f0dd12ec5c
The first attempt missed changing test files for tools (update_llc_test_checks.py). Original commit message: This implements the main suggested change from issue #56498. Using the shorter (non-extending) instruction with only -Oz ("minsize") rather than -Os ("optsize") is left as a possible follow-up. As noted in the bug report, the zero-extending load may have shorter latency/better throughput across a wide range of x86 micro-arches, and it avoids a potential false dependency. The cost is an extra instruction byte. This could cause perf ups and downs from secondary effects, but I don't think it is possible to account for those in advance, and that will likely also depend on exact micro-arch. This does bring LLVM x86 codegen more in line with existing gcc codegen, so if problems are exposed they are more likely to occur for both compilers. Differential Revision: https://reviews.llvm.org/D129775
18 lines
536 B
LLVM
18 lines
536 B
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc -mtriple=x86_64-- -mcpu=generic < %s | FileCheck %s
|
|
|
|
; Verify that the DAGCombiner doesn't wrongly remove the 'and' from the dag.
|
|
|
|
define i8 @foo(ptr %V) {
|
|
; CHECK-LABEL: foo:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: movzbl 2(%rdi), %eax
|
|
; CHECK-NEXT: andb $95, %al
|
|
; CHECK-NEXT: retq
|
|
%V3i8 = load <3 x i8>, ptr %V, align 4
|
|
%t0 = and <3 x i8> %V3i8, <i8 undef, i8 undef, i8 95>
|
|
%t1 = extractelement <3 x i8> %t0, i64 2
|
|
ret i8 %t1
|
|
}
|
|
|