[PATCH] D131672: [instcombine] Optimise for zero initialisation of product given fast flags are enabled

Zain Jaffal via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Tue Aug 16 09:14:42 PDT 2022


zjaffal updated this revision to Diff 453038.
zjaffal added a comment.

Updating D131672 <https://reviews.llvm.org/D131672>: [instcombine] Optimise for zero initialisation of product given fast flags are enabled


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D131672/new/

https://reviews.llvm.org/D131672

Files:
  llvm/lib/Analysis/ValueTracking.cpp
  llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
  llvm/test/Transforms/InstCombine/remove-loop-phi-fastmul.ll


Index: llvm/test/Transforms/InstCombine/remove-loop-phi-fastmul.ll
===================================================================
--- llvm/test/Transforms/InstCombine/remove-loop-phi-fastmul.ll
+++ llvm/test/Transforms/InstCombine/remove-loop-phi-fastmul.ll
@@ -6,15 +6,11 @@
 ; CHECK-NEXT:    br label [[FOR_BODY:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[I_02:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ [[INC:%.*]], [[FOR_BODY]] ]
-; CHECK-NEXT:    [[F_PROD_01:%.*]] = phi double [ 0.000000e+00, [[ENTRY]] ], [ [[MUL:%.*]], [[FOR_BODY]] ]
-; CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [1000 x double], ptr [[ARR_D:%.*]], i64 0, i64 [[I_02]]
-; CHECK-NEXT:    [[TMP0:%.*]] = load double, ptr [[ARRAYIDX]], align 8
-; CHECK-NEXT:    [[MUL]] = fmul fast double [[F_PROD_01]], [[TMP0]]
 ; CHECK-NEXT:    [[INC]] = add i64 [[I_02]], 1
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i64 [[INC]], 1000
 ; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[END:%.*]]
 ; CHECK:       end:
-; CHECK-NEXT:    ret double [[MUL]]
+; CHECK-NEXT:    ret double 0.000000e+00
 ;
 entry:
   br label %for.body
@@ -40,15 +36,11 @@
 ; CHECK-NEXT:    br label [[FOR_BODY:%.*]]
 ; CHECK:       for.body:
 ; CHECK-NEXT:    [[I_02:%.*]] = phi i64 [ 0, [[ENTRY:%.*]] ], [ [[INC:%.*]], [[FOR_BODY]] ]
-; CHECK-NEXT:    [[F_PROD_01:%.*]] = phi double [ 0.000000e+00, [[ENTRY]] ], [ [[MUL:%.*]], [[FOR_BODY]] ]
-; CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds [1000 x double], ptr [[ARR_D:%.*]], i64 0, i64 [[I_02]]
-; CHECK-NEXT:    [[TMP0:%.*]] = load double, ptr [[ARRAYIDX]], align 8
-; CHECK-NEXT:    [[MUL]] = fmul nnan nsz double [[F_PROD_01]], [[TMP0]]
 ; CHECK-NEXT:    [[INC]] = add i64 [[I_02]], 1
 ; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i64 [[INC]], 1000
 ; CHECK-NEXT:    br i1 [[CMP]], label [[FOR_BODY]], label [[END:%.*]]
 ; CHECK:       end:
-; CHECK-NEXT:    ret double [[MUL]]
+; CHECK-NEXT:    ret double 0.000000e+00
 ;
 entry:
   br label %for.body
Index: llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
===================================================================
--- llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
+++ llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
@@ -15,6 +15,7 @@
 #include "llvm/ADT/APInt.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/Analysis/InstructionSimplify.h"
+#include "llvm/Analysis/ValueTracking.h"
 #include "llvm/IR/BasicBlock.h"
 #include "llvm/IR/Constant.h"
 #include "llvm/IR/Constants.h"
@@ -672,6 +673,15 @@
     }
   }
 
+  // Given NoNans and NoSignedZeros are enabled
+  // and we have a phi node initalised to zero
+  // We can replace fmul operation within a loop by zero
+  PHINode *PN = nullptr;
+  Value *Start = nullptr, *Step = nullptr;
+  if (matchSimpleRecurrence(&I, PN, Start, Step) && I.hasNoNaNs() &&
+      I.hasNoSignedZeros() && match(Start, m_Zero()))
+    return replaceInstUsesWith(I, Start);
+
   return nullptr;
 }
 
Index: llvm/lib/Analysis/ValueTracking.cpp
===================================================================
--- llvm/lib/Analysis/ValueTracking.cpp
+++ llvm/lib/Analysis/ValueTracking.cpp
@@ -6502,7 +6502,8 @@
     case Instruction::Sub:
     case Instruction::And:
     case Instruction::Or:
-    case Instruction::Mul: {
+    case Instruction::Mul:
+    case Instruction::FMul: {
       Value *LL = LU->getOperand(0);
       Value *LR = LU->getOperand(1);
       // Find a recurrence.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D131672.453038.patch
Type: text/x-patch
Size: 3449 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20220816/73892403/attachment.bin>


More information about the llvm-commits mailing list