[llvm] r366678 - [X86][SSE] Add EltsFromConsecutiveLoads test case identified in rL366501

[Simon Pilgrim via llvm-commits]

Author: rksimon
Date: Mon Jul 22 05:17:56 2019
New Revision: 366678

URL: http://llvm.org/viewvc/llvm-project?rev=366678&view=rev
Log:
[X86][SSE] Add EltsFromConsecutiveLoads test case identified in rL366501

Test case that led to rL366441 being reverted at rL366501

Modified:
    llvm/trunk/test/CodeGen/X86/load-partial.ll

Modified: llvm/trunk/test/CodeGen/X86/load-partial.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/load-partial.ll?rev=366678&r1=366677&r2=366678&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/load-partial.ll (original)
+++ llvm/trunk/test/CodeGen/X86/load-partial.ll Mon Jul 22 05:17:56 2019
@@ -177,3 +177,62 @@ define <4 x double> @load_double4_0u2u(d
   %7 = shufflevector <4 x double> %6, <4 x double> undef, <4 x i32> <i32 0, i32 0, i32 2, i32 2>
   ret <4 x double> %7
 }
+
+; Test case identified in rL366501
+ at h = local_unnamed_addr global i8 0, align 1
+define i32 @load_partial_illegal_type() {
+; SSE2-LABEL: load_partial_illegal_type:
+; SSE2:       # %bb.0:
+; SSE2-NEXT:    movzwl {{.*}}(%rip), %eax
+; SSE2-NEXT:    movd %eax, %xmm0
+; SSE2-NEXT:    movdqa %xmm0, %xmm1
+; SSE2-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]
+; SSE2-NEXT:    punpcklwd {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3]
+; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[1,1,0,3]
+; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
+; SSE2-NEXT:    movl $2, %eax
+; SSE2-NEXT:    movd %eax, %xmm1
+; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[3,0]
+; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
+; SSE2-NEXT:    andps {{.*}}(%rip), %xmm0
+; SSE2-NEXT:    packuswb %xmm0, %xmm0
+; SSE2-NEXT:    packuswb %xmm0, %xmm0
+; SSE2-NEXT:    movd %xmm0, %eax
+; SSE2-NEXT:    retq
+;
+; SSSE3-LABEL: load_partial_illegal_type:
+; SSSE3:       # %bb.0:
+; SSSE3-NEXT:    movzwl {{.*}}(%rip), %eax
+; SSSE3-NEXT:    movd %eax, %xmm0
+; SSSE3-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],mem[0],xmm0[1],mem[1],xmm0[2],mem[2],xmm0[3],mem[3]
+; SSSE3-NEXT:    movd %xmm0, %eax
+; SSSE3-NEXT:    retq
+;
+; SSE41-LABEL: load_partial_illegal_type:
+; SSE41:       # %bb.0:
+; SSE41-NEXT:    movzwl {{.*}}(%rip), %eax
+; SSE41-NEXT:    movd %eax, %xmm0
+; SSE41-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[0,1,2,3,1],zero,zero,zero,xmm0[u,u,u,u,u,u,u,u]
+; SSE41-NEXT:    movl $2, %eax
+; SSE41-NEXT:    pinsrd $2, %eax, %xmm0
+; SSE41-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[0,4,8,u,u,u,u,u,u,u,u,u,u,u,u,u]
+; SSE41-NEXT:    movd %xmm0, %eax
+; SSE41-NEXT:    retq
+;
+; AVX-LABEL: load_partial_illegal_type:
+; AVX:       # %bb.0:
+; AVX-NEXT:    movzwl {{.*}}(%rip), %eax
+; AVX-NEXT:    vmovd %eax, %xmm0
+; AVX-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,2,3,1],zero,zero,zero,xmm0[u,u,u,u,u,u,u,u]
+; AVX-NEXT:    movl $2, %eax
+; AVX-NEXT:    vpinsrd $2, %eax, %xmm0, %xmm0
+; AVX-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,4,8,u,u,u,u,u,u,u,u,u,u,u,u,u]
+; AVX-NEXT:    vmovd %xmm0, %eax
+; AVX-NEXT:    retq
+  %1 = load <2 x i8>, <2 x i8>* bitcast (i8* @h to <2 x i8>*), align 1
+  %2 = shufflevector <2 x i8> %1, <2 x i8> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
+  %3 = insertelement <4 x i8> %2, i8 2, i32 2
+  %4 = bitcast <4 x i8> %3 to i32
+  ret i32 %4
+}
+