From mboxrd@z Thu Jan 1 00:00:00 1970 From: Lee Schermerhorn Date: Fri, 22 Aug 2008 17:11:00 -0400 Message-Id: <20080822211100.29898.60644.sendpatchset@murky.usa.hp.com> In-Reply-To: <20080822211028.29898.82599.sendpatchset@murky.usa.hp.com> References: <20080822211028.29898.82599.sendpatchset@murky.usa.hp.com> Subject: [PATCH 5/7] Mlock: update locked_vm on munmap() of mlocked() region. Sender: owner-linux-mm@kvack.org Return-Path: To: akpm@linux-foundation.org Cc: riel@redhat.com, linux-mm , kosaki.motohiro@jp.fujitsu.com, Eric.Whitney@hp.com List-ID: against patch: mmap-handle-mlocked-pages-during-map-remap-unmap.patch munlock_vma_pages_range() clears VM_LOCKED for munlock_vma_page(), et al to work. This causes remove_vma_list(), called from do_munmap(), to skip updating locked_vm. We don't want to restore the VM_LOCKED in munlock_vma_pages_range() because the pages are still on the lru. If vmscan attempts to reclaim any of these pages before we get a chance to unmap them, try_to_un{lock|map}() may mlock them again. This will result in freeing an mlocked page. Add comment block to munlock_vma_pages_range() to explain this to future would be callers. Move the accounting of locked_vm from remove_vma_list() to the munlock loop in do_munmap(). This is where the pages are munlocked and VM_LOCKED is cleared. Note that remove_vma_list() is a helper function for do_munmap(), called only from there. Signed-off-by: Lee Schermerhorn mm/mlock.c | 17 ++++++++++++++++- mm/mmap.c | 6 +++--- 2 files changed, 19 insertions(+), 4 deletions(-) Index: linux-2.6.27-rc4-mmotm/mm/mlock.c =================================================================== --- linux-2.6.27-rc4-mmotm.orig/mm/mlock.c 2008-08-21 12:04:06.000000000 -0400 +++ linux-2.6.27-rc4-mmotm/mm/mlock.c 2008-08-22 15:43:39.000000000 -0400 @@ -305,7 +305,22 @@ no_mlock: /* - * munlock all pages in the vma range. For mremap(), munmap() and exit(). + * munlock_vma_pages_range() - munlock all pages in the vma range.' + * @vma - vma containing range to be munlock()ed. + * @start - start address in @vma of the range + * @end - end of range in @vma. + * + * For mremap(), munmap() and exit(). + * + * Called with @vma VM_LOCKED. + * + * Returns with VM_LOCKED cleared. Callers must be prepared to + * deal with this. + * + * We don't save and restore VM_LOCKED here because pages are + * still on lru. In unmap path, pages might be scanned by reclaim + * and re-mlocked by try_to_{munlock|unmap} before we unmap and + * free them. This will result in freeing mlocked pages. */ void munlock_vma_pages_range(struct vm_area_struct *vma, unsigned long start, unsigned long end) Index: linux-2.6.27-rc4-mmotm/mm/mmap.c =================================================================== --- linux-2.6.27-rc4-mmotm.orig/mm/mmap.c 2008-08-22 09:19:10.000000000 -0400 +++ linux-2.6.27-rc4-mmotm/mm/mmap.c 2008-08-22 15:21:40.000000000 -0400 @@ -1752,8 +1752,6 @@ static void remove_vma_list(struct mm_st long nrpages = vma_pages(vma); mm->total_vm -= nrpages; - if (vma->vm_flags & VM_LOCKED) - mm->locked_vm -= nrpages; vm_stat_account(mm, vma->vm_flags, vma->vm_file, -nrpages); vma = remove_vma(vma); } while (vma); @@ -1924,8 +1922,10 @@ int do_munmap(struct mm_struct *mm, unsi if (mm->locked_vm) { struct vm_area_struct *tmp = vma; while (tmp && tmp->vm_start < end) { - if (tmp->vm_flags & VM_LOCKED) + if (tmp->vm_flags & VM_LOCKED) { + mm->locked_vm -= vma_pages(tmp); munlock_vma_pages_all(tmp); + } tmp = tmp->vm_next; } } -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org