The patch titled Subject: mm: memory: merge shared-writable dirtying branches in do_wp_page() has been added to the -mm tree. Its filename is mm-memory-merge-shared-writable-dirtying-branches-in-do_wp_page.patch This patch should soon appear at http://ozlabs.org/~akpm/mmots/broken-out/mm-memory-merge-shared-writable-dirtying-branches-in-do_wp_page.patch and later at http://ozlabs.org/~akpm/mmotm/broken-out/mm-memory-merge-shared-writable-dirtying-branches-in-do_wp_page.patch Before you just go and hit "reply", please: a) Consider who else should be cc'ed b) Prefer to cc a suitable mailing list as well c) Ideally: find the original patch on the mailing list and do a reply-to-all to that, adding suitable additional cc's *** Remember to use Documentation/SubmitChecklist when testing your code *** The -mm tree is included into linux-next and is updated there every 3-4 working days ------------------------------------------------------ From: Johannes Weiner <hannes@xxxxxxxxxxx> Subject: mm: memory: merge shared-writable dirtying branches in do_wp_page() Whether there is a vm_ops->page_mkwrite or not, the page dirtying is pretty much the same. Make sure the page references are the same in both cases, then merge the two branches. It's tempting to go even further and page-lock the !page_mkwrite case, to get it in line with everybody else setting the page table and thus further simplify the model. But that's not quite compelling enough to justify dropping the pte lock, then relocking and verifying the entry for filesystems without ->page_mkwrite, which notably includes tmpfs. Leave it for now and lock the page late in the !page_mkwrite case. Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx> Acked-by: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx> Reviewed-by: Jan Kara <jack@xxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> --- mm/memory.c | 48 +++++++++++++++++------------------------------- 1 file changed, 17 insertions(+), 31 deletions(-) diff -puN mm/memory.c~mm-memory-merge-shared-writable-dirtying-branches-in-do_wp_page mm/memory.c --- a/mm/memory.c~mm-memory-merge-shared-writable-dirtying-branches-in-do_wp_page +++ a/mm/memory.c @@ -2033,7 +2033,7 @@ static int do_wp_page(struct mm_struct * pte_t entry; int ret = 0; int page_mkwrite = 0; - struct page *dirty_page = NULL; + bool dirty_shared = false; unsigned long mmun_start = 0; /* For mmu_notifiers */ unsigned long mmun_end = 0; /* For mmu_notifiers */ struct mem_cgroup *memcg; @@ -2084,6 +2084,7 @@ static int do_wp_page(struct mm_struct * unlock_page(old_page); } else if (unlikely((vma->vm_flags & (VM_WRITE|VM_SHARED)) == (VM_WRITE|VM_SHARED))) { + page_cache_get(old_page); /* * Only catch write-faults on shared writable pages, * read-only shared pages can get COWed by @@ -2091,7 +2092,7 @@ static int do_wp_page(struct mm_struct * */ if (vma->vm_ops && vma->vm_ops->page_mkwrite) { int tmp; - page_cache_get(old_page); + pte_unmap_unlock(page_table, ptl); tmp = do_page_mkwrite(vma, old_page, address); if (unlikely(!tmp || (tmp & @@ -2111,11 +2112,10 @@ static int do_wp_page(struct mm_struct * unlock_page(old_page); goto unlock; } - page_mkwrite = 1; } - dirty_page = old_page; - get_page(dirty_page); + + dirty_shared = true; reuse: /* @@ -2134,43 +2134,29 @@ reuse: pte_unmap_unlock(page_table, ptl); ret |= VM_FAULT_WRITE; - if (!dirty_page) - return ret; - - if (!page_mkwrite) { + if (dirty_shared) { struct address_space *mapping; int dirtied; - lock_page(dirty_page); - dirtied = set_page_dirty(dirty_page); - VM_BUG_ON_PAGE(PageAnon(dirty_page), dirty_page); - mapping = dirty_page->mapping; - unlock_page(dirty_page); + if (!page_mkwrite) + lock_page(old_page); - if (dirtied && mapping) { - /* - * Some device drivers do not set page.mapping - * but still dirty their pages - */ - balance_dirty_pages_ratelimited(mapping); - } + dirtied = set_page_dirty(old_page); + VM_BUG_ON_PAGE(PageAnon(old_page), old_page); + mapping = old_page->mapping; + unlock_page(old_page); + page_cache_release(old_page); - file_update_time(vma->vm_file); - } - put_page(dirty_page); - if (page_mkwrite) { - struct address_space *mapping = dirty_page->mapping; - - set_page_dirty(dirty_page); - unlock_page(dirty_page); - page_cache_release(dirty_page); - if (mapping) { + if ((dirtied || page_mkwrite) && mapping) { /* * Some device drivers do not set page.mapping * but still dirty their pages */ balance_dirty_pages_ratelimited(mapping); } + + if (!page_mkwrite) + file_update_time(vma->vm_file); } return ret; _ Patches currently in -mm which might be from hannes@xxxxxxxxxxx are origin.patch mm-page_alloc-embed-oom-killing-naturally-into-allocation-slowpath.patch mm-page_alloc-embed-oom-killing-naturally-into-allocation-slowpath-update.patch mm-page_alloc-embed-oom-killing-naturally-into-allocation-slowpath-fix.patch mm-page_allocc-__alloc_pages_nodemask-dont-alter-arg-gfp_mask.patch mm-page_allocc-__alloc_pages_nodemask-dont-alter-arg-gfp_mask-fix.patch mm-protect-set_page_dirty-from-ongoing-truncation.patch mm-memory-remove-vm_file-check-on-shared-writable-vmas.patch mm-memory-merge-shared-writable-dirtying-branches-in-do_wp_page.patch -- To unsubscribe from this list: send the line "unsubscribe mm-commits" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html