From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Liu Date: Mon, 29 Jul 2013 18:50:10 +0800 Subject: [Ocfs2-devel] NULL pointer dereference at ocfs2_dir_foreach_blk_id In-Reply-To: <1394429.Cu7O6gPXEE@o3-3> References: <1394429.Cu7O6gPXEE@o3-3> Message-ID: <51F648E2.8030704@oracle.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ocfs2-devel@oss.oracle.com Hi David, Thanks for your report, could you try the fix below? From: Jie Liu This patch fix an NULL pointer deference while removing an empty directory, which was introduced by commits: commit: 3704412bdbf37ec836152f571ac74fe72220c05a [readdir] convert ocfs2 BUG: unable to handle kernel NULL pointer dereference at (null) IP: [< (null)>] (null) PGD 6da85067 PUD 6da89067 PMD 0 Oops: 0010 [#1] SMP CPU: 0 PID: 6564 Comm: rmdir Tainted: G O 3.11.0-rc1 #4 RIP: 0010:[<0000000000000000>] [< (null)>] (null) Call Trace: [] ? ocfs2_dir_foreach_blk_id+0x17e/0x220 [ocfs2] [] ocfs2_dir_foreach+0x49/0x50 [ocfs2] [] ocfs2_empty_dir+0x12c/0x3e0 [ocfs2] [] ocfs2_unlink+0x56e/0xc10 [ocfs2] [] vfs_rmdir+0xd5/0x140 [] do_rmdir+0x1cb/0x1e0 [] ? lockdep_sys_exit_thunk+0x35/0x67 [] ? trace_hardirqs_on_thunk+0x3a/0x3f [] SyS_rmdir+0x16/0x20 [] system_call_fastpath+0x16/0x1b Code: Bad RIP value. RIP [< (null)>] (null) RSP CR2: 0000000000000000 ---[ end trace dbb276999e4cdc71 ]--- Reported-by: David Weber Cc: Al Viro Signed-off-by: Jie Liu --- fs/ocfs2/dir.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index eb760d8..c91d986 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -2153,10 +2153,12 @@ int ocfs2_empty_dir(struct inode *inode) { int ret; struct ocfs2_empty_dir_priv priv = { - .ctx.actor = ocfs2_empty_dir_filldir + .ctx.actor = ocfs2_empty_dir_filldir, + .ctx.pos = 0, }; - memset(&priv, 0, sizeof(priv)); + memset(&priv + sizeof(struct dir_context), 0, + sizeof(priv) - sizeof(struct dir_context)); if (ocfs2_dir_indexed(inode)) { ret = ocfs2_empty_dir_dx(inode, &priv); -- 1.7.9.5 On 07/25/2013 06:13 PM, David Weber wrote: > Hi, > > we reproducibly run into a NULL pointer dereference in OCFS2 on Linux 3.11.0- > rc2 > It always happens if we try to copy or delete directories. > > The Filesystem was created with: > mkfs.ocfs2 -b 4K -C 1M -J block64 -L kvm-images -T vmstore /dev/drbd0 > > cat /etc/ocfs2/cluster.conf : > cluster: > heartbeat_mode = global > node_count = 2 > name = kvm > > node: > number = 0 > cluster = kvm > ip_port = 7777 > ip_address = 192.168.100.229 > name = dinah > > node: > number = 1 > cluster = kvm > ip_port = 7777 > ip_address = 192.168.100.228 > name = alice > > > dmesg: > [ 42.191816] BUG: unable to handle kernel NULL pointer dereference at (null) > [ 42.192753] IP: [< (null)>] (null) > [ 42.193348] PGD 79c1f9067 PUD 79c38a067 PMD 0 > [ 42.193913] Oops: 0010 [#1] SMP > [ 42.194338] Modules linked in: ebtable_nat ebtables ocfs2_stack_o2cb bridge > stp llc kvm_intel kvm drbd lru_cache dlm sctp libcrc32c ocfs2_dlm ocfs2_dlmfs > ocfs2 ocfs2_stackglue ocfs2_nodemanager configfs e1000e > [ 42.196944] CPU: 1 PID: 2392 Comm: rm Not tainted 3.11.0-rc2 #19 > [ 42.197617] Hardware name: Supermicro X8DT6/X8DT6, BIOS 2.0a 09/14/2010 > [ 42.198389] task: ffff880799d06320 ti: ffff88079c664000 task.ti: ffff88079c664000 > [ 42.199251] RIP: 0010:[<0000000000000000>] [< (null)>] (null) > [ 42.200545] RSP: 0018:ffff88079c665c30 EFLAGS: 00010293 > [ 42.201394] RAX: 0000000000000002 RBX: 0000000000000010 RCX: 0000000000000000 > [ 42.202190] RDX: 0000000000000001 RSI: ffff88079290c0d4 RDI: ffff88079c665ce8 > [ 42.202995] RBP: ffff88079c665ca8 R08: 00000000000ea90e R09: 0000000000000004 > [ 42.203794] R10: 0000000000000000 R11: 0000000000000000 R12: ffff88079290c0c0 > [ 42.204600] R13: ffff88079c665ce8 R14: ffff88079290c0c8 R15: ffff8807960ba598 > [ 42.205406] FS: 00007f4b6d259700(0000) GS:ffff8807a1220000(0000) > knlGS:0000000000000000 > [ 42.206115] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b > [ 42.206453] CR2: 0000000000000000 CR3: 000000079b02a000 CR4: 00000000000007e0 > [ 42.207045] Stack: > [ 42.207266] ffffffffa00d8947 ffff88079c665e18 ffff8807960ba2c0 ffff88079c665c78 > [ 42.208184] ffff88079290c0c0 ffff88079290c000 ffff88079c665cc0 ffff8807960ba598 > [ 42.209089] ffff8807961269c0 ffff88079c665c88 ffff8807960ba598 00000000ffffffd9 > [ 42.210298] Call Trace: > [ 42.210579] [] ? ocfs2_dir_foreach_blk_id+0x169/0x212 [ocfs2] > [ 42.211419] [] ocfs2_dir_foreach+0x3a/0x3e [ocfs2] > [ 42.212145] [] ocfs2_empty_dir+0x148/0x391 [ocfs2] > [ 42.212880] [] ocfs2_unlink+0x567/0xbc3 [ocfs2] > [ 42.213573] [] ? __ocfs2_cluster_unlock.isra.41+0x89/0xbb [ocfs2] > [ 42.214463] [] vfs_rmdir+0xb0/0xfe > [ 42.215039] [] do_rmdir+0x143/0x19b > [ 42.215611] [] ? task_work_run+0x86/0xac > [ 42.216232] [] SyS_unlinkat+0x25/0x27 > [ 42.216818] [] system_call_fastpath+0x16/0x1b > [ 42.217487] Code: Bad RIP value. > [ 42.217901] RIP [< (null)>] (null) > [ 42.218509] RSP > [ 42.218896] CR2: 0000000000000000 > [ 42.219352] ---[ end trace 1c32c45da41ce169 ]--- > > > The dereference happens here: > Reading symbols from /usr/src/linux-3.11-rc2/fs/ocfs2/dir.o...done. > (gdb) list *(ocfs2_dir_foreach_blk_id+0x169) > 0x4497 is in ocfs2_dir_foreach_blk_id (fs/ocfs2/dir.c:1820). > 1815 unsigned char d_type = DT_UNKNOWN; > 1816 > 1817 if (de->file_type < OCFS2_FT_MAX) > 1818 d_type = ocfs2_filetype_table[de->file_type]; > 1819 > 1820 if (!dir_emit(ctx, de->name, de->name_len, > 1821 le64_to_cpu(de->inode), d_type)) > 1822 goto out; > 1823 } > 1824 ctx->pos += le16_to_cpu(de->rec_len); > > Thanks in advance! > > Cheers, > David > > _______________________________________________ > Ocfs2-devel mailing list > Ocfs2-devel at oss.oracle.com > https://oss.oracle.com/mailman/listinfo/ocfs2-devel