Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.
Comment: Describe ported patch commit comments more fully

...

For patches backported from master to a maintenance branch:Lustre maintenance branch (e.g. b2_5) there are some conventions to follow so that the changes/fixes can more easily be tracked across branches. The best method for porting a patch from one branch to another is to use git cherry-pick {commit_hash} on the branch where you want the patch to land.  This will apply the whole patch (as best as is able, and show conflicts where needed), copy the commit message, preserve the original patch author.  With luck, there will not be any patch conflicts and no further work is needed.  If necessary, the patch conflicts need to be resolved before committing the patch.  For the commit message:

  • the entire commit message, including the summary line, should be copied from the source change
  • the Signed-off-by: line of the original author should also be kept
  • the person who commits the ported change should add a Signed-off-by: line with their name and email following the original one
  • the original Change-Id: line can be kept (newer Gerrit versions can handle the same Change-Id: line on different branches
  • the original Reviewed-by: lines can be kept, and Gerrit will automatically add them as reviewers to the new patch
  • the Tested-by: Maloo and Tested-by: Jenkins lines should be removed from the new commit message, though any Tested-by: lines from real people can be kept
  • the "Reviewed-on: http://review.whamcloud.com/nnnnn" line should be changed to "Lustre-change: http://review.whamcloud.com/nnnnn" (please use the "canonical" gerrit URL as shown)
  • the "cherry picked from commit abcdef1234567890" line should be changed to  "Lustre-commit: abcdef1234567890"
  • any non-trivial changes to the original patch should be described after the original Signed-off-by: and Lustre-commit: lines before your own Signed-off-by: line
  • if two or more patches are being combined into a single patch (only in case there are bugs in the original patch that were fixed in later commits) the full commit message from each patch should be kept, along with the Signed-off-by: and Lustre-commit: and Lustre-change: lines of each commit

:

No Format
LU-4725 mdt: child-parent lock ordering in rename
    
Change rename so that it always has parent-child lock ordering,
otherwise it may deadlock with other operations.
    
Lustre-commit: 4e308ef74f271ec7e19917e3c0f88586537582c3
Lustre-change: http://review.whamcloud.com/9538
    
LU-4725 obd: lu_object_find_at hung
    
lu_object_find_at hangs if called two times and object is removed
in between. It makes mdt_rename_sanity not workable for rename.
Change mdt_rename_sanity to work on existing object.
    
Lustre-commit: b7c72ec1ddeda2cf94ea151f974d3f94e3a7d1ed
Lustre-change: http://review.whamcloud.com/10484
Xyratex-bug-id: MRP-1700
    
Test-Parameters: alwaysuploadlogs \
envdefinitions=SLOW=yes,ENABLE_QUOTA=yes,ONLY=54 \
ossjob=lustre-b2_4 mdsjob=lustre-b2_4 ossbuildno=73 mdsbuildno=73 \
testlist=sanityn
    
Signed-off-by: Vitaly Fertman <vitaly_fertman@xyratex.com>
Signed-off-by: Rahul Deshmukh <rahul_deshmukh@xyratex.com>
No Format
LU-4579 ldlm: Properly display time on lock timeout
    
When lock timeout happens, need to printtime since last lock
activity, not since when the lock was granted.
As such set l_last_activity to current time when sending callbacks too.
    
Lustre-change: http://review.whamcloud.com/10601
Lustre-commit: f3864905b26a90ed3a9bb6f99357264fa40015c4
    
Change-Id: I31d0de6d9add67865cfcb16af903647f7cb932a0
Ic9ce52bfcd8788834347fba155cc8c6be674dcd8

For patches ported from master to the upstream kernel are treated similarly (keep all comments and Signed-off-by: lines from the original patch) add new Signed-off-by

...

For patches ported from master to the upstream kernel: and comments afterward, but replace the Lustre-commit: line (which doesn't mean anything in the upstream kernel git) with Intel-bug-id: {jira_URL} so that the original bug can still be identified:

No Format
lustre/llite: simplify dentry revalidate
    
Lustre client dentry validation is protected by LDLM lock, so
any time a dentry is found, it's valid and no need to revalidate
from MDS, and even it does, there is race that it may be
invalidated after revalidation is finished.
    
Signed-off-by: Lai Siyao <lai.siyao@intel.com>
Reviewed-on: http://review.whamcloud.com/7475
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-3544
Signed-off-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

...