DAOS-18196 object: large stack for collective object RPC ULT #17147
Conversation
|
Ticket title is 'daos_engine segfaults when we perform concurrent IO operations on multiple pools on mdonssd testing.' |
Nasf-Fan
force-pushed
the
Nasf-Fan/DAOS-18196_1
branch
from
2aadeff to
536e91e
Compare
Nasf-Fan
changed the title
Nasf-Fan
force-pushed
the
Nasf-Fan/DAOS-18196_1
branch
from
536e91e to
563beed
Compare
|
Test stage Unit Test bdev on EL 8.8 completed with status FAILURE. https://jenkins-3.daos.hpc.amslabs.hpecorp.net/job/daos-stack/job/daos/job/PR-17147/5/display/redirect |
|
Test stage Unit Test bdev with memcheck on EL 8.8 completed with status FAILURE. https://jenkins-3.daos.hpc.amslabs.hpecorp.net/job/daos-stack/job/daos/job/PR-17147/5/display/redirect |
|
Test stage Unit Test on EL 8.8 completed with status FAILURE. https://jenkins-3.daos.hpc.amslabs.hpecorp.net/job/daos-stack/job/daos/job/PR-17147/5/display/redirect |
|
Test stage Unit Test with memcheck on EL 8.8 completed with status FAILURE. https://jenkins-3.daos.hpc.amslabs.hpecorp.net/job/daos-stack/job/daos/job/PR-17147/5/display/redirect |
|
Test stage NLT on EL 8.8 completed with status FAILURE. https://jenkins-3.daos.hpc.amslabs.hpecorp.net/job/daos-stack/job/daos/job/PR-17147/5/display/redirect |
To avoid potential ULT stack overflow. Allow-unstable-test: true Signed-off-by: Fan Yong <[email protected]>
Nasf-Fan
force-pushed
the
Nasf-Fan/DAOS-18196_1
branch
from
563beed to
5d3541e
Compare
wangshilong
left a comment
wangshilong
left a comment
There was a problem hiding this comment.
Just curious that did we confirm the PR help fix the issue?
|
|
||
| rc = dss_thread_collective_reduce(&coll_ops, &coll_args, DSS_USE_CURRENT_ULT); | ||
| rc = dss_thread_collective_reduce(&coll_ops, &coll_args, | ||
| DSS_ULT_DEEP_STACK | DSS_USE_CURRENT_ULT); |
There was a problem hiding this comment.
I think this one line change is the whole purpose of this PR. Other changes looks not necessary (and not correct) to me.
BTW, due to the DSS_USE_CURRENT_ULT flag, the collective function executed on current xstream won't be able to run in deep stack. I think that's something needs be fixed.
There was a problem hiding this comment.
I think this one line change is the whole purpose of this PR. Other changes looks not necessary (and not correct) to me.
No, current ULT itself is also part of the collective operation, it also needs deep stack when it is created. That is the changes for others in this patch.
BTW, due to the DSS_USE_CURRENT_ULT flag, the collective function executed on current xstream won't be able to run in deep stack. I think that's something needs be fixed.
If current ULT can do the task by itself, why need to create new ULT on the same XS?
To avoid potential ULT stack overflow.
Allow-unstable-test: true
Steps for the author:
After all prior steps are complete: