[x86] Clean up a couple of test cases with the new update script. Split

one test case that is only partially tested in 32-bits into two test
cases so that the script doesn't generate massive spews of tests for the
cases we don't care about.

llvm-svn: 229955

diff --git a/llvm/test/CodeGen/X86/bswap-vector.ll b/llvm/test/CodeGen/X86/bswap-vector.ll
index 9dc960d..0351b22 100644 (file)
--- a/llvm/test/CodeGen/X86/bswap-vector.ll
+++ b/llvm/test/CodeGen/X86/bswap-vector.ll
@@ -1,7 +1,8 @@
-; RUN: llc < %s -mcpu=x86-64 | FileCheck %s -check-prefix=CHECK-NOSSSE3
-; RUN: llc < %s -mcpu=core2 | FileCheck %s -check-prefix=CHECK-SSSE3
-; RUN: llc < %s -mcpu=core-avx2 | FileCheck %s -check-prefix=CHECK-AVX2
-; RUN: llc < %s -mcpu=core-avx2 -x86-experimental-vector-widening-legalization | FileCheck %s -check-prefix=CHECK-WIDE-AVX2
+; RUN: llc < %s -mcpu=x86-64 | FileCheck %s --check-prefix=CHECK-NOSSSE3
+; RUN: llc < %s -mcpu=core2 | FileCheck %s --check-prefix=CHECK-SSSE3
+; RUN: llc < %s -mcpu=core-avx2 | FileCheck %s --check-prefix=CHECK-AVX2
+; RUN: llc < %s -mcpu=core-avx2 -x86-experimental-vector-widening-legalization | FileCheck %s --check-prefix=CHECK-WIDE-AVX2
+
 target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128"
 target triple = "x86_64-unknown-linux-gnu"
 
@@ -9,165 +10,381 @@ declare <8 x i16> @llvm.bswap.v8i16(<8 x i16>)
 declare <4 x i32> @llvm.bswap.v4i32(<4 x i32>)
 declare <2 x i64> @llvm.bswap.v2i64(<2 x i64>)
 
-define <8 x i16> @test1(<8 x i16> %v) #0 {
+define <8 x i16> @test1(<8 x i16> %v) {
+; CHECK-NOSSSE3-LABEL: test1:
+; CHECK-NOSSSE3:       # BB#0: # %entry
+; CHECK-NOSSSE3-NEXT:    pextrw $7, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm1
+; CHECK-NOSSSE3-NEXT:    pextrw $3, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1],xmm2[2],xmm1[2],xmm2[3],xmm1[3]
+; CHECK-NOSSSE3-NEXT:    pextrw $5, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm3
+; CHECK-NOSSSE3-NEXT:    pextrw $1, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm1
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1],xmm1[2],xmm3[2],xmm1[3],xmm3[3]
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
+; CHECK-NOSSSE3-NEXT:    pextrw $6, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    pextrw $2, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm3
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm3 = xmm3[0],xmm2[0],xmm3[1],xmm2[1],xmm3[2],xmm2[2],xmm3[3],xmm2[3]
+; CHECK-NOSSSE3-NEXT:    pextrw $4, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm0
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1],xmm0[2],xmm3[2],xmm0[3],xmm3[3]
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; CHECK-NOSSSE3-NEXT:    retq
+;
+; CHECK-SSSE3-LABEL: test1:
+; CHECK-SSSE3:       # BB#0: # %entry
+; CHECK-SSSE3-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14]
+; CHECK-SSSE3-NEXT:    retq
+;
+; CHECK-AVX2-LABEL: test1:
+; CHECK-AVX2:       # BB#0: # %entry
+; CHECK-AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14]
+; CHECK-AVX2-NEXT:    retq
+;
+; CHECK-WIDE-AVX2-LABEL: test1:
+; CHECK-WIDE-AVX2:       # BB#0: # %entry
+; CHECK-WIDE-AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14]
+; CHECK-WIDE-AVX2-NEXT:    retq
 entry:
   %r = call <8 x i16> @llvm.bswap.v8i16(<8 x i16> %v)
   ret <8 x i16> %r
-
-; CHECK-NOSSSE3-LABEL: @test1
-; CHECK-NOSSSE3: rolw
-; CHECK-NOSSSE3: rolw
-; CHECK-NOSSSE3: rolw
-; CHECK-NOSSSE3: rolw
-; CHECK-NOSSSE3: rolw
-; CHECK-NOSSSE3: rolw
-; CHECK-NOSSSE3: rolw
-; CHECK-NOSSSE3: rolw
-; CHECK-NOSSSE3: retq
-
-; CHECK-SSSE3-LABEL: @test1
-; CHECK-SSSE3: pshufb
-; CHECK-SSSE3-NEXT: retq
-
-; CHECK-AVX2-LABEL: @test1
-; CHECK-AVX2: vpshufb
-; CHECK-AVX2-NEXT: retq
-
-; CHECK-WIDE-AVX2-LABEL: @test1
-; CHECK-WIDE-AVX2: vpshufb
-; CHECK-WIDE-AVX2-NEXT: retq
 }
 
-define <4 x i32> @test2(<4 x i32> %v) #0 {
+define <4 x i32> @test2(<4 x i32> %v) {
+; CHECK-NOSSSE3-LABEL: test2:
+; CHECK-NOSSSE3:       # BB#0: # %entry
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[3,1,2,3]
+; CHECK-NOSSSE3-NEXT:    movd %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm1
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm2 = xmm0[1,1,2,3]
+; CHECK-NOSSSE3-NEXT:    movd %xmm2, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm1
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm0
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
+; CHECK-NOSSSE3-NEXT:    movdqa %xmm1, %xmm0
+; CHECK-NOSSSE3-NEXT:    retq
+;
+; CHECK-SSSE3-LABEL: test2:
+; CHECK-SSSE3:       # BB#0: # %entry
+; CHECK-SSSE3-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
+; CHECK-SSSE3-NEXT:    retq
+;
+; CHECK-AVX2-LABEL: test2:
+; CHECK-AVX2:       # BB#0: # %entry
+; CHECK-AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
+; CHECK-AVX2-NEXT:    retq
+;
+; CHECK-WIDE-AVX2-LABEL: test2:
+; CHECK-WIDE-AVX2:       # BB#0: # %entry
+; CHECK-WIDE-AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
+; CHECK-WIDE-AVX2-NEXT:    retq
 entry:
   %r = call <4 x i32> @llvm.bswap.v4i32(<4 x i32> %v)
   ret <4 x i32> %r
-
-; CHECK-NOSSSE3-LABEL: @test2
-; CHECK-NOSSSE3: bswapl
-; CHECK-NOSSSE3: bswapl
-; CHECK-NOSSSE3: bswapl
-; CHECK-NOSSSE3: bswapl
-; CHECK-NOSSSE3: retq
-
-; CHECK-SSSE3-LABEL: @test2
-; CHECK-SSSE3: pshufb
-; CHECK-SSSE3-NEXT: retq
-
-; CHECK-AVX2-LABEL: @test2
-; CHECK-AVX2: vpshufb
-; CHECK-AVX2-NEXT: retq
-
-; CHECK-WIDE-AVX2-LABEL: @test2
-; CHECK-WIDE-AVX2: vpshufb
-; CHECK-WIDE-AVX2-NEXT: retq
 }
 
-define <2 x i64> @test3(<2 x i64> %v) #0 {
+define <2 x i64> @test3(<2 x i64> %v) {
+; CHECK-NOSSSE3-LABEL: test3:
+; CHECK-NOSSSE3:       # BB#0: # %entry
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %rax
+; CHECK-NOSSSE3-NEXT:    bswapq %rax
+; CHECK-NOSSSE3-NEXT:    movd %rax, %xmm1
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %rax
+; CHECK-NOSSSE3-NEXT:    bswapq %rax
+; CHECK-NOSSSE3-NEXT:    movd %rax, %xmm0
+; CHECK-NOSSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
+; CHECK-NOSSSE3-NEXT:    movdqa %xmm1, %xmm0
+; CHECK-NOSSSE3-NEXT:    retq
+;
+; CHECK-SSSE3-LABEL: test3:
+; CHECK-SSSE3:       # BB#0: # %entry
+; CHECK-SSSE3-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8]
+; CHECK-SSSE3-NEXT:    retq
+;
+; CHECK-AVX2-LABEL: test3:
+; CHECK-AVX2:       # BB#0: # %entry
+; CHECK-AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8]
+; CHECK-AVX2-NEXT:    retq
+;
+; CHECK-WIDE-AVX2-LABEL: test3:
+; CHECK-WIDE-AVX2:       # BB#0: # %entry
+; CHECK-WIDE-AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8]
+; CHECK-WIDE-AVX2-NEXT:    retq
 entry:
   %r = call <2 x i64> @llvm.bswap.v2i64(<2 x i64> %v)
   ret <2 x i64> %r
-
-; CHECK-NOSSSE3-LABEL: @test3
-; CHECK-NOSSSE3: bswapq
-; CHECK-NOSSSE3: bswapq
-; CHECK-NOSSSE3: retq
-
-; CHECK-SSSE3-LABEL: @test3
-; CHECK-SSSE3: pshufb
-; CHECK-SSSE3-NEXT: retq
-
-; CHECK-AVX2-LABEL: @test3
-; CHECK-AVX2: vpshufb
-; CHECK-AVX2-NEXT: retq
-
-; CHECK-WIDE-AVX2-LABEL: @test3
-; CHECK-WIDE-AVX2: vpshufb
-; CHECK-WIDE-AVX2-NEXT: retq
 }
 
 declare <16 x i16> @llvm.bswap.v16i16(<16 x i16>)
 declare <8 x i32> @llvm.bswap.v8i32(<8 x i32>)
 declare <4 x i64> @llvm.bswap.v4i64(<4 x i64>)
 
-define <16 x i16> @test4(<16 x i16> %v) #0 {
+define <16 x i16> @test4(<16 x i16> %v) {
+; CHECK-NOSSSE3-LABEL: test4:
+; CHECK-NOSSSE3:       # BB#0: # %entry
+; CHECK-NOSSSE3-NEXT:    pextrw $7, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    pextrw $3, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm3
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm3 = xmm3[0],xmm2[0],xmm3[1],xmm2[1],xmm3[2],xmm2[2],xmm3[3],xmm2[3]
+; CHECK-NOSSSE3-NEXT:    pextrw $5, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm4
+; CHECK-NOSSSE3-NEXT:    pextrw $1, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm4[0],xmm2[1],xmm4[1],xmm2[2],xmm4[2],xmm2[3],xmm4[3]
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1],xmm2[2],xmm3[2],xmm2[3],xmm3[3]
+; CHECK-NOSSSE3-NEXT:    pextrw $6, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm3
+; CHECK-NOSSSE3-NEXT:    pextrw $2, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm4
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm4 = xmm4[0],xmm3[0],xmm4[1],xmm3[1],xmm4[2],xmm3[2],xmm4[3],xmm3[3]
+; CHECK-NOSSSE3-NEXT:    pextrw $4, %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm3
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm0
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1],xmm0[2],xmm3[2],xmm0[3],xmm3[3]
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1],xmm0[2],xmm4[2],xmm0[3],xmm4[3]
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3]
+; CHECK-NOSSSE3-NEXT:    pextrw $7, %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    pextrw $3, %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm3
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm3 = xmm3[0],xmm2[0],xmm3[1],xmm2[1],xmm3[2],xmm2[2],xmm3[3],xmm2[3]
+; CHECK-NOSSSE3-NEXT:    pextrw $5, %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm4
+; CHECK-NOSSSE3-NEXT:    pextrw $1, %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm4[0],xmm2[1],xmm4[1],xmm2[2],xmm4[2],xmm2[3],xmm4[3]
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1],xmm2[2],xmm3[2],xmm2[3],xmm3[3]
+; CHECK-NOSSSE3-NEXT:    pextrw $6, %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm3
+; CHECK-NOSSSE3-NEXT:    pextrw $2, %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm4
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm4 = xmm4[0],xmm3[0],xmm4[1],xmm3[1],xmm4[2],xmm3[2],xmm4[3],xmm3[3]
+; CHECK-NOSSSE3-NEXT:    pextrw $4, %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm3
+; CHECK-NOSSSE3-NEXT:    movd %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    rolw $8, %ax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm1
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1],xmm1[2],xmm3[2],xmm1[3],xmm3[3]
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm4[0],xmm1[1],xmm4[1],xmm1[2],xmm4[2],xmm1[3],xmm4[3]
+; CHECK-NOSSSE3-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3]
+; CHECK-NOSSSE3-NEXT:    retq
+;
+; CHECK-SSSE3-LABEL: test4:
+; CHECK-SSSE3:       # BB#0: # %entry
+; CHECK-SSSE3-NEXT:    movdqa {{.*#+}} xmm2 = [1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14]
+; CHECK-SSSE3-NEXT:    pshufb %xmm2, %xmm0
+; CHECK-SSSE3-NEXT:    pshufb %xmm2, %xmm1
+; CHECK-SSSE3-NEXT:    retq
+;
+; CHECK-AVX2-LABEL: test4:
+; CHECK-AVX2:       # BB#0: # %entry
+; CHECK-AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14,17,16,19,18,21,20,23,22,25,24,27,26,29,28,31,30]
+; CHECK-AVX2-NEXT:    retq
+;
+; CHECK-WIDE-AVX2-LABEL: test4:
+; CHECK-WIDE-AVX2:       # BB#0: # %entry
+; CHECK-WIDE-AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14,17,16,19,18,21,20,23,22,25,24,27,26,29,28,31,30]
+; CHECK-WIDE-AVX2-NEXT:    retq
 entry:
   %r = call <16 x i16> @llvm.bswap.v16i16(<16 x i16> %v)
   ret <16 x i16> %r
-
-; CHECK-SSSE3-LABEL: @test4
-; CHECK-SSSE3: pshufb
-; CHECK-SSSE3: pshufb
-; CHECK-SSSE3-NEXT: retq
-
-; CHECK-AVX2-LABEL: @test4
-; CHECK-AVX2: vpshufb
-; CHECK-AVX2-NEXT: retq
-
-; CHECK-WIDE-AVX2-LABEL: @test4
-; CHECK-WIDE-AVX2: vpshufb
-; CHECK-WIDE-AVX2-NEXT: retq
 }
 
-define <8 x i32> @test5(<8 x i32> %v) #0 {
+define <8 x i32> @test5(<8 x i32> %v) {
+; CHECK-NOSSSE3-LABEL: test5:
+; CHECK-NOSSSE3:       # BB#0: # %entry
+; CHECK-NOSSSE3-NEXT:    movdqa %xmm0, %xmm2
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm2[3,1,2,3]
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm0
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm3 = xmm2[1,1,2,3]
+; CHECK-NOSSSE3-NEXT:    movd %xmm3, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm3
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm3 = xmm3[0],xmm0[0],xmm3[1],xmm0[1]
+; CHECK-NOSSSE3-NEXT:    movd %xmm2, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm0
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm2 = xmm2[2,3,0,1]
+; CHECK-NOSSSE3-NEXT:    movd %xmm2, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1]
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm2 = xmm1[3,1,2,3]
+; CHECK-NOSSSE3-NEXT:    movd %xmm2, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm3 = xmm1[1,1,2,3]
+; CHECK-NOSSSE3-NEXT:    movd %xmm3, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm3
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm3 = xmm3[0],xmm2[0],xmm3[1],xmm2[1]
+; CHECK-NOSSSE3-NEXT:    movd %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
+; CHECK-NOSSSE3-NEXT:    movd %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm1
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1]
+; CHECK-NOSSSE3-NEXT:    movdqa %xmm2, %xmm1
+; CHECK-NOSSSE3-NEXT:    retq
+;
+; CHECK-SSSE3-LABEL: test5:
+; CHECK-SSSE3:       # BB#0: # %entry
+; CHECK-SSSE3-NEXT:    movdqa {{.*#+}} xmm2 = [3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
+; CHECK-SSSE3-NEXT:    pshufb %xmm2, %xmm0
+; CHECK-SSSE3-NEXT:    pshufb %xmm2, %xmm1
+; CHECK-SSSE3-NEXT:    retq
+;
+; CHECK-AVX2-LABEL: test5:
+; CHECK-AVX2:       # BB#0: # %entry
+; CHECK-AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12,19,18,17,16,23,22,21,20,27,26,25,24,31,30,29,28]
+; CHECK-AVX2-NEXT:    retq
+;
+; CHECK-WIDE-AVX2-LABEL: test5:
+; CHECK-WIDE-AVX2:       # BB#0: # %entry
+; CHECK-WIDE-AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12,19,18,17,16,23,22,21,20,27,26,25,24,31,30,29,28]
+; CHECK-WIDE-AVX2-NEXT:    retq
 entry:
   %r = call <8 x i32> @llvm.bswap.v8i32(<8 x i32> %v)
   ret <8 x i32> %r
-
-; CHECK-SSSE3-LABEL: @test5
-; CHECK-SSSE3: pshufb
-; CHECK-SSSE3: pshufb
-; CHECK-SSSE3-NEXT: retq
-
-; CHECK-AVX2-LABEL: @test5
-; CHECK-AVX2: vpshufb
-; CHECK-AVX2-NEXT: retq
-
-; CHECK-WIDE-AVX2-LABEL: @test5
-; CHECK-WIDE-AVX2: vpshufb
-; CHECK-WIDE-AVX2-NEXT: retq
 }
 
-define <4 x i64> @test6(<4 x i64> %v) #0 {
+define <4 x i64> @test6(<4 x i64> %v) {
+; CHECK-NOSSSE3-LABEL: test6:
+; CHECK-NOSSSE3:       # BB#0: # %entry
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %rax
+; CHECK-NOSSSE3-NEXT:    bswapq %rax
+; CHECK-NOSSSE3-NEXT:    movd %rax, %xmm2
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %rax
+; CHECK-NOSSSE3-NEXT:    bswapq %rax
+; CHECK-NOSSSE3-NEXT:    movd %rax, %xmm0
+; CHECK-NOSSSE3-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm0[0]
+; CHECK-NOSSSE3-NEXT:    movd %xmm1, %rax
+; CHECK-NOSSSE3-NEXT:    bswapq %rax
+; CHECK-NOSSSE3-NEXT:    movd %rax, %xmm3
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[2,3,0,1]
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %rax
+; CHECK-NOSSSE3-NEXT:    bswapq %rax
+; CHECK-NOSSSE3-NEXT:    movd %rax, %xmm0
+; CHECK-NOSSSE3-NEXT:    punpcklqdq {{.*#+}} xmm3 = xmm3[0],xmm0[0]
+; CHECK-NOSSSE3-NEXT:    movdqa %xmm2, %xmm0
+; CHECK-NOSSSE3-NEXT:    movdqa %xmm3, %xmm1
+; CHECK-NOSSSE3-NEXT:    retq
+;
+; CHECK-SSSE3-LABEL: test6:
+; CHECK-SSSE3:       # BB#0: # %entry
+; CHECK-SSSE3-NEXT:    movdqa {{.*#+}} xmm2 = [7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8]
+; CHECK-SSSE3-NEXT:    pshufb %xmm2, %xmm0
+; CHECK-SSSE3-NEXT:    pshufb %xmm2, %xmm1
+; CHECK-SSSE3-NEXT:    retq
+;
+; CHECK-AVX2-LABEL: test6:
+; CHECK-AVX2:       # BB#0: # %entry
+; CHECK-AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8,23,22,21,20,19,18,17,16,31,30,29,28,27,26,25,24]
+; CHECK-AVX2-NEXT:    retq
+;
+; CHECK-WIDE-AVX2-LABEL: test6:
+; CHECK-WIDE-AVX2:       # BB#0: # %entry
+; CHECK-WIDE-AVX2-NEXT:    vpshufb {{.*#+}} ymm0 = ymm0[7,6,5,4,3,2,1,0,15,14,13,12,11,10,9,8,23,22,21,20,19,18,17,16,31,30,29,28,27,26,25,24]
+; CHECK-WIDE-AVX2-NEXT:    retq
 entry:
   %r = call <4 x i64> @llvm.bswap.v4i64(<4 x i64> %v)
   ret <4 x i64> %r
-
-; CHECK-SSSE3-LABEL: @test6
-; CHECK-SSSE3: pshufb
-; CHECK-SSSE3: pshufb
-; CHECK-SSSE3-NEXT: retq
-
-; CHECK-AVX2-LABEL: @test6
-; CHECK-AVX2: vpshufb
-; CHECK-AVX2-NEXT: retq
-
-; CHECK-WIDE-AVX2-LABEL: @test6
-; CHECK-WIDE-AVX2: vpshufb
-; CHECK-WIDE-AVX2-NEXT: retq
 }
 
 declare <4 x i16> @llvm.bswap.v4i16(<4 x i16>)
 
-define <4 x i16> @test7(<4 x i16> %v) #0 {
+define <4 x i16> @test7(<4 x i16> %v) {
+; CHECK-NOSSSE3-LABEL: test7:
+; CHECK-NOSSSE3:       # BB#0: # %entry
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[3,1,2,3]
+; CHECK-NOSSSE3-NEXT:    movd %xmm1, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm1
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm2 = xmm0[1,1,2,3]
+; CHECK-NOSSSE3-NEXT:    movd %xmm2, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm2
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm1
+; CHECK-NOSSSE3-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
+; CHECK-NOSSSE3-NEXT:    movd %xmm0, %eax
+; CHECK-NOSSSE3-NEXT:    bswapl %eax
+; CHECK-NOSSSE3-NEXT:    movd %eax, %xmm0
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; CHECK-NOSSSE3-NEXT:    punpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
+; CHECK-NOSSSE3-NEXT:    psrld $16, %xmm1
+; CHECK-NOSSSE3-NEXT:    movdqa %xmm1, %xmm0
+; CHECK-NOSSSE3-NEXT:    retq
+;
+; CHECK-SSSE3-LABEL: test7:
+; CHECK-SSSE3:       # BB#0: # %entry
+; CHECK-SSSE3-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
+; CHECK-SSSE3-NEXT:    psrld $16, %xmm0
+; CHECK-SSSE3-NEXT:    retq
+;
+; CHECK-AVX2-LABEL: test7:
+; CHECK-AVX2:       # BB#0: # %entry
+; CHECK-AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
+; CHECK-AVX2-NEXT:    vpsrld $16, %xmm0, %xmm0
+; CHECK-AVX2-NEXT:    retq
+;
+; CHECK-WIDE-AVX2-LABEL: test7:
+; CHECK-WIDE-AVX2:       # BB#0: # %entry
+; CHECK-WIDE-AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[1,0,3,2,5,4,7,6,9,8,11,10,13,12,15,14]
+; CHECK-WIDE-AVX2-NEXT:    retq
 entry:
   %r = call <4 x i16> @llvm.bswap.v4i16(<4 x i16> %v)
   ret <4 x i16> %r
-
-; CHECK-SSSE3-LABEL: @test7
-; CHECK-SSSE3: pshufb
-; CHECK-SSSE3: psrld $16
-; CHECK-SSSE3-NEXT: retq
-
-; CHECK-AVX2-LABEL: @test7
-; CHECK-AVX2: vpshufb
-; CHECK-AVX2: vpsrld $16
-; CHECK-AVX2-NEXT: retq
-
-; CHECK-WIDE-AVX2-LABEL: @test7
-; CHECK-WIDE-AVX2: vpshufb
-; CHECK-WIDE-AVX2-NEXT: retq
 }
-
-attributes #0 = { nounwind uwtable }
-

diff --git a/llvm/test/CodeGen/X86/mmx-arg-passing-x86-64.ll b/llvm/test/CodeGen/X86/mmx-arg-passing-x86-64.ll
new file mode 100644 (file)
index 0000000..b4a51e4
--- /dev/null
+++ b/llvm/test/CodeGen/X86/mmx-arg-passing-x86-64.ll
@@ -0,0 +1,64 @@
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mattr=+mmx,+sse2 | FileCheck %s --check-prefix=X86-64
+;
+; On Darwin x86-64, v8i8, v4i16, v2i32 values are passed in XMM[0-7].
+; On Darwin x86-64, v1i64 values are passed in 64-bit GPRs.
+
+@g_v8qi = external global <8 x i8>
+
+define void @t3() nounwind  {
+; X86-64-LABEL: t3:
+; X86-64:       ## BB#0:
+; X86-64-NEXT:    movq _g_v8qi@{{.*}}(%rip), %rax
+; X86-64-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; X86-64-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; X86-64-NEXT:    pand {{.*}}(%rip), %xmm0
+; X86-64-NEXT:    packuswb %xmm0, %xmm0
+; X86-64-NEXT:    movd %xmm0, %rax
+; X86-64-NEXT:    movd %rax, %xmm0
+; X86-64-NEXT:    movb $1, %al
+; X86-64-NEXT:    jmp _pass_v8qi ## TAILCALL
+  %tmp3 = load <8 x i8>* @g_v8qi, align 8
+  %tmp3a = bitcast <8 x i8> %tmp3 to x86_mmx
+  %tmp4 = tail call i32 (...)* @pass_v8qi( x86_mmx %tmp3a ) nounwind
+  ret void
+}
+
+define void @t4(x86_mmx %v1, x86_mmx %v2) nounwind  {
+; X86-64-LABEL: t4:
+; X86-64:       ## BB#0:
+; X86-64-NEXT:    movdq2q %xmm1, %mm0
+; X86-64-NEXT:    movq %mm0, -{{[0-9]+}}(%rsp)
+; X86-64-NEXT:    movdq2q %xmm0, %mm0
+; X86-64-NEXT:    movq %mm0, -{{[0-9]+}}(%rsp)
+; X86-64-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
+; X86-64-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
+; X86-64-NEXT:    paddb %xmm0, %xmm1
+; X86-64-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
+; X86-64-NEXT:    pand {{.*}}(%rip), %xmm1
+; X86-64-NEXT:    packuswb %xmm1, %xmm1
+; X86-64-NEXT:    movd %xmm1, %rax
+; X86-64-NEXT:    movd %rax, %xmm0
+; X86-64-NEXT:    movb $1, %al
+; X86-64-NEXT:    jmp _pass_v8qi ## TAILCALL
+  %v1a = bitcast x86_mmx %v1 to <8 x i8>
+  %v2b = bitcast x86_mmx %v2 to <8 x i8>
+  %tmp3 = add <8 x i8> %v1a, %v2b
+  %tmp3a = bitcast <8 x i8> %tmp3 to x86_mmx
+  %tmp4 = tail call i32 (...)* @pass_v8qi( x86_mmx %tmp3a ) nounwind
+  ret void
+}
+
+define void @t5() nounwind  {
+; X86-64-LABEL: t5:
+; X86-64:       ## BB#0:
+; X86-64-NEXT:    pushq %rax
+; X86-64-NEXT:    xorl %edi, %edi
+; X86-64-NEXT:    callq _pass_v1di
+; X86-64-NEXT:    popq %rax
+; X86-64-NEXT:    retq
+  call void @pass_v1di( <1 x i64> zeroinitializer )
+  ret void
+}
+
+declare i32 @pass_v8qi(...)
+declare void @pass_v1di(<1 x i64>)

diff --git a/llvm/test/CodeGen/X86/mmx-arg-passing.ll b/llvm/test/CodeGen/X86/mmx-arg-passing.ll
index 728c668..4e00310 100644 (file)
--- a/llvm/test/CodeGen/X86/mmx-arg-passing.ll
+++ b/llvm/test/CodeGen/X86/mmx-arg-passing.ll
@@ -1,5 +1,5 @@
-; RUN: llc < %s -mtriple=i386-apple-darwin -mattr=+mmx | FileCheck %s -check-prefix=X86-32
-; RUN: llc < %s -mtriple=x86_64-apple-darwin -mattr=+mmx,+sse2 | FileCheck %s -check-prefix=X86-64
+; RUN: llc < %s -mtriple=i386-apple-darwin -mattr=+mmx | FileCheck %s --check-prefix=X86-32
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mattr=+mmx,+sse2 | FileCheck %s --check-prefix=X86-64
 ;
 ; On Darwin x86-32, v8i8, v4i16, v2i32 values are passed in MM[0-2].
 ; On Darwin x86-32, v1i64 values are passed in memory.  In this example, they
@@ -10,64 +10,40 @@
 @u1 = external global x86_mmx
 
 define void @t1(x86_mmx %v1) nounwind  {
-       store x86_mmx %v1, x86_mmx* @u1, align 8
-       ret void
-
 ; X86-32-LABEL: t1:
-; X86-32: movq %mm0
-
+; X86-32:       ## BB#0:
+; X86-32-NEXT:    movl L_u1$non_lazy_ptr, %eax
+; X86-32-NEXT:    movq %mm0, (%eax)
+; X86-32-NEXT:    retl
+;
 ; X86-64-LABEL: t1:
-; X86-64: movdq2q %xmm0
-; X86-64: movq %mm0
+; X86-64:       ## BB#0:
+; X86-64-NEXT:    movdq2q %xmm0, %mm0
+; X86-64-NEXT:    movq _u1@{{.*}}(%rip), %rax
+; X86-64-NEXT:    movq %mm0, (%rax)
+; X86-64-NEXT:    retq
+       store x86_mmx %v1, x86_mmx* @u1, align 8
+       ret void
 }
 
 @u2 = external global x86_mmx
 
 define void @t2(<1 x i64> %v1) nounwind  {
+; X86-32-LABEL: t2:
+; X86-32:       ## BB#0:
+; X86-32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X86-32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X86-32-NEXT:    movl L_u2$non_lazy_ptr, %edx
+; X86-32-NEXT:    movl %ecx, 4(%edx)
+; X86-32-NEXT:    movl %eax, (%edx)
+; X86-32-NEXT:    retl
+;
+; X86-64-LABEL: t2:
+; X86-64:       ## BB#0:
+; X86-64-NEXT:    movq _u2@{{.*}}(%rip), %rax
+; X86-64-NEXT:    movq %rdi, (%rax)
+; X86-64-NEXT:    retq
         %tmp = bitcast <1 x i64> %v1 to x86_mmx
        store x86_mmx %tmp, x86_mmx* @u2, align 8
        ret void
-
-; X86-32-LABEL: t2:
-; X86-32: movl 4(%esp)
-; X86-32: movl 8(%esp)
-
-; X86-64-LABEL: t2:
-; X86-64: movq %rdi
-}
-
-@g_v8qi = external global <8 x i8>
-
-define void @t3() nounwind  {
-; X86-64-LABEL: t3:
-; X86-64-NOT: movdq2q
-; X86-64: punpcklbw
-  %tmp3 = load <8 x i8>* @g_v8qi, align 8
-  %tmp3a = bitcast <8 x i8> %tmp3 to x86_mmx
-  %tmp4 = tail call i32 (...)* @pass_v8qi( x86_mmx %tmp3a ) nounwind
-  ret void
 }
-
-define void @t4(x86_mmx %v1, x86_mmx %v2) nounwind  {
-; X86-64-LABEL: t4:
-; X86-64: movdq2q
-; X86-64: movdq2q
-; X86-64-NOT: movdq2q
-  %v1a = bitcast x86_mmx %v1 to <8 x i8>
-  %v2b = bitcast x86_mmx %v2 to <8 x i8>
-  %tmp3 = add <8 x i8> %v1a, %v2b
-  %tmp3a = bitcast <8 x i8> %tmp3 to x86_mmx
-  %tmp4 = tail call i32 (...)* @pass_v8qi( x86_mmx %tmp3a ) nounwind
-  ret void
-}
-
-define void @t5() nounwind  {
-; X86-64-LABEL: t5:
-; X86-64-NOT: movdq2q
-; X86-64: xorl  %edi, %edi
-  call void @pass_v1di( <1 x i64> zeroinitializer )
-  ret void
-}
-
-declare i32 @pass_v8qi(...)
-declare void @pass_v1di(<1 x i64>)