X86: cvtpi2ps is just an SSE instruction with MMX operands. It has no AVX equivalent.

Give it the right register format so we can also emit it when AVX is enabled.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@183971 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/X86/X86InstrMMX.td b/lib/Target/X86/X86InstrMMX.td
index 07314a092c8b5cfd7180f384a59a10a4f517c669..cb129562aa90d24bb7928c462e2dd2e1788fa09b 100644 (file)
--- a/lib/Target/X86/X86InstrMMX.td
+++ b/lib/Target/X86/X86InstrMMX.td
@@ -189,13 +189,14 @@ multiclass sse12_cvt_pint<bits<8> opc, RegisterClass SrcRC, RegisterClass DstRC,
 multiclass sse12_cvt_pint_3addr<bits<8> opc, RegisterClass SrcRC,
                     RegisterClass DstRC, Intrinsic Int, X86MemOperand x86memop,
                     PatFrag ld_frag, string asm, Domain d> {
-  def irr : PI<opc, MRMSrcReg, (outs DstRC:$dst),(ins DstRC:$src1, SrcRC:$src2),
-              asm, [(set DstRC:$dst, (Int DstRC:$src1, SrcRC:$src2))], 
-              NoItinerary, d>;
-  def irm : PI<opc, MRMSrcMem, (outs DstRC:$dst),
-                   (ins DstRC:$src1, x86memop:$src2), asm,
-              [(set DstRC:$dst, (Int DstRC:$src1, (ld_frag addr:$src2)))], 
-              NoItinerary, d>;
+  def irr : MMXPI<opc, MRMSrcReg, (outs DstRC:$dst),
+                  (ins DstRC:$src1, SrcRC:$src2), asm,
+                  [(set DstRC:$dst, (Int DstRC:$src1, SrcRC:$src2))],
+                  NoItinerary, d>;
+  def irm : MMXPI<opc, MRMSrcMem, (outs DstRC:$dst),
+                  (ins DstRC:$src1, x86memop:$src2), asm,
+                  [(set DstRC:$dst, (Int DstRC:$src1, (ld_frag addr:$src2)))],
+                  NoItinerary, d>;
 }
 
 //===----------------------------------------------------------------------===//

diff --git a/test/CodeGen/X86/mmx-builtins.ll b/test/CodeGen/X86/mmx-builtins.ll
index a8d33f43da013960f75937a19ff4ec0879e9ff2e..f5b3f765fe6705a4b1d1854d3370e1ff4246da2e 100644 (file)
--- a/test/CodeGen/X86/mmx-builtins.ll
+++ b/test/CodeGen/X86/mmx-builtins.ll
@@ -1337,3 +1337,11 @@ entry:
   %7 = extractelement <1 x i64> %6, i32 0
   ret i64 %7
 }
+
+define <4 x float> @test89(<4 x float> %a, x86_mmx %b) nounwind {
+; CHECK: cvtpi2ps
+  %c = tail call <4 x float> @llvm.x86.sse.cvtpi2ps(<4 x float> %a, x86_mmx %b)
+  ret <4 x float> %c
+}
+
+declare <4 x float> @llvm.x86.sse.cvtpi2ps(<4 x float>, x86_mmx) nounwind readnone