7c51c310ad
When we vectorize loads or stores we only keep the address of the first lane. The rest may become dead. This patch adds the address operands of vectorized loads or stores to the dead candidates set, such that they get erased if dead.
33 lines
1.2 KiB
LLVM
33 lines
1.2 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5
|
|
; RUN: opt -passes=sandbox-vectorizer -sbvec-vec-reg-bits=1024 -sbvec-allow-non-pow2 -sbvec-passes="bottom-up-vec<>" %s -S | FileCheck %s
|
|
|
|
|
|
declare void @foo()
|
|
define void @slice_seeds(ptr %ptr, float %val) {
|
|
; CHECK-LABEL: define void @slice_seeds(
|
|
; CHECK-SAME: ptr [[PTR:%.*]], float [[VAL:%.*]]) {
|
|
; CHECK-NEXT: [[PTR0:%.*]] = getelementptr float, ptr [[PTR]], i32 0
|
|
; CHECK-NEXT: [[PTR2:%.*]] = getelementptr float, ptr [[PTR]], i32 2
|
|
; CHECK-NEXT: [[LD2:%.*]] = load float, ptr [[PTR2]], align 4
|
|
; CHECK-NEXT: store float [[LD2]], ptr [[PTR2]], align 4
|
|
; CHECK-NEXT: call void @foo()
|
|
; CHECK-NEXT: [[VECL:%.*]] = load <2 x float>, ptr [[PTR0]], align 4
|
|
; CHECK-NEXT: store <2 x float> [[VECL]], ptr [[PTR0]], align 4
|
|
; CHECK-NEXT: ret void
|
|
;
|
|
%ptr0 = getelementptr float, ptr %ptr, i32 0
|
|
%ptr1 = getelementptr float, ptr %ptr, i32 1
|
|
%ptr2 = getelementptr float, ptr %ptr, i32 2
|
|
|
|
%ld2 = load float, ptr %ptr2
|
|
store float %ld2, ptr %ptr2
|
|
; This call blocks scheduling of all 3 stores.
|
|
call void @foo()
|
|
|
|
%ld0 = load float, ptr %ptr0
|
|
%ld1 = load float, ptr %ptr1
|
|
store float %ld0, ptr %ptr0
|
|
store float %ld1, ptr %ptr1
|
|
ret void
|
|
}
|