[llvm-commits] [llvm] r145403 - in /llvm/trunk: lib/Target/X86/X86InstrSSE.td test/CodeGen/X86/avx-arith.ll

[Elena Demikhovsky]

Author: delena
Date: Tue Nov 29 09:00:45 2011
New Revision: 145403

URL: http://llvm.org/viewvc/llvm-project?rev=145403&view=rev
Log:
Fixed vsqrt.ss intrinsic usage - order of input operands was wrong.
Added a test.
Thanks Bruno for reviewing the patch.


Modified:
    llvm/trunk/lib/Target/X86/X86InstrSSE.td
    llvm/trunk/test/CodeGen/X86/avx-arith.ll

Modified: llvm/trunk/lib/Target/X86/X86InstrSSE.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86InstrSSE.td?rev=145403&r1=145402&r2=145403&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86InstrSSE.td (original)
+++ llvm/trunk/lib/Target/X86/X86InstrSSE.td Tue Nov 29 09:00:45 2011
@@ -2872,7 +2872,7 @@
                 !strconcat(OpcodeStr,
                            "ss\t{$src2, $src1, $dst|$dst, $src1, $src2}"), []>;
   def SSm_Int : SSI<opc, MRMSrcMem, (outs VR128:$dst),
-                (ins ssmem:$src1, VR128:$src2),
+                (ins VR128:$src1, ssmem:$src2),
                 !strconcat(OpcodeStr,
                            "ss\t{$src2, $src1, $dst|$dst, $src1, $src2}"), []>;
 }

Modified: llvm/trunk/test/CodeGen/X86/avx-arith.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/avx-arith.ll?rev=145403&r1=145402&r2=145403&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/avx-arith.ll (original)
+++ llvm/trunk/test/CodeGen/X86/avx-arith.ll Tue Nov 29 09:00:45 2011
@@ -259,3 +259,14 @@
   ret <4 x i64> %x
 }
 
+declare <4 x float> @llvm.x86.sse.sqrt.ss(<4 x float>) nounwind readnone
+
+define <4 x float> @int_sqrt_ss() {
+; CHECK: int_sqrt_ss
+; CHECK: vsqrtss
+ %x0 = load float addrspace(1)* undef, align 8
+ %x1 = insertelement <4 x float> undef, float %x0, i32 0
+ %x2 = call <4 x float> @llvm.x86.sse.sqrt.ss(<4 x float> %x1) nounwind
+ ret <4 x float> %x2
+}
+