Remove the pmulld intrinsic and autoupdate it as a vector multiply.

Rewrite the pmulld patterns, and make sure that they fold in loads of
arguments into the instruction.

llvm-svn: 99910

1 files changed, 16 insertions, 0 deletions

diff --git a/llvm/test/CodeGen/X86/pmulld.ll b/llvm/test/CodeGen/X86/pmulld.ll
new file mode 100644
index 00000000000..3ef594112b4
--- /dev/null
+++ b/llvm/test/CodeGen/X86/pmulld.ll
@@ -0,0 +1,16 @@
+; RUN: llc < %s -march=x86-64 -mattr=+sse41 -asm-verbose=0 | FileCheck %s
+
+define <4 x i32> @test1(<4 x i32> %A, <4 x i32> %B) nounwind {
+; CHECK: test1:
+; CHECK-NEXT: pmulld
+  %C = mul <4 x i32> %A, %B
+  ret <4 x i32> %C
+}
+
+define <4 x i32> @test1a(<4 x i32> %A, <4 x i32> *%Bp) nounwind {
+; CHECK: test1a:
+; CHECK-NEXT: pmulld
+  %B = load <4 x i32>* %Bp
+  %C = mul <4 x i32> %A, %B
+  ret <4 x i32> %C
+}