Move the plugin used in the unittest to test Inline Advisor Plugins
into a separate folder to clean up the cmake file for the analysis
tests.
Details
Diff Detail
- Repository
- rG LLVM Github Monorepo
Event Timeline
This patch addresses the cmake structure concerns mentioned in https://reviews.llvm.org/D139644
lgtm, but please also wait for @thakis 's feedback.
Should the other case at https://github.com/llvm/llvm-project/blob/main/llvm/unittests/Passes/CMakeLists.txt follow suit with a similar design?
Thanks, this looks like a big improvement to me :)
llvm/unittests/Analysis/CMakeLists.txt | ||
---|---|---|
62 | (as mentioned elsewhere, we shouldn't use the _with_input_files variant, but that's unrelated here :) ) |
Could one of you two push it for me, I don't have commit access.
Username: ibricchi
Email: ibricchi@student.ethz.ch
Sorry to raise this issue late (due to the holidays), but it seems this change causes a test failure on AIX
shard JSON output does not exist: /scratch/powerllvm/powerllvm_env/aix-ppc64/clang-ppc64-aix/build/unittests/Analysis/./AnalysisTests-LLVM-Unit-26345894-43-81.json
https://lab.llvm.org/buildbot/#/builders/214/builds/5000/steps/6/logs/FAIL__LLVM-Unit__81
Do you have any idea why?
Hi, sorry for the delay.
I see the error but I'm not sure what it could be, would you be able to tag someone who might know more about why this only fails on AXI @Jake-Egan?
Is there any way of testing changes on buildbot so that I can try and try and correct the behaviour?
It seems this problem is hard to investigate without having access to an AIX machine, so we are looking into it.
llvm/unittests/Analysis/CMakeLists.txt | ||
---|---|---|
67–68 | Is the WIN32 check not redundant when we check for AIX on the next line? Other than that you should be good to commit again, thanks! |
I'm not 100% sure on the overlap between the two conditions, I just went off what was there before I changed anything. But I've ammended the code to only check for AIX.
I don't have commit access, but if this looks good to you @Jake-Egan would you be able to commit the fix.
ibricchi <ibricchi@student.ethz.ch>
(as mentioned elsewhere, we shouldn't use the _with_input_files variant, but that's unrelated here :) )