[PATCH] D127765: modify the insertion of freeze instructions for SLU pass

[Ruobing Han via Phabricator via llvm-commits]

drcut updated this revision to Diff 436850.
drcut added a comment.

Fix bug for hoisting freeze instruction


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D127765/new/

https://reviews.llvm.org/D127765

Files:
  llvm/lib/Transforms/Scalar/SimpleLoopUnswitch.cpp
  llvm/test/Transforms/SimpleLoopUnswitch/exponential-nontrivial-unswitch-nested.ll
  llvm/test/Transforms/SimpleLoopUnswitch/nontrivial-unswitch.ll


Index: llvm/test/Transforms/SimpleLoopUnswitch/nontrivial-unswitch.ll
===================================================================
--- llvm/test/Transforms/SimpleLoopUnswitch/nontrivial-unswitch.ll
+++ llvm/test/Transforms/SimpleLoopUnswitch/nontrivial-unswitch.ll
@@ -2437,8 +2437,9 @@
   br label %bb3
 ; CHECK-NOT:     br i1 %a
 ;
-; CHECK:         %[[FROZEN:.+]] = freeze i1 %a
-; CHECK-NEXT:    br i1 %[[FROZEN]], label %[[BB_SPLIT_US:.*]], label %[[BB_SPLIT:.*]]
+; CHECK:         %[[FROZENA:.+]] = freeze i1 %a
+; CHECK:         %[[FROZENB:.+]] = freeze i1 %b
+; CHECK-NEXT:    br i1 %[[FROZENA]], label %[[BB_SPLIT_US:.*]], label %[[BB_SPLIT:.*]]
 ;
 ; CHECK-NOT:     br i1 %a
 ; CHECK-NOT:     br i1 %b
Index: llvm/test/Transforms/SimpleLoopUnswitch/exponential-nontrivial-unswitch-nested.ll
===================================================================
--- llvm/test/Transforms/SimpleLoopUnswitch/exponential-nontrivial-unswitch-nested.ll
+++ llvm/test/Transforms/SimpleLoopUnswitch/exponential-nontrivial-unswitch-nested.ll
@@ -35,6 +35,11 @@
 ; RUN:     -passes='loop-mssa(licm,simple-loop-unswitch<nontrivial>),print<loops>' -disable-output 2>&1 | \
 ; RUN:     sort -b -k 1 | FileCheck %s --check-prefixes=LOOP-UNSCALE4-DIV1
 ;
+; RUN: opt < %s -enable-unswitch-cost-multiplier=true \
+; RUN:     -unswitch-num-initial-unscaled-candidates=4 -unswitch-siblings-toplevel-div=1 \
+; RUN:     -passes='loop-unswitch-func,print<loops>' -disable-output 2>&1 | \
+; RUN:     sort -b -k 1 | FileCheck %s --check-prefixes=LOOP-UNSCALE4-DIV1
+;
 ; NB: sort -b is essential here and below, otherwise blanks might lead to different
 ; order depending on locale.
 ;
@@ -43,6 +48,11 @@
 ; RUN:     -passes='loop-mssa(licm,simple-loop-unswitch<nontrivial>),print<loops>' -disable-output 2>&1 | \
 ; RUN:     sort -b -k 1 | FileCheck %s --check-prefixes=LOOP-UNSCALE4-DIV2
 ;
+; RUN: opt < %s -enable-unswitch-cost-multiplier=true \
+; RUN:     -unswitch-num-initial-unscaled-candidates=4 -unswitch-siblings-toplevel-div=2 \
+; RUN:     -passes='loop-unswitch-func,print<loops>' -disable-output 2>&1 | \
+; RUN:     sort -b -k 1 | FileCheck %s --check-prefixes=LOOP-UNSCALE4-DIV2
+;
 ; Get
 ;    2^(num conds) == 2^5 = 32
 ; loop nests when cost multiplier is disabled:
@@ -51,6 +61,10 @@
 ; RUN:     -passes='loop-mssa(licm,simple-loop-unswitch<nontrivial>),print<loops>' -disable-output 2>&1 | \
 ; RUN:	   sort -b -k 1 | FileCheck %s --check-prefixes=LOOP32
 ;
+; RUN: opt < %s -enable-unswitch-cost-multiplier=false \
+; RUN:     -passes='loop-unswitch-func,print<loops>' -disable-output 2>&1 | \
+; RUN:	   sort -b -k 1 | FileCheck %s --check-prefixes=LOOP32
+;
 ; Single loop nest, not unswitched
 ; LOOP1:     Loop at depth 1 containing:
 ; LOOP1:     Loop at depth 2 containing:
Index: llvm/lib/Transforms/Scalar/SimpleLoopUnswitch.cpp
===================================================================
--- llvm/lib/Transforms/Scalar/SimpleLoopUnswitch.cpp
+++ llvm/lib/Transforms/Scalar/SimpleLoopUnswitch.cpp
@@ -2238,8 +2238,19 @@
       BI->setSuccessor(1 - ClonedSucc, LoopPH);
       Value *Cond = skipTrivialSelect(BI->getCondition());
       if (InsertFreeze) {
-        if (!isGuaranteedNotToBeUndefOrPoison(Cond, &AC, BI, &DT))
-          Cond = new FreezeInst(Cond, Cond->getName() + ".fr", BI);
+        if (!isGuaranteedNotToBeUndefOrPoison(Cond, &AC, BI, &DT)) {
+          // directly insert the freeze inst to the place it will be hoisted by
+          // LICM, so that we do not require LICM as a following pass
+          Instruction *InsertPoint = BI;
+          while (Loop *OuterLoop = LI.getLoopFor(InsertPoint->getParent())) {
+            if (OuterLoop->isLoopInvariant(Cond)) {
+              InsertPoint = OuterLoop->getLoopPreheader()->getTerminator();
+            } else {
+              break;
+            }
+          }
+          Cond = new FreezeInst(Cond, Cond->getName() + ".fr", InsertPoint);
+        }
       }
       BI->setCondition(Cond);
       DTUpdates.push_back({DominatorTree::Insert, SplitBB, ClonedPH});


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D127765.436850.patch
Type: text/x-patch
Size: 4073 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20220614/d65eb014/attachment.bin>