[X86] Remove the X86ISD::MOVLHPD. Lowering doesn't use it and it's not a real instruc...

It was used in patterns, but we had the exact same patterns with Unpckl as well. So now just use Unpckl in the instruction patterns.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@313506 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/X86/X86ISelLowering.cpp b/lib/Target/X86/X86ISelLowering.cpp
index d327c98be69302ad48bad26df7ebd58b24bc67ab..89fff7d06e59f7289c35eb948a9a9bc0dba7a819 100644 (file)
--- a/lib/Target/X86/X86ISelLowering.cpp
+++ b/lib/Target/X86/X86ISelLowering.cpp
@@ -4246,7 +4246,6 @@ static bool isTargetShuffle(unsigned Opcode) {
   case X86ISD::VSHLDQ:
   case X86ISD::VSRLDQ:
   case X86ISD::MOVLHPS:
-  case X86ISD::MOVLHPD:
   case X86ISD::MOVHLPS:
   case X86ISD::MOVLPS:
   case X86ISD::MOVLPD:
@@ -5714,7 +5713,6 @@ static bool getTargetShuffleMask(SDNode *N, MVT VT, bool AllowSentinelZero,
     DecodeMOVDDUPMask(VT, Mask);
     IsUnary = true;
     break;
-  case X86ISD::MOVLHPD:
   case X86ISD::MOVLPD:
   case X86ISD::MOVLPS:
     // Not yet implemented
@@ -24630,7 +24628,6 @@ const char *X86TargetLowering::getTargetNodeName(unsigned Opcode) const {
   case X86ISD::SHUFP:              return "X86ISD::SHUFP";
   case X86ISD::SHUF128:            return "X86ISD::SHUF128";
   case X86ISD::MOVLHPS:            return "X86ISD::MOVLHPS";
-  case X86ISD::MOVLHPD:            return "X86ISD::MOVLHPD";
   case X86ISD::MOVHLPS:            return "X86ISD::MOVHLPS";
   case X86ISD::MOVLPS:             return "X86ISD::MOVLPS";
   case X86ISD::MOVLPD:             return "X86ISD::MOVLPD";

diff --git a/lib/Target/X86/X86ISelLowering.h b/lib/Target/X86/X86ISelLowering.h
index 71caa562eec393b12b5f477ec9d11ddc73acde0e..8f60a33e8d3c599d3c42427773526789ab537932 100644 (file)
--- a/lib/Target/X86/X86ISelLowering.h
+++ b/lib/Target/X86/X86ISelLowering.h
@@ -394,7 +394,6 @@ namespace llvm {
       MOVSHDUP,
       MOVSLDUP,
       MOVLHPS,
-      MOVLHPD,
       MOVHLPS,
       MOVLPS,
       MOVLPD,

diff --git a/lib/Target/X86/X86InstrAVX512.td b/lib/Target/X86/X86InstrAVX512.td
index 4c16e140c2b314fc08b110ff39f14f26f00f970b..4698d2856a2d03e034fca8a25cb421fbea452b7d 100644 (file)
--- a/lib/Target/X86/X86InstrAVX512.td
+++ b/lib/Target/X86/X86InstrAVX512.td
@@ -6075,7 +6075,7 @@ multiclass avx512_mov_hilo_packed<bits<8> opc, string OpcodeStr, SDNode OpNode,
 
 defm VMOVHPSZ128 : avx512_mov_hilo_packed<0x16, "vmovhps", X86Movlhps,
                                   v4f32x_info>, EVEX_CD8<32, CD8VT2>, PS;
-defm VMOVHPDZ128 : avx512_mov_hilo_packed<0x16, "vmovhpd", X86Movlhpd,
+defm VMOVHPDZ128 : avx512_mov_hilo_packed<0x16, "vmovhpd", X86Unpckl,
                                   v2f64x_info>, EVEX_CD8<64, CD8VT1>, PD, VEX_W;
 defm VMOVLPSZ128 : avx512_mov_hilo_packed<0x12, "vmovlps", X86Movlps,
                                   v4f32x_info>, EVEX_CD8<32, CD8VT2>, PS;
@@ -6091,9 +6091,6 @@ let Predicates = [HasAVX512] in {
                (bc_v4f32 (v2i64 (X86vzload addr:$src2)))),
           (VMOVHPSZ128rm VR128X:$src1, addr:$src2)>;
   // VMOVHPD patterns
-  def : Pat<(v2f64 (X86Unpckl VR128X:$src1,
-                    (scalar_to_vector (loadf64 addr:$src2)))),
-           (VMOVHPDZ128rm VR128X:$src1, addr:$src2)>;
   def : Pat<(v2f64 (X86Unpckl VR128X:$src1,
                     (bc_v2f64 (v2i64 (scalar_to_vector (loadi64 addr:$src2)))))),
            (VMOVHPDZ128rm VR128X:$src1, addr:$src2)>;

diff --git a/lib/Target/X86/X86InstrSSE.td b/lib/Target/X86/X86InstrSSE.td
index 22a605c4bef599e6a363a623262e8e08818095fa..2523d6252c1e0fd84a5aaea3e8c4eaeb1ab07b49 100644 (file)
--- a/lib/Target/X86/X86InstrSSE.td
+++ b/lib/Target/X86/X86InstrSSE.td
@@ -995,7 +995,7 @@ let Predicates = [UseSSE2] in {
 //===----------------------------------------------------------------------===//
 
 let AddedComplexity = 20 in {
-  defm MOVH : sse12_mov_hilo_packed<0x16, X86Movlhps, X86Movlhpd, "movhp",
+  defm MOVH : sse12_mov_hilo_packed<0x16, X86Movlhps, X86Unpckl, "movhp",
                                     IIC_SSE_MOV_LH>;
 }
 
@@ -1037,14 +1037,6 @@ let Predicates = [UseAVX] in {
                  (bc_v4f32 (v2i64 (X86vzload addr:$src2)))),
             (VMOVHPSrm VR128:$src1, addr:$src2)>;
 
-  // FIXME: Instead of X86Unpckl, there should be a X86Movlhpd here, the problem
-  // is during lowering, where it's not possible to recognize the load fold
-  // cause it has two uses through a bitcast. One use disappears at isel time
-  // and the fold opportunity reappears.
-  def : Pat<(v2f64 (X86Unpckl VR128:$src1,
-                      (scalar_to_vector (loadf64 addr:$src2)))),
-            (VMOVHPDrm VR128:$src1, addr:$src2)>;
-
   // Also handle an i64 load because that may get selected as a faster way to
   // load the data.
   def : Pat<(v2f64 (X86Unpckl VR128:$src1,
@@ -1075,14 +1067,6 @@ let Predicates = [UseSSE1] in {
 let Predicates = [UseSSE2] in {
   // MOVHPD patterns
 
-  // FIXME: Instead of X86Unpckl, there should be a X86Movlhpd here, the problem
-  // is during lowering, where it's not possible to recognize the load fold
-  // cause it has two uses through a bitcast. One use disappears at isel time
-  // and the fold opportunity reappears.
-  def : Pat<(v2f64 (X86Unpckl VR128:$src1,
-                      (scalar_to_vector (loadf64 addr:$src2)))),
-            (MOVHPDrm VR128:$src1, addr:$src2)>;
-
   // Also handle an i64 load because that may get selected as a faster way to
   // load the data.
   def : Pat<(v2f64 (X86Unpckl VR128:$src1,