Messages in this thread Patch in this message | | | Date | Tue, 14 Oct 2014 14:48:28 +0300 | From | "Kirill A. Shutemov" <> | Subject | Re: [BUG] mm, thp: khugepaged can't allocate on requested node when confined to a cpuset |
| |
On Wed, Oct 08, 2014 at 02:10:50PM -0500, Alex Thorlton wrote: > Hey everyone, > > I've run into a some frustrating behavior from the khugepaged thread, > that I'm hoping to get sorted out. It appears that if you pin > khugepaged to a cpuset (i.e. node 0),
Why whould you want to pin khugpeaged? Is there a valid use-case? Looks like userspace shoots to its leg.
> and it begins scanning/collapsing pages for a process on a cpuset that > doesn't have any memory nodes in common with kugepaged (i.e. node 1), > then the collapsed pages will all be allocated khugepaged's node (in > this case node 0), clearly breaking the cpuset boundary set up for the > process in question. > > I'm aware that there are some known issues with khugepaged performing > off-node allocations in certain situations, but I believe this is a bit > of a special circumstance since, in this situation, there's no way for > khugepaged to perform an allocation on the desired node. > > The problem really stems from the way that we determine the allowed > memory nodes in get_page_from_freelist. When we call down to > cpuset_zone_allowed_softwall, we check current->mems_allowed to > determine what nodes we're allowed on. In the case of khugepaged, we'll > be making allocations for the mm of the process we're collapsing for, > but we'll be checking the mems_allowed of khugepaged, which can > obviously cause some problems.
Is there a reason why we should respect cpuset limitation for kernel threads?
Should we bypass cpuset for PF_KTHREAD completely?
diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 736d8e1b6381..03a74878ad46 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1960,6 +1960,9 @@ get_page_from_freelist(gfp_t gfp_mask, nodemask_t *nodemask, unsigned int order, zonelist_scan: zonelist_rescan = false; + /* Bypass cpuset limitation if allocate from kernel thread context */ + if (current->flags & PF_KTHREAD) + alloc_flags &= ~ALLOC_CPUSET; /* * Scan zonelist, looking for a zone with enough free. * See also __cpuset_node_allowed_softwall() comment in kernel/cpuset.c. -- Kirill A. Shutemov
| |