[llvm] 4bbf371 - [SLP][NFC]Improve compile-time by using map {TreeEntry *, Instruction *}

[Alexey Bataev via llvm-commits]

Author: Alexey Bataev
Date: 2023-07-18T13:24:55-07:00
New Revision: 4bbf37199c776304981fcd9107492e87d733dde6

URL: https://github.com/llvm/llvm-project/commit/4bbf37199c776304981fcd9107492e87d733dde6
DIFF: https://github.com/llvm/llvm-project/commit/4bbf37199c776304981fcd9107492e87d733dde6.diff

LOG: [SLP][NFC]Improve compile-time by using map {TreeEntry *, Instruction *}
in getLastInstructionInBundle(), NFC.

Instead of building EntryToLastInstruction before the vectorization,
build it automatically during the calls to getLastInstructionInBundle()
function.

Added: 
    

Modified: 
    llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp

Removed: 
    


################################################################################
diff  --git a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
index 9e179b80298177..32279ed71f3dfb 100644
--- a/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
+++ b/llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
@@ -9015,6 +9015,9 @@ void BoUpSLP::reorderInputsAccordingToOpcode(
 }
 
 Instruction &BoUpSLP::getLastInstructionInBundle(const TreeEntry *E) {
+  auto &Res = EntryToLastInstruction.FindAndConstruct(E);
+  if (Res.second)
+    return *Res.second;
   // Get the basic block this bundle is in. All instructions in the bundle
   // should be in this block (except for extractelement-like instructions with
   // constant indeces).
@@ -9029,7 +9032,7 @@ Instruction &BoUpSLP::getLastInstructionInBundle(const TreeEntry *E) {
            isVectorLikeInstWithConstOps(I);
   }));
 
-  auto &&FindLastInst = [E, Front, this, &BB]() {
+  auto FindLastInst = [&]() {
     Instruction *LastInst = Front;
     for (Value *V : E->Scalars) {
       auto *I = dyn_cast<Instruction>(V);
@@ -9065,7 +9068,7 @@ Instruction &BoUpSLP::getLastInstructionInBundle(const TreeEntry *E) {
     return LastInst;
   };
 
-  auto &&FindFirstInst = [E, Front, this]() {
+  auto FindFirstInst = [&]() {
     Instruction *FirstInst = Front;
     for (Value *V : E->Scalars) {
       auto *I = dyn_cast<Instruction>(V);
@@ -9105,7 +9108,6 @@ Instruction &BoUpSLP::getLastInstructionInBundle(const TreeEntry *E) {
   if (doesNotNeedToSchedule(E->Scalars) ||
       (E->State != TreeEntry::NeedToGather &&
        all_of(E->Scalars, isVectorLikeInstWithConstOps))) {
-    Instruction *InsertInst;
     if ((E->getOpcode() == Instruction::GetElementPtr &&
          any_of(E->Scalars,
                 [](Value *V) {
@@ -9114,15 +9116,12 @@ Instruction &BoUpSLP::getLastInstructionInBundle(const TreeEntry *E) {
         all_of(E->Scalars, [](Value *V) {
           return !isVectorLikeInstWithConstOps(V) && isUsedOutsideBlock(V);
         }))
-      InsertInst = FindLastInst();
+      Res.second = FindLastInst();
     else
-      InsertInst = FindFirstInst();
-    return *InsertInst;
+      Res.second = FindFirstInst();
+    return *Res.second;
   }
 
-  // The last instruction in the bundle in program order.
-  Instruction *LastInst = nullptr;
-
   // Find the last instruction. The common case should be that BB has been
   // scheduled, and the last instruction is VL.back(). So we start with
   // VL.back() and iterate over schedule data until we reach the end of the
@@ -9135,7 +9134,7 @@ Instruction &BoUpSLP::getLastInstructionInBundle(const TreeEntry *E) {
     if (Bundle && Bundle->isPartOfBundle())
       for (; Bundle; Bundle = Bundle->NextInBundle)
         if (Bundle->OpValue == Bundle->Inst)
-          LastInst = Bundle->Inst;
+          Res.second = Bundle->Inst;
   }
 
   // LastInst can still be null at this point if there's either not an entry
@@ -9156,15 +9155,15 @@ Instruction &BoUpSLP::getLastInstructionInBundle(const TreeEntry *E) {
   // not ideal. However, this should be exceedingly rare since it requires that
   // we both exit early from buildTree_rec and that the bundle be out-of-order
   // (causing us to iterate all the way to the end of the block).
-  if (!LastInst)
-    LastInst = FindLastInst();
-  assert(LastInst && "Failed to find last instruction in bundle");
-  return *LastInst;
+  if (!Res.second)
+    Res.second = FindLastInst();
+  assert(Res.second && "Failed to find last instruction in bundle");
+  return *Res.second;
 }
 
 void BoUpSLP::setInsertPointAfterBundle(const TreeEntry *E) {
   auto *Front = E->getMainOp();
-  Instruction *LastInst = EntryToLastInstruction.lookup(E);
+  Instruction *LastInst = &getLastInstructionInBundle(E);
   assert(LastInst && "Failed to find last instruction in bundle");
   // If the instruction is PHI, set the insert point after all the PHIs.
   bool IsPHI = isa<PHINode>(LastInst);
@@ -9685,7 +9684,7 @@ Value *BoUpSLP::vectorizeOperand(TreeEntry *E, unsigned NodeIdx) {
   IRBuilder<>::InsertPointGuard Guard(Builder);
   if (E->getOpcode() != Instruction::InsertElement &&
       E->getOpcode() != Instruction::PHI) {
-    Instruction *LastInst = EntryToLastInstruction.lookup(E);
+    Instruction *LastInst = &getLastInstructionInBundle(E);
     assert(LastInst && "Failed to find last instruction in bundle");
     Builder.SetInsertPoint(LastInst);
   }
@@ -10704,18 +10703,9 @@ Value *BoUpSLP::vectorizeTree(
   for (auto &BSIter : BlocksSchedules) {
     scheduleBlock(BSIter.second.get());
   }
-
-  // Pre-gather last instructions.
-  for (const std::unique_ptr<TreeEntry> &E : VectorizableTree) {
-    if ((E->State == TreeEntry::NeedToGather &&
-         (!E->getMainOp() || E->Idx > 0)) ||
-        (E->State != TreeEntry::NeedToGather &&
-         E->getOpcode() == Instruction::ExtractValue) ||
-        E->getOpcode() == Instruction::InsertElement)
-        continue;
-    Instruction *LastInst = &getLastInstructionInBundle(E.get());
-    EntryToLastInstruction.try_emplace(E.get(), LastInst);
-  }
+  // Clean Entry-to-LastInstruction table. It can be affected after scheduling,
+  // need to rebuild it.
+  EntryToLastInstruction.clear();
 
   Builder.SetInsertPoint(ReductionRoot ? ReductionRoot
                                        : &F->getEntryBlock().front());