (either because
dirty_background_ratio is exceeded or because we need to start flushing
old inodes). If so, just do background write back.
This way, bdi_start_background_writeback() just needs to wake up the
flusher thread. It will do background writeback as soon as there is no
other work.
This is a preparatory patch for the next patch which stops background
writeback as soon as there is other work to do.
Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Wu Fengguang <fengguang.wu@intel.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Jan Engelhardt <jengelh@medozas.de> Cc: Jens Axboe <axboe@kernel.dk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Mel Gorman [Thu, 13 Jan 2011 23:45:43 +0000 (15:45 -0800)]
mm: vmstat: use a single setter function and callback for adjusting percpu thresholds
reduce_pgdat_percpu_threshold() and restore_pgdat_percpu_threshold() exist
to adjust the per-cpu vmstat thresholds while kswapd is awake to avoid
errors due to counter drift. The functions duplicate some code so this
patch replaces them with a single set_pgdat_percpu_threshold() that takes
a callback function to calculate the desired threshold as a parameter.
Mel Gorman [Thu, 13 Jan 2011 23:45:41 +0000 (15:45 -0800)]
mm: page allocator: adjust the per-cpu counter threshold when memory is low
Commit 13b4c97 ("calculate a better estimate of NR_FREE_PAGES when memory
is low") noted that watermarks were based on the vmstat NR_FREE_PAGES. To
avoid synchronization overhead, these counters are maintained on a per-cpu
basis and drained both periodically and when a threshold is above a
threshold. On large CPU systems, the difference between the estimate and
real value of NR_FREE_PAGES can be very high. The system can get into a
case where pages are allocated far below the min watermark potentially
causing livelock issues. The commit solved the problem by taking a better
reading of NR_FREE_PAGES when memory was low.
Unfortately, as reported by Shaohua Li this accurate reading can consume a
large amount of CPU time on systems with many sockets due to cache line
bouncing. This patch takes a different approach. For large machines
where counter drift might be unsafe and while kswapd is awake, the per-cpu
thresholds for the target pgdat are reduced to limit the level of drift to
what should be a safe level. This incurs a performance penalty in heavy
memory pressure by a factor that depends on the workload and the machine
but the machine should function correctly without accidentally exhausting
all memory on a node. There is an additional cost when kswapd wakes and
sleeps but the event is not expected to be frequent - in Shaohua's test
case, there was one recorded sleep and wake event at least.
To ensure that kswapd wakes up, a safe version of zone_watermark_ok() is
introduced that takes a more accurate reading of NR_FREE_PAGES when called
from wakeup_kswapd, when deciding whether it is really safe to go back to
sleep in sleeping_prematurely() and when deciding if a zone is really
balanced or not in balance_pgdat(). We are still using an expensive
function but limiting how often it is called.
When the test case is reproduced, the time spent in the watermark
functions is reduced. The following report is on the percentage of time
spent cumulatively spent in the functions zone_nr_free_pages(),
zone_watermark_ok(), __zone_watermark_ok(), zone_watermark_ok_safe(),
zone_page_state_snapshot(), zone_page_state().
vanilla 11.6615%
disable-threshold 0.2584%
David said:
: We had to pull 13b4c970 "mm: page allocator: calculate a better estimate
: of NR_FREE_PAGES when memory is low and kswapd is awake" from 2.6.36
: internally because tests showed that it would cause the machine to stall
: as the result of heavy kswapd activity. I merged it back with this fix as
: it is pending in the -mm tree and it solves the issue we were seeing, so I
: definitely think this should be pushed to -stable (and I would seriously
: consider it for 2.6.37 inclusion even at this late date).
Signed-off-by: Mel Gorman <mel@csn.ul.ie> Reported-by: Shaohua Li <shaohua.li@intel.com> Reviewed-by: Christoph Lameter <cl@linux.com> Tested-by: Nicolas Bareil <nico@chdir.org> Cc: David Rientjes <rientjes@google.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Cc: <stable@kernel.org> [2.6.37.1, 2.6.36.x] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Dave Jones [Thu, 13 Jan 2011 23:45:40 +0000 (15:45 -0800)]
sched: remove long deprecated CLONE_STOPPED flag
This warning was added in commit c1781789874f ("clone: prepare to recycle
CLONE_STOPPED") three years ago. 2.6.26 came and went. As far as I know,
no-one is actually using CLONE_STOPPED.
Signed-off-by: Dave Jones <davej@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ingo Molnar <mingo@elte.hu> Cc: Tejun Heo <tj@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Claudio Scordino [Thu, 13 Jan 2011 23:45:39 +0000 (15:45 -0800)]
atmel_serial: fix RTS high after initialization in RS485 mode
When working in RS485 mode, the atmel_serial driver keeps RTS high after
the initialization of the serial port. It goes low only after the first
character has been sent.
Eric Dumazet [Thu, 13 Jan 2011 23:45:38 +0000 (15:45 -0800)]
irq: use per_cpu kstat_irqs
Use modern per_cpu API to increment {soft|hard}irq counters, and use
per_cpu allocation for (struct irq_desc)->kstats_irq instead of an array.
This gives better SMP/NUMA locality and saves few instructions per irq.
With small nr_cpuids values (8 for example), kstats_irq was a small array
(less than L1_CACHE_BYTES), potentially source of false sharing.
In the !CONFIG_SPARSE_IRQ case, remove the huge, NUMA/cache unfriendly
kstat_irqs_all[NR_IRQS][NR_CPUS] array.
Note: we still populate kstats_irq for all possible irqs in
early_irq_init(). We probably could use on-demand allocations. (Code
included in alloc_descs()). Problem is not all IRQS are used with a prior
alloc_descs() call.
kstat_irqs_this_cpu() is not used anymore, remove it.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Reviewed-by: Christoph Lameter <cl@linux.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Andi Kleen <andi@firstfloor.org> Cc: Tejun Heo <tj@kernel.org> Cc: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Bruce Chang [Thu, 13 Jan 2011 23:45:37 +0000 (15:45 -0800)]
MAINTAINERS: update entries affecting VIA Technologies
Since the original maintainer-Joseph Chan (josephchan@via.com.tw) doesn't
handle the Linux driver for VIA now, I would like to request to update the
maintainer for the SD/MMC CARD CONTROLLER DRIVER and VIA
UNICHROME(PRO)/CHROME9 FRAMEBUFFER DRIVER before we find a better one.
Signed-off-by: Bruce Chang <brucechang@via.com.tw> Signed-off-by: Florian Tobias Schandinat <FlorianSchandinat@gmx.de> Cc: Joseph Chan <JosephChan@via.com.tw> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Harald Welte <HaraldWelte@viatech.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* git://git.kernel.org/pub/scm/linux/kernel/git/agk/linux-2.6-dm: (32 commits)
dm: raid456 basic support
dm: per target unplug callback support
dm: introduce target callbacks and congestion callback
dm mpath: delay activate_path retry on SCSI_DH_RETRY
dm: remove superfluous irq disablement in dm_request_fn
dm log: use PTR_ERR value instead of ENOMEM
dm snapshot: avoid storing private suspended state
dm snapshot: persistent make metadata_wq multithreaded
dm: use non reentrant workqueues if equivalent
dm: convert workqueues to alloc_ordered
dm stripe: switch from local workqueue to system_wq
dm: dont use flush_scheduled_work
dm snapshot: remove unused dm_snapshot queued_bios_work
dm ioctl: suppress needless warning messages
dm crypt: add loop aes iv generator
dm crypt: add multi key capability
dm crypt: add post iv call to iv generator
dm crypt: use io thread for reads only if mempool exhausted
dm crypt: scale to multiple cpus
dm crypt: simplify compatible table output
...
Linus Torvalds [Fri, 14 Jan 2011 01:30:20 +0000 (17:30 -0800)]
Merge branch 'for-linus' of git://neil.brown.name/md
* 'for-linus' of git://neil.brown.name/md:
md: Fix removal of extra drives when converting RAID6 to RAID5
md: range check slot number when manually adding a spare.
md/raid5: handle manually-added spares in start_reshape.
md: fix sync_completed reporting for very large drives (>2TB)
md: allow suspend_lo and suspend_hi to decrease as well as increase.
md: Don't let implementation detail of curr_resync leak out through sysfs.
md: separate meta and data devs
md-new-param-to_sync_page_io
md-new-param-to-calc_dev_sboffset
md: Be more careful about clearing flags bit in ->recovery
md: md_stop_writes requires mddev_lock.
md/raid5: use sysfs_notify_dirent_safe to avoid NULL pointer
md: Ensure no IO request to get md device before it is properly initialised.
md: Fix single printks with multiple KERN_<level>s
md: fix regression resulting in delays in clearing bits in a bitmap
md: fix regression with re-adding devices to arrays with no metadata
Linus Torvalds [Fri, 14 Jan 2011 01:19:38 +0000 (17:19 -0800)]
ecryptfs: fix broken build
Stephen Rothwell reports that the vfs merge broke the build of ecryptfs.
The breakage comes from commit 9ff5dbd8b57a ("sanitize ecryptfs
->mount()") which was obviously not even build tested. Tssk, tssk, Al.
This is the minimal build fixup for the situation, although I don't have
a filesystem to actually test it with.
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
NeilBrown [Thu, 13 Jan 2011 22:14:34 +0000 (09:14 +1100)]
md/raid5: handle manually-added spares in start_reshape.
It is possible to manually add spares to specific slots before
starting a reshape.
raid5_start_reshape should recognised this possibility and include
it in the accounting.
NeilBrown [Thu, 13 Jan 2011 22:14:34 +0000 (09:14 +1100)]
md: allow suspend_lo and suspend_hi to decrease as well as increase.
The sysfs attributes 'suspend_lo' and 'suspend_hi' describe a region
to which read/writes are suspended so that the under lying data can be
manipulated without user-space noticing.
Currently the window they describe can only move forwards along the
device. However this is an unnecessary restriction which will cause
problems with planned developments.
So relax this restriction and allow these endpoints to move
arbitrarily.
NeilBrown [Thu, 13 Jan 2011 22:14:34 +0000 (09:14 +1100)]
md: Don't let implementation detail of curr_resync leak out through sysfs.
mddev->curr_resync has artificial values of '1' and '2' which are used
by the code which ensures only one resync is happening at a time on
any given device.
These values are internal and should never be exposed to user-space
(except when translated appropriately as in the 'pending' status in
/proc/mdstat).
Unfortunately they are as ->curr_resync is assigned to
->curr_resync_completed and that value is directly visible through
sysfs.
So change the assignments to ->curr_resync_completed to get the same
valued from elsewhere in a form that doesn't have the magic '1' or '2'
values.
Jonathan Brassow [Thu, 13 Jan 2011 22:14:33 +0000 (09:14 +1100)]
md-new-param-to_sync_page_io
Add new parameter to 'sync_page_io'.
The new parameter allows us to distinguish between metadata and data
operations. This becomes important later when we add the ability to
use separate devices for data and metadata.
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Jonathan Brassow [Thu, 13 Jan 2011 22:14:33 +0000 (09:14 +1100)]
md-new-param-to-calc_dev_sboffset
When we allow for separate devices for data and metadata
in a later patch, we will need to be able to calculate
the superblock offset based on more than the bdev.
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
NeilBrown [Thu, 13 Jan 2011 22:14:33 +0000 (09:14 +1100)]
md: Be more careful about clearing flags bit in ->recovery
Setting ->recovery to 0 is generally not a good idea as it could clear
bits that shouldn't be cleared. In particular, MD_RECOVERY_FROZEN
should only be cleared on explicit request from user-space.
So when we need to clear things, just clear the bits that need
clearing.
As there are a few different places which reap a resync process - and
some do an incomplte job - factor out the code for doing the from
md_check_recovery and call that function instead of open coding part
of it.
Signed-off-by: NeilBrown <neilb@suse.de> Reported-by: Jonathan Brassow <jbrassow@redhat.com>
NeilBrown [Thu, 13 Jan 2011 22:14:33 +0000 (09:14 +1100)]
md: md_stop_writes requires mddev_lock.
As md_stop_writes manipulates the sync_thread and calls md_update_sb,
it need to be called with mddev_lock held.
In all internal cases it is, but the symbol is exported for dm-raid to
call and in that case the lock won't be help.
Do make an exported version which takes the lock, and an internal
version which does not.
Jonathan Brassow [Thu, 13 Jan 2011 22:14:33 +0000 (09:14 +1100)]
md/raid5: use sysfs_notify_dirent_safe to avoid NULL pointer
With the module parameter 'start_dirty_degraded' set,
raid5_spare_active() previously called sysfs_notify_dirent() with a NULL
argument (rdev->sysfs_state) when a rebuild finished.
Signed-off-by: Jonathan Brassow <jbrassow@redhat.com> Signed-off-by: Mike Snitzer <snitzer@redhat.com>
NeilBrown [Thu, 13 Jan 2011 22:14:33 +0000 (09:14 +1100)]
md: Ensure no IO request to get md device before it is properly initialised.
When an md device is in the process of coming on line it is possible
for an IO request (typically a partition table probe) to get through
before the array is fully initialised, which can cause unexpected
behaviour (e.g. a crash).
So explicitly record when the array is ready for IO and don't allow IO
through until then.
There is no possibility for a similar problem when the array is going
off-line as there must only be one 'open' at that time, and it is busy
off-lining the array and so cannot send IO requests. So no memory
barrier is needed in md_stop()
This has been a bug since commit 7d82a58ebec in 2.6.30 which
introduced md_make_request. Before then, each personality would
register its own make_request_fn when it was ready.
This is suitable for any stable kernel from 2.6.30.y onwards.
NeilBrown [Thu, 13 Jan 2011 22:13:53 +0000 (09:13 +1100)]
md: fix regression resulting in delays in clearing bits in a bitmap
commit 294097819fd4 (2.6.37-rc4) fixed a problem were md_thread would
sometimes call the ->run function at a bad time.
If an error is detected during array start up after the md_thread has
been started, the md_thread is killed. This resulted in the ->run
function being called once. However the array may not be in a state
that it is safe to call ->run.
However the fix imposed meant that ->run was not called on a timeout.
This means that when an array goes idle, bitmap bits do not get
cleared promptly. While the array is busy the bits will still be
cleared when appropriate so this is not very serious. There is no
risk to data.
Change the test so that we only avoid calling ->run when the thread
is being stopped. This more explicitly addresses the problem situation.
This is suitable for 2.6.37-stable and any -stable kernel to which 294097819fd4 was applied.