[X86] Simplify the CHECK lines in vector-reduce-and/or/xor-widen.ll in similar way...

Forgot to do the widen forms when I was doing the others.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@362310 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/test/CodeGen/X86/vector-reduce-and-widen.ll b/test/CodeGen/X86/vector-reduce-and-widen.ll
index a00f81cc63d17603d85ea242bec915b9eb2c082f..16c45da9c338b7964dca3897a2a4dd2ba6de98fe 100644 (file)
--- a/test/CodeGen/X86/vector-reduce-and-widen.ll
+++ b/test/CodeGen/X86/vector-reduce-and-widen.ll
@@ -3,8 +3,8 @@
 ; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+sse4.1 | FileCheck %s --check-prefix=ALL --check-prefix=SSE --check-prefix=SSE41
 ; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX1
 ; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX2
-; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw | FileCheck %s --check-prefix=ALL --check-prefix=AVX512 --check-prefix=AVX512BW
-; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw,+avx512vl | FileCheck %s --check-prefix=ALL --check-prefix=AVX512 --check-prefix=AVX512VL
+; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX512
+; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw,+avx512vl | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX512
 
 ;
 ; vXi64
@@ -24,13 +24,6 @@ define i64 @test_v2i64(<2 x i64> %a0) {
 ; AVX-NEXT:    vpand %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovq %xmm0, %rax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i64:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovq %xmm0, %rax
-; AVX512-NEXT:    retq
   %1 = call i64 @llvm.experimental.vector.reduce.and.i64.v2i64(<2 x i64> %a0)
   ret i64 %1
 }
@@ -200,13 +193,6 @@ define i32 @test_v2i32(<2 x i32> %a0) {
 ; AVX-NEXT:    vpand %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i32:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    retq
   %1 = call i32 @llvm.experimental.vector.reduce.and.i32.v2i32(<2 x i32> %a0)
   ret i32 %1
 }
@@ -229,15 +215,6 @@ define i32 @test_v4i32(<4 x i32> %a0) {
 ; AVX-NEXT:    vpand %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v4i32:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    retq
   %1 = call i32 @llvm.experimental.vector.reduce.and.i32.v4i32(<4 x i32> %a0)
   ret i32 %1
 }
@@ -434,14 +411,6 @@ define i16 @test_v2i16(<2 x i16> %a0) {
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    # kill: def $ax killed $ax killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i16:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    # kill: def $ax killed $ax killed $eax
-; AVX512-NEXT:    retq
   %1 = call i16 @llvm.experimental.vector.reduce.and.i16.v2i16(<2 x i16> %a0)
   ret i16 %1
 }
@@ -467,16 +436,6 @@ define i16 @test_v4i16(<4 x i16> %a0) {
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    # kill: def $ax killed $ax killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v4i16:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    # kill: def $ax killed $ax killed $eax
-; AVX512-NEXT:    retq
   %1 = call i16 @llvm.experimental.vector.reduce.and.i16.v4i16(<4 x i16> %a0)
   ret i16 %1
 }
@@ -506,18 +465,6 @@ define i16 @test_v8i16(<8 x i16> %a0) {
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    # kill: def $ax killed $ax killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v8i16:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    # kill: def $ax killed $ax killed $eax
-; AVX512-NEXT:    retq
   %1 = call i16 @llvm.experimental.vector.reduce.and.i16.v8i16(<8 x i16> %a0)
   ret i16 %1
 }
@@ -762,14 +709,6 @@ define i8 @test_v2i8(<2 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.and.i8.v2i8(<2 x i8> %a0)
   ret i8 %1
 }
@@ -808,16 +747,6 @@ define i8 @test_v4i8(<4 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v4i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.and.i8.v4i8(<4 x i8> %a0)
   ret i8 %1
 }
@@ -862,18 +791,6 @@ define i8 @test_v8i8(<8 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v8i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.and.i8.v8i8(<8 x i8> %a0)
   ret i8 %1
 }
@@ -924,20 +841,6 @@ define i8 @test_v16i8(<16 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v16i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpand %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.and.i8.v16i8(<16 x i8> %a0)
   ret i8 %1
 }

diff --git a/test/CodeGen/X86/vector-reduce-or-widen.ll b/test/CodeGen/X86/vector-reduce-or-widen.ll
index ef060acfb67829daa406fc7d3063f3b4307370d2..b5abc094dfdc98357b8eef9f6015e8b01448cc12 100644 (file)
--- a/test/CodeGen/X86/vector-reduce-or-widen.ll
+++ b/test/CodeGen/X86/vector-reduce-or-widen.ll
@@ -3,8 +3,8 @@
 ; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+sse4.1 | FileCheck %s --check-prefix=ALL --check-prefix=SSE --check-prefix=SSE41
 ; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX1
 ; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX2
-; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw | FileCheck %s --check-prefix=ALL --check-prefix=AVX512 --check-prefix=AVX512BW
-; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw,+avx512vl | FileCheck %s --check-prefix=ALL --check-prefix=AVX512 --check-prefix=AVX512VL
+; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX512
+; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw,+avx512vl | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX512
 
 ;
 ; vXi64
@@ -24,13 +24,6 @@ define i64 @test_v2i64(<2 x i64> %a0) {
 ; AVX-NEXT:    vpor %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovq %xmm0, %rax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i64:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovq %xmm0, %rax
-; AVX512-NEXT:    retq
   %1 = call i64 @llvm.experimental.vector.reduce.or.i64.v2i64(<2 x i64> %a0)
   ret i64 %1
 }
@@ -200,13 +193,6 @@ define i32 @test_v2i32(<2 x i32> %a0) {
 ; AVX-NEXT:    vpor %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i32:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    retq
   %1 = call i32 @llvm.experimental.vector.reduce.or.i32.v2i32(<2 x i32> %a0)
   ret i32 %1
 }
@@ -229,15 +215,6 @@ define i32 @test_v4i32(<4 x i32> %a0) {
 ; AVX-NEXT:    vpor %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v4i32:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    retq
   %1 = call i32 @llvm.experimental.vector.reduce.or.i32.v4i32(<4 x i32> %a0)
   ret i32 %1
 }
@@ -434,14 +411,6 @@ define i16 @test_v2i16(<2 x i16> %a0) {
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    # kill: def $ax killed $ax killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i16:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    # kill: def $ax killed $ax killed $eax
-; AVX512-NEXT:    retq
   %1 = call i16 @llvm.experimental.vector.reduce.or.i16.v2i16(<2 x i16> %a0)
   ret i16 %1
 }
@@ -467,16 +436,6 @@ define i16 @test_v4i16(<4 x i16> %a0) {
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    # kill: def $ax killed $ax killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v4i16:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    # kill: def $ax killed $ax killed $eax
-; AVX512-NEXT:    retq
   %1 = call i16 @llvm.experimental.vector.reduce.or.i16.v4i16(<4 x i16> %a0)
   ret i16 %1
 }
@@ -506,18 +465,6 @@ define i16 @test_v8i16(<8 x i16> %a0) {
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    # kill: def $ax killed $ax killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v8i16:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    # kill: def $ax killed $ax killed $eax
-; AVX512-NEXT:    retq
   %1 = call i16 @llvm.experimental.vector.reduce.or.i16.v8i16(<8 x i16> %a0)
   ret i16 %1
 }
@@ -762,14 +709,6 @@ define i8 @test_v2i8(<2 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.or.i8.v2i8(<2 x i8> %a0)
   ret i8 %1
 }
@@ -808,16 +747,6 @@ define i8 @test_v4i8(<4 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v4i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.or.i8.v4i8(<4 x i8> %a0)
   ret i8 %1
 }
@@ -862,18 +791,6 @@ define i8 @test_v8i8(<8 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v8i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.or.i8.v8i8(<8 x i8> %a0)
   ret i8 %1
 }
@@ -924,20 +841,6 @@ define i8 @test_v16i8(<16 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v16i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.or.i8.v16i8(<16 x i8> %a0)
   ret i8 %1
 }

diff --git a/test/CodeGen/X86/vector-reduce-xor-widen.ll b/test/CodeGen/X86/vector-reduce-xor-widen.ll
index 62bbba55f5ca16b0e64624a2eb4aa2fe50293846..7486a190f9f8669ef1cce191d1df9cf7c4aea8e3 100644 (file)
--- a/test/CodeGen/X86/vector-reduce-xor-widen.ll
+++ b/test/CodeGen/X86/vector-reduce-xor-widen.ll
@@ -3,8 +3,8 @@
 ; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+sse4.1 | FileCheck %s --check-prefix=ALL --check-prefix=SSE --check-prefix=SSE41
 ; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX1
 ; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX2
-; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw | FileCheck %s --check-prefix=ALL --check-prefix=AVX512 --check-prefix=AVX512BW
-; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw,+avx512vl | FileCheck %s --check-prefix=ALL --check-prefix=AVX512 --check-prefix=AVX512VL
+; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX512
+; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f,+avx512bw,+avx512vl | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX512
 
 ;
 ; vXi64
@@ -24,13 +24,6 @@ define i64 @test_v2i64(<2 x i64> %a0) {
 ; AVX-NEXT:    vpxor %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovq %xmm0, %rax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i64:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovq %xmm0, %rax
-; AVX512-NEXT:    retq
   %1 = call i64 @llvm.experimental.vector.reduce.xor.i64.v2i64(<2 x i64> %a0)
   ret i64 %1
 }
@@ -200,13 +193,6 @@ define i32 @test_v2i32(<2 x i32> %a0) {
 ; AVX-NEXT:    vpxor %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i32:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    retq
   %1 = call i32 @llvm.experimental.vector.reduce.xor.i32.v2i32(<2 x i32> %a0)
   ret i32 %1
 }
@@ -229,15 +215,6 @@ define i32 @test_v4i32(<4 x i32> %a0) {
 ; AVX-NEXT:    vpxor %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v4i32:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    retq
   %1 = call i32 @llvm.experimental.vector.reduce.xor.i32.v4i32(<4 x i32> %a0)
   ret i32 %1
 }
@@ -434,14 +411,6 @@ define i16 @test_v2i16(<2 x i16> %a0) {
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    # kill: def $ax killed $ax killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i16:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    # kill: def $ax killed $ax killed $eax
-; AVX512-NEXT:    retq
   %1 = call i16 @llvm.experimental.vector.reduce.xor.i16.v2i16(<2 x i16> %a0)
   ret i16 %1
 }
@@ -467,16 +436,6 @@ define i16 @test_v4i16(<4 x i16> %a0) {
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    # kill: def $ax killed $ax killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v4i16:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    # kill: def $ax killed $ax killed $eax
-; AVX512-NEXT:    retq
   %1 = call i16 @llvm.experimental.vector.reduce.xor.i16.v4i16(<4 x i16> %a0)
   ret i16 %1
 }
@@ -506,18 +465,6 @@ define i16 @test_v8i16(<8 x i16> %a0) {
 ; AVX-NEXT:    vmovd %xmm0, %eax
 ; AVX-NEXT:    # kill: def $ax killed $ax killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v8i16:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vmovd %xmm0, %eax
-; AVX512-NEXT:    # kill: def $ax killed $ax killed $eax
-; AVX512-NEXT:    retq
   %1 = call i16 @llvm.experimental.vector.reduce.xor.i16.v8i16(<8 x i16> %a0)
   ret i16 %1
 }
@@ -762,14 +709,6 @@ define i8 @test_v2i8(<2 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v2i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.xor.i8.v2i8(<2 x i8> %a0)
   ret i8 %1
 }
@@ -808,16 +747,6 @@ define i8 @test_v4i8(<4 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v4i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.xor.i8.v4i8(<4 x i8> %a0)
   ret i8 %1
 }
@@ -862,18 +791,6 @@ define i8 @test_v8i8(<8 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v8i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.xor.i8.v8i8(<8 x i8> %a0)
   ret i8 %1
 }
@@ -924,20 +841,6 @@ define i8 @test_v16i8(<16 x i8> %a0) {
 ; AVX-NEXT:    vpextrb $0, %xmm0, %eax
 ; AVX-NEXT:    # kill: def $al killed $al killed $eax
 ; AVX-NEXT:    retq
-;
-; AVX512-LABEL: test_v16i8:
-; AVX512:       # %bb.0:
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3]
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrld $16, %xmm0, %xmm1
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpsrlw $8, %xmm0, %xmm1
-; AVX512-NEXT:    vpxor %xmm1, %xmm0, %xmm0
-; AVX512-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX512-NEXT:    # kill: def $al killed $al killed $eax
-; AVX512-NEXT:    retq
   %1 = call i8 @llvm.experimental.vector.reduce.xor.i8.v16i8(<16 x i8> %a0)
   ret i8 %1
 }