[PATCH] D63448: [ARM] MVE shuffle broadcast costs

[Dave Green via Phabricator via llvm-commits]

dmgreen created this revision.
dmgreen added reviewers: t.p.northover, samparker, SjoerdMeijer, simon_tatham.
Herald added subscribers: hiraditya, kristof.beyls, javed.absar.
Herald added a project: LLVM.
dmgreen added a parent revision: D63447: [ARM] Put some of the TTI costmodel behind hasNeon calls..

A VDUP will perform a vector broadcast in a single instruction. Update the cost model for MVE accordingly.

Code originally by David Sherwood.


https://reviews.llvm.org/D63448

Files:
  llvm/lib/Target/ARM/ARMTargetTransformInfo.cpp
  llvm/test/Analysis/CostModel/ARM/shuffle.ll


Index: llvm/test/Analysis/CostModel/ARM/shuffle.ll
===================================================================
--- llvm/test/Analysis/CostModel/ARM/shuffle.ll
+++ llvm/test/Analysis/CostModel/ARM/shuffle.ll
@@ -5,16 +5,16 @@
 define void @broadcast() {
 ; CHECK-MVE-LABEL: 'broadcast'
 ; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %v7 = shufflevector <2 x i8> undef, <2 x i8> undef, <2 x i32> zeroinitializer
-; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %v8 = shufflevector <4 x i8> undef, <4 x i8> undef, <4 x i32> zeroinitializer
-; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %v9 = shufflevector <8 x i8> undef, <8 x i8> undef, <8 x i32> zeroinitializer
-; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 17 for instruction: %v10 = shufflevector <16 x i8> undef, <16 x i8> undef, <16 x i32> zeroinitializer
+; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %v8 = shufflevector <4 x i8> undef, <4 x i8> undef, <4 x i32> zeroinitializer
+; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %v9 = shufflevector <8 x i8> undef, <8 x i8> undef, <8 x i32> zeroinitializer
+; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %v10 = shufflevector <16 x i8> undef, <16 x i8> undef, <16 x i32> zeroinitializer
 ; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %v11 = shufflevector <2 x i16> undef, <2 x i16> undef, <2 x i32> zeroinitializer
-; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %v12 = shufflevector <4 x i16> undef, <4 x i16> undef, <4 x i32> zeroinitializer
-; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 9 for instruction: %v13 = shufflevector <8 x i16> undef, <8 x i16> undef, <8 x i32> zeroinitializer
+; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %v12 = shufflevector <4 x i16> undef, <4 x i16> undef, <4 x i32> zeroinitializer
+; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %v13 = shufflevector <8 x i16> undef, <8 x i16> undef, <8 x i32> zeroinitializer
 ; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %v14 = shufflevector <2 x i32> undef, <2 x i32> undef, <2 x i32> zeroinitializer
-; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %v15 = shufflevector <4 x i32> undef, <4 x i32> undef, <4 x i32> zeroinitializer
-; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 3 for instruction: %v16 = shufflevector <2 x float> undef, <2 x float> undef, <2 x i32> zeroinitializer
-; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 5 for instruction: %v17 = shufflevector <4 x float> undef, <4 x float> undef, <4 x i32> zeroinitializer
+; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %v15 = shufflevector <4 x i32> undef, <4 x i32> undef, <4 x i32> zeroinitializer
+; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %v16 = shufflevector <2 x float> undef, <2 x float> undef, <2 x i32> zeroinitializer
+; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 1 for instruction: %v17 = shufflevector <4 x float> undef, <4 x float> undef, <4 x i32> zeroinitializer
 ; CHECK-MVE-NEXT:  Cost Model: Unknown cost for instruction: %v18 = shufflevector <8 x half> undef, <8 x half> undef, <4 x i32> zeroinitializer
 ; CHECK-MVE-NEXT:  Cost Model: Found an estimated cost of 0 for instruction: ret void
 ;
Index: llvm/lib/Target/ARM/ARMTargetTransformInfo.cpp
===================================================================
--- llvm/lib/Target/ARM/ARMTargetTransformInfo.cpp
+++ llvm/lib/Target/ARM/ARMTargetTransformInfo.cpp
@@ -513,6 +513,23 @@
         return LT.first * Entry->Cost;
     }
   }
+  if (ST->hasMVEIntegerOps()) {
+    if (Kind == TTI::SK_Broadcast) {
+      static const CostTblEntry MVEDupTbl[] = {
+          // VDUP handles these cases.
+          {ISD::VECTOR_SHUFFLE, MVT::v4i32, 1},
+          {ISD::VECTOR_SHUFFLE, MVT::v8i16, 1},
+          {ISD::VECTOR_SHUFFLE, MVT::v16i8, 1},
+          {ISD::VECTOR_SHUFFLE, MVT::v4f32, 1},
+          {ISD::VECTOR_SHUFFLE, MVT::v8f16, 1}};
+
+      std::pair<int, MVT> LT = TLI->getTypeLegalizationCost(DL, Tp);
+
+      if (const auto *Entry = CostTableLookup(MVEDupTbl, ISD::VECTOR_SHUFFLE,
+                                              LT.second))
+        return LT.first * Entry->Cost;
+    }
+  }
   return BaseT::getShuffleCost(Kind, Tp, Index, SubTp);
 }
 


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D63448.205127.patch
Type: text/x-patch
Size: 4582 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190617/f3fb979e/attachment.bin>