Details
-
Bug
-
Resolution: Fixed
-
Major
-
Lustre 2.7.0
-
None
-
el6.6 kernels
-
3
-
17328
Description
new warnings occasionally seen in ldiskfs due to checking code new in upstream ext4 code for el6.6.
stack traces look like ldiskfs_truncate() ..... -> ldiskfs_flush_unwritten_io()
lustre calls some ext4/ldiskfs entry points without locking i_mutex. It must do so in order to avoid deadlocks. This triggers a WARN_ON_ONCE() check that is in el6.6 and newer kernels. ldiskfs patch series for future kernels took care of this, but the patch series for el6.6 did not. Need to add a patch to remove the check for ldiskfs on el6.6
Attachments
Activity
| Link | New: This issue is duplicated by ISBC-26 [ ISBC-26 ] |
| Link | New: This issue is duplicated by DDN-182 [ DDN-182 ] |
| Resolution | New: Fixed [ 1 ] | |
| Status | Original: Open [ 1 ] | New: Resolved [ 5 ] |
| Fix Version/s | New: Lustre 2.7.0 [ 10631 ] |
Oleg Drokin (oleg.drokin@intel.com) merged in patch http://review.whamcloud.com/13604/
Subject: LU-6199 ldiskfs: delete bad WARN_ON_ONCE from ldiskfs
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 6c55653ed07c4e4c98efdad7693a9543a41d591b
| Link | New: This issue is related to INTL-193 [ INTL-193 ] |
| Affects Version/s | New: Lustre 2.7.0 [ 10631 ] |
| Description |
Original:
new warnings occasionally seen in ldiskfs due to checking code new in upstream ext4 code for el6.6. stack traces look like ldiskfs_truncate() ..... -> ldiskfs_flush_unwritten_io() lustre calls some ext4/ldiskfs entry point without locking i_mutex. It must do so in order to avoid deadlocks. This triggers a WARN_ON_ONCE() check that is in el6.6 and newer kernels. ldiskfs patch series for future kernels took care of this, but the patch series for el6.6 did not. Need to add a patch to remove the check for ldiskfs on el6.6 |
New:
new warnings occasionally seen in ldiskfs due to checking code new in upstream ext4 code for el6.6. stack traces look like ldiskfs_truncate() ..... -> ldiskfs_flush_unwritten_io() lustre calls some ext4/ldiskfs entry points without locking i_mutex. It must do so in order to avoid deadlocks. This triggers a WARN_ON_ONCE() check that is in el6.6 and newer kernels. ldiskfs patch series for future kernels took care of this, but the patch series for el6.6 did not. Need to add a patch to remove the check for ldiskfs on el6.6 |
Bob Glossman (bob.glossman@intel.com) uploaded a new patch: http://review.whamcloud.com/13604
Subject: LU-6199 ldiskfs: delete bad WARN_ON_ONCE from ldiskfs
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 4dd2041c74c58e83910964607d13ca18560a1bb1
Patch landed to Master.