Overview
Request 1120553 revoked
- Created by badshah400
- In state revoked
- Open review for factory-staging
Request History
badshah400 created request
factory-auto added opensuse-review-team as a reviewer
Please review sources
factory-auto accepted review
Check script succeeded
licensedigger accepted review
ok
dimstar accepted review
anag+factory set openSUSE:Factory:Staging:L as a staging project
Being evaluated by staging project "openSUSE:Factory:Staging:L"
anag+factory accepted review
Picked "openSUSE:Factory:Staging:L"
anag+factory added factory-staging as a reviewer
Being evaluated by group "factory-staging"
anag+factory accepted review
Unstaged from project "openSUSE:Factory:Staging:L"
anag+factory set openSUSE:Factory:Staging:E as a staging project
Being evaluated by staging project "openSUSE:Factory:Staging:E"
anag+factory accepted review
Picked "openSUSE:Factory:Staging:E"
dimstar declined review
declinging - seem comments/discussion
Change is incorrect, see my comments in the original SR.
dimstar declined request
declinging - seem comments/discussion
Change is incorrect, see my comments in the original SR.
anag+factory reopened request
Reopened via staging workflow.
anag+factory added factory-staging as a reviewer
Being evaluated by group "factory-staging"
anag+factory accepted review
Unstaged from project "openSUSE:Factory:Staging:E"
anag+factory declined request
Declined via staging workflow.
badshah400 revoked request
@dimstar
Change is incorrect, see my comments in the original SR.
@StefanBruens - thank you for catching this! The reason why Bernhard did this: he's working on making builds immutable. The reason that he looked at openblas was therefore likely that it had triggered his detection. So this might indicate that this setting is not taken from the env variable but from some detection at build time. This may be an oversight in the spec file or a bug in the code. Since the fix is applied to the pc file, it should not be relevant to the openblas build therefore I would argue for the latter. Unfortunately, I don't have time to look into this right now as I have more urgent mess sort out ATM :(
The rationale behind this is clear to me, I have done my fair share of reproducible build fixes myself ;-)
Your assumption this has been picked up from the build host instead of the taking up the supplied NUM_THREADS build parameter sounds likely.
If this number is pushed to dependent packages you might end up with packages which only support up to 4 threads. The trivial fix is to just use "%{num_threads}" in the sed command, but an upstream bug report should be opened as well.
Correct, doing this would be a cheap workaround - it won't help the openblas project if there was an issue here.
@anag+factory - please unstage