Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

This revision was automatically updated to reflect the committed changes. Closed by commit rL277166: Initial support for vectorization using svml (short vector math library). (authored by mmasten). Changed prior to commit: https://reviews.llvm.org/D19544?vs=65152&id=66141#toc Repository: rL

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

mmasten added a comment. I was just recently given commit privileges, so I can do it. Thanks Hal. https://reviews.llvm.org/D19544 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

hfinkel added a comment. In https://reviews.llvm.org/D19544#492258, @mmasten wrote: > I think this is just saying that some of the weird types are not supported on > all targets. For now, is it ok to proceed with checking this code in? Correct. In https://reviews.llvm.org/D19544#493403, @mmas

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

mmasten added a comment. Thanks Michael. The tests have been updated. Matt https://reviews.llvm.org/D19544 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

mmasten updated this revision to Diff 65152. https://reviews.llvm.org/D19544 Files: include/llvm/Analysis/TargetLibraryInfo.h lib/Analysis/TargetLibraryInfo.cpp test/Transforms/LoopVectorize/X86/svml-calls.ll Index: lib/Analysis/TargetLibraryInfo.cpp ===

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

mzolotukhin accepted this revision. mzolotukhin added a comment. This revision is now accepted and ready to land. LGTM with a small nit: could you please run `opt -instnamer` on your test (it'll replace `%0`, `%1`,... with `%tmp0`, `%tmp1` etc, making it easier to modify test in future)? Thanks

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

mmasten added a comment. I think this is just saying that some of the weird types are not supported on all targets. For now, is it ok to proceed with checking this code in? Thanks, Matt https://reviews.llvm.org/D19544 ___ cfe-commits mailing list

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

mmasten updated this revision to Diff 64571. https://reviews.llvm.org/D19544 Files: include/llvm/Analysis/TargetLibraryInfo.h lib/Analysis/TargetLibraryInfo.cpp test/Transforms/LoopVectorize/X86/svml-calls.ll Index: lib/Analysis/TargetLibraryInfo.cpp ===

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

spatel added a subscriber: davide. spatel added a comment. In https://reviews.llvm.org/D19544#488589, @mmasten wrote: > In the process of writing test cases, I noticed that a loop with a call to > llvm.log.f32 was not getting vectorized due to cost modeling. When forcing > vectorization on the

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

mmasten added a comment. In the process of writing test cases, I noticed that a loop with a call to llvm.log.f32 was not getting vectorized due to cost modeling. When forcing vectorization on the loop and throwing -fveclib=SVML, the loop was vectorized with a widened intrinsic instead of the sv

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

hfinkel added a comment. In https://reviews.llvm.org/D19544#485677, @mmasten wrote: > Thanks for reviewing. One concern I have going forward is the number of > entries that will appear in the switch statement inside > addVectorizableFunctionsFromVecLib(). I assume that the right thing to do is

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

mmasten added a comment. Thanks for reviewing. One concern I have going forward is the number of entries that will appear in the switch statement inside addVectorizableFunctionsFromVecLib(). I assume that the right thing to do is to replace this with something that is TableGen'd? Also, I just w

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

hfinkel added a comment. In https://reviews.llvm.org/D19544#484439, @spatel wrote: > Hi Matt - > > This looks like the right first step in the path that Hal suggested, except I > think we need a test case for each function that you want to enable. Please > see test/Transforms/LoopVectorize/X86/

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

spatel added a comment. Hi Matt - This looks like the right first step in the path that Hal suggested, except I think we need a test case for each function that you want to enable. Please see test/Transforms/LoopVectorize/X86/veclib-calls.ll as a reference for how to do that. https://reviews

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

mmasten added a comment. Hello all, Just wanted to see if you guys have some time to review. Thanks, Matt https://reviews.llvm.org/D19544 ___ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-c

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

mmasten added a subscriber: cfe-commits. mmasten updated this revision to Diff 61110. http://reviews.llvm.org/D19544 Files: include/clang/Frontend/CodeGenOptions.def include/clang/Frontend/CodeGenOptions.h lib/CodeGen/BackendUtil.cpp lib/Frontend/CompilerInvocation.cpp Index: lib/CodeGen

Re: [PATCH] D19544: Pass for translating math intrinsics to math library calls.

mmasten updated this revision to Diff 6. http://reviews.llvm.org/D19544 Files: include/llvm/Analysis/TargetLibraryInfo.h lib/Analysis/TargetLibraryInfo.cpp Index: lib/Analysis/TargetLibraryInfo.cpp === --- lib/Analysis/Targe