Thu, Nov 25
Abandoning since we're removing support for the 32 bit multilib build in favour of building for a real 32 bit triple. See D114473.
I am going to make the 32 bit CI build a soft failure until we remove it entirely and ship this patch. See D114473 for removing the 32 bit multilib support.
ddress review comments
Fix clang-format issue
vectorization works fine without any of this as I was hoping, declaring bankruptcy here, thanks!
Update commit message and improve the release note.
Nice simplification, thanks!
This will prob. also need to be rebased since I couldn't land the parent.
I cannot seem to be able to arc patch this and submit it for you.
[runtimes] Remove support for building in 32 bits mode on a 64 bit target
I think this needs to be clarified. It is about the GCC multilib style -m32 (inherited to Clang), not about the general cross-compilation feature.
Your revert breaks normal building: http://188.8.131.52/linux/61593/step_4.txt
Thanks for decomposing into an easy to review PR, LGTM!
Thanks a lot for your time and help @MyDeveloperDay
Can you explain why you think "the implementation did not do that"? Lines 12 and 13 in the old test (before your changes) cover that case...
This revision isn't properly based -- please rebase on the current master tip and update.
I thought about this again and just wanted to make sure that I did not misunderstand your previous comment.
So, if this also works for in another architecture (like x86_64 or Arm on Linux or Mac), then it should be fine to merge.
Do you mean that it would be preferable for other targets to test this patch as well to ensure it is a reasonable fix? Would this be necessary as I'm using a hard-coding a few of the Linux libm produced values (that produces what is already expected in MultiSource/Benchmarks/MiBench/telecomm-FFT/telecomm-fft.reference_output)?
Can you please provide an explanation of what this change does and why? It would be useful to reviewers that might not have all the context (like myself).
Removing from the runtimes review queue -- please ping me directly if you need my attention on this.
Seem to be already fixed on the upstream default branch
Removed some tests already disabled by D114481.
Added missing #include-s
Fold in a change to 2010-05-26-DotDebugLoc.ll -- I wasn't sure why it needed a change, however it turns out to be spurious too. In DBG_VALUE mode, a DBG_VALUE is issued after a copy of the argument $rdi to $rbx, but in instr-ref mode that copy is deferred until after an instruction that clobbers $rdi. Hence, we expect the location-list MCSymbol to come after one additional copy.
IMO these tests would benefit little as they're really targetted at LiveDebugValues, and shouldn't really be .ll tests. On the other hand, given that there isn't actually a test difference between the two implementations, it won't hurt to add some extra RUN lines, WDYT?
Clarify TracksDebugUserValues documenting comment.
- Rename DebugValuesAllocated -> TracksDebugUserValues
- MIR key for TracksDebugUserValues
- TracksDebugUserValues tests for MachineVerifier
- LiveDebugValues test to ensure undefined debug uses delete DBG_VALUE instructions
Passed internal CI