Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

TCP/TEST: Fix simultaneous ep close with ucp_hello_world - v1.11.x #7251

Merged

Conversation

yosefe
Copy link
Contributor

@yosefe yosefe commented Aug 17, 2021

Why

Backport #7224

No conflicts, except didn't port a whitespace fix in ucp_ep.c since it's in a function that does not exist in v1.11.x

Fixes an assertion failure when running ucp_hello_world over TCP
transport.

UCP: When both sides close their TCP endpoints, one side can receive
connection reset event while it's trying to close-flush its endpoint. We
should not try to invoke user error callback in such case. Instead, the
close operation should complete with status CONNECTION_RESET.

UCT/TCP: Need to purge outstanding PUT operations when getting an error.

Test: Run ucp_hello_world over several transports. Currently it used
TCP only when ran inside a docker, so issue was not detected.
Fix barrier in hello_world test to prevent failures.
@yosefe yosefe changed the title TCP/TEST: Fix simultaneous ep close with ucp_hello_world TCP/TEST: Fix simultaneous ep close with ucp_hello_world - v1.11.x Aug 17, 2021
@yosefe yosefe added this to the v1.11.1 milestone Aug 17, 2021
@yosefe yosefe merged commit 5d8c109 into openucx:v1.11.x Aug 18, 2021
@yosefe yosefe deleted the topic/tcp-test-fix-simultaneous-ep-close-v1.11.x branch August 18, 2021 13:02
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants