Bug 1423373
| Summary: | Crash in index xlator because of race in inode_ctx_set and inode_ref | |||
|---|---|---|---|---|
| Product: | [Community] GlusterFS | Reporter: | Poornima G <pgurusid> | |
| Component: | core | Assignee: | Poornima G <pgurusid> | |
| Status: | CLOSED CURRENTRELEASE | QA Contact: | ||
| Severity: | unspecified | Docs Contact: | ||
| Priority: | unspecified | |||
| Version: | mainline | CC: | bugs | |
| Target Milestone: | --- | |||
| Target Release: | --- | |||
| Hardware: | Unspecified | |||
| OS: | Unspecified | |||
| Whiteboard: | ||||
| Fixed In Version: | glusterfs-3.11.0 | Doc Type: | If docs needed, set a value | |
| Doc Text: | Story Points: | --- | ||
| Clone Of: | ||||
| : | 1423385 (view as bug list) | Environment: | ||
| Last Closed: | 2017-05-30 18:44:10 UTC | Type: | Bug | |
| Regression: | --- | Mount Type: | --- | |
| Documentation: | --- | CRM: | ||
| Verified Versions: | Category: | --- | ||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | ||
| Cloudforms Team: | --- | Target Upstream Version: | ||
| Embargoed: | ||||
| Bug Depends On: | ||||
| Bug Blocks: | 1423385 | |||
REVIEW: https://review.gluster.org/16622 (libglusterfs: Fix a crash due to race between inode_ctx_set and inode_ref) posted (#4) for review on master by Poornima G (pgurusid) COMMIT: https://review.gluster.org/16622 committed in master by Pranith Kumar Karampuri (pkarampu) ------ commit 4badd7334d345b1c27f528140e4c28afaf1f32f6 Author: Poornima G <pgurusid> Date: Wed Feb 15 11:18:31 2017 +0530 libglusterfs: Fix a crash due to race between inode_ctx_set and inode_ref Issue: Currently inode ref count is gaurded by inode_table->lock, and inode_ctx is gauarded by inode->lock. With the new patch [1] inode_ref was modified to change the inode_ctx to track the ref count per xlator. Thus inode_ref performed under inode_table->lock is modifying inode_ctx which has to be modified only under inode->lock Solution: When a inode is created, inode_ctx holder is allocated for all the xlators. Hence in case of inode_ctx_set instead of using the first free index in inode ctx holder, we can have predecided index for every xlator in the graph. Credits Pranith K <pkarampu> [1] http://review.gluster.org/13736 Change-Id: I1bfe111c211fcc4fcd761bba01dc87c4c69b5170 BUG: 1423373 Signed-off-by: Poornima G <pgurusid> Reviewed-on: https://review.gluster.org/16622 Smoke: Gluster Build System <jenkins.org> NetBSD-regression: NetBSD Build System <jenkins.org> Reviewed-by: Niels de Vos <ndevos> CentOS-regression: Gluster Build System <jenkins.org> Reviewed-by: Pranith Kumar Karampuri <pkarampu> This bug is getting closed because a release has been made available that should address the reported issue. In case the problem is still not fixed with glusterfs-3.11.0, please open a new bug report. glusterfs-3.11.0 has been announced on the Gluster mailinglists [1], packages for several distributions should become available in the near future. Keep an eye on the Gluster Users mailinglist [2] and the update infrastructure for your distribution. [1] http://lists.gluster.org/pipermail/announce/2017-May/000073.html [2] https://www.gluster.org/pipermail/gluster-users/ |
Description of problem: Some regression test cases in afr were failing because of this bug. RCA: Currently inode ref count is gaurded by inode_table->lock, and inode_ctx is gauarded by inode->lock. With the new patch [1] inode_ref was modified to change the inode_ctx to track the ref count per xlator. Thus inode_ref performed under inode_table->lock is modifying inode_ctx which has to be modified only under inode->lock Solution: When a inode is created, inode_ctx holder is allocated for all the xlators. Hence in case of inode_ctx_set instead of using the first free index in inode ctx holder, we can have predecided index for every xlator in the graph. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info: