提交 · ad8100e0d20e0123def9f83c040b68c96c8638f0 · openeuler / Kernel

27 6月, 2006 40 次提交

ocfs2: use GFP_NOFS in some dlm operations · ad8100e0

由 Kurt Hackel 提交于 5月 01, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

ad8100e0

ocfs2: wait for recovery when starting lock mastery · b7084ab5

由 Kurt Hackel 提交于 5月 01, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

b7084ab5

K
ocfs2: continue recovery when a dead node is encountered · c27069e6
由 Kurt Hackel 提交于 5月 01, 2006
```
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>
```
c27069e6
K
ocfs2: remove unneccesary spin_unlock() in dlm_remaster_locks() · 67a18741
由 Kurt Hackel 提交于 5月 01, 2006
```
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>
```
67a18741

ocfs2: dlm_remaster_locks() should never exit without completing · 6a413211

由 Kurt Hackel 提交于 5月 01, 2006

We cannot restart recovery. Once we begin to recover a node, keep the state
of the recovery intact and follow through, regardless of any other node
deaths that may occur.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

6a413211

ocfs2: special case recovery lock in dlmlock_remote() · c8df412e

由 Kurt Hackel 提交于 5月 01, 2006

If the previous master of the recovery lock dies, let calc_usage take it
down completely and let the caller completely redo the dlmlock() call.
Otherwise, there will never be an opportunity to re-master the lockres and
recovery wont be able to progress.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

c8df412e

ocfs2: pending mastery asserts and migrations should block each other · 36407488

由 Kurt Hackel 提交于 5月 01, 2006

Use the existing structure for blocking migrations when ASTs are pending to
achieve the same result. If we can catch the assert before it goes on the
wire, just cancel it and let the migration continue.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

36407488

ocfs2: temporarily disable automatic lock migration · c87a9ae7

由 Kurt Hackel 提交于 5月 01, 2006

Now we never change the owner of a lock resource until unmount or node
death. This will be re-enabled once some issues in the algorithm used have
been resolved.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

c87a9ae7

ocfs2: do not unconditionally purge the lockres in dlmlock_remote() · 2abaf97e

由 Kurt Hackel 提交于 5月 01, 2006

In dlmlock_remote(), do not call purge_lockres until the lock resource
actually changes. otherwise, the mastery info on the lockres will go away
underneath the caller.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

2abaf97e

ocfs2: increase backoff before waiting for recovery · aa087b84

由 Kurt Hackel 提交于 5月 01, 2006

When mastering non-recovery lock resources, additional time was frequently
needed to allow the disk heartbeat to catch up with the network timeout. the
recovery lock resource is time critical and avoids this path.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

aa087b84

ocfs2: have dlm_pre_master_reco_lockres() ignore dead nodes · f42a100b

由 Kurt Hackel 提交于 5月 01, 2006

Recovery will spin in dlm_pre_master_reco_lockres if we do not ignore
timed-out network responses from dead nodes.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

f42a100b

K
ocfs2: give the dlm dirty list a reference on the lockres · 6ff06a93
由 Kurt Hackel 提交于 5月 01, 2006
```
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>
```
6ff06a93

ocfs2: teach dlm_restart_lock_mastery() to wait on recovery · e7e69eb3

由 Kurt Hackel 提交于 5月 01, 2006

Change behavior of dlm_restart_lock_mastery() when a node goes down.  Dump
all responses that have been collected and start over.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

e7e69eb3

ocfs2: gracefully handle stale create_lock messages. · e4eb0368

由 Kurt Hackel 提交于 5月 01, 2006

This is an error on the sending side, so gracefully error out on the
receiving end.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

e4eb0368

ocfs2: update lvb immediately during recovery · ccd8b1f9

由 Kurt Hackel 提交于 5月 01, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

ccd8b1f9

ocfs2: do not send master requests to localhost · 588e0090

由 Kurt Hackel 提交于 5月 01, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

588e0090

ocfs2: purge lockres' sooner · 8b219809

由 Kurt Hackel 提交于 5月 01, 2006

Immediately purge a lockress that the local node is not the master of.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

8b219809

ocfs2: dump mismatching migrated lvbs before BUG() · 343e26a4

由 Kurt Hackel 提交于 5月 01, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

343e26a4

ocfs2: make dlm recovery finalization 2 stage · 466d1a45

由 Kurt Hackel 提交于 5月 01, 2006

Makes it easier for the recovery process to deal with node death.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

466d1a45

ocfs2: dlm recovery / lockres reference count fix · 69d72b06

由 Kurt Hackel 提交于 5月 01, 2006

Take a reference on lockres structures while they are on the recovery list.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

69d72b06

ocfs2: better error handling during assert master message · a9ee4c8a

由 Kurt Hackel 提交于 4月 27, 2006

handle errors during lock assert master by either killing self or other node
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

a9ee4c8a

K
ocfs2: dump lockres info before we BUG() on a bad reference · a7f90d83
由 Kurt Hackel 提交于 4月 27, 2006
```
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>
```
a7f90d83

ocfs2: do LVB puts in place · c0a8520c

由 Mark Fasheh 提交于 4月 27, 2006

Don't wait until the AST will be fired to do the LVB copy into the lock
resource.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

c0a8520c

ocfs2: mle ref count debugging · aa852354

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

aa852354

ocfs2: allow for an assert message during lock mastery · dc2ed195

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

dc2ed195

ocfs2: take mle reference during migration · 2d1a868c

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

2d1a868c

ocfs2: properly initialize the mle structure · 41b8c8a1

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

41b8c8a1

ocfs2: detach mle from heartbeat events · da01ad05

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

da01ad05

ocfs2: mle ref counting fixes · a2bf0477

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

a2bf0477

ocfs2: better mle debugging · 95883719

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

95883719

ocfs2: clean up recovery related messages · d6dea6e9

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

d6dea6e9

ocfs2: handle network errors during recovery · 29c0fa0f

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

29c0fa0f

ocfs2: only recover one dead node at a time · c3187ce5

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

c3187ce5

ocfs2: Better tracking for recovery state changes · ab27eb6f

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

ab27eb6f

ocfs2: Fix empty lvb check · 8bc674cb

由 Kurt Hackel 提交于 4月 27, 2006

The check for an empty lvb should check the entire buffer not just the first
byte.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

8bc674cb

ocfs2: fix inverted logic in dlm_is_node_dead · aba9aac7

由 Kurt Hackel 提交于 4月 27, 2006

Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

aba9aac7

ocfs2: recheck lockres master before sending an unlock request. · 2580a580

由 Kurt Hackel 提交于 4月 27, 2006

Recovery may have happened and it may now be mastered locally.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

2580a580

ocfs2: add a small delay after a failed migration · 8d79d088

由 Kurt Hackel 提交于 4月 27, 2006

Otherwise we risk starving other threads.
Signed-off-by: NKurt Hackel <kurt.hackel@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

8d79d088

M
ocfs2: silence a compile warning in dlm_alloc_pagevec() · 685f1adb
由 Mark Fasheh 提交于 3月 23, 2006
```
Reported by Andrew Morton.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>
```
685f1adb

[PATCH] ocfs2: Alloc at least a page for the DLM hash · c8f33b6e

由 Joel Becker 提交于 3月 16, 2006

The OCFS2 DLM allocates a number of pages for a hash to lookup locks.
There was a bug where a PAGE_SIZE bigger than the hash size (eg, 64K
pages) would result in zero pages allocated.
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

c8f33b6e

openeuler / Kernel 1 年多 前同步成功

openeuler / Kernel
1 年多前同步成功