[PATCH] D43974: [AArch64] Adjust the cost of integer vector division

[Evandro Menezes via Phabricator via llvm-commits]

evandro created this revision.
evandro added reviewers: t.p.northover, mssimpso.
Herald added subscribers: llvm-commits, hiraditya, kristof.beyls, javed.absar, rengolin.

Since there is no legal instruction for integer vector division, factor in the cost of singling out each element to be used with the scalar division instruction.


Repository:
  rL LLVM

https://reviews.llvm.org/D43974

Files:
  llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp


Index: llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
===================================================================
--- llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
+++ llvm/lib/Target/AArch64/AArch64TargetTransformInfo.cpp
@@ -493,32 +493,49 @@
 
   int ISD = TLI->InstructionOpcodeToISD(Opcode);
 
-  if (ISD == ISD::SDIV &&
-      Opd2Info == TargetTransformInfo::OK_UniformConstantValue &&
-      Opd2PropInfo == TargetTransformInfo::OP_PowerOf2) {
-    // On AArch64, scalar signed division by constants power-of-two are
-    // normally expanded to the sequence ADD + CMP + SELECT + SRA.
-    // The OperandValue properties many not be same as that of previous
-    // operation; conservatively assume OP_None.
-    Cost += getArithmeticInstrCost(Instruction::Add, Ty, Opd1Info, Opd2Info,
-                                   TargetTransformInfo::OP_None,
-                                   TargetTransformInfo::OP_None);
-    Cost += getArithmeticInstrCost(Instruction::Sub, Ty, Opd1Info, Opd2Info,
-                                   TargetTransformInfo::OP_None,
-                                   TargetTransformInfo::OP_None);
-    Cost += getArithmeticInstrCost(Instruction::Select, Ty, Opd1Info, Opd2Info,
-                                   TargetTransformInfo::OP_None,
-                                   TargetTransformInfo::OP_None);
-    Cost += getArithmeticInstrCost(Instruction::AShr, Ty, Opd1Info, Opd2Info,
-                                   TargetTransformInfo::OP_None,
-                                   TargetTransformInfo::OP_None);
-    return Cost;
-  }
-
   switch (ISD) {
   default:
     return Cost + BaseT::getArithmeticInstrCost(Opcode, Ty, Opd1Info, Opd2Info,
                                                 Opd1PropInfo, Opd2PropInfo);
+  case ISD::SDIV:
+    if (Opd2Info == TargetTransformInfo::OK_UniformConstantValue &&
+        Opd2PropInfo == TargetTransformInfo::OP_PowerOf2) {
+      // On AArch64, scalar signed division by constants power-of-two are
+      // normally expanded to the sequence ADD + CMP + SELECT + SRA.
+      // The OperandValue properties many not be same as that of previous
+      // operation; conservatively assume OP_None.
+      Cost += getArithmeticInstrCost(Instruction::Add, Ty, Opd1Info, Opd2Info,
+                                     TargetTransformInfo::OP_None,
+                                     TargetTransformInfo::OP_None);
+      Cost += getArithmeticInstrCost(Instruction::Sub, Ty, Opd1Info, Opd2Info,
+                                     TargetTransformInfo::OP_None,
+                                     TargetTransformInfo::OP_None);
+      Cost += getArithmeticInstrCost(Instruction::Select, Ty, Opd1Info, Opd2Info,
+                                     TargetTransformInfo::OP_None,
+                                     TargetTransformInfo::OP_None);
+      Cost += getArithmeticInstrCost(Instruction::AShr, Ty, Opd1Info, Opd2Info,
+                                     TargetTransformInfo::OP_None,
+                                     TargetTransformInfo::OP_None);
+      return Cost;
+    }
+    LLVM_FALLTHROUGH;
+  case ISD::UDIV:
+    Cost += BaseT::getArithmeticInstrCost(Opcode, Ty, Opd1Info, Opd2Info,
+                                          Opd1PropInfo, Opd2PropInfo);
+    if (Ty->isVectorTy()) {
+      // On AArch64, vector divisions are not supported natively and are
+      // expanded into scalar divisions of each pair of elements.
+      // Conservatively assume OP_None.
+      Cost += getArithmeticInstrCost(Instruction::ExtractElement, Ty, Opd1Info,
+                                     Opd2Info, Opd1PropInfo, Opd2PropInfo);
+      Cost += getArithmeticInstrCost(Instruction::InsertElement, Ty, Opd1Info,
+                                     Opd2Info, Opd1PropInfo, Opd2PropInfo);
+      // TODO: if one of the arguments is a scalar, then it's not necessary to
+      // double the cost of handling the vector elements.
+      Cost += Cost;
+    }
+    return Cost;
+
   case ISD::ADD:
   case ISD::MUL:
   case ISD::XOR:


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D43974.136606.patch
Type: text/x-patch
Size: 4079 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20180301/3ab89975/attachment.bin>