53b2a7bb74ad141950aa0fe56cc8b906a07086a0 - llvm-project/llvm

commit	53b2a7bb74ad141950aa0fe56cc8b906a07086a0	[log] [tgz]
author	Simon Pilgrim <llvm-dev@redking.me.uk>	Thu Apr 15 13:55:32 2021 +0100
committer	Copybara-Service <copybara-worker@google.com>	Fri Apr 16 12:56:18 2021 -0700
tree	c83e477197ec616d7c7a1672a6b29231de23ba4e
parent	0558ccb3eeb309fcb340d8dd1409ced78c289d05 [diff]

[X86] combineCMP - fold cmpEQ/NE(TRUNC(X),0) -> cmpEQ/NE(X,0)

If we are truncating from a i32 source before comparing the result against zero, then see if we can directly compare the source value against zero.

If the upper (truncated) bits are known to be zero then we can compare against that, hopefully increasing the chances of us folding the compare into a EFLAG result of the source's operation.

Fixes PR49028.

Differential Revision: https://reviews.llvm.org/D100491

GitOrigin-RevId: 9d57a77b81265c650c75511115067588790d7ac0

lib/Target/X86/X86ISelLowering.cpp[diff]
test/CodeGen/X86/and-with-overflow.ll[diff]
test/CodeGen/X86/pr34137.ll[diff]
test/CodeGen/X86/pr49028.ll[diff]
test/CodeGen/X86/setcc-logic.ll[diff]
test/CodeGen/X86/vector-compare-any_of.ll[diff]
test/CodeGen/X86/vector-reduce-and-cmp.ll[diff]
test/CodeGen/X86/vector-reduce-or-bool.ll[diff]

8 files changed

tree: c83e477197ec616d7c7a1672a6b29231de23ba4e