[PATCH 3.10 027/268] ocfs2: fix crash caused by stale lvb with fsdlm plugin

From: Willy Tarreau
Date: Mon Jun 19 2017 - 14:40:50 EST

From: Eric Ren <zren@xxxxxxxx>

commit e7ee2c089e94067d68475990bdeed211c8852917 upstream.

The crash happens rather often when we reset some cluster nodes while
nodes contend fiercely to do truncate and append.

The crash backtrace is below:

dlm: C21CBDA5E0774F4BA5A9D4F317717495: dlm_recover_grant 1 locks on 971 resources
dlm: C21CBDA5E0774F4BA5A9D4F317717495: dlm_recover 9 generation 5 done: 4 ms
ocfs2: Begin replay journal (node 318952601, slot 2) on device (253,18)
ocfs2: End replay journal (node 318952601, slot 2) on device (253,18)
ocfs2: Beginning quota recovery on device (253,18) for slot 2
ocfs2: Finishing quota recovery on device (253,18) for slot 2
(truncate,30154,1):ocfs2_truncate_file:470 ERROR: bug expression: le64_to_cpu(fe->i_size) != i_size_read(inode)
(truncate,30154,1):ocfs2_truncate_file:470 ERROR: Inode 290321, inode i_size = 732 != di i_size = 937, i_flags = 0x1
------------[ cut here ]------------
kernel BUG at /usr/src/linux/fs/ocfs2/file.c:470!
invalid opcode: 0000 [#1] SMP
Modules linked in: ocfs2_stack_user(OEN) ocfs2(OEN) ocfs2_nodemanager ocfs2_stackglue(OEN) quota_tree dlm(OEN) configfs fuse sd_mod iscsi_tcp libiscsi_tcp libiscsi scsi_transport_iscsi af_packet iscsi_ibft iscsi_boot_sysfs softdog xfs libcrc32c ppdev parport_pc pcspkr parport joydev virtio_balloon virtio_net i2c_piix4 acpi_cpufreq button processor ext4 crc16 jbd2 mbcache ata_generic cirrus virtio_blk ata_piix drm_kms_helper ahci syscopyarea libahci sysfillrect sysimgblt fb_sys_fops ttm floppy libata drm virtio_pci virtio_ring uhci_hcd virtio ehci_hcd usbcore serio_raw usb_common sg dm_multipath dm_mod scsi_dh_rdac scsi_dh_emc scsi_dh_alua scsi_mod autofs4
Supported: No, Unsupported modules are loaded
CPU: 1 PID: 30154 Comm: truncate Tainted: G OE N 4.4.21-69-default #1
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.8.1-0-g4adadbd-20151112_172657-sheep25 04/01/2014
task: ffff88004ff6d240 ti: ffff880074e68000 task.ti: ffff880074e68000
RIP: 0010:[<ffffffffa05c8c30>] [<ffffffffa05c8c30>] ocfs2_truncate_file+0x640/0x6c0 [ocfs2]
RSP: 0018:ffff880074e6bd50 EFLAGS: 00010282
RAX: 0000000000000074 RBX: 000000000000029e RCX: 0000000000000000
RDX: 0000000000000001 RSI: 0000000000000246 RDI: 0000000000000246
RBP: ffff880074e6bda8 R08: 000000003675dc7a R09: ffffffff82013414
R10: 0000000000034c50 R11: 0000000000000000 R12: ffff88003aab3448
R13: 00000000000002dc R14: 0000000000046e11 R15: 0000000000000020
FS: 00007f839f965700(0000) GS:ffff88007fc80000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 00007f839f97e000 CR3: 0000000036723000 CR4: 00000000000006e0
Call Trace:
ocfs2_setattr+0x698/0xa90 [ocfs2]
notify_change+0x1ae/0x380
do_truncate+0x5e/0x90
do_sys_ftruncate.constprop.11+0x108/0x160
entry_SYSCALL_64_fastpath+0x12/0x6d
Code: 24 28 ba d6 01 00 00 48 c7 c6 30 43 62 a0 8b 41 2c 89 44 24 08 48 8b 41 20 48 c7 c1 78 a3 62 a0 48 89 04 24 31 c0 e8 a0 97 f9 ff <0f> 0b 3d 00 fe ff ff 0f 84 ab fd ff ff 83 f8 fc 0f 84 a2 fd ff
RIP [<ffffffffa05c8c30>] ocfs2_truncate_file+0x640/0x6c0 [ocfs2]

It's because ocfs2_inode_lock() get us stale LVB in which the i_size is
not equal to the disk i_size. We mistakenly trust the LVB because the
underlaying fsdlm dlm_lock() doesn't set lkb_sbflags with
DLM_SBF_VALNOTVALID properly for us. But, why?

The current code tries to downconvert lock without DLM_LKF_VALBLK flag
to tell o2cb don't update RSB's LVB if it's a PR->NULL conversion, even
if the lock resource type needs LVB. This is not the right way for
fsdlm.

The fsdlm plugin behaves different on DLM_LKF_VALBLK, it depends on
DLM_LKF_VALBLK to decide if we care about the LVB in the LKB. If
DLM_LKF_VALBLK is not set, fsdlm will skip recovering RSB's LVB from
this lkb and set the right DLM_SBF_VALNOTVALID appropriately when node
failure happens.

The following diagram briefly illustrates how this crash happens:

RSB1 is inode metadata lock resource with LOCK_TYPE_USES_LVB;

The 1st round:

Node1 Node2
RSB1: PR
RSB1(master): NULL->EX
ocfs2_downconvert_lock(PR->NULL, set_lvb==0)
ocfs2_dlm_lock(no DLM_LKF_VALBLK)

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

dlm_lock(no DLM_LKF_VALBLK)
convert_lock(overwrite lkb->lkb_exflags
with no DLM_LKF_VALBLK)

RSB1: NULL RSB1: EX
reset Node2
dlm_recover_rsbs()
recover_lvb()

/* The LVB is not trustable if the node with EX fails and
* no lock >= PR is left. We should set RSB_VALNOTVALID for RSB1.
*/

if(!(kb_exflags & DLM_LKF_VALBLK)) /* This means we miss the chance to
return; * to invalid the LVB here.
*/

The 2nd round:

Node 1 Node2
RSB1(become master from recovery)

ocfs2_setattr()
ocfs2_inode_lock(NULL->EX)
/* dlm_lock() return the stale lvb without setting DLM_SBF_VALNOTVALID */
ocfs2_meta_lvb_is_trustable() return 1 /* so we don't refresh inode from disk */
ocfs2_truncate_file()
mlog_bug_on_msg(disk isize != i_size_read(inode)) /* crash! */

The fix is quite straightforward. We keep to set DLM_LKF_VALBLK flag
for dlm_lock() if the lock resource type needs LVB and the fsdlm plugin
is uesed.

Link: http://lkml.kernel.org/r/1481275846-6604-1-git-send-email-zren@xxxxxxxx
Signed-off-by: Eric Ren <zren@xxxxxxxx>
Reviewed-by: Joseph Qi <jiangqi903@xxxxxxxxx>
Cc: Mark Fasheh <mfasheh@xxxxxxxxxxx>
Cc: Joel Becker <jlbec@xxxxxxxxxxxx>
Cc: Junxiao Bi <junxiao.bi@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Signed-off-by: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Signed-off-by: Willy Tarreau <w@xxxxxx>
---
fs/ocfs2/dlmglue.c | 10 ++++++++++
fs/ocfs2/stackglue.c | 6 ++++++
fs/ocfs2/stackglue.h | 3 +++
3 files changed, 19 insertions(+)

diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c
index b294deb..cf88dd4 100644
--- a/fs/ocfs2/dlmglue.c
+++ b/fs/ocfs2/dlmglue.c
@@ -3264,6 +3264,16 @@ static int ocfs2_downconvert_lock(struct ocfs2_super *osb,
mlog(ML_BASTS, "lockres %s, level %d => %d\n", lockres->l_name,
lockres->l_level, new_level);

+ /*
+ * On DLM_LKF_VALBLK, fsdlm behaves differently with o2cb. It always
+ * expects DLM_LKF_VALBLK being set if the LKB has LVB, so that
+ * we can recover correctly from node failure. Otherwise, we may get
+ * invalid LVB in LKB, but without DLM_SBF_VALNOTVALIDÂbeing set.
+ */
+ if (!ocfs2_is_o2cb_active() &&
+ lockres->l_ops->flags & LOCK_TYPE_USES_LVB)
+ lvb = 1;
+
if (lvb)
dlm_flags |= DLM_LKF_VALBLK;

diff --git a/fs/ocfs2/stackglue.c b/fs/ocfs2/stackglue.c
index 39abf89..88610b3 100644
--- a/fs/ocfs2/stackglue.c
+++ b/fs/ocfs2/stackglue.c
@@ -48,6 +48,12 @@ static char ocfs2_hb_ctl_path[OCFS2_MAX_HB_CTL_PATH] = "/sbin/ocfs2_hb_ctl";
*/
static struct ocfs2_stack_plugin *active_stack;

+inline int ocfs2_is_o2cb_active(void)
+{
+ return !strcmp(active_stack->sp_name, OCFS2_STACK_PLUGIN_O2CB);
+}
+EXPORT_SYMBOL_GPL(ocfs2_is_o2cb_active);
+
static struct ocfs2_stack_plugin *ocfs2_stack_lookup(const char *name)
{
struct ocfs2_stack_plugin *p;
diff --git a/fs/ocfs2/stackglue.h b/fs/ocfs2/stackglue.h
index 1ec56fd..fa49d8a 100644
--- a/fs/ocfs2/stackglue.h
+++ b/fs/ocfs2/stackglue.h
@@ -289,4 +289,7 @@ void ocfs2_stack_glue_set_max_proto_version(struct ocfs2_protocol_version *max_p
int ocfs2_stack_glue_register(struct ocfs2_stack_plugin *plugin);
void ocfs2_stack_glue_unregister(struct ocfs2_stack_plugin *plugin);

+/* In ocfs2_downconvert_lock(), we need to know which stack we are using */
+int ocfs2_is_o2cb_active(void);
+
#endif /* STACKGLUE_H */
--
2.8.0.rc2.1.gbe9624a