Re: [PATCH] mm/hugetlbfs: Unmap pages if page fault raced with hole punch

2016-01-07 Thread Mike Kravetz
On 01/07/2016 12:06 AM, Hillf Danton wrote: >> >> Page faults can race with fallocate hole punch. If a page fault happens >> between the unmap and remove operations, the page is not removed and >> remains within the hole. This is not the desired behavior. The race >> is difficult to detect in

Re: [PATCH] mm/hugetlbfs: Unmap pages if page fault raced with hole punch

2016-01-07 Thread Hillf Danton
> > Page faults can race with fallocate hole punch. If a page fault happens > between the unmap and remove operations, the page is not removed and > remains within the hole. This is not the desired behavior. The race > is difficult to detect in user level code as even in the non-race > case, a

Re: [PATCH] mm/hugetlbfs: Unmap pages if page fault raced with hole punch

2016-01-07 Thread Hillf Danton
> > Page faults can race with fallocate hole punch. If a page fault happens > between the unmap and remove operations, the page is not removed and > remains within the hole. This is not the desired behavior. The race > is difficult to detect in user level code as even in the non-race > case, a

Re: [PATCH] mm/hugetlbfs: Unmap pages if page fault raced with hole punch

2016-01-07 Thread Mike Kravetz
On 01/07/2016 12:06 AM, Hillf Danton wrote: >> >> Page faults can race with fallocate hole punch. If a page fault happens >> between the unmap and remove operations, the page is not removed and >> remains within the hole. This is not the desired behavior. The race >> is difficult to detect in

[PATCH] mm/hugetlbfs: Unmap pages if page fault raced with hole punch

2016-01-06 Thread Mike Kravetz
Page faults can race with fallocate hole punch. If a page fault happens between the unmap and remove operations, the page is not removed and remains within the hole. This is not the desired behavior. The race is difficult to detect in user level code as even in the non-race case, a page within

[PATCH] mm/hugetlbfs: Unmap pages if page fault raced with hole punch

2016-01-06 Thread Mike Kravetz
Page faults can race with fallocate hole punch. If a page fault happens between the unmap and remove operations, the page is not removed and remains within the hole. This is not the desired behavior. The race is difficult to detect in user level code as even in the non-race case, a page within