]>
Commit | Line | Data |
---|---|---|
80997bc3 MK |
1 | .\" This man page is Copyright (C) 2006 Andi Kleen <ak@muc.de>. |
2 | .\" Permission is granted to distribute possibly modified copies | |
3 | .\" of this page provided the header is included verbatim, | |
4 | .\" and in case of nontrivial modification author and date | |
5 | .\" of the modification is added to the header. | |
6 | .\" 2008, mtk, various edits | |
7 | .TH getcpu 2 2008-06-03 "Linux" "Linux Programmer's Manual" | |
8 | .SH NAME | |
9 | getcpu \- determine CPU and NUMA node on which the calling thread is running | |
10 | .SH SYNOPSIS | |
11 | .nf | |
12 | .B #include <linux/getcpu.h> | |
13 | .sp | |
14 | .BI "int getcpu(unsigned *" cpu ", unsigned *" node \ | |
15 | ", struct getcpu_cache *" tcache ); | |
16 | .fi | |
17 | .SH DESCRIPTION | |
18 | The | |
19 | .BR getcpu () | |
20 | system call identifies the processor and node on which the calling | |
21 | thread or process is currently running and writes them into the | |
22 | integers pointed to by the | |
23 | .I cpu | |
24 | and | |
25 | .I node | |
26 | arguments. | |
27 | The processor is a unique small integer identifying a CPU. | |
28 | The node is a unique small identifier identifying a NUMA node. | |
29 | When either | |
30 | .I cpu | |
31 | or | |
32 | .I node | |
33 | is NULL nothing is written to the respective pointer. | |
34 | ||
35 | The third argument to this system call is nowadays unused. | |
36 | ||
37 | The information placed in | |
38 | .I cpu | |
39 | is only guaranteed to be current at the time of the call: | |
40 | unless the CPU affinity has been fixed using | |
41 | .BR sched_setaffinity (2), | |
42 | the kernel might change the CPU at any time. | |
43 | (Normally this does not happen | |
44 | because the scheduler tries to minimize movements between CPUs to | |
45 | keep caches hot, but it is possible.) | |
46 | The caller must be prepared to handle the situation when | |
47 | .I cpu | |
48 | and | |
49 | .I node | |
50 | are no longer the current CPU and node. | |
51 | .SH VERSIONS | |
52 | .BR getcpu (2) | |
53 | was added in kernel 2.6.19 for x86_64 and i386. | |
54 | .SH NOTES | |
55 | Linux makes a best effort to make this call as fast possible. | |
56 | The intention of | |
57 | .I getcpu(2) | |
58 | is to allow programs to make optimizations with per-CPU data | |
59 | or for NUMA optimization. | |
60 | ||
61 | The | |
62 | .I tcache | |
63 | argument is unused since Linux 2.6.24. | |
64 | .\" commit 4307d1e5ada595c87f9a4d16db16ba5edb70dcb1 | |
65 | .\" Author: Ingo Molnar <mingo@elte.hu> | |
66 | .\" Date: Wed Nov 7 18:37:48 2007 +0100 | |
67 | .\" x86: ignore the sys_getcpu() tcache parameter | |
68 | In earlier kernels, | |
69 | if this argument was non-NULL, | |
70 | then it specified a pointer to a caller-allocated buffer in thread-local | |
71 | storage that was used to provide a caching mechanism for | |
72 | .BR getcpu (). | |
73 | Use of the cache could speed | |
74 | .BR getcpu () | |
75 | calls, at the cost that there was a very small chance that | |
76 | the returned information would be out of date. | |
77 | The caching mechanism was considered to cause problems when | |
78 | migrating threads between CPUs, and so the argument is now ignored. | |
79 | .\" | |
80 | .\" ===== Before kernel 2.6.24: ===== | |
81 | .\" .I tcache | |
82 | .\" is a pointer to a | |
83 | .\" .IR "struct getcpu_cache" | |
84 | .\" that is used as a cache by | |
85 | .\" .BR getcpu (). | |
86 | .\" The caller should put the cache into a thread-local variable | |
87 | .\" if the process is multithreaded, | |
88 | .\" because the cache cannot be shared between different threads. | |
89 | .\" .I tcache | |
90 | .\" can be NULL. | |
91 | .\" If it is not NULL | |
92 | .\" .BR getcpu () | |
93 | .\" will use it to speed up operation. | |
94 | .\" The information inside the cache is private to the system call | |
95 | .\" and should not be accessed by the user program. | |
96 | .\" The information placed in the cache can change between kernel releases. | |
97 | .\" | |
98 | .\" When no cache is specified | |
99 | .\" .BR getcpu () | |
100 | .\" will be slower, | |
101 | .\" but always retrieve the current CPU and node information. | |
102 | .\" With a cache | |
103 | .\" .BR getcpu () | |
104 | .\" is faster. | |
105 | .\" However, the cached information is only updated once per jiffy (see | |
106 | .\" .BR time (7)). | |
107 | .\" This means that the information could theoretically be out of date, | |
108 | .\" although in practice the scheduler's attempt to maintain | |
109 | .\" soft CPU affinity means that the information is unlikely to change | |
110 | .\" over the course of the caching interval. | |
111 | .SH SEE ALSO | |
112 | .\" FIXME(mk) add SEE ALSO entries in other pages, pointing to this page | |
113 | .BR mbind (2), | |
114 | .BR set_mempolicy (2), | |
115 | .BR sched_setaffinity (2) |