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       #include <linux/getcpu.h>
11
12       int getcpu(unsigned *cpu, unsigned *node, struct getcpu_cache *tcache);
13
14       Note: There is no glibc wrapper for this system call; see NOTES.
15

DESCRIPTION

17       The getcpu() system call identifies the processor and node on which the
18       calling thread or process is currently running and writes them into the
19       integers pointed to by the cpu and node arguments.  The processor is  a
20       unique  small  integer  identifying  a CPU.  The node is a unique small
21       identifier identifying a NUMA node.  When either cpu or  node  is  NULL
22       nothing is written to the respective pointer.
23
24       The  third  argument to this system call is nowadays unused, and should
25       be specified as NULL unless portability to Linux 2.6.23 or  earlier  is
26       required (see NOTES).
27
28       The  information  placed in cpu is guaranteed to be current only at the
29       time of the  call:  unless  the  CPU  affinity  has  been  fixed  using
30       sched_setaffinity(2),  the  kernel  might  change  the CPU at any time.
31       (Normally this does not happen because the scheduler tries to  minimize
32       movements  between  CPUs  to keep caches hot, but it is possible.)  The
33       caller must allow for the possibility that the information returned  in
34       cpu and node is no longer current by the time the call returns.
35

RETURN VALUE

37       On  success, 0 is returned.  On error, -1 is returned, and errno is set
38       appropriately.
39

ERRORS

41       EFAULT Arguments point outside the calling process's address space.
42

VERSIONS

44       getcpu() was added in kernel 2.6.19 for x86-64 and i386.
45

CONFORMING TO

47       getcpu() is Linux-specific.
48

NOTES

50       Linux makes a best effort to make this call as fast as  possible.   (On
51       some architectures, this is done via an implementation in the vdso(7).)
52       The intention of getcpu() is to allow programs  to  make  optimizations
53       with per-CPU data or for NUMA optimization.
54
55       Glibc  does  not  provide a wrapper for this system call; call it using
56       syscall(2); or use sched_getcpu(3) instead.
57
58       The tcache argument is unused since Linux 2.6.24.  In earlier  kernels,
59       if this argument was non-NULL, then it specified a pointer to a caller-
60       allocated buffer in thread-local storage that was  used  to  provide  a
61       caching  mechanism for getcpu().  Use of the cache could speed getcpu()
62       calls, at the cost that there was a very small chance that the returned
63       information would be out of date.  The caching mechanism was considered
64       to cause problems when migrating threads between CPUs, and so the argu‐
65       ment is now ignored.
66

SEE ALSO

68       mbind(2),   sched_setaffinity(2),   set_mempolicy(2),  sched_getcpu(3),
69       cpuset(7), vdso(7)
70

COLOPHON

72       This page is part of release 4.15 of the Linux  man-pages  project.   A
73       description  of  the project, information about reporting bugs, and the
74       latest    version    of    this    page,    can     be     found     at
75       https://www.kernel.org/doc/man-pages/.
76
77
78
79Linux                             2017-09-15                         GETCPU(2)
Impressum