From 848d7af956442fbf08310cc2d094035802fbe6ea Mon Sep 17 00:00:00 2001
From: Yingwei Zheng <dtcxzyw2333@gmail.com>
Date: Sat, 6 Jan 2024 02:32:57 +0800
Subject: [CVP] Improve the value solving of select at use (#76700)

This patch improves the value solving of select at use if the condition
is an icmp and we know the result of comparison from
`LVI->getPredicateAt`.

Compile-time impact:
http://llvm-compile-time-tracker.com/compare.php?from=7e405eb722e40c79b7726201d0f76b5dab34ba0f&to=3c315b1ddcb0ad82554b33f08b9356679fae4bb7&stat=instructions:u

|stage1-O3|stage1-ReleaseThinLTO|stage1-ReleaseLTO-g|stage1-O0-g|stage2-O3|stage2-O0-g|stage2-clang|
|--|--|--|--|--|--|--|
|-0.01%|+0.01%|-0.00%|-0.00%|-0.08%|+0.02%|-0.01%|
---
 .../Scalar/CorrelatedValuePropagation.cpp          | 51 +++++++++++-----------
 .../CorrelatedValuePropagation/select.ll           | 23 ++++++++++
 2 files changed, 49 insertions(+), 25 deletions(-)

diff --git a/llvm/lib/Transforms/Scalar/CorrelatedValuePropagation.cpp b/llvm/lib/Transforms/Scalar/CorrelatedValuePropagation.cpp
index c44d374..9235850 100644
--- a/llvm/lib/Transforms/Scalar/CorrelatedValuePropagation.cpp
+++ b/llvm/lib/Transforms/Scalar/CorrelatedValuePropagation.cpp
@@ -94,6 +94,31 @@ STATISTIC(NumUDivURemsNarrowedExpanded,
           "Number of bound udiv's/urem's expanded");
 STATISTIC(NumZExt, "Number of non-negative deductions");
 
+static Constant *getConstantAt(Value *V, Instruction *At, LazyValueInfo *LVI) {
+  if (Constant *C = LVI->getConstant(V, At))
+    return C;
+
+  // TODO: The following really should be sunk inside LVI's core algorithm, or
+  // at least the outer shims around such.
+  auto *C = dyn_cast<CmpInst>(V);
+  if (!C)
+    return nullptr;
+
+  Value *Op0 = C->getOperand(0);
+  Constant *Op1 = dyn_cast<Constant>(C->getOperand(1));
+  if (!Op1)
+    return nullptr;
+
+  LazyValueInfo::Tristate Result = LVI->getPredicateAt(
+      C->getPredicate(), Op0, Op1, At, /*UseBlockValue=*/false);
+  if (Result == LazyValueInfo::Unknown)
+    return nullptr;
+
+  return (Result == LazyValueInfo::True)
+             ? ConstantInt::getTrue(C->getContext())
+             : ConstantInt::getFalse(C->getContext());
+}
+
 static bool processSelect(SelectInst *S, LazyValueInfo *LVI) {
   if (S->getType()->isVectorTy() || isa<Constant>(S->getCondition()))
     return false;
@@ -106,7 +131,7 @@ static bool processSelect(SelectInst *S, LazyValueInfo *LVI) {
       C = LVI->getConstantOnEdge(S->getCondition(), PN->getIncomingBlock(U),
                                  I->getParent(), I);
     else
-      C = LVI->getConstant(S->getCondition(), I);
+      C = getConstantAt(S->getCondition(), I, LVI);
 
     auto *CI = dyn_cast_or_null<ConstantInt>(C);
     if (!CI)
@@ -1109,30 +1134,6 @@ static bool processAnd(BinaryOperator *BinOp, LazyValueInfo *LVI) {
   return true;
 }
 
-
-static Constant *getConstantAt(Value *V, Instruction *At, LazyValueInfo *LVI) {
-  if (Constant *C = LVI->getConstant(V, At))
-    return C;
-
-  // TODO: The following really should be sunk inside LVI's core algorithm, or
-  // at least the outer shims around such.
-  auto *C = dyn_cast<CmpInst>(V);
-  if (!C) return nullptr;
-
-  Value *Op0 = C->getOperand(0);
-  Constant *Op1 = dyn_cast<Constant>(C->getOperand(1));
-  if (!Op1) return nullptr;
-
-  LazyValueInfo::Tristate Result = LVI->getPredicateAt(
-      C->getPredicate(), Op0, Op1, At, /*UseBlockValue=*/false);
-  if (Result == LazyValueInfo::Unknown)
-    return nullptr;
-
-  return (Result == LazyValueInfo::True) ?
-    ConstantInt::getTrue(C->getContext()) :
-    ConstantInt::getFalse(C->getContext());
-}
-
 static bool runImpl(Function &F, LazyValueInfo *LVI, DominatorTree *DT,
                     const SimplifyQuery &SQ) {
   bool FnChanged = false;
diff --git a/llvm/test/Transforms/CorrelatedValuePropagation/select.ll b/llvm/test/Transforms/CorrelatedValuePropagation/select.ll
index 28a8516..9842328 100644
--- a/llvm/test/Transforms/CorrelatedValuePropagation/select.ll
+++ b/llvm/test/Transforms/CorrelatedValuePropagation/select.ll
@@ -372,4 +372,27 @@ define i64 @select_cond_may_undef(i32 %a) {
   ret i64 %max
 }
 
+define i32 @test_solve_select_at_use(i32 %a, i32 %b, i32 %c) {
+; CHECK-LABEL: define i32 @test_solve_select_at_use
+; CHECK-SAME: (i32 [[A:%.*]], i32 [[B:%.*]], i32 [[C:%.*]]) {
+; CHECK-NEXT:  entry:
+; CHECK-NEXT:    [[CMP:%.*]] = icmp slt i32 [[A]], 0
+; CHECK-NEXT:    [[COND:%.*]] = icmp sgt i32 [[A]], -1
+; CHECK-NEXT:    br i1 [[COND]], label [[IF_THEN:%.*]], label [[IF_ELSE:%.*]]
+; CHECK:       if.then:
+; CHECK-NEXT:    ret i32 [[C]]
+; CHECK:       if.else:
+; CHECK-NEXT:    ret i32 [[B]]
+;
+entry:
+  %cmp = icmp slt i32 %a, 0
+  %retval = select i1 %cmp, i32 %b, i32 %c
+  %cond = icmp sgt i32 %a, -1
+  br i1 %cond, label %if.then, label %if.else
+if.then:
+  ret i32 %retval
+if.else:
+  ret i32 %retval
+}
+
 !0 = !{}
-- 
cgit v1.1