projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: eaf7329) | patch
Revert SVML support for sqrt
authorWenlei He <aktoon@gmail.com>
Mon, 5 Oct 2020 15:12:54 +0000 (08:12 -0700)
committerWenlei He <aktoon@gmail.com>
Mon, 5 Oct 2020 15:13:11 +0000 (08:13 -0700)
commit89e8a8b223b2e20bb63e930ddb78cb80a3ed45a2
treee757b9ccc3eacfc02a2821faeefa0ab3cc50ec36tree | snapshot
parenteaf73293cb6b8d45dd85ffced57aea7ad4177754commit | diff
Revert SVML support for sqrt

As was brought up in D87169 by @craig.topper we shouldn't map llvm.sqrt to svml since there is a faster native instruction.
https://software.intel.com/sites/landingpage/IntrinsicsGuide/#text=_mm_sqrt_p&expand=5824,5823,5356,5823,5825,5365,5356

Reviewed By: craig.topper

Differential Revision: https://reviews.llvm.org/D88620
llvm/include/llvm/Analysis/VecFuncs.def diff | blob | history
llvm/test/Transforms/LoopVectorize/X86/svml-calls.ll diff | blob | history
Domain: System / Toolchain;