projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: e2bb977) | patch
[InstCombine] fold fake floating point vector extract to shift+trunc.
authorjacquesguan <Jianjian.Guan@streamcomputing.com>
Fri, 26 Aug 2022 07:17:02 +0000 (15:17 +0800)
committerjacquesguan <Jianjian.Guan@streamcomputing.com>
Tue, 30 Aug 2022 02:12:16 +0000 (10:12 +0800)
commitdf525c7705d8f57b660bac18c9dd800a64b1917e
tree32e297eedb72b1e9c2994d8b0640dc27c316c767tree | snapshot
parente2bb9774b173487cebec9cb8b0568c3d9b4d55f8commit | diff
[InstCombine] fold fake floating point vector extract to shift+trunc.

This patch supports the FP part of D111082.

Differential Revision: https://reviews.llvm.org/D125750
llvm/lib/Transforms/InstCombine/InstCombineVectorOps.cpp diff | blob | history
llvm/test/Transforms/InstCombine/extractelement.ll diff | blob | history
Domain: System / Toolchain;