1GETCPU(2)                  Linux Programmer's Manual                 GETCPU(2)
2
3
4

NAME

6       getcpu  -  determine  CPU  and NUMA node on which the calling thread is
7       running
8

SYNOPSIS

10       #define _GNU_SOURCE             /* See feature_test_macros(7) */
11       #include <sched.h>
12
13       int getcpu(unsigned int *cpu, unsigned int *node);
14

DESCRIPTION

16       The getcpu() system call identifies the processor and node on which the
17       calling thread or process is currently running and writes them into the
18       integers pointed to by the cpu and node arguments.  The processor is  a
19       unique  small  integer  identifying  a CPU.  The node is a unique small
20       identifier identifying a NUMA node.  When either cpu or  node  is  NULL
21       nothing is written to the respective pointer.
22
23       The  information  placed in cpu is guaranteed to be current only at the
24       time of the  call:  unless  the  CPU  affinity  has  been  fixed  using
25       sched_setaffinity(2),  the  kernel  might  change  the CPU at any time.
26       (Normally this does not happen because the scheduler tries to  minimize
27       movements  between  CPUs  to keep caches hot, but it is possible.)  The
28       caller must allow for the possibility that the information returned  in
29       cpu and node is no longer current by the time the call returns.
30

RETURN VALUE

32       On  success, 0 is returned.  On error, -1 is returned, and errno is set
33       to indicate the error.
34

ERRORS

36       EFAULT Arguments point outside the calling process's address space.
37

VERSIONS

39       getcpu() was added in kernel 2.6.19 for x86-64 and i386.  Library  sup‐
40       port  was added in glibc 2.29 (Earlier glibc versions did not provide a
41       wrapper for this system call, necessitating the use of syscall(2).)
42

CONFORMING TO

44       getcpu() is Linux-specific.
45

NOTES

47       Linux makes a best effort to make this call as fast as  possible.   (On
48       some architectures, this is done via an implementation in the vdso(7).)
49       The intention of getcpu() is to allow programs  to  make  optimizations
50       with per-CPU data or for NUMA optimization.
51
52   C library/kernel differences
53       The kernel system call has a third argument:
54
55           int getcpu(unsigned int *cpu, unsigned int *node,
56                      struct getcpu_cache *tcache);
57
58       The  tcache  argument  is unused since Linux 2.6.24, and (when invoking
59       the system call directly) should be specified as NULL, unless portabil‐
60       ity to Linux 2.6.23 or earlier is required.
61
62       In  Linux 2.6.23 and earlier, if the tcache argument was non-NULL, then
63       it specified a pointer to a  caller-allocated  buffer  in  thread-local
64       storage that was used to provide a caching mechanism for getcpu().  Use
65       of the cache could speed getcpu() calls, at the cost that there  was  a
66       very  small  chance that the returned information would be out of date.
67       The caching mechanism was considered to cause problems  when  migrating
68       threads between CPUs, and so the argument is now ignored.
69

SEE ALSO

71       mbind(2),   sched_setaffinity(2),   set_mempolicy(2),  sched_getcpu(3),
72       cpuset(7), vdso(7)
73

COLOPHON

75       This page is part of release 5.12 of the Linux  man-pages  project.   A
76       description  of  the project, information about reporting bugs, and the
77       latest    version    of    this    page,    can     be     found     at
78       https://www.kernel.org/doc/man-pages/.
79
80
81
82Linux                             2021-03-22                         GETCPU(2)
Impressum