From e244ff61fb05bb677f72b9a0f5e3e22b2ca1114e Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Sun, 22 Apr 2018 12:13:41 +0200 Subject: [PATCH] 4.4-stable patches added patches: ext4-bugfix-for-mmaped-pages-in-mpage_release_unused_pages.patch --- ...-pages-in-mpage_release_unused_pages.patch | 72 +++++++++++++++++++ queue-4.4/series | 1 + 2 files changed, 73 insertions(+) create mode 100644 queue-4.4/ext4-bugfix-for-mmaped-pages-in-mpage_release_unused_pages.patch diff --git a/queue-4.4/ext4-bugfix-for-mmaped-pages-in-mpage_release_unused_pages.patch b/queue-4.4/ext4-bugfix-for-mmaped-pages-in-mpage_release_unused_pages.patch new file mode 100644 index 00000000000..43a8b743e6b --- /dev/null +++ b/queue-4.4/ext4-bugfix-for-mmaped-pages-in-mpage_release_unused_pages.patch @@ -0,0 +1,72 @@ +From 4e800c0359d9a53e6bf0ab216954971b2515247f Mon Sep 17 00:00:00 2001 +From: wangguang +Date: Thu, 15 Sep 2016 11:32:46 -0400 +Subject: ext4: bugfix for mmaped pages in mpage_release_unused_pages() +MIME-Version: 1.0 +Content-Type: text/plain; charset=UTF-8 +Content-Transfer-Encoding: 8bit + +From: wangguang + +commit 4e800c0359d9a53e6bf0ab216954971b2515247f upstream. + +Pages clear buffers after ext4 delayed block allocation failed, +However, it does not clean its pte_dirty flag. +if the pages unmap ,in cording to the pte_dirty , +unmap_page_range may try to call __set_page_dirty, + +which may lead to the bugon at +mpage_prepare_extent_to_map:head = page_buffers(page);. + +This patch just call clear_page_dirty_for_io to clean pte_dirty +at mpage_release_unused_pages for pages mmaped. + +Steps to reproduce the bug: + +（1） mmap a file in ext4 + addr = (char *)mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_SHARED, + fd, 0); + memset(addr, 'i', 4096); + +（2） return EIO at + + ext4_writepages->mpage_map_and_submit_extent->mpage_map_one_extent + +which causes this log message to be print: + + ext4_msg(sb, KERN_CRIT, + "Delayed block allocation failed for " + "inode %lu at logical offset %llu with" + " max blocks %u with error %d", + inode->i_ino, + (unsigned long long)map->m_lblk, + (unsigned)map->m_len, -err); + +(3）Unmap the addr cause warning at + + __set_page_dirty:WARN_ON_ONCE(warn && !PageUptodate(page)); + +(4) wait for a minute,then bugon happen. + +Cc: stable@vger.kernel.org +Signed-off-by: wangguang +Signed-off-by: Theodore Ts'o +[@nathanchance: Resolved conflict from lack of 09cbfeaf1a5a6] +Signed-off-by: Nathan Chancellor +Signed-off-by: Greg Kroah-Hartman + +--- + fs/ext4/inode.c | 2 ++ + 1 file changed, 2 insertions(+) + +--- a/fs/ext4/inode.c ++++ b/fs/ext4/inode.c +@@ -1515,6 +1515,8 @@ static void mpage_release_unused_pages(s + BUG_ON(!PageLocked(page)); + BUG_ON(PageWriteback(page)); + if (invalidate) { ++ if (page_mapped(page)) ++ clear_page_dirty_for_io(page); + block_invalidatepage(page, 0, PAGE_CACHE_SIZE); + ClearPageUptodate(page); + } diff --git a/queue-4.4/series b/queue-4.4/series index 81bcbb5f2cb..2d0271516ba 100644 --- a/queue-4.4/series +++ b/queue-4.4/series @@ -92,3 +92,4 @@ don-t-leak-mnt_internal-away-from-internal-mounts.patch autofs-mount-point-create-should-honour-passed-in-mode.patch mm-allow-gfp_-fs-io-for-page_cache_read-page-cache-allocation.patch mm-filemap.c-fix-null-pointer-in-page_cache_tree_insert.patch +ext4-bugfix-for-mmaped-pages-in-mpage_release_unused_pages.patch -- 2.47.3