
Some of the tests from the X86 directory can be generalized to improve coverage for other architectures
37 lines
1.8 KiB
LLVM
37 lines
1.8 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
|
|
; RUN: %if x86-registered-target %{ opt -passes=slp-vectorizer -S -slp-threshold=-10 -mtriple=x86_64-unknown-linux-gnu < %s | FileCheck %s %}
|
|
; RUN: %if aarch64-registered-target %{ opt -passes=slp-vectorizer -S -slp-threshold=-10 -mtriple=aarch64-unknown-linux-gnu < %s | FileCheck %s %}
|
|
|
|
define void @test(i8 %0) {
|
|
; CHECK-LABEL: define void @test(
|
|
; CHECK-SAME: i8 [[TMP0:%.*]]) {
|
|
; CHECK-NEXT: entry:
|
|
; CHECK-NEXT: [[TMP1:%.*]] = insertelement <2 x i8> <i8 0, i8 poison>, i8 [[TMP0]], i32 1
|
|
; CHECK-NEXT: [[TMP2:%.*]] = sext <2 x i8> [[TMP1]] to <2 x i32>
|
|
; CHECK-NEXT: [[TMP3:%.*]] = mul <2 x i8> [[TMP1]], zeroinitializer
|
|
; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i8> [[TMP3]], i32 0
|
|
; CHECK-NEXT: [[TMP5:%.*]] = zext i8 [[TMP4]] to i32
|
|
; CHECK-NEXT: [[TMP6:%.*]] = extractelement <2 x i8> [[TMP3]], i32 1
|
|
; CHECK-NEXT: [[TMP7:%.*]] = zext i8 [[TMP6]] to i32
|
|
; CHECK-NEXT: [[ADD:%.*]] = or i32 [[TMP5]], [[TMP7]]
|
|
; CHECK-NEXT: [[SHR:%.*]] = lshr i32 [[ADD]], 1
|
|
; CHECK-NEXT: [[CONV9:%.*]] = trunc i32 [[SHR]] to i8
|
|
; CHECK-NEXT: store i8 [[CONV9]], ptr null, align 1
|
|
; CHECK-NEXT: [[TMP8:%.*]] = shufflevector <2 x i32> [[TMP2]], <2 x i32> poison, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
|
|
; CHECK-NEXT: ret void
|
|
;
|
|
entry:
|
|
%conv3 = sext i8 %0 to i32
|
|
%conv7 = sext i8 0 to i32
|
|
%conv = zext i16 0 to i32
|
|
%mul = mul i32 %conv3, %conv
|
|
%conv6 = zext i16 0 to i32
|
|
%mul8 = mul i32 %conv7, %conv6
|
|
%add = or i32 %mul8, %mul
|
|
%shr = lshr i32 %add, 1
|
|
%conv9 = trunc i32 %shr to i8
|
|
store i8 %conv9, ptr null, align 1
|
|
%broadcast.splatinsert = insertelement <8 x i32> poison, i32 %conv3, i64 0
|
|
ret void
|
|
}
|