From 6916c75fbae29e7b282448fce5e82777c410df42 Mon Sep 17 00:00:00 2001
From: Bruno Cardoso Lopes <bruno.cardoso@gmail.com>
Date: Mon, 23 Feb 2015 15:23:06 +0000
Subject: [PATCH] [X86][MMX] Add tests for pslli, psrli and psrai intrinsics

Add tests to cover the RR form of the pslli, psrli and psrai intrinsics.
In the next commit, the loads are going to be folded and the
instructions use the RM form.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230224 91177308-0d34-0410-b5e6-96231b3b80d8
---
 test/CodeGen/X86/mmx-fold-load.ll | 145 ++++++++++++++++++++++++++++++
 1 file changed, 145 insertions(+)
 create mode 100644 test/CodeGen/X86/mmx-fold-load.ll

diff --git a/test/CodeGen/X86/mmx-fold-load.ll b/test/CodeGen/X86/mmx-fold-load.ll
new file mode 100644
index 00000000000..861706fc704
--- /dev/null
+++ b/test/CodeGen/X86/mmx-fold-load.ll
@@ -0,0 +1,145 @@
+; RUN: llc < %s -march=x86-64 -mattr=+mmx,+sse2 | FileCheck %s
+
+define i64 @t0(<1 x i64>* %a, i32* %b) {
+; CHECK-LABEL: t0:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    movq (%rdi), %mm0
+; CHECK-NEXT:    movd (%rsi), %mm1
+; CHECK-NEXT:    psllq %mm1, %mm0
+; CHECK-NEXT:    movd %mm0, %rax
+; CHECK-NEXT:    retq
+entry:
+  %0 = bitcast <1 x i64>* %a to x86_mmx*
+  %1 = load x86_mmx* %0, align 8
+  %2 = load i32* %b, align 4
+  %3 = tail call x86_mmx @llvm.x86.mmx.pslli.q(x86_mmx %1, i32 %2)
+  %4 = bitcast x86_mmx %3 to i64
+  ret i64 %4
+}
+declare x86_mmx @llvm.x86.mmx.pslli.q(x86_mmx, i32)
+
+define i64 @t1(<1 x i64>* %a, i32* %b) {
+; CHECK-LABEL: t1:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    movq (%rdi), %mm0
+; CHECK-NEXT:    movd (%rsi), %mm1
+; CHECK-NEXT:    psrlq %mm1, %mm0
+; CHECK-NEXT:    movd %mm0, %rax
+; CHECK-NEXT:    retq
+entry:
+  %0 = bitcast <1 x i64>* %a to x86_mmx*
+  %1 = load x86_mmx* %0, align 8
+  %2 = load i32* %b, align 4
+  %3 = tail call x86_mmx @llvm.x86.mmx.psrli.q(x86_mmx %1, i32 %2)
+  %4 = bitcast x86_mmx %3 to i64
+  ret i64 %4
+}
+declare x86_mmx @llvm.x86.mmx.psrli.q(x86_mmx, i32)
+
+define i64 @t2(<1 x i64>* %a, i32* %b) {
+; CHECK-LABEL: t2:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    movq (%rdi), %mm0
+; CHECK-NEXT:    movd (%rsi), %mm1
+; CHECK-NEXT:    psllw %mm1, %mm0
+; CHECK-NEXT:    movd %mm0, %rax
+; CHECK-NEXT:    retq
+entry:
+  %0 = bitcast <1 x i64>* %a to x86_mmx*
+  %1 = load x86_mmx* %0, align 8
+  %2 = load i32* %b, align 4
+  %3 = tail call x86_mmx @llvm.x86.mmx.pslli.w(x86_mmx %1, i32 %2)
+  %4 = bitcast x86_mmx %3 to i64
+  ret i64 %4
+}
+declare x86_mmx @llvm.x86.mmx.pslli.w(x86_mmx, i32)
+
+define i64 @t3(<1 x i64>* %a, i32* %b) {
+; CHECK-LABEL: t3:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    movq (%rdi), %mm0
+; CHECK-NEXT:    movd (%rsi), %mm1
+; CHECK-NEXT:    psrlw %mm1, %mm0
+; CHECK-NEXT:    movd %mm0, %rax
+; CHECK-NEXT:    retq
+entry:
+  %0 = bitcast <1 x i64>* %a to x86_mmx*
+  %1 = load x86_mmx* %0, align 8
+  %2 = load i32* %b, align 4
+  %3 = tail call x86_mmx @llvm.x86.mmx.psrli.w(x86_mmx %1, i32 %2)
+  %4 = bitcast x86_mmx %3 to i64
+  ret i64 %4
+}
+declare x86_mmx @llvm.x86.mmx.psrli.w(x86_mmx, i32)
+
+define i64 @t4(<1 x i64>* %a, i32* %b) {
+; CHECK-LABEL: t4:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    movq (%rdi), %mm0
+; CHECK-NEXT:    movd (%rsi), %mm1
+; CHECK-NEXT:    pslld %mm1, %mm0
+; CHECK-NEXT:    movd %mm0, %rax
+; CHECK-NEXT:    retq
+entry:
+  %0 = bitcast <1 x i64>* %a to x86_mmx*
+  %1 = load x86_mmx* %0, align 8
+  %2 = load i32* %b, align 4
+  %3 = tail call x86_mmx @llvm.x86.mmx.pslli.d(x86_mmx %1, i32 %2)
+  %4 = bitcast x86_mmx %3 to i64
+  ret i64 %4
+}
+declare x86_mmx @llvm.x86.mmx.pslli.d(x86_mmx, i32)
+
+define i64 @t5(<1 x i64>* %a, i32* %b) {
+; CHECK-LABEL: t5:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    movq (%rdi), %mm0
+; CHECK-NEXT:    movd (%rsi), %mm1
+; CHECK-NEXT:    psrld %mm1, %mm0
+; CHECK-NEXT:    movd %mm0, %rax
+; CHECK-NEXT:    retq
+entry:
+  %0 = bitcast <1 x i64>* %a to x86_mmx*
+  %1 = load x86_mmx* %0, align 8
+  %2 = load i32* %b, align 4
+  %3 = tail call x86_mmx @llvm.x86.mmx.psrli.d(x86_mmx %1, i32 %2)
+  %4 = bitcast x86_mmx %3 to i64
+  ret i64 %4
+}
+declare x86_mmx @llvm.x86.mmx.psrli.d(x86_mmx, i32)
+
+define i64 @t6(<1 x i64>* %a, i32* %b) {
+; CHECK-LABEL: t6:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    movq (%rdi), %mm0
+; CHECK-NEXT:    movd (%rsi), %mm1
+; CHECK-NEXT:    psraw %mm1, %mm0
+; CHECK-NEXT:    movd %mm0, %rax
+; CHECK-NEXT:    retq
+entry:
+  %0 = bitcast <1 x i64>* %a to x86_mmx*
+  %1 = load x86_mmx* %0, align 8
+  %2 = load i32* %b, align 4
+  %3 = tail call x86_mmx @llvm.x86.mmx.psrai.w(x86_mmx %1, i32 %2)
+  %4 = bitcast x86_mmx %3 to i64
+  ret i64 %4
+}
+declare x86_mmx @llvm.x86.mmx.psrai.w(x86_mmx, i32)
+
+define i64 @t7(<1 x i64>* %a, i32* %b) {
+; CHECK-LABEL: t7:
+; CHECK:       ## BB#0: ## %entry
+; CHECK-NEXT:    movq (%rdi), %mm0
+; CHECK-NEXT:    movd (%rsi), %mm1
+; CHECK-NEXT:    psrad %mm1, %mm0
+; CHECK-NEXT:    movd %mm0, %rax
+; CHECK-NEXT:    retq
+entry:
+  %0 = bitcast <1 x i64>* %a to x86_mmx*
+  %1 = load x86_mmx* %0, align 8
+  %2 = load i32* %b, align 4
+  %3 = tail call x86_mmx @llvm.x86.mmx.psrai.d(x86_mmx %1, i32 %2)
+  %4 = bitcast x86_mmx %3 to i64
+  ret i64 %4
+}
+declare x86_mmx @llvm.x86.mmx.psrai.d(x86_mmx, i32)
-- 
2.34.1