[PATCH] D132555: [instcombine] Optimise for zero initialisation of integer product

[Zain Jaffal via Phabricator via llvm-commits]

zjaffal created this revision.
zjaffal added a reviewer: fhahn.
Herald added a subscriber: hiraditya.
Herald added a project: All.
zjaffal requested review of this revision.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

Similar to the work on `D131672` we optimise integer product if the accumlator value of a loop is initalised to zero


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D132555

Files:
  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
@@ -280,15 +280,11 @@
 ; 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 i32 [ 0, [[ENTRY]] ], [ [[MUL:%.*]], [[FOR_BODY]] ]
-; CHECK-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, ptr [[ARR_D:%.*]], i64 [[I_02]]
-; CHECK-NEXT:    [[TMP0:%.*]] = load i32, ptr [[ARRAYIDX]], align 4
-; CHECK-NEXT:    [[MUL]] = mul nsw i32 [[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 i32 [[MUL]]
+; CHECK-NEXT:    ret i32 0
 ;
 entry:
   br label %for.body
Index: llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
===================================================================
--- llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
+++ llvm/lib/Transforms/InstCombine/InstCombineMulDivRem.cpp
@@ -424,6 +424,14 @@
     I.setHasNoUnsignedWrap(true);
   }
 
+  // Optimise loop multiplication when the inital accumulator value is zero.
+  // When we have a PHINode initalised to zero and we repeatedly multiply it
+  // with array elements we can remove the loop.
+  PHINode *PN = nullptr;
+  Value *Start = nullptr, *Step = nullptr;
+  if (matchSimpleRecurrence(&I, PN, Start, Step) && match(Start, m_Zero()))
+    return replaceInstUsesWith(I, Start);
+
   return Changed ? &I : nullptr;
 }
 


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D132555.455182.patch
Type: text/x-patch
Size: 1820 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20220824/34807d78/attachment.bin>