man5/slabinfo.5

   1 .\" Copyright (c) 2001 Andreas Dilger (adilger@turbolinux.com)
   2 .\"
   3 .\" Permission is granted to make and distribute verbatim copies of this
   4 .\" manual provided the copyright notice and this permission notice are
   5 .\" preserved on all copies.
   6 .\"
   7 .\" Permission is granted to copy and distribute modified versions of this
   8 .\" manual under the conditions for verbatim copying, provided that the
   9 .\" entire resulting derived work is distributed under the terms of a
  10 .\" permission notice identical to this one.
  11 .\"
  12 .\" Since the Linux kernel and libraries are constantly changing, this
  13 .\" manual page may be incorrect or out-of-date.  The author(s) assume no
  14 .\" responsibility for errors or omissions, or for damages resulting from
  15 .\" the use of the information contained herein.  The author(s) may not
  16 .\" have taken the same level of care in the production of this manual,
  17 .\" which is licensed free of charge, as they might when working
  18 .\" professionally.
  19 .\"
  20 .\" Formatted or processed versions of this manual, if unaccompanied by
  21 .\" the source, must acknowledge the copyright and authors of this work.
  22 .\"
  23 .TH SLABINFO 5 2001-06-19 "" "Linux manual"
  24 .SH NAME
  25 /proc/slabinfo \- Kernel slab allocator statistics
  26 .SH SYNOPSIS
  27 .B cat /proc/slabinfo
  28 .SH DESCRIPTION
  29 Frequently used objects in the Linux kernel
  30 (buffer heads, inodes, dentries, etc.)
  31 have their own cache. The file
  32 .I /proc/slabinfo
  33 gives statistics. For example:
  34 .LP
  35 .RS
  36 .nf
  37 % cat /proc/slabinfo
  38 slabinfo \- version: 1.1
  39 kmem_cache            60     78    100    2    2    1
  40 blkdev_requests     5120   5120     96  128  128    1
  41 mnt_cache             20     40     96    1    1    1
  42 inode_cache         7005  14792    480 1598 1849    1
  43 dentry_cache        5469   5880    128  183  196    1
  44 filp                 726    760     96   19   19    1
  45 buffer_head        67131  71240     96 1776 1781    1
  46 vm_area_struct      1204   1652     64   23   28    1
  47 \&...
  48 size-8192              1     17   8192    1   17    2
  49 size-4096             41     73   4096   41   73    1
  50 \&...
  51 .fi
  52 .RE
  53 .LP
  54 For each slab cache, the cache name, the number of currently
  55 active objects, the total number of available objects, the
  56 size of each object in bytes, the number of pages with at
  57 least one active object, the total number of allocated pages,
  58 and the number of pages per slab are given.
  59
  60 Note that because of object alignment and slab cache overhead,
  61 objects are not normally packed tightly into pages.  Pages with
  62 even one in-use object are considered in-use and cannot be
  63 freed.
  64
  65 Kernels compiled with slab cache statistics will also have
  66 "(statistics)" in the first line of output, and will have 5
  67 additional columns, namely: the high water mark of active
  68 objects; the number of times objects have been allocated;
  69 the number of times the cache has grown (new pages added
  70 to this cache); the number of times the cache has been
  71 reaped (unused pages removed from this cache); and the
  72 number of times there was an error allocating new pages
  73 to this cache.  If slab cache statistics are not enabled
  74 for this kernel, these columns will not be shown.
  75
  76 SMP systems will also have "(SMP)" in the first line of
  77 output, and will have two additional columns for each slab,
  78 reporting the slab allocation policy for the CPU-local
  79 cache (to reduce the need for inter-CPU synchronization
  80 when allocating objects from the cache).  The first column
  81 is the per-CPU limit: the maximum number of objects that
  82 will be cached for each CPU.  The second column is the
  83 batchcount: the maximum number of free objects in the
  84 global cache that will be transferred to the per-CPU cache
  85 if it is empty, or the number of objects to be returned
  86 to the global cache if the per-CPU cache is full.
  87
  88 If both slab cache statistics and SMP are defined, there
  89 will be four additional columns, reporting the per-CPU
  90 cache statistics.  The first two are the per-CPU cache
  91 allocation hit and miss counts: the number of times an
  92 object was or was not available in the per-CPU cache
  93 for allocation.  The next two are the per-CPU cache free
  94 hit and miss counts: the number of times a freed object
  95 could or could not fit within the per-CPU cache limit,
  96 before flushing objects to the global cache.
  97
  98 It is possible to tune the SMP per-CPU slab cache limit
  99 and batchcount via:
 100
 101 .nf
 102 echo "\fIcache_name limit batchcount\fP" > /proc/slabinfo
 103 .fi
 104
 105 .SH AVAILABILITY
 106 .I /proc/slabinfo
 107 exists since Linux 2.1.23.
 108 SMP per-CPU caches exist since Linux 2.4.0-test3.
 109
 110 .SH FILES
 111 .I <linux/slab.h>