diff options
| author | Li Zefan <lizf@cn.fujitsu.com> | 2008-09-13 05:33:09 -0400 |
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2008-09-13 17:41:51 -0400 |
| commit | 024994310e10cd7632f43b6873558820308c1af1 (patch) | |
| tree | 8afc91968962ce0adc322e6763b99df3fc70a41e /Documentation | |
| parent | 4e74339af6a59c061cf02f1ac497766bca1de19a (diff) | |
cpuset: hotplug documentation fix
If all the cpus in a cpuset are offlined, the tasks in it will be moved to
the nearest ancestor with non-empty cpus.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Acked-by: Paul Jackson <pj@sgi.com>
Cc: Paul Menage <menage@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Diffstat (limited to 'Documentation')
| -rw-r--r-- | Documentation/cpusets.txt | 18 |
1 files changed, 10 insertions, 8 deletions
diff --git a/Documentation/cpusets.txt b/Documentation/cpusets.txt index 1f5a924d1e5..47e568a9370 100644 --- a/Documentation/cpusets.txt +++ b/Documentation/cpusets.txt | |||
| @@ -635,14 +635,16 @@ prior 'mems' setting, will not be moved. | |||
| 635 | 635 | ||
| 636 | There is an exception to the above. If hotplug functionality is used | 636 | There is an exception to the above. If hotplug functionality is used |
| 637 | to remove all the CPUs that are currently assigned to a cpuset, | 637 | to remove all the CPUs that are currently assigned to a cpuset, |
| 638 | then the kernel will automatically update the cpus_allowed of all | 638 | then all the tasks in that cpuset will be moved to the nearest ancestor |
| 639 | tasks attached to CPUs in that cpuset to allow all CPUs. When memory | 639 | with non-empty cpus. But the moving of some (or all) tasks might fail if |
| 640 | hotplug functionality for removing Memory Nodes is available, a | 640 | cpuset is bound with another cgroup subsystem which has some restrictions |
| 641 | similar exception is expected to apply there as well. In general, | 641 | on task attaching. In this failing case, those tasks will stay |
| 642 | the kernel prefers to violate cpuset placement, over starving a task | 642 | in the original cpuset, and the kernel will automatically update |
| 643 | that has had all its allowed CPUs or Memory Nodes taken offline. User | 643 | their cpus_allowed to allow all online CPUs. When memory hotplug |
| 644 | code should reconfigure cpusets to only refer to online CPUs and Memory | 644 | functionality for removing Memory Nodes is available, a similar exception |
| 645 | Nodes when using hotplug to add or remove such resources. | 645 | is expected to apply there as well. In general, the kernel prefers to |
| 646 | violate cpuset placement, over starving a task that has had all | ||
| 647 | its allowed CPUs or Memory Nodes taken offline. | ||
| 646 | 648 | ||
| 647 | There is a second exception to the above. GFP_ATOMIC requests are | 649 | There is a second exception to the above. GFP_ATOMIC requests are |
| 648 | kernel internal allocations that must be satisfied, immediately. | 650 | kernel internal allocations that must be satisfied, immediately. |