[PATCH] D122126: [LoopVectorize] Don't interleave when the number of runtime checks exceeds the threshold

[Mindong Chen via Phabricator via cfe-commits]

This revision was landed with ongoing or failed builds.
This revision was automatically updated to reflect the committed changes.
Closed by commit rG3ed9f603fd59: [LoopVectorize] Don't interleave when the number of runtime checks exceeds the… (authored by TiehuZhang, committed by mdchen).

Changed prior to commit:
  https://reviews.llvm.org/D122126?vs=429110&id=430696#toc

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D122126/new/

https://reviews.llvm.org/D122126

Files:
  llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h
  llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
  llvm/test/Transforms/LoopVectorize/PowerPC/interleaved-pointer-runtime-check-unprofitable.ll


Index: llvm/test/Transforms/LoopVectorize/PowerPC/interleaved-pointer-runtime-check-unprofitable.ll
===================================================================
--- llvm/test/Transforms/LoopVectorize/PowerPC/interleaved-pointer-runtime-check-unprofitable.ll
+++ llvm/test/Transforms/LoopVectorize/PowerPC/interleaved-pointer-runtime-check-unprofitable.ll
@@ -8,7 +8,7 @@
 ; (e.g., runtime-memory-check-threshold, default 8).
 
 ; CHECK-LABEL: @eddy_diff_caleddy_
-; CHECK: vector.memcheck
+; CHECK-NOT: vector.memcheck
 
 define fastcc void @eddy_diff_caleddy_(i64* %wet_cl, i64 %0, i32 %ncol.cast.val) {
 entry:
Index: llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
===================================================================
--- llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
+++ llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
@@ -7492,6 +7492,14 @@
   return VectorizationFactor::Disabled();
 }
 
+bool LoopVectorizationPlanner::requiresTooManyRuntimeChecks() const {
+  unsigned NumRuntimePointerChecks = Requirements.getNumRuntimePointerChecks();
+  return (NumRuntimePointerChecks >
+              VectorizerParams::RuntimeMemoryCheckThreshold &&
+          !Hints.allowReordering()) ||
+         NumRuntimePointerChecks > PragmaVectorizeMemoryCheckThreshold;
+}
+
 Optional<VectorizationFactor>
 LoopVectorizationPlanner::plan(ElementCount UserVF, unsigned UserIC) {
   assert(OrigLoop->isInnermost() && "Inner loop expected.");
@@ -7560,30 +7568,7 @@
     return VectorizationFactor::Disabled();
 
   // Select the optimal vectorization factor.
-  auto SelectedVF = CM.selectVectorizationFactor(VFCandidates);
-
-  // Check if it is profitable to vectorize with runtime checks.
-  unsigned NumRuntimePointerChecks = Requirements.getNumRuntimePointerChecks();
-  if (SelectedVF.Width.getKnownMinValue() > 1 && NumRuntimePointerChecks) {
-    bool PragmaThresholdReached =
-        NumRuntimePointerChecks > PragmaVectorizeMemoryCheckThreshold;
-    bool ThresholdReached =
-        NumRuntimePointerChecks > VectorizerParams::RuntimeMemoryCheckThreshold;
-    if ((ThresholdReached && !Hints.allowReordering()) ||
-        PragmaThresholdReached) {
-      ORE->emit([&]() {
-        return OptimizationRemarkAnalysisAliasing(
-                   DEBUG_TYPE, "CantReorderMemOps", OrigLoop->getStartLoc(),
-                   OrigLoop->getHeader())
-               << "loop not vectorized: cannot prove it is safe to reorder "
-                  "memory operations";
-      });
-      LLVM_DEBUG(dbgs() << "LV: Too many memory checks needed.\n");
-      Hints.emitRemarkWithHints();
-      return VectorizationFactor::Disabled();
-    }
-  }
-  return SelectedVF;
+  return CM.selectVectorizationFactor(VFCandidates);
 }
 
 VPlan &LoopVectorizationPlanner::getBestPlanFor(ElementCount VF) const {
@@ -10499,6 +10484,18 @@
   unsigned IC = 1;
 
   if (MaybeVF) {
+    if (LVP.requiresTooManyRuntimeChecks()) {
+      ORE->emit([&]() {
+        return OptimizationRemarkAnalysisAliasing(
+                   DEBUG_TYPE, "CantReorderMemOps", L->getStartLoc(),
+                   L->getHeader())
+               << "loop not vectorized: cannot prove it is safe to reorder "
+                  "memory operations";
+      });
+      LLVM_DEBUG(dbgs() << "LV: Too many memory checks needed.\n");
+      Hints.emitRemarkWithHints();
+      return false;
+    }
     VF = *MaybeVF;
     // Select the interleave count.
     IC = CM.selectInterleaveCount(VF.Width, *VF.Cost.getValue());
Index: llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h
===================================================================
--- llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h
+++ llvm/lib/Transforms/Vectorize/LoopVectorizationPlanner.h
@@ -320,6 +320,9 @@
   getDecisionAndClampRange(const std::function<bool(ElementCount)> &Predicate,
                            VFRange &Range);
 
+  /// Check if the number of runtime checks exceeds the threshold.
+  bool requiresTooManyRuntimeChecks() const;
+
 protected:
   /// Collect the instructions from the original loop that would be trivially
   /// dead in the vectorized loop if generated.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D122126.430696.patch
Type: text/x-patch
Size: 4159 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20220519/31c8a7c9/attachment.bin>