Provide /proc/sys/vm/stat_refresh to force an immediate update of
per-cpu into global vmstats: useful to avoid a sleep(2) or whatever
before checking counts when testing. Originally added to work around a
bug which left counts stranded indefinitely on a cpu going idle (an
inaccuracy magnified when small below-batch numbers represent "huge"
amounts of memory), but I believe that bug is now fixed: nonetheless,
this is still a useful knob.
Its schedule_on_each_cpu() is probably too expensive just to fold into
reading /proc/meminfo itself: give this mode 0600 to prevent abuse.
Allow a write or a read to do the same: nothing to read, but "grep -h
Shmem /proc/sys/vm/stat_refresh /proc/meminfo" is convenient. Oh, and
since global_page_state() itself is careful to disguise any underflow as
0, hack in an "Invalid argument" and pr_warn() if a counter is negative
after the refresh - this helped to fix a misaccounting of
NR_ISOLATED_FILE in my migration code.
But on recent kernels, I find that NR_ALLOC_BATCH and NR_PAGES_SCANNED
often go negative some of the time. I have not yet worked out why, but
have no evidence that it's actually harmful. Punt for the moment by
just ignoring the anomaly on those.
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Andres Lagar-Cavilla <andreslc@google.com>
Cc: Yang Shi <yang.shi@linaro.org>
Cc: Ning Qu <quning@gmail.com>
Cc: Mel Gorman <mgorman@techsingularity.net>
Cc: Andres Lagar-Cavilla <andreslc@google.com>
Cc: Konstantin Khlebnikov <koct9i@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In machines with 140G of memory and enterprise flash storage, we have
seen read and write bursts routinely exceed the kswapd watermarks and
cause thundering herds in direct reclaim. Unfortunately, the only way
to tune kswapd aggressiveness is through adjusting min_free_kbytes - the
system's emergency reserves - which is entirely unrelated to the
system's latency requirements. In order to get kswapd to maintain a
250M buffer of free memory, the emergency reserves need to be set to 1G.
That is a lot of memory wasted for no good reason.
On the other hand, it's reasonable to assume that allocation bursts and
overall allocation concurrency scale with memory capacity, so it makes
sense to make kswapd aggressiveness a function of that as well.
Change the kswapd watermark scale factor from the currently fixed 25% of
the tunable emergency reserve to a tunable 0.1% of memory.
Beyond 1G of memory, this will produce bigger watermark steps than the
current formula in default settings. Ensure that the new formula never
chooses steps smaller than that, i.e. 25% of the emergency reserve.
On a 140G machine, this raises the default watermark steps - the
distance between min and low, and low and high - from 16M to 143M.
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Mel Gorman <mgorman@suse.de>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: David Rientjes <rientjes@google.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Merge first patch-bomb from Andrew Morton:
- A few hotfixes which missed 4.4 becasue I was asleep. cc'ed to
-stable
- A few misc fixes
- OCFS2 updates
- Part of MM. Including pretty large changes to page-flags handling
and to thp management which have been buffered up for 2-3 cycles now.
I have a lot of MM material this time.
[ It turns out the THP part wasn't quite ready, so that got dropped from
this series - Linus ]
* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (117 commits)
zsmalloc: reorganize struct size_class to pack 4 bytes hole
mm/zbud.c: use list_last_entry() instead of list_tail_entry()
zram/zcomp: do not zero out zcomp private pages
zram: pass gfp from zcomp frontend to backend
zram: try vmalloc() after kmalloc()
zram/zcomp: use GFP_NOIO to allocate streams
mm: add tracepoint for scanning pages
drivers/base/memory.c: fix kernel warning during memory hotplug on ppc64
mm/page_isolation: use macro to judge the alignment
mm: fix noisy sparse warning in LIBCFS_ALLOC_PRE()
mm: rework virtual memory accounting
include/linux/memblock.h: fix ordering of 'flags' argument in comments
mm: move lru_to_page to mm_inline.h
Documentation/filesystems: describe the shared memory usage/accounting
memory-hotplug: don't BUG() in register_memory_resource()
hugetlb: make mm and fs code explicitly non-modular
mm/swapfile.c: use list_for_each_entry_safe in free_swap_count_continuations
mm: /proc/pid/clear_refs: no need to clear VM_SOFTDIRTY in clear_soft_dirty_pmd()
mm: make sure isolate_lru_page() is never called for tail page
vmstat: make vmstat_updater deferrable again and shut down on idle
...
Pull trivial tree updates from Jiri Kosina.
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial:
floppy: make local variable non-static
exynos: fixes an incorrect header guard
dt-bindings: fixes some incorrect header guards
cpufreq-dt: correct dead link in documentation
cpufreq: ARM big LITTLE: correct dead link in documentation
treewide: Fix typos in printk
Documentation: filesystem: Fix typo in fs/eventfd.c
fs/super.c: use && instead of & for warn_on condition
Documentation: fix sysfs-ptp
lib: scatterlist: fix Kconfig description
Address Space Layout Randomization (ASLR) provides a barrier to
exploitation of user-space processes in the presence of security
vulnerabilities by making it more difficult to find desired code/data
which could help an attack. This is done by adding a random offset to
the location of regions in the process address space, with a greater
range of potential offset values corresponding to better protection/a
larger search-space for brute force, but also to greater potential for
fragmentation.
The offset added to the mmap_base address, which provides the basis for
the majority of the mappings for a process, is set once on process exec
in arch_pick_mmap_layout() and is done via hard-coded per-arch values,
which reflect, hopefully, the best compromise for all systems. The
trade-off between increased entropy in the offset value generation and
the corresponding increased variability in address space fragmentation
is not absolute, however, and some platforms may tolerate higher amounts
of entropy. This patch introduces both new Kconfig values and a sysctl
interface which may be used to change the amount of entropy used for
offset generation on a system.
The direct motivation for this change was in response to the
libstagefright vulnerabilities that affected Android, specifically to
information provided by Google's project zero at:
http://googleprojectzero.blogspot.com/2015/09/stagefrightened.html
The attack presented therein, by Google's project zero, specifically
targeted the limited randomness used to generate the offset added to the
mmap_base address in order to craft a brute-force-based attack.
Concretely, the attack was against the mediaserver process, which was
limited to respawning every 5 seconds, on an arm device. The hard-coded
8 bits used resulted in an average expected success rate of defeating
the mmap ASLR after just over 10 minutes (128 tries at 5 seconds a
piece). With this patch, and an accompanying increase in the entropy
value to 16 bits, the same attack would take an average expected time of
over 45 hours (32768 tries), which makes it both less feasible and more
likely to be noticed.
The introduced Kconfig and sysctl options are limited by per-arch
minimum and maximum values, the minimum of which was chosen to match the
current hard-coded value and the maximum of which was chosen so as to
give the greatest flexibility without generating an invalid mmap_base
address, generally a 3-4 bits less than the number of bits in the
user-space accessible virtual address space.
When decided whether or not to change the default value, a system
developer should consider that mmap_base address could be placed
anywhere up to 2^(value) bits away from the non-randomized location,
which would introduce variable-sized areas above and below the mmap_base
address such that the maximum vm_area_struct size may be reduced,
preventing very large allocations.
This patch (of 4):
ASLR only uses as few as 8 bits to generate the random offset for the
mmap base address on 32 bit architectures. This value was chosen to
prevent a poorly chosen value from dividing the address space in such a
way as to prevent large allocations. This may not be an issue on all
platforms. Allow the specification of a minimum number of bits so that
platforms desiring greater ASLR protection may determine where to place
the trade-off.
Signed-off-by: Daniel Cashman <dcashman@google.com>
Cc: Russell King <linux@arm.linux.org.uk>
Acked-by: Kees Cook <keescook@chromium.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: David Rientjes <rientjes@google.com>
Cc: Mark Salyzyn <salyzyn@android.com>
Cc: Jeff Vander Stoep <jeffv@google.com>
Cc: Nick Kralevich <nnk@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Hector Marco-Gisbert <hecmargi@upv.es>
Cc: Borislav Petkov <bp@suse.de>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
s/avaiable/available/g
This fixup is already in scripts/spelling.txt.
The fix in Documentation/ABI/testing/sysfs-ptp affects documentation of
a /sys entry: the /sys entry itself is correct.
Signed-off-by: Chris Dunlop <chris@onthe.net.au>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
The origin document references to cap_vm_enough_memory is because
cap_vm_enough_memory invoked __vm_enough_memory before and it no longer
does now.
Signed-off-by: Chun Chen <ramichen@tencent.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The comment says that the per-cpu batchsize and zone watermarks are
determined by present_pages which is definitely wrong, they are both
calculated from managed_pages. Fix it.
Signed-off-by: Yaowei Bai <bywxiaobai@163.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
/proc/extfrag_index does not exist. This file is in debugfs. Fix the
description of extfrag_threshold to reflect this.
Signed-off-by: Rabin Vincent <rabin.vincent@axis.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Currently, pages which are marked as unevictable are protected from
compaction, but not from other types of migration. The POSIX real time
extension explicitly states that mlock() will prevent a major page
fault, but the spirit of this is that mlock() should give a process the
ability to control sources of latency, including minor page faults.
However, the mlock manpage only explicitly says that a locked page will
not be written to swap and this can cause some confusion. The
compaction code today does not give a developer who wants to avoid swap
but wants to have large contiguous areas available any method to achieve
this state. This patch introduces a sysctl for controlling compaction
behavior with respect to the unevictable lru. Users who demand no page
faults after a page is present can set compact_unevictable_allowed to 0
and users who need the large contiguous areas can enable compaction on
locked memory by leaving the default value of 1.
To illustrate this problem I wrote a quick test program that mmaps a
large number of 1MB files filled with random data. These maps are
created locked and read only. Then every other mmap is unmapped and I
attempt to allocate huge pages to the static huge page pool. When the
compact_unevictable_allowed sysctl is 0, I cannot allocate hugepages
after fragmenting memory. When the value is set to 1, allocations
succeed.
Signed-off-by: Eric B Munson <emunson@akamai.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Acked-by: Christoph Lameter <cl@linux.com>
Acked-by: David Rientjes <rientjes@google.com>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Christoph Lameter <cl@linux.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Mel Gorman <mgorman@suse.de>
Cc: David Rientjes <rientjes@google.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>
Merge second set of updates from Andrew Morton:
"More of MM"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>: (83 commits)
mm/nommu.c: fix arithmetic overflow in __vm_enough_memory()
mm/mmap.c: fix arithmetic overflow in __vm_enough_memory()
vmstat: Reduce time interval to stat update on idle cpu
mm/page_owner.c: remove unnecessary stack_trace field
Documentation/filesystems/proc.txt: describe /proc/<pid>/map_files
mm: incorporate read-only pages into transparent huge pages
vmstat: do not use deferrable delayed work for vmstat_update
mm: more aggressive page stealing for UNMOVABLE allocations
mm: always steal split buddies in fallback allocations
mm: when stealing freepages, also take pages created by splitting buddy page
mincore: apply page table walker on do_mincore()
mm: /proc/pid/clear_refs: avoid split_huge_page()
mm: pagewalk: fix misbehavior of walk_page_range for vma(VM_PFNMAP)
mempolicy: apply page table walker on queue_pages_range()
arch/powerpc/mm/subpage-prot.c: use walk->vma and walk_page_vma()
memcg: cleanup preparation for page table walk
numa_maps: remove numa_maps->vma
numa_maps: fix typo in gather_hugetbl_stats
pagemap: use walk->vma instead of calling find_vma()
clear_refs: remove clear_refs_private->vma and introduce clear_refs_test_walk()
...
Dave noticed that unprivileged process can allocate significant amount of
memory -- >500 MiB on x86_64 -- and stay unnoticed by oom-killer and
memory cgroup. The trick is to allocate a lot of PMD page tables. Linux
kernel doesn't account PMD tables to the process, only PTE.
The use-cases below use few tricks to allocate a lot of PMD page tables
while keeping VmRSS and VmPTE low. oom_score for the process will be 0.
#include <errno.h>
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#include <sys/mman.h>
#include <sys/prctl.h>
#define PUD_SIZE (1UL << 30)
#define PMD_SIZE (1UL << 21)
#define NR_PUD 130000
int main(void)
{
char *addr = NULL;
unsigned long i;
prctl(PR_SET_THP_DISABLE);
for (i = 0; i < NR_PUD ; i++) {
addr = mmap(addr + PUD_SIZE, PUD_SIZE, PROT_WRITE|PROT_READ,
MAP_ANONYMOUS|MAP_PRIVATE, -1, 0);
if (addr == MAP_FAILED) {
perror("mmap");
break;
}
*addr = 'x';
munmap(addr, PMD_SIZE);
mmap(addr, PMD_SIZE, PROT_WRITE|PROT_READ,
MAP_ANONYMOUS|MAP_PRIVATE|MAP_FIXED, -1, 0);
if (addr == MAP_FAILED)
perror("re-mmap"), exit(1);
}
printf("PID %d consumed %lu KiB in PMD page tables\n",
getpid(), i * 4096 >> 10);
return pause();
}
The patch addresses the issue by account PMD tables to the process the
same way we account PTE.
The main place where PMD tables is accounted is __pmd_alloc() and
free_pmd_range(). But there're few corner cases:
- HugeTLB can share PMD page tables. The patch handles by accounting
the table to all processes who share it.
- x86 PAE pre-allocates few PMD tables on fork.
- Architectures with FIRST_USER_ADDRESS > 0. We need to adjust sanity
check on exit(2).
Accounting only happens on configuration where PMD page table's level is
present (PMD is not folded). As with nr_ptes we use per-mm counter. The
counter value is used to calculate baseline for badness score by
oom-killer.
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Reported-by: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Hugh Dickins <hughd@google.com>
Reviewed-by: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: David Rientjes <rientjes@google.com>
Tested-by: Sedat Dilek <sedat.dilek@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch fix a spelling typo in Documentation/sysctl/vm.txt
Signed-off-by: Masanari Iida <standby24x7@gmail.com>
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Oleg reports a division by zero error on zero-length write() to the
percpu_pagelist_fraction sysctl:
divide error: 0000 [#1] SMP DEBUG_PAGEALLOC
CPU: 1 PID: 9142 Comm: badarea_io Not tainted 3.15.0-rc2-vm-nfs+ #19
Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011
task: ffff8800d5aeb6e0 ti: ffff8800d87a2000 task.ti: ffff8800d87a2000
RIP: 0010: percpu_pagelist_fraction_sysctl_handler+0x84/0x120
RSP: 0018:ffff8800d87a3e78 EFLAGS: 00010246
RAX: 0000000000000f89 RBX: ffff88011f7fd000 RCX: 0000000000000000
RDX: 0000000000000000 RSI: 0000000000000001 RDI: 0000000000000010
RBP: ffff8800d87a3e98 R08: ffffffff81d002c8 R09: ffff8800d87a3f50
R10: 000000000000000b R11: 0000000000000246 R12: 0000000000000060
R13: ffffffff81c3c3e0 R14: ffffffff81cfddf8 R15: ffff8801193b0800
FS: 00007f614f1e9740(0000) GS:ffff88011f440000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00007f614f1fa000 CR3: 00000000d9291000 CR4: 00000000000006e0
Call Trace:
proc_sys_call_handler+0xb3/0xc0
proc_sys_write+0x14/0x20
vfs_write+0xba/0x1e0
SyS_write+0x46/0xb0
tracesys+0xe1/0xe6
However, if the percpu_pagelist_fraction sysctl is set by the user, it
is also impossible to restore it to the kernel default since the user
cannot write 0 to the sysctl.
This patch allows the user to write 0 to restore the default behavior.
It still requires a fraction equal to or larger than 8, however, as
stated by the documentation for sanity. If a value in the range [1, 7]
is written, the sysctl will return EINVAL.
This successfully solves the divide by zero issue at the same time.
Signed-off-by: David Rientjes <rientjes@google.com>
Reported-by: Oleg Drokin <green@linuxhacker.ru>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Existing description is worded in a way which almost encourages setting of
vfs_cache_pressure above 100, possibly way above it.
Users are left in a dark what this numeric value is - an int? a
percentage? what the scale is?
As a result, we are getting reports about noticeable performance
degradation from users who have set vfs_cache_pressure to ridiculously
high values - because they thought there is no downside to it.
Via code inspection it's obvious that this value is treated as a
percentage. This patch changes text to reflect this fact, and adds a
cautionary paragraph advising against setting vfs_cache_pressure sky high.
Signed-off-by: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When it was introduced, zone_reclaim_mode made sense as NUMA distances
punished and workloads were generally partitioned to fit into a NUMA
node. NUMA machines are now common but few of the workloads are
NUMA-aware and it's routine to see major performance degradation due to
zone_reclaim_mode being enabled but relatively few can identify the
problem.
Those that require zone_reclaim_mode are likely to be able to detect
when it needs to be enabled and tune appropriately so lets have a
sensible default for the bulk of users.
This patch (of 2):
zone_reclaim_mode causes processes to prefer reclaiming memory from
local node instead of spilling over to other nodes. This made sense
initially when NUMA machines were almost exclusively HPC and the
workload was partitioned into nodes. The NUMA penalties were
sufficiently high to justify reclaiming the memory. On current machines
and workloads it is often the case that zone_reclaim_mode destroys
performance but not all users know how to detect this. Favour the
common case and disable it by default. Users that are sophisticated
enough to know they need zone_reclaim_mode will detect it.
Signed-off-by: Mel Gorman <mgorman@suse.de>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Zhang Yanfei <zhangyanfei@cn.fujitsu.com>
Acked-by: Michal Hocko <mhocko@suse.cz>
Reviewed-by: Christoph Lameter <cl@linux.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
There is plenty of anecdotal evidence and a load of blog posts
suggesting that using "drop_caches" periodically keeps your system
running in "tip top shape". Perhaps adding some kernel documentation
will increase the amount of accurate data on its use.
If we are not shrinking caches effectively, then we have real bugs.
Using drop_caches will simply mask the bugs and make them harder to
find, but certainly does not fix them, nor is it an appropriate
"workaround" to limit the size of the caches. On the contrary, there
have been bug reports on issues that turned out to be misguided use of
cache dropping.
Dropping caches is a very drastic and disruptive operation that is good
for debugging and running tests, but if it creates bug reports from
production use, kernel developers should be aware of its use.
Add a bit more documentation about it, a syslog message to track down
abusers, and vmstat drop counters to help analyze problem reports.
[akpm@linux-foundation.org: checkpatch fixes]
[hannes@cmpxchg.org: add runtime suppression control]
Signed-off-by: Dave Hansen <dave@linux.vnet.ibm.com>
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
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>
Prior to commit fe35004fbf ("mm: avoid swapping out with
swappiness==0") setting swappiness to 0, reclaim code could still evict
recently used user anonymous memory to swap even though there is a
significant amount of RAM used for page cache.
The behaviour of setting swappiness to 0 has since changed. When set,
the reclaim code does not initiate swap until the amount of free pages
and file-backed pages, is less than the high water mark in a zone.
Let's update the documentation to reflect this.
[akpm@linux-foundation.org: remove comma, per Randy]
Signed-off-by: Aaron Tomlin <atomlin@redhat.com>
Acked-by: Rik van Riel <riel@redhat.com>
Acked-by: Bryn M. Reeves <bmr@redhat.com>
Cc: Satoru Moriya <satoru.moriya@hds.com>
Cc: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Some applications that run on HPC clusters are designed around the
availability of RAM and the overcommit ratio is fine tuned to get the
maximum usage of memory without swapping. With growing memory, the
1%-of-all-RAM grain provided by overcommit_ratio has become too coarse
for these workload (on a 2TB machine it represents no less than 20GB).
This patch adds the new overcommit_kbytes sysctl variable that allow a
much finer grain.
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: fix nommu build]
Signed-off-by: Jerome Marchand <jmarchan@redhat.com>
Cc: Dave Hansen <dave.hansen@linux.intel.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Now dirty_background_ratio/dirty_ratio contains a percentage of total
avaiable memory, which contains free pages and reclaimable pages. The
number of these pages is not equal to the number of total system memory.
But they are described as a percentage of total system memory in
Documentation/sysctl/vm.txt. So we need to fix them to avoid
misunderstanding.
Signed-off-by: Zheng Liu <wenqing.lz@taobao.com>
Cc: Rob Landley <rob@landley.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Now hugepage migration is enabled, although restricted on pmd-based
hugepages for now (due to lack of testing.) So we should allocate
migratable hugepages from ZONE_MOVABLE if possible.
This patch makes GFP flags in hugepage allocation dependent on migration
support, not only the value of hugepages_treat_as_movable. It provides no
change on the behavior for architectures which do not support hugepage
migration,
Signed-off-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Acked-by: Andi Kleen <ak@linux.intel.com>
Reviewed-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Cc: Hillf Danton <dhillf@gmail.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Hugh Dickins <hughd@google.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Rik van Riel <riel@redhat.com>
Cc: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The default zonelist order selecter will select "node" order if any nodes
DMA zone comprises greater than 70% of its local memory instead of 60%,
according to default_zonelist_order::low_kmem_size > total * 70/100.
Signed-off-by: Wanpeng Li <liwanp@linux.vnet.ibm.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add an admin_reserve_kbytes knob to allow admins to change the hardcoded
memory reserve to something other than 3%, which may be multiple
gigabytes on large memory systems. Only about 8MB is necessary to
enable recovery in the default mode, and only a few hundred MB are
required even when overcommit is disabled.
This affects OVERCOMMIT_GUESS and OVERCOMMIT_NEVER.
admin_reserve_kbytes is initialized to min(3% free pages, 8MB)
I arrived at 8MB by summing the RSS of sshd or login, bash, and top.
Please see first patch in this series for full background, motivation,
testing, and full changelog.
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: make init_admin_reserve() static]
Signed-off-by: Andrew Shewmaker <agshew@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add user_reserve_kbytes knob.
Limit the growth of the memory reserved for other user processes to
min(3% current process size, user_reserve_pages). Only about 8MB is
necessary to enable recovery in the default mode, and only a few hundred
MB are required even when overcommit is disabled.
user_reserve_pages defaults to min(3% free pages, 128MB)
I arrived at 128MB by taking the max VSZ of sshd, login, bash, and top ...
then adding the RSS of each.
This only affects OVERCOMMIT_NEVER mode.
Background
1. user reserve
__vm_enough_memory reserves a hardcoded 3% of the current process size for
other applications when overcommit is disabled. This was done so that a
user could recover if they launched a memory hogging process. Without the
reserve, a user would easily run into a message such as:
bash: fork: Cannot allocate memory
2. admin reserve
Additionally, a hardcoded 3% of free memory is reserved for root in both
overcommit 'guess' and 'never' modes. This was intended to prevent a
scenario where root-cant-log-in and perform recovery operations.
Note that this reserve shrinks, and doesn't guarantee a useful reserve.
Motivation
The two hardcoded memory reserves should be updated to account for current
memory sizes.
Also, the admin reserve would be more useful if it didn't shrink too much.
When the current code was originally written, 1GB was considered
"enterprise". Now the 3% reserve can grow to multiple GB on large memory
systems, and it only needs to be a few hundred MB at most to enable a user
or admin to recover a system with an unwanted memory hogging process.
I've found that reducing these reserves is especially beneficial for a
specific type of application load:
* single application system
* one or few processes (e.g. one per core)
* allocating all available memory
* not initializing every page immediately
* long running
I've run scientific clusters with this sort of load. A long running job
sometimes failed many hours (weeks of CPU time) into a calculation. They
weren't initializing all of their memory immediately, and they weren't
using calloc, so I put systems into overcommit 'never' mode. These
clusters run diskless and have no swap.
However, with the current reserves, a user wishing to allocate as much
memory as possible to one process may be prevented from using, for
example, almost 2GB out of 32GB.
The effect is less, but still significant when a user starts a job with
one process per core. I have repeatedly seen a set of processes
requesting the same amount of memory fail because one of them could not
allocate the amount of memory a user would expect to be able to allocate.
For example, Message Passing Interfce (MPI) processes, one per core. And
it is similar for other parallel programming frameworks.
Changing this reserve code will make the overcommit never mode more useful
by allowing applications to allocate nearly all of the available memory.
Also, the new admin_reserve_kbytes will be safer than the current behavior
since the hardcoded 3% of available memory reserve can shrink to something
useless in the case where applications have grabbed all available memory.
Risks
* "bash: fork: Cannot allocate memory"
The downside of the first patch-- which creates a tunable user reserve
that is only used in overcommit 'never' mode--is that an admin can set
it so low that a user may not be able to kill their process, even if
they already have a shell prompt.
Of course, a user can get in the same predicament with the current 3%
reserve--they just have to launch processes until 3% becomes negligible.
* root-cant-log-in problem
The second patch, adding the tunable rootuser_reserve_pages, allows
the admin to shoot themselves in the foot by setting it too small. They
can easily get the system into a state where root-can't-log-in.
However, the new admin_reserve_kbytes will be safer than the current
behavior since the hardcoded 3% of available memory reserve can shrink
to something useless in the case where applications have grabbed all
available memory.
Alternatives
* Memory cgroups provide a more flexible way to limit application memory.
Not everyone wants to set up cgroups or deal with their overhead.
* We could create a fourth overcommit mode which provides smaller reserves.
The size of useful reserves may be drastically different depending
on the whether the system is embedded or enterprise.
* Force users to initialize all of their memory or use calloc.
Some users don't want/expect the system to overcommit when they malloc.
Overcommit 'never' mode is for this scenario, and it should work well.
The new user and admin reserve tunables are simple to use, with low
overhead compared to cgroups. The patches preserve current behavior where
3% of memory is less than 128MB, except that the admin reserve doesn't
shrink to an unusable size under pressure. The code allows admins to tune
for embedded and enterprise usage.
FAQ
* How is the root-cant-login problem addressed?
What happens if admin_reserve_pages is set to 0?
Root is free to shoot themselves in the foot by setting
admin_reserve_kbytes too low.
On x86_64, the minimum useful reserve is:
8MB for overcommit 'guess'
128MB for overcommit 'never'
admin_reserve_pages defaults to min(3% free memory, 8MB)
So, anyone switching to 'never' mode needs to adjust
admin_reserve_pages.
* How do you calculate a minimum useful reserve?
A user or the admin needs enough memory to login and perform
recovery operations, which includes, at a minimum:
sshd or login + bash (or some other shell) + top (or ps, kill, etc.)
For overcommit 'guess', we can sum resident set sizes (RSS)
because we only need enough memory to handle what the recovery
programs will typically use. On x86_64 this is about 8MB.
For overcommit 'never', we can take the max of their virtual sizes (VSZ)
and add the sum of their RSS. We use VSZ instead of RSS because mode
forces us to ensure we can fulfill all of the requested memory allocations--
even if the programs only use a fraction of what they ask for.
On x86_64 this is about 128MB.
When swap is enabled, reserves are useful even when they are as
small as 10MB, regardless of overcommit mode.
When both swap and overcommit are disabled, then the admin should
tune the reserves higher to be absolutley safe. Over 230MB each
was safest in my testing.
* What happens if user_reserve_pages is set to 0?
Note, this only affects overcomitt 'never' mode.
Then a user will be able to allocate all available memory minus
admin_reserve_kbytes.
However, they will easily see a message such as:
"bash: fork: Cannot allocate memory"
And they won't be able to recover/kill their application.
The admin should be able to recover the system if
admin_reserve_kbytes is set appropriately.
* What's the difference between overcommit 'guess' and 'never'?
"Guess" allows an allocation if there are enough free + reclaimable
pages. It has a hardcoded 3% of free pages reserved for root.
"Never" allows an allocation if there is enough swap + a configurable
percentage (default is 50) of physical RAM. It has a hardcoded 3% of
free pages reserved for root, like "Guess" mode. It also has a
hardcoded 3% of the current process size reserved for additional
applications.
* Why is overcommit 'guess' not suitable even when an app eventually
writes to every page? It takes free pages, file pages, available
swap pages, reclaimable slab pages into consideration. In other words,
these are all pages available, then why isn't overcommit suitable?
Because it only looks at the present state of the system. It
does not take into account the memory that other applications have
malloced, but haven't initialized yet. It overcommits the system.
Test Summary
There was little change in behavior in the default overcommit 'guess'
mode with swap enabled before and after the patch. This was expected.
Systems run most predictably (i.e. no oom kills) in overcommit 'never'
mode with swap enabled. This also allowed the most memory to be allocated
to a user application.
Overcommit 'guess' mode without swap is a bad idea. It is easy to
crash the system. None of the other tested combinations crashed.
This matches my experience on the Roadrunner supercomputer.
Without the tunable user reserve, a system in overcommit 'never' mode
and without swap does not allow the admin to recover, although the
admin can.
With the new tunable reserves, a system in overcommit 'never' mode
and without swap can be configured to:
1. maximize user-allocatable memory, running close to the edge of
recoverability
2. maximize recoverability, sacrificing allocatable memory to
ensure that a user cannot take down a system
Test Description
Fedora 18 VM - 4 x86_64 cores, 5725MB RAM, 4GB Swap
System is booted into multiuser console mode, with unnecessary services
turned off. Caches were dropped before each test.
Hogs are user memtester processes that attempt to allocate all free memory
as reported by /proc/meminfo
In overcommit 'never' mode, memory_ratio=100
Test Results
3.9.0-rc1-mm1
Overcommit | Swap | Hogs | MB Got/Wanted | OOMs | User Recovery | Admin Recovery
---------- ---- ---- ------------- ---- ------------- --------------
guess yes 1 5432/5432 no yes yes
guess yes 4 5444/5444 1 yes yes
guess no 1 5302/5449 no yes yes
guess no 4 - crash no no
never yes 1 5460/5460 1 yes yes
never yes 4 5460/5460 1 yes yes
never no 1 5218/5432 no no yes
never no 4 5203/5448 no no yes
3.9.0-rc1-mm1-tunablereserves
User and Admin Recovery show their respective reserves, if applicable.
Overcommit | Swap | Hogs | MB Got/Wanted | OOMs | User Recovery | Admin Recovery
---------- ---- ---- ------------- ---- ------------- --------------
guess yes 1 5419/5419 no - yes 8MB yes
guess yes 4 5436/5436 1 - yes 8MB yes
guess no 1 5440/5440 * - yes 8MB yes
guess no 4 - crash - no 8MB no
* process would successfully mlock, then the oom killer would pick it
never yes 1 5446/5446 no 10MB yes 20MB yes
never yes 4 5456/5456 no 10MB yes 20MB yes
never no 1 5387/5429 no 128MB no 8MB barely
never no 1 5323/5428 no 226MB barely 8MB barely
never no 1 5323/5428 no 226MB barely 8MB barely
never no 1 5359/5448 no 10MB no 10MB barely
never no 1 5323/5428 no 0MB no 10MB barely
never no 1 5332/5428 no 0MB no 50MB yes
never no 1 5293/5429 no 0MB no 90MB yes
never no 1 5001/5427 no 230MB yes 338MB yes
never no 4* 4998/5424 no 230MB yes 338MB yes
* more memtesters were launched, able to allocate approximately another 100MB
Future Work
- Test larger memory systems.
- Test an embedded image.
- Test other architectures.
- Time malloc microbenchmarks.
- Would it be useful to be able to set overcommit policy for
each memory cgroup?
- Some lines are slightly above 80 chars.
Perhaps define a macro to convert between pages and kb?
Other places in the kernel do this.
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: make init_user_reserve() static]
Signed-off-by: Andrew Shewmaker <agshew@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The pdflush thread is long gone, however we still mention it incorrectly in the
kernel documentation. This patch fixes the situation.
Cc: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The number of ptes and swap entries are used in the oom killer's badness
heuristic, so they should be shown in the tasklist dump.
This patch adds those fields and replaces cpu and oom_adj values that are
currently emitted. Cpu isn't interesting and oom_adj is deprecated and
will be removed later this year, the same information is already displayed
as oom_score_adj which is used internally.
At the same time, make the documentation a little more clear to state this
information is helpful to determine why the oom killer chose the task it
did to kill.
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: 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>
Since per-BDI flusher threads were introduced in 2.6, the pdflush
mechanism is not used any more. But the old interface exported through
/proc/sys/vm/nr_pdflush_threads still exists and is obviously useless.
For back-compatibility, printk warning information and return 2 to notify
the users that the interface is removed.
Signed-off-by: Wanpeng Li <liwp@linux.vnet.ibm.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Fix of the documentation of /proc/sys/vm/page-cluster to match the
behavior of the code and add some comments about what the tunable will
change in that behavior.
Signed-off-by: Christian Ehrhardt <ehrhardt@linux.vnet.ibm.com>
Acked-by: Jens Axboe <axboe@kernel.dk>
Reviewed-by: Minchan Kim <minchan@kernel.org>
Cc: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When dirty_ratio or dirty_bytes is written the other parameter is disabled
and set to 0 (in dirty_bytes_handler() / dirty_ratio_handler()).
We do the same for dirty_background_ratio and dirty_background_bytes.
However, in the sysctl documentation, we say that the counterpart becomes
a function of the old value, that is not correct.
Clarify the documentation reporting the actual behaviour.
Reviewed-by: Greg Thelen <gthelen@google.com>
Acked-by: David Rientjes <rientjes@google.com>
Signed-off-by: Andrea Righi <arighi@develer.com>
Cc: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The oom killer tasklist dump, enabled with the oom_dump_tasks sysctl, is
very helpful information in diagnosing why a user's task has been killed.
It emits useful information such as each eligible thread's memory usage
that can determine why the system is oom, so it should be enabled by
default.
Signed-off-by: David Rientjes <rientjes@google.com>
Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: 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>
The kernel applies some heuristics when deciding if memory should be
compacted or reclaimed to satisfy a high-order allocation. One of these
is based on the fragmentation. If the index is below 500, memory will not
be compacted. This choice is arbitrary and not based on data. To help
optimise the system and set a sensible default for this value, this patch
adds a sysctl extfrag_threshold. The kernel will only compact memory if
the fragmentation index is above the extfrag_threshold.
[randy.dunlap@oracle.com: Fix build errors when proc fs is not configured]
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Minchan Kim <minchan.kim@gmail.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: 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>
Add a proc file /proc/sys/vm/compact_memory. When an arbitrary value is
written to the file, all zones are compacted. The expected user of such a
trigger is a job scheduler that prepares the system before the target
application runs.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Acked-by: Rik van Riel <riel@redhat.com>
Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Reviewed-by: Minchan Kim <minchan.kim@gmail.com>
Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Reviewed-by: Christoph Lameter <cl@linux-foundation.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Presently, if panic_on_oom=2, the whole system panics even if the oom
happend in some special situation (as cpuset, mempolicy....). Then,
panic_on_oom=2 means painc_on_oom_always.
Now, memcg doesn't check panic_on_oom flag. This patch adds a check.
BTW, how it's useful ?
kdump+panic_on_oom=2 is the last tool to investigate what happens in
oom-ed system. When a task is killed, the sysytem recovers and there will
be few hint to know what happnes. In mission critical system, oom should
never happen. Then, panic_on_oom=2+kdump is useful to avoid next OOM by
knowing precise information via snapshot.
TODO:
- For memcg, it's for isolate system's memory usage, oom-notiifer and
freeze_at_oom (or rest_at_oom) should be implemented. Then, management
daemon can do similar jobs (as kdump) or taking snapshot per cgroup.
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Balbir Singh <balbir@linux.vnet.ibm.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Nick Piggin <npiggin@suse.de>
Reviewed-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'hwpoison' of git://git.kernel.org/pub/scm/linux/kernel/git/ak/linux-mce-2.6: (21 commits)
HWPOISON: Enable error_remove_page on btrfs
HWPOISON: Add simple debugfs interface to inject hwpoison on arbitary PFNs
HWPOISON: Add madvise() based injector for hardware poisoned pages v4
HWPOISON: Enable error_remove_page for NFS
HWPOISON: Enable .remove_error_page for migration aware file systems
HWPOISON: The high level memory error handler in the VM v7
HWPOISON: Add PR_MCE_KILL prctl to control early kill behaviour per process
HWPOISON: shmem: call set_page_dirty() with locked page
HWPOISON: Define a new error_remove_page address space op for async truncation
HWPOISON: Add invalidate_inode_page
HWPOISON: Refactor truncate to allow direct truncating of page v2
HWPOISON: check and isolate corrupted free pages v2
HWPOISON: Handle hardware poisoned pages in try_to_unmap
HWPOISON: Use bitmask/action code for try_to_unmap behaviour
HWPOISON: x86: Add VM_FAULT_HWPOISON handling to x86 page fault handler v2
HWPOISON: Add poison check to page fault handling
HWPOISON: Add basic support for poisoned pages in fault handler v3
HWPOISON: Add new SIGBUS error codes for hardware poison signals
HWPOISON: Add support for poison swap entries v2
HWPOISON: Export some rmap vma locking to outside world
...
Reported-by: Christian Thaeter <ct@pipapo.org>
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Add the high level memory handler that poisons pages
that got corrupted by hardware (typically by a two bit flip in a DIMM
or a cache) on the Linux level. The goal is to prevent everyone
from accessing these pages in the future.
This done at the VM level by marking a page hwpoisoned
and doing the appropriate action based on the type of page
it is.
The code that does this is portable and lives in mm/memory-failure.c
To quote the overview comment:
High level machine check handler. Handles pages reported by the
hardware as being corrupted usually due to a 2bit ECC memory or cache
failure.
This focuses on pages detected as corrupted in the background.
When the current CPU tries to consume corruption the currently
running process can just be killed directly instead. This implies
that if the error cannot be handled for some reason it's safe to
just ignore it because no corruption has been consumed yet. Instead
when that happens another machine check will happen.
Handles page cache pages in various states. The tricky part
here is that we can access any page asynchronous to other VM
users, because memory failures could happen anytime and anywhere,
possibly violating some of their assumptions. This is why this code
has to be extremely careful. Generally it tries to use normal locking
rules, as in get the standard locks, even if that means the
error handling takes potentially a long time.
Some of the operations here are somewhat inefficient and have non
linear algorithmic complexity, because the data structures have not
been optimized for this case. This is in particular the case
for the mapping from a vma to a process. Since this case is expected
to be rare we hope we can get away with this.
There are in principle two strategies to kill processes on poison:
- just unmap the data and wait for an actual reference before
killing
- kill as soon as corruption is detected.
Both have advantages and disadvantages and should be used
in different situations. Right now both are implemented and can
be switched with a new sysctl vm.memory_failure_early_kill
The default is early kill.
The patch does some rmap data structure walking on its own to collect
processes to kill. This is unusual because normally all rmap data structure
knowledge is in rmap.c only. I put it here for now to keep
everything together and rmap knowledge has been seeping out anyways
Includes contributions from Johannes Weiner, Chris Mason, Fengguang Wu,
Nick Piggin (who did a lot of great work) and others.
Cc: npiggin@suse.de
Cc: riel@redhat.com
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Acked-by: Rik van Riel <riel@redhat.com>
Reviewed-by: Hidehiro Kawai <hidehiro.kawai.ez@hitachi.com>
A bug was brought to my attention against a distro kernel but it affects
mainline and I believe problems like this have been reported in various
guises on the mailing lists although I don't have specific examples at the
moment.
The reported problem was that malloc() stalled for a long time (minutes in
some cases) if a large tmpfs mount was occupying a large percentage of
memory overall. The pages did not get cleaned or reclaimed by
zone_reclaim() because the zone_reclaim_mode was unsuitable, but the lists
are uselessly scanned frequencly making the CPU spin at near 100%.
This patchset intends to address that bug and bring the behaviour of
zone_reclaim() more in line with expectations which were noticed during
investigation. It is based on top of mmotm and takes advantage of
Kosaki's work with respect to zone_reclaim().
Patch 1 fixes the heuristics that zone_reclaim() uses to determine if the
scan should go ahead. The broken heuristic is what was causing the
malloc() stall as it uselessly scanned the LRU constantly. Currently,
zone_reclaim is assuming zone_reclaim_mode is 1 and historically it
could not deal with tmpfs pages at all. This fixes up the heuristic so
that an unnecessary scan is more likely to be correctly avoided.
Patch 2 notes that zone_reclaim() returning a failure automatically means
the zone is marked full. This is not always true. It could have
failed because the GFP mask or zone_reclaim_mode were unsuitable.
Patch 3 introduces a counter zreclaim_failed that will increment each
time the zone_reclaim scan-avoidance heuristics fail. If that
counter is rapidly increasing, then zone_reclaim_mode should be
set to 0 as a temporarily resolution and a bug reported because
the scan-avoidance heuristic is still broken.
This patch:
On NUMA machines, the administrator can configure zone_reclaim_mode that
is a more targetted form of direct reclaim. On machines with large NUMA
distances for example, a zone_reclaim_mode defaults to 1 meaning that
clean unmapped pages will be reclaimed if the zone watermarks are not
being met.
There is a heuristic that determines if the scan is worthwhile but the
problem is that the heuristic is not being properly applied and is
basically assuming zone_reclaim_mode is 1 if it is enabled. The lack of
proper detection can manfiest as high CPU usage as the LRU list is scanned
uselessly.
Historically, once enabled it was depending on NR_FILE_PAGES which may
include swapcache pages that the reclaim_mode cannot deal with. Patch
vmscan-change-the-number-of-the-unmapped-files-in-zone-reclaim.patch by
Kosaki Motohiro noted that zone_page_state(zone, NR_FILE_PAGES) included
pages that were not file-backed such as swapcache and made a calculation
based on the inactive, active and mapped files. This is far superior when
zone_reclaim==1 but if RECLAIM_SWAP is set, then NR_FILE_PAGES is a
reasonable starting figure.
This patch alters how zone_reclaim() works out how many pages it might be
able to reclaim given the current reclaim_mode. If RECLAIM_SWAP is set in
the reclaim_mode it will either consider NR_FILE_PAGES as potential
candidates or else use NR_{IN}ACTIVE}_PAGES-NR_FILE_MAPPED to discount
swapcache and other non-file-backed pages. If RECLAIM_WRITE is not set,
then NR_FILE_DIRTY number of pages are not candidates. If RECLAIM_SWAP is
not set, then NR_FILE_MAPPED are not.
[kosaki.motohiro@jp.fujitsu.com: Estimate unmapped pages minus tmpfs pages]
[fengguang.wu@intel.com: Fix underflow problem in Kosaki's estimate]
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Reviewed-by: Rik van Riel <riel@redhat.com>
Acked-by: Christoph Lameter <cl@linux-foundation.org>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
ALLOC_WMARK_MIN, ALLOC_WMARK_LOW and ALLOC_WMARK_HIGH determin whether
pages_min, pages_low or pages_high is used as the zone watermark when
allocating the pages. Two branches in the allocator hotpath determine
which watermark to use.
This patch uses the flags as an array index into a watermark array that is
indexed with WMARK_* defines accessed via helpers. All call sites that
use zone->pages_* are updated to use the helpers for accessing the values
and the array offsets for setting.
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Reviewed-by: Christoph Lameter <cl@linux-foundation.org>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Cc: Dave Hansen <dave@linux.vnet.ibm.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>
This reverts commit fafd688e4c.
Work is progressing to switch away from pdflush as the process backing
for flushing out dirty data. So it seems pointless to add more knobs
to control pdflush threads. The original author of the patch did not
have any specific use cases for adding the knobs, so we can easily
revert this before 2.6.30 to avoid having to maintain this API
forever.
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
Add /proc entries to give the admin the ability to control the minimum and
maximum number of pdflush threads. This allows finer control of pdflush
on both large and small machines.
The rationale is simply one size does not fit all. Admins on large and/or
small systems may want to tune the min/max pdflush thread count to best
suit their needs. Right now the min/max is hardcoded to 2/8. While
probably a fair estimate for smaller machines, large machines with large
numbers of CPUs and large numbers of filesystems/block devices may benefit
from larger numbers of threads working on different block devices.
Even if the background flushing algorithm is radically changed, it is
still likely that multiple threads will be involved and admins would still
desire finer control on the min/max other than to have to recompile the
kernel.
The patch adds '/proc/sys/vm/nr_pdflush_threads_min' and
'/proc/sys/vm/nr_pdflush_threads_max' with r/w permissions.
The minimum value for nr_pdflush_threads_min is 1 and the maximum value is
the current value of nr_pdflush_threads_max. This minimum is required
since additional thread creation is performed in a pdflush thread itself.
The minimum value for nr_pdflush_threads_max is the current value of
nr_pdflush_threads_min and the maximum value can be 1000.
Documentation/sysctl/vm.txt is also updated.
[akpm@linux-foundation.org: fix comment, fix whitespace, use __read_mostly]
Signed-off-by: Peter W Morreale <pmorreale@novell.com>
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>
Update Documentation/sysctl/vm.txt and Documentation/filesystems/proc.txt.
More specifically, the section on /proc/sys/vm in
Documentation/filesystems/proc.txt was removed and a link to
Documentation/sysctl/vm.txt added.
Most of the verbiage from proc.txt was simply moved in vm.txt, with new
addtional text for "swappiness" and "stat_interval".
Signed-off-by: Peter W Morreale <pmorreale@novell.com>
Acked-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
NOMMU mmap allocates a piece of memory for an mmap that's rounded up in size to
the nearest power-of-2 number of pages. Currently it then discards the excess
pages back to the page allocator, making that memory available for use by other
things. This can, however, cause greater amount of fragmentation.
To counter this, a sysctl is added in order to fine-tune the trimming
behaviour. The default behaviour remains to trim pages aggressively, while
this can either be disabled completely or set to a higher page-granular
watermark in order to have finer-grained control.
vm region vm_top bits taken from an earlier patch by David Howells.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-by: Mike Frysinger <vapier.adi@gmail.com>
This change introduces two new sysctls to /proc/sys/vm:
dirty_background_bytes and dirty_bytes.
dirty_background_bytes is the counterpart to dirty_background_ratio and
dirty_bytes is the counterpart to dirty_ratio.
With growing memory capacities of individual machines, it's no longer
sufficient to specify dirty thresholds as a percentage of the amount of
dirtyable memory over the entire system.
dirty_background_bytes and dirty_bytes specify quantities of memory, in
bytes, that represent the dirty limits for the entire system. If either
of these values is set, its value represents the amount of dirty memory
that is needed to commence either background or direct writeback.
When a `bytes' or `ratio' file is written, its counterpart becomes a
function of the written value. For example, if dirty_bytes is written to
be 8096, 8K of memory is required to commence direct writeback.
dirty_ratio is then functionally equivalent to 8K / the amount of
dirtyable memory:
dirtyable_memory = free pages + mapped pages + file cache
dirty_background_bytes = dirty_background_ratio * dirtyable_memory
-or-
dirty_background_ratio = dirty_background_bytes / dirtyable_memory
AND
dirty_bytes = dirty_ratio * dirtyable_memory
-or-
dirty_ratio = dirty_bytes / dirtyable_memory
Only one of dirty_background_bytes and dirty_background_ratio may be
specified at a time, and only one of dirty_bytes and dirty_ratio may be
specified. When one sysctl is written, the other appears as 0 when read.
The `bytes' files operate on a page size granularity since dirty limits
are compared with ZVC values, which are in page units.
Prior to this change, the minimum dirty_ratio was 5 as implemented by
get_dirty_limits() although /proc/sys/vm/dirty_ratio would show any user
written value between 0 and 100. This restriction is maintained, but
dirty_bytes has a lower limit of only one page.
Also prior to this change, the dirty_background_ratio could not equal or
exceed dirty_ratio. This restriction is maintained in addition to
restricting dirty_background_bytes. If either background threshold equals
or exceeds that of the dirty threshold, it is implicitly set to half the
dirty threshold.
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Christoph Lameter <cl@linux-foundation.org>
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: Andrea Righi <righi.andrea@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Hugh Dickins