github-mirror/linux
1
0
Fork 0
mirror of https://github.com/torvalds/linux.git synced 2026-06-08 14:42:37 +02:00
Code Issues Actions Packages Projects Releases Wiki Activity Graph
Linux kernel source tree
787,639 Commits 1 Branch 934 Tags 20 GiB
C 97%
Assembly 1%
Shell 0.6%
Rust 0.5%
Python 0.4%
Other 0.3%
43f47331a4
Go to file
Greg Thelen 43f47331a4 mm: writeback: use exact memcg dirty counts 
commit 0b3d6e6f2d upstream.

Since commit a983b5ebee ("mm: memcontrol: fix excessive complexity in
memory.stat reporting") memcg dirty and writeback counters are managed
as:

 1) per-memcg per-cpu values in range of [-32..32]

 2) per-memcg atomic counter

When a per-cpu counter cannot fit in [-32..32] it's flushed to the
atomic.  Stat readers only check the atomic.  Thus readers such as
balance_dirty_pages() may see a nontrivial error margin: 32 pages per
cpu.

Assuming 100 cpus:
   4k x86 page_size:  13 MiB error per memcg
  64k ppc page_size: 200 MiB error per memcg

Considering that dirty+writeback are used together for some decisions the
errors double.

This inaccuracy can lead to undeserved oom kills.  One nasty case is
when all per-cpu counters hold positive values offsetting an atomic
negative value (i.e.  per_cpu[*]=32, atomic=n_cpu*-32).
balance_dirty_pages() only consults the atomic and does not consider
throttling the next n_cpu*32 dirty pages.  If the file_lru is in the
13..200 MiB range then there's absolutely no dirty throttling, which
burdens vmscan with only dirty+writeback pages thus resorting to oom
kill.

It could be argued that tiny containers are not supported, but it's more
subtle.  It's the amount the space available for file lru that matters.
If a container has memory.max-200MiB of non reclaimable memory, then it
will also suffer such oom kills on a 100 cpu machine.

The following test reliably ooms without this patch.  This patch avoids
oom kills.

  $ cat test
  mount -t cgroup2 none /dev/cgroup
  cd /dev/cgroup
  echo +io +memory > cgroup.subtree_control
  mkdir test
  cd test
  echo 10M > memory.max
  (echo $BASHPID > cgroup.procs && exec /memcg-writeback-stress /foo)
  (echo $BASHPID > cgroup.procs && exec dd if=/dev/zero of=/foo bs=2M count=100)

  $ cat memcg-writeback-stress.c
  /*
   * Dirty pages from all but one cpu.
   * Clean pages from the non dirtying cpu.
   * This is to stress per cpu counter imbalance.
   * On a 100 cpu machine:
   * - per memcg per cpu dirty count is 32 pages for each of 99 cpus
   * - per memcg atomic is -99*32 pages
   * - thus the complete dirty limit: sum of all counters 0
   * - balance_dirty_pages() only sees atomic count -99*32 pages, which
   *   it max()s to 0.
   * - So a workload can dirty -99*32 pages before balance_dirty_pages()
   *   cares.
   */
  #define _GNU_SOURCE
  #include <err.h>
  #include <fcntl.h>
  #include <sched.h>
  #include <stdlib.h>
  #include <stdio.h>
  #include <sys/stat.h>
  #include <sys/sysinfo.h>
  #include <sys/types.h>
  #include <unistd.h>

  static char *buf;
  static int bufSize;

  static void set_affinity(int cpu)
  {
  	cpu_set_t affinity;

  	CPU_ZERO(&affinity);
  	CPU_SET(cpu, &affinity);
  	if (sched_setaffinity(0, sizeof(affinity), &affinity))
  		err(1, "sched_setaffinity");
  }

  static void dirty_on(int output_fd, int cpu)
  {
  	int i, wrote;

  	set_affinity(cpu);
  	for (i = 0; i < 32; i++) {
  		for (wrote = 0; wrote < bufSize; ) {
  			int ret = write(output_fd, buf+wrote, bufSize-wrote);
  			if (ret == -1)
  				err(1, "write");
  			wrote += ret;
  		}
  	}
  }

  int main(int argc, char **argv)
  {
  	int cpu, flush_cpu = 1, output_fd;
  	const char *output;

  	if (argc != 2)
  		errx(1, "usage: output_file");

  	output = argv[1];
  	bufSize = getpagesize();
  	buf = malloc(getpagesize());
  	if (buf == NULL)
  		errx(1, "malloc failed");

  	output_fd = open(output, O_CREAT|O_RDWR);
  	if (output_fd == -1)
  		err(1, "open(%s)", output);

  	for (cpu = 0; cpu < get_nprocs(); cpu++) {
  		if (cpu != flush_cpu)
  			dirty_on(output_fd, cpu);
  	}

  	set_affinity(flush_cpu);
  	if (fsync(output_fd))
  		err(1, "fsync(%s)", output);
  	if (close(output_fd))
  		err(1, "close(%s)", output);
  	free(buf);
  }

Make balance_dirty_pages() and wb_over_bg_thresh() work harder to
collect exact per memcg counters.  This avoids the aforementioned oom
kills.

This does not affect the overhead of memory.stat, which still reads the
single atomic counter.

Why not use percpu_counter? memcg already handles cpus going offline, so
no need for that overhead from percpu_counter.  And the percpu_counter
spinlocks are more heavyweight than is required.

It probably also makes sense to use exact dirty and writeback counters
in memcg oom reports.  But that is saved for later.

Link: http://lkml.kernel.org/r/20190329174609.164344-1-gthelen@google.com
Signed-off-by: Greg Thelen <gthelen@google.com>
Reviewed-by: Roman Gushchin <guro@fb.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: <stable@vger.kernel.org>	[4.16+]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2019-04-17 08:38:51 +02:00
arch kvm: svm: fix potential get_num_contig_pages overflow 2019-04-17 08:38:50 +02:00
block block, bfq: fix in-service-queue check for queue merging 2019-04-05 22:33:12 +02:00
certs export.h: remove VMLINUX_SYMBOL() and VMLINUX_SYMBOL_STR() 2018-08-22 23:21:44 +09:00
crypto crypto: testmgr - skip crc32c context test for ahash algorithms 2019-03-23 20:09:55 +01:00
Documentation ARM: 8833/1: Ensure that NEON code always compiles with Clang 2019-04-05 22:33:08 +02:00
drivers drm/udl: add a release method and delay modeset teardown 2019-04-17 08:38:50 +02:00
firmware kbuild: remove all dummy assignments to obj- 2017-11-18 11:46:06 +09:00
fs f2fs: UBSAN: set boolean value iostat_enable correctly 2019-04-05 22:33:14 +02:00
include mm: writeback: use exact memcg dirty counts 2019-04-17 08:38:51 +02:00
init Revert "mm: use early_pfn_to_nid in page_ext_init" 2019-03-23 20:09:46 +01:00
ipc ipc/shm.c: use ERR_CAST() for shm_lock() error return 2018-10-05 16:32:04 -07:00
kernel alarmtimer: Return correct remaining time 2019-04-17 08:38:50 +02:00
lib lib/string.c: implement a basic bcmp 2019-04-17 08:38:45 +02:00
LICENSES LICENSES: Remove CC-BY-SA-4.0 license text 2018-10-18 11:28:50 +02:00
mm mm: writeback: use exact memcg dirty counts 2019-04-17 08:38:51 +02:00
net netfilter: nfnetlink_cttimeout: fetch timeouts for udplite and gre, too 2019-04-17 08:38:46 +02:00
samples samples: mei: use /dev/mei0 instead of /dev/mei 2019-02-15 08:10:11 +01:00
scripts kbuild: deb-pkg: fix bindeb-pkg breakage when O= is used 2019-04-17 08:38:45 +02:00
security selinux: do not override context on context mounts 2019-04-05 22:33:16 +02:00
sound ALSA: hda - Add two more machines to the power_save_blacklist 2019-04-17 08:38:48 +02:00
tools net/sched: act_sample: fix divide by zero in the traffic path 2019-04-17 08:38:41 +02:00
usr initramfs: move gen_initramfs_list.sh from scripts/ to usr/ 2018-08-22 23:21:44 +09:00
virt KVM: Reject device ioctls from processes other than the VM's creator 2019-04-03 06:26:29 +02:00
.clang-format clang-format: Set IndentWrappedFunctionNames false 2018-08-01 18:38:51 +02:00
.cocciconfig
.get_maintainer.ignore
.gitattributes .gitattributes: set git diff driver for C source code files 2016-10-07 18:46:30 -07:00
.gitignore Kbuild updates for v4.17 (2nd) 2018-04-15 17:21:30 -07:00
.mailmap libnvdimm-for-4.19_misc 2018-08-25 18:13:10 -07:00
COPYING COPYING: use the new text with points to the license files 2018-03-23 12:41:45 -06:00
CREDITS 9p: remove Ron Minnich from MAINTAINERS 2018-08-17 16:20:26 -07:00
Kbuild Kbuild updates for v4.15 2017-11-17 17:45:29 -08:00
Kconfig kconfig: move the "Executable file formats" menu to fs/Kconfig.binfmt 2018-08-02 08:06:55 +09:00
MAINTAINERS MAINTAINERS: Add Sasha as a stable branch maintainer 2018-12-01 09:37:25 +01:00
Makefile kbuild: clang: choose GCC_TOOLCHAIN_DIR not on LD 2019-04-17 08:38:45 +02:00
README Docs: Added a pointer to the formatted docs to README 2018-03-21 09:02:53 -06:00

README 

Linux kernel
============

There are several guides for kernel developers and users. These guides can
be rendered in a number of formats, like HTML and PDF. Please read
Documentation/admin-guide/README.rst first.

In order to build the documentation, use ``make htmldocs`` or
``make pdfdocs``.  The formatted documentation can also be read online at:

    https://www.kernel.org/doc/html/latest/

There are various text files in the Documentation/ subdirectory,
several of them using the Restructured Text markup notation.
See Documentation/00-INDEX for a list of what is contained in each file.

Please read the Documentation/process/changes.rst file, as it contains the
requirements for building and running the kernel, and information about
the problems which may result by upgrading your kernel.