[Bug malloc/17195] New: excessive heap trimming in threaded programs

[Bug malloc/17195] New: excessive heap trimming in threaded programs

[jtaylor.debian at googlemail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=17195

            Bug ID: 17195
           Summary: excessive heap trimming in threaded programs
           Product: glibc
           Version: 2.19
            Status: NEW
          Severity: normal
          Priority: P2
         Component: malloc
          Assignee: unassigned at sourceware dot org
          Reporter: jtaylor.debian at googlemail dot com

using glibc 2.19 from ubuntu 14.04 amd64 the following simple program causes a
huge amount of page faults due to the heap being trimmed on every second free()
even though heap trimming was disabled and the mmap threshold is set to the
maximum:

#include <stdlib.h>
#include <string.h>
#include <stdio.h>
#include <malloc.h>

int main(int argc, const char *argv[])
{
    mallopt (M_TRIM_THRESHOLD, -1);
    mallopt (M_MMAP_THRESHOLD, 1024*1024*32);
#pragma omp parallel for num_threads(2)
    for (size_t i = 0; i < 1000; i++) {
        size_t n = 1024*1024*16;
        double * m = malloc(n);
        memset(m, 1, n);
        free(m);
    }
    return 0;
}

gcc test.c -std=c99 -g -O2 -fopenmp
perf stat -e page-faults ./a.out 
         2,036,312 page-faults         

strace -f -e madvise ./a.out 2>&1 | grep MADV_DONTNEED | wc -l
500

when setting the number of threads to 1 it reuses the memory as expected:
perf stat -e page-faults ./a.out 
             4,271 page-faults

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug malloc/17195] excessive heap trimming in threaded programs even when disabled with mallopt

[jtaylor.debian at googlemail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=17195

Julian Taylor <jtaylor.debian at googlemail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|excessive heap trimming in  |excessive heap trimming in
                   |threaded programs           |threaded programs even when
                   |                            |disabled with mallopt

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug malloc/17195] excessive heap trimming in threaded programs even when disabled with mallopt

[neleai at seznam dot cz]

https://sourceware.org/bugzilla/show_bug.cgi?id=17195

Ondrej Bilka <neleai at seznam dot cz> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |neleai at seznam dot cz
         Resolution|---                         |DUPLICATE

--- Comment #1 from Ondrej Bilka <neleai at seznam dot cz> ---
Yes that is known issue

*** This bug has been marked as a duplicate of bug 1128 ***

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug malloc/17195] excessive heap trimming in threaded programs even when disabled with mallopt

[jtaylor.debian at googlemail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=17195

--- Comment #2 from Julian Taylor <jtaylor.debian at googlemail dot com> ---
my issue is not with sbrk but with madvise in threaded applications which
cannot be worked around by using mallopt like in the other bug.
the trimming can be disabled in single threaded applications but not
multithreaded ones.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug malloc/17195] excessive heap trimming in threaded programs even when disabled with mallopt

[jtaylor.debian at googlemail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=17195

--- Comment #3 from Julian Taylor <jtaylor.debian at googlemail dot com> ---
a patch has been posted on the list (not by me):
https://sourceware.org/ml/libc-alpha/2015-02/msg00168.html

I still think the duplicate is wrong, this is about madvise in threads, not
sbrk.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug malloc/17195] excessive heap trimming in threaded programs even when disabled with mallopt

[siddhesh at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=17195

Siddhesh Poyarekar <siddhesh at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
                 CC|                            |siddhesh at redhat dot com
         Resolution|DUPLICATE                   |---

--- Comment #4 from Siddhesh Poyarekar <siddhesh at redhat dot com> ---
Right, it's not a duplicate.  I'll have the patch pushed soon.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug malloc/17195] excessive heap trimming in threaded programs even when disabled with mallopt

[cvs-commit at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=17195

--- Comment #5 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> ---
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GNU C Library master sources".

The branch, master has been updated
       via  c26efef9798914e208329c0e8c3c73bb1135d9e3 (commit)
      from  a3d9ab5070b56b49aa91be2887fa5b118012b2cd (commit)

Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.

- Log -----------------------------------------------------------------
https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=c26efef9798914e208329c0e8c3c73bb1135d9e3

commit c26efef9798914e208329c0e8c3c73bb1135d9e3
Author: Mel Gorman <mgorman@suse.de>
Date:   Thu Apr 2 12:14:14 2015 +0530

    malloc: Consistently apply trim_threshold to all heaps [BZ #17195]

    Trimming heaps is a balance between saving memory and the system overhead
    required to update page tables and discard allocated pages. The malloc
    option M_TRIM_THRESHOLD is a tunable that users are meant to use to decide
    where this balance point is but it is only applied to the main arena.

    For scalability reasons, glibc malloc has per-thread heaps but these are
    shrunk with madvise() if there is one page free at the top of the heap.
    In some circumstances this can lead to high system overhead if a thread
    has a control flow like

        while (data_to_process) {
            buf = malloc(large_size);
            do_stuff();
            free(buf);
        }

    For a large size, the free() will call madvise (pagetable teardown, page
    free and TLB flush) every time followed immediately by a malloc (fault,
    kernel page alloc, zeroing and charge accounting). The kernel overhead
    can dominate such a workload.

    This patch allows the user to tune when madvise gets called by applying
    the trim threshold to the per-thread heaps and using similar logic to the
    main arena when deciding whether to shrink. Alternatively if the dynamic
    brk/mmap threshold gets adjusted then the new values will be obeyed by
    the per-thread heaps.

    Bug 17195 was a test case motivated by a problem encountered in scientific
    applications written in python that performance badly due to high page
fault
    overhead. The basic operation of such a program was posted by Julian Taylor
    https://sourceware.org/ml/libc-alpha/2015-02/msg00373.html

    With this patch applied, the overhead is eliminated. All numbers in this
    report are in seconds and were recorded by running Julian's program 30
    times.

    pyarray
                                     glibc               madvise
                                      2.21                    v2
    System  min             1.81 (  0.00%)        0.00 (100.00%)
    System  mean            1.93 (  0.00%)        0.02 ( 99.20%)
    System  stddev          0.06 (  0.00%)        0.01 ( 88.99%)
    System  max             2.06 (  0.00%)        0.03 ( 98.54%)
    Elapsed min             3.26 (  0.00%)        2.37 ( 27.30%)
    Elapsed mean            3.39 (  0.00%)        2.41 ( 28.84%)
    Elapsed stddev          0.14 (  0.00%)        0.02 ( 82.73%)
    Elapsed max             4.05 (  0.00%)        2.47 ( 39.01%)

                   glibc     madvise
                    2.21          v2
    User          141.86      142.28
    System         57.94        0.60
    Elapsed       102.02       72.66

    Note that almost a minutes worth of system time is eliminted and the
    program completes 28% faster on average.

    To illustrate the problem without python this is a basic test-case for
    the worst case scenario where every free is a madvise followed by a an
alloc

    /* gcc bench-free.c -lpthread -o bench-free */
    static int num = 1024;

    void __attribute__((noinline,noclone)) dostuff (void *p)
    {
    }

    void *worker (void *data)
    {
      int i;

      for (i = num; i--;)
        {
          void *m = malloc (48*4096);
          dostuff (m);
          free (m);
        }

      return NULL;
    }

    int main()
    {
      int i;
      pthread_t t;
      void *ret;
      if (pthread_create (&t, NULL, worker, NULL))
        exit (2);
      if (pthread_join (t, &ret))
        exit (3);
      return 0;
    }

    Before the patch, this resulted in 1024 calls to madvise. With the patch
applied,
    madvise is called twice because the default trim threshold is high enough
to avoid
    this.

    This a more complex case where there is a mix of frees. It's simply a
different worker
    function for the test case above

    void *worker (void *data)
    {
      int i;
      int j = 0;
      void *free_index[num];

      for (i = num; i--;)
        {
          void *m = malloc ((i % 58) *4096);
          dostuff (m);
          if (i % 2 == 0) {
            free (m);
          } else {
            free_index[j++] = m;
          }
        }
      for (; j >= 0; j--)
        {
          free(free_index[j]);
        }

      return NULL;
    }

    glibc 2.21 calls malloc 90305 times but with the patch applied, it's
    called 13438. Increasing the trim threshold will decrease the number of
    times it's called with the option of eliminating the overhead.

    ebizzy is meant to generate a workload resembling common web application
    server workloads. It is threaded with a large working set that at its core
    has an allocation, do_stuff, free loop that also hits this case. The
primary
    metric of the benchmark is records processed per second. This is running on
    my desktop which is a single socket machine with an I7-4770 and 8 cores.
    Each thread count was run for 30 seconds. It was only run once as the
    performance difference is so high that the variation is insignificant.

                    glibc 2.21              patch
    threads 1            10230              44114
    threads 2            19153              84925
    threads 4            34295             134569
    threads 8            51007             183387

    Note that the saving happens to be a concidence as the size allocated
    by ebizzy was less than the default threshold. If a different number of
    chunks were specified then it may also be necessary to tune the threshold
    to compensate

    This is roughly quadrupling the performance of this benchmark. The
difference in
    system CPU usage illustrates why.

    ebizzy running 1 thread with glibc 2.21
    10230 records/s 306904
    real 30.00 s
    user  7.47 s
    sys  22.49 s

    22.49 seconds was spent in the kernel for a workload runinng 30 seconds.
With the
    patch applied

    ebizzy running 1 thread with patch applied
    44126 records/s 1323792
    real 30.00 s
    user 29.97 s
    sys   0.00 s

    system CPU usage was zero with the patch applied. strace shows that glibc
    running this workload calls madvise approximately 9000 times a second. With
    the patch applied madvise was called twice during the workload (or 0.06
    times per second).

    2015-02-10  Mel Gorman  <mgorman@suse.de>

      [BZ #17195]
      * malloc/arena.c (free): Apply trim threshold to per-thread heaps
        as well as the main arena.

-----------------------------------------------------------------------

Summary of changes:
 ChangeLog      |    6 ++++++
 NEWS           |   12 ++++++------
 malloc/arena.c |   13 ++++++++++---
 3 files changed, 22 insertions(+), 9 deletions(-)

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug malloc/17195] excessive heap trimming in threaded programs even when disabled with mallopt

[siddhesh at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=17195

Siddhesh Poyarekar <siddhesh at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
         Resolution|---                         |FIXED

--- Comment #6 from Siddhesh Poyarekar <siddhesh at redhat dot com> ---
Fixed in master.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug malloc/17195] excessive heap trimming in threaded programs even when disabled with mallopt

[cvs-commit at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=17195

--- Comment #8 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> ---
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GNU C Library master sources".

The branch, master has been updated
       via  e4bc326dbbf7328775fe7dd39de1178821363e0a (commit)
      from  58a3a98d8f3488b659318b1a1c6efc169a6f06bf (commit)

Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.

- Log -----------------------------------------------------------------
https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=e4bc326dbbf7328775fe7dd39de1178821363e0a

commit e4bc326dbbf7328775fe7dd39de1178821363e0a
Author: Carlos O'Donell <carlos@systemhalted.org>
Date:   Wed Oct 7 22:21:36 2015 -0400

    malloc: Consistently apply trim_threshold to all heaps (Bug 17195)

    In the per-thread arenas we apply trim_threshold-based checks
    to the extra space between the pad and the top_area. This isn't
    quite accurate and instead we should be harmonizing with the way
    in which trim_treshold is applied everywhere else like sysrtim
    and _int_free. The trimming check should be based on the size of
    the top chunk and only the size of the top chunk. The following
    patch harmonizes the trimming and make it consistent for the main
    arena and thread arenas.

    In the old code a large padding request might have meant that
    trimming was not triggered. Now trimming is considered first based
    on the chunk, then the pad is subtracted, and the remainder trimmed.
    This is how all the other trimmings operate. I didn't measure the
    performance difference of this change because it corrects what I
    consider to be a behavioural anomaly. We'll need some profile driven
    optimization to make this code better, and even there Ondrej and
    others have better ideas on how to speedup malloc.

    Tested on x86_64 with no regressions. Already reviewed by Siddhesh
    Poyarekar and Mel Gorman here and discussed here:
    https://sourceware.org/ml/libc-alpha/2015-05/msg00002.html

-----------------------------------------------------------------------

Summary of changes:
 ChangeLog      |    6 ++++++
 malloc/arena.c |   10 ++++++++--
 2 files changed, 14 insertions(+), 2 deletions(-)

-- 
You are receiving this mail because:
You are on the CC list for the bug.