jiridanek edited a comment on pull request #1351:
URL: https://github.com/apache/qpid-dispatch/pull/1351#issuecomment-1030404166


   > And to your point (re: usefulness): it takes way too long to get feedback 
from travis for pull request verification. Checking the history we're talking 
~5 hours plus to get through travis. GHA runs the same tests in under 30 
minutes - which gives me enough confidence that my PR is solid enough for 
review.
   
   Meaning that Travis is useless for PRs, regardless of what we do or do not 
do in this PR. The build wait queue is always going to be long.
   
   > Where I think travis can give us value is testing alternate platforms and 
tooling, like ARM and clang, which aren't currently covered in GHA. Also travis 
seems to have a knack for finding TSAN issues better than GHA - for whatever 
reason. Let's leverage that strength as well.
   
   Agreed.
   
   > TSAN checking, esp. on non-x86_64 architectures: arm64 and s390x
   
   I wonder whether the Tsan on non-x86 is somehow different than the x86 
variant. I don't think so. The checks the software performs are still the same, 
what's different is that the synchronization issues may also have more visible 
actual consequences, but that is irrespective of the sanitizer. In fact, tsan 
can make e.g. arm less sensitive to synch issues:
   
   > *force_seq_cst_atomics* This flag turns all atomic operations into 
sequentially consistent operations. This is useful to check whether a race is 
caused by incorrect memory ordering or not.
   > -- https://github.com/google/sanitizers/wiki/ThreadSanitizerFlags
   
   There is a fundamentally different ASAN for arm64, 
https://stackoverflow.com/search?q=user%3A1047788+aarch64+deleted%3Ayes


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]



---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to