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

DESCRIPTION

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

VERSIONS

33       getcpu() was added in kernel 2.6.19 for x86_64 and i386.
34

CONFORMING TO

36       getcpu() is Linux specific.
37

NOTES

39       Linux  makes  a  best  effort  to make this call as fast possible.  The
40       intention of getcpu() is to allow programs to make  optimizations  with
41       per-CPU data or for NUMA optimization.
42
43       Glibc  does  not  provide a wrapper for this system call; call it using
44       syscall(2); or use sched_getcpu(3) instead.
45
46       The tcache argument is unused since Linux 2.6.24.  In earlier  kernels,
47       if this argument was non-NULL, then it specified a pointer to a caller-
48       allocated buffer in thread-local storage that was  used  to  provide  a
49       caching  mechanism for getcpu().  Use of the cache could speed getcpu()
50       calls, at the cost that there was a very small chance that the returned
51       information would be out of date.  The caching mechanism was considered
52       to cause problems when migrating threads between CPUs, and so the argu‐
53       ment is now ignored.
54

SEE ALSO

56       mbind(2),   sched_setaffinity(2),   set_mempolicy(2),  sched_getcpu(3),
57       cpuset(7)
58

COLOPHON

60       This page is part of release 3.25 of the Linux  man-pages  project.   A
61       description  of  the project, and information about reporting bugs, can
62       be found at http://www.kernel.org/doc/man-pages/.
63
64
65
66Linux                             2008-06-03                         GETCPU(2)
Impressum