Alexey Sachkov
Alexey Sachkov
> @krasznaa on the contrary, I think buffer is ok. But I find it suspicious, that you capture a contexpr static variable by value. I'm not sure it is allowed....
This is a draft PR, which is intended to be a proof of concept that we can fix debug info issues (known at the moment) caused by integration footer presence...
I will move this back to draft state to prevent accidental merge, because I've realized that I don't have a proper test for one of the main fixes in this...
> I will move this back to draft state to prevent accidental merge, because I've realized that I don't have a proper test for one of the main fixes in...
Moved back to draft, because the issue with checksums is still present: it seems like more changes in `CGDebugInfo.cpp` is needed
/verify with intel/llvm-test-suite#970
> Do I understand correctly this won't impact the scenario with the "full" `-O0` compilation due to no DAE metadata being provided by the "early optimization" part of the pipeline?...
/verify with intel/llvm-test-suite#970
CUDA LLVM test suite failure doesn't seem to be caused by this patch and it was already recorded in #6463