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

2015-11-13 Thread Hugh Dickins
On Tue, 10 Nov 2015, Mike Kravetz wrote: > On 11/09/2015 02:55 PM, Mike Kravetz wrote: > > On 11/08/2015 11:42 PM, Hugh Dickins wrote: > >> On Fri, 30 Oct 2015, Mike Kravetz wrote: > >>> > >>> The 'next = start' code is actually from the original truncate_hugepages > >>> routine. This functionalit

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

2015-11-10 Thread Mike Kravetz
On 11/09/2015 02:55 PM, Mike Kravetz wrote: > On 11/08/2015 11:42 PM, Hugh Dickins wrote: >> On Fri, 30 Oct 2015, Mike Kravetz wrote: >>> >>> The 'next = start' code is actually from the original truncate_hugepages >>> routine. This functionality was combined with that needed for hole punch >>> to

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

2015-11-09 Thread Mike Kravetz
On 11/08/2015 11:42 PM, Hugh Dickins wrote: > On Fri, 30 Oct 2015, Mike Kravetz wrote: >> >> The 'next = start' code is actually from the original truncate_hugepages >> routine. This functionality was combined with that needed for hole punch >> to create remove_inode_hugepages(). >> >> The followi

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

2015-11-08 Thread Hugh Dickins
On Fri, 30 Oct 2015, Mike Kravetz wrote: > > The 'next = start' code is actually from the original truncate_hugepages > routine. This functionality was combined with that needed for hole punch > to create remove_inode_hugepages(). > > The following code was in truncate_hugepages: > > next

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

2015-10-30 Thread Mike Kravetz
On 10/30/2015 09:45 AM, Mike Kravetz wrote: > On 10/29/2015 08:32 PM, Hugh Dickins wrote: >> On Thu, 29 Oct 2015, Mike Kravetz wrote: >> >>> This patch is a combination of: >>> [PATCH v2 4/4] mm/hugetlb: Unmap pages to remove if page fault raced >>> with hole punch and, >>> [PATCH] mm/hugetlb:

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

2015-10-30 Thread Mike Kravetz
On 10/29/2015 08:32 PM, Hugh Dickins wrote: > On Thu, 29 Oct 2015, Mike Kravetz wrote: > >> This patch is a combination of: >> [PATCH v2 4/4] mm/hugetlb: Unmap pages to remove if page fault raced >> with hole punch and, >> [PATCH] mm/hugetlb: i_mmap_lock_write before unmapping in >> rem

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

2015-10-29 Thread Hugh Dickins
On Thu, 29 Oct 2015, Mike Kravetz wrote: > This patch is a combination of: > [PATCH v2 4/4] mm/hugetlb: Unmap pages to remove if page fault raced > with hole punch and, > [PATCH] mm/hugetlb: i_mmap_lock_write before unmapping in > remove_inode_hugepages > This patch can replace the en

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

2015-10-29 Thread Mike Kravetz
This patch is a combination of: [PATCH v2 4/4] mm/hugetlb: Unmap pages to remove if page fault raced with hole punch and, [PATCH] mm/hugetlb: i_mmap_lock_write before unmapping in remove_inode_hugepages This patch can replace the entire series: [PATCH v2 0/4] hugetlbfs fallocate ho