linux-stable.git/mm/slab.c, branch linux-3.6.y

slab: fix the DEADLOCK issue on l3 alien lock

2012-10-12T20:50:17+00:00

commit 947ca1856a7e60aa6d20536785e6a42dff25aa6e upstream.

DEADLOCK will be report while running a kernel with NUMA and LOCKDEP enabled,
the process of this fake report is:

	   kmem_cache_free()	//free obj in cachep
	-> cache_free_alien()	//acquire cachep's l3 alien lock
	-> __drain_alien_cache()
	-> free_block()
	-> slab_destroy()
	-> kmem_cache_free()	//free slab in cachep->slabp_cache
	-> cache_free_alien()	//acquire cachep->slabp_cache's l3 alien lock

Since the cachep and cachep->slabp_cache's l3 alien are in the same lock class,
fake report generated.

This should not happen since we already have init_lock_keys() which will
reassign the lock class for both l3 list and l3 alien.

However, init_lock_keys() was invoked at a wrong position which is before we
invoke enable_cpucache() on each cache.

Since until set slab_state to be FULL, we won't invoke enable_cpucache()
on caches to build their l3 alien while creating them, so although we invoked
init_lock_keys(), the l3 alien lock class won't change since we don't have
them until invoked enable_cpucache() later.

This patch will invoke init_lock_keys() after we done enable_cpucache()
instead of before to avoid the fake DEADLOCK report.

Michael traced the problem back to a commit in release 3.0.0:

commit 30765b92ada267c5395fc788623cb15233276f5c
Author: Peter Zijlstra 
Date:   Thu Jul 28 23:22:56 2011 +0200

    slab, lockdep: Annotate the locks before using them

    Fernando found we hit the regular OFF_SLAB 'recursion' before we
    annotate the locks, cure this.

    The relevant portion of the stack-trace:

    > [    0.000000]  [] rt_spin_lock+0x50/0x56
    > [    0.000000]  [] __cache_free+0x43/0xc3
    > [    0.000000]  [] kmem_cache_free+0x6c/0xdc
    > [    0.000000]  [] slab_destroy+0x4f/0x53
    > [    0.000000]  [] free_block+0x94/0xc1
    > [    0.000000]  [] do_tune_cpucache+0x10b/0x2bb
    > [    0.000000]  [] enable_cpucache+0x7b/0xa7
    > [    0.000000]  [] kmem_cache_init_late+0x1f/0x61
    > [    0.000000]  [] start_kernel+0x24c/0x363
    > [    0.000000]  [] i386_start_kernel+0xa9/0xaf

    Reported-by: Fernando Lopez-Lezcano 
    Acked-by: Pekka Enberg 
    Signed-off-by: Peter Zijlstra 
    Link: http://lkml.kernel.org/r/1311888176.2617.379.camel@laptop
    Signed-off-by: Ingo Molnar 

The commit moved init_lock_keys() before we build up the alien, so we
failed to reclass it.

Acked-by: Christoph Lameter 
Tested-by: Paul E. McKenney 
Signed-off-by: Michael Wang 
Signed-off-by: Pekka Enberg 
Signed-off-by: Greg Kroah-Hartman

slab: fix starting index for finding another object

2012-09-17T22:00:38+00:00

In array cache, there is a object at index 0, check it.

Signed-off-by: Joonsoo Kim 
Signed-off-by: Mel Gorman 
Cc: David Miller 
Cc: Chuck Lever 
Cc: David Rientjes 
Cc: Pekka Enberg 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

slab: do ClearSlabPfmemalloc() for all pages of slab

2012-09-17T22:00:38+00:00

Right now, we call ClearSlabPfmemalloc() for first page of slab when we
clear SlabPfmemalloc flag.  This is fine for most swap-over-network use
cases as it is expected that order-0 pages are in use.  Unfortunately it
is possible that that __ac_put_obj() checks SlabPfmemalloc on a tail
page and while this is harmless, it is sloppy.  This patch ensures that
the head page is always used.

This problem was originally identified by Joonsoo Kim.

[js1304@gmail.com: Original implementation and problem identification]
Signed-off-by: Mel Gorman 
Cc: David Miller 
Cc: Chuck Lever 
Cc: Joonsoo Kim 
Cc: David Rientjes 
Cc: Pekka Enberg 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm, slab: lock the correct nodelist after reenabling irqs

2012-08-29T18:32:21+00:00

cache_grow() can reenable irqs so the cpu (and node) can change, so ensure
that we take list_lock on the correct nodelist.

This fixes an issue with commit 072bb0aa5e06 ("mm: sl[au]b: add
knowledge of PFMEMALLOC reserve pages") where list_lock for the wrong
node was taken after growing the cache.

Reported-and-tested-by: Haggai Eran 
Signed-off-by: David Rientjes 
Signed-off-by: Linus Torvalds

mm: micro-optimise slab to avoid a function call

2012-08-01T01:42:46+00:00

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

Signed-off-by: Mel Gorman 
Cc: David Miller 
Cc: Neil Brown 
Cc: Peter Zijlstra 
Cc: Mike Christie 
Cc: Eric B Munson 
Cc: Eric Dumazet 
Cc: Sebastian Andrzej Siewior 
Cc: Mel Gorman 
Cc: Christoph Lameter 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm: introduce __GFP_MEMALLOC to allow access to emergency reserves

2012-08-01T01:42:45+00:00

__GFP_MEMALLOC will allow the allocation to disregard the watermarks, much
like PF_MEMALLOC.  It allows one to pass along the memalloc state in
object related allocation flags as opposed to task related flags, such as
sk->sk_allocation.  This removes the need for ALLOC_PFMEMALLOC as callers
using __GFP_MEMALLOC can get the ALLOC_NO_WATERMARK flag which is now
enough to identify allocations related to page reclaim.

Signed-off-by: Peter Zijlstra 
Signed-off-by: Mel Gorman 
Cc: David Miller 
Cc: Neil Brown 
Cc: Mike Christie 
Cc: Eric B Munson 
Cc: Eric Dumazet 
Cc: Sebastian Andrzej Siewior 
Cc: Mel Gorman 
Cc: Christoph Lameter 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm: sl[au]b: add knowledge of PFMEMALLOC reserve pages

2012-08-01T01:42:45+00:00

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

The Linux Terminal Server Project recommends the use of the Network Block
Device (NBD) for swap according to the manual at
https://sourceforge.net/projects/ltsp/files/Docs-Admin-Guide/LTSPManual.pdf/download
There is also documentation and tutorials on how to setup swap over NBD at
places like https://help.ubuntu.com/community/UbuntuLTSP/EnableNBDSWAP The
nbd-client also documents the use of NBD as swap.  Despite this, the fact
is that a machine using NBD for swap can deadlock within minutes if swap
is used intensively.  This patch series addresses the problem.

The core issue is that network block devices do not use mempools like
normal block devices do.  As the host cannot control where they receive
packets from, they cannot reliably work out in advance how much memory
they might need.  Some years ago, Peter Zijlstra developed a series of
patches that supported swap over an NFS that at least one distribution is
carrying within their kernels.  This patch series borrows very heavily
from Peter's work to support swapping over NBD as a pre-requisite to
supporting swap-over-NFS.  The bulk of the complexity is concerned with
preserving memory that is allocated from the PFMEMALLOC reserves for use
by the network layer which is needed for both NBD and NFS.

Patch 1 adds knowledge of the PFMEMALLOC reserves to SLAB and SLUB to
	preserve access to pages allocated under low memory situations
	to callers that are freeing memory.

Patch 2 optimises the SLUB fast path to avoid pfmemalloc checks

Patch 3 introduces __GFP_MEMALLOC to allow access to the PFMEMALLOC
	reserves without setting PFMEMALLOC.

Patch 4 opens the possibility for softirqs to use PFMEMALLOC reserves
	for later use by network packet processing.

Patch 5 only sets page->pfmemalloc when ALLOC_NO_WATERMARKS was required

Patch 6 ignores memory policies when ALLOC_NO_WATERMARKS is set.

Patches 7-12 allows network processing to use PFMEMALLOC reserves when
	the socket has been marked as being used by the VM to clean pages. If
	packets are received and stored in pages that were allocated under
	low-memory situations and are unrelated to the VM, the packets
	are dropped.

	Patch 11 reintroduces __skb_alloc_page which the networking
	folk may object to but is needed in some cases to propogate
	pfmemalloc from a newly allocated page to an skb. If there is a
	strong objection, this patch can be dropped with the impact being
	that swap-over-network will be slower in some cases but it should
	not fail.

Patch 13 is a micro-optimisation to avoid a function call in the
	common case.

Patch 14 tags NBD sockets as being SOCK_MEMALLOC so they can use
	PFMEMALLOC if necessary.

Patch 15 notes that it is still possible for the PFMEMALLOC reserve
	to be depleted. To prevent this, direct reclaimers get throttled on
	a waitqueue if 50% of the PFMEMALLOC reserves are depleted.  It is
	expected that kswapd and the direct reclaimers already running
	will clean enough pages for the low watermark to be reached and
	the throttled processes are woken up.

Patch 16 adds a statistic to track how often processes get throttled

Some basic performance testing was run using kernel builds, netperf on
loopback for UDP and TCP, hackbench (pipes and sockets), iozone and
sysbench.  Each of them were expected to use the sl*b allocators
reasonably heavily but there did not appear to be significant performance
variances.

For testing swap-over-NBD, a machine was booted with 2G of RAM with a
swapfile backed by NBD.  8*NUM_CPU processes were started that create
anonymous memory mappings and read them linearly in a loop.  The total
size of the mappings were 4*PHYSICAL_MEMORY to use swap heavily under
memory pressure.

Without the patches and using SLUB, the machine locks up within minutes
and runs to completion with them applied.  With SLAB, the story is
different as an unpatched kernel run to completion.  However, the patched
kernel completed the test 45% faster.

MICRO
                                         3.5.0-rc2 3.5.0-rc2
					 vanilla     swapnbd
Unrecognised test vmscan-anon-mmap-write
MMTests Statistics: duration
Sys Time Running Test (seconds)             197.80    173.07
User+Sys Time Running Test (seconds)        206.96    182.03
Total Elapsed Time (seconds)               3240.70   1762.09

This patch: mm: sl[au]b: add knowledge of PFMEMALLOC reserve pages

Allocations of pages below the min watermark run a risk of the machine
hanging due to a lack of memory.  To prevent this, only callers who have
PF_MEMALLOC or TIF_MEMDIE set and are not processing an interrupt are
allowed to allocate with ALLOC_NO_WATERMARKS.  Once they are allocated to
a slab though, nothing prevents other callers consuming free objects
within those slabs.  This patch limits access to slab pages that were
alloced from the PFMEMALLOC reserves.

When this patch is applied, pages allocated from below the low watermark
are returned with page->pfmemalloc set and it is up to the caller to
determine how the page should be protected.  SLAB restricts access to any
page with page->pfmemalloc set to callers which are known to able to
access the PFMEMALLOC reserve.  If one is not available, an attempt is
made to allocate a new page rather than use a reserve.  SLUB is a bit more
relaxed in that it only records if the current per-CPU page was allocated
from PFMEMALLOC reserve and uses another partial slab if the caller does
not have the necessary GFP or process flags.  This was found to be
sufficient in tests to avoid hangs due to SLUB generally maintaining
smaller lists than SLAB.

In low-memory conditions it does mean that !PFMEMALLOC allocators can fail
a slab allocation even though free objects are available because they are
being preserved for callers that are freeing pages.

[a.p.zijlstra@chello.nl: Original implementation]
[sebastian@breakpoint.cc: Correct order of page flag clearing]
Signed-off-by: Mel Gorman 
Cc: David Miller 
Cc: Neil Brown 
Cc: Peter Zijlstra 
Cc: Mike Christie 
Cc: Eric B Munson 
Cc: Eric Dumazet 
Cc: Sebastian Andrzej Siewior 
Cc: Mel Gorman 
Cc: Christoph Lameter 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mm, sl[aou]b: Move kmem_cache_create mutex handling to common code

2012-07-09T09:13:42+00:00

Move the mutex handling into the common kmem_cache_create()
function.

Then we can also move more checks out of SLAB's kmem_cache_create()
into the common code.

Reviewed-by: Glauber Costa 
Signed-off-by: Christoph Lameter 
Signed-off-by: Pekka Enberg

mm, sl[aou]b: Use a common mutex definition

2012-07-09T09:13:41+00:00

Use the mutex definition from SLAB and make it the common way to take a sleeping lock.

This has the effect of using a mutex instead of a rw semaphore for SLUB.

SLOB gains the use of a mutex for kmem_cache_create serialization.
Not needed now but SLOB may acquire some more features later (like slabinfo
/ sysfs support) through the expansion of the common code that will
need this.

Reviewed-by: Glauber Costa 
Reviewed-by: Joonsoo Kim 
Signed-off-by: Christoph Lameter 
Signed-off-by: Pekka Enberg

mm, sl[aou]b: Common definition for boot state of the slab allocators

2012-07-09T09:13:35+00:00

All allocators have some sort of support for the bootstrap status.

Setup a common definition for the boot states and make all slab
allocators use that definition.

Reviewed-by: Glauber Costa 
Reviewed-by: Joonsoo Kim 
Signed-off-by: Christoph Lameter 
Signed-off-by: Pekka Enberg