[llvm] [Transforms] Mirror optimizeStrRChr with optimizeStrChr (PR #77685)

[via llvm-commits]

https://github.com/AtariDreams updated https://github.com/llvm/llvm-project/pull/77685

>From 791e4974781c340436a135f2ae0ca51e985aa8c5 Mon Sep 17 00:00:00 2001
From: Rose <83477269+AtariDreams at users.noreply.github.com>
Date: Wed, 10 Jan 2024 15:44:36 -0500
Subject: [PATCH 1/2] [Transforms] Mirror optimizeStrRChr with optimizeStrChr

strchr and strrchr have a lot in common. Basically, they can have most optimizations applied to them, only difference being we work backwards, and the "end" of the string being the first argument, and the "start" being the null terminator. Basically, we can do similar transformations with the same transformations and checks, especially if memrchr does the same as memchr but backwards from the null terminator.
---
 .../lib/Transforms/Utils/SimplifyLibCalls.cpp | 55 +++++++++++++++----
 1 file changed, 44 insertions(+), 11 deletions(-)

diff --git a/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp b/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp
index a7cd68e860e467..d0da4771f698e6 100644
--- a/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp
+++ b/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp
@@ -483,8 +483,7 @@ Value *LibCallSimplifier::optimizeStrChr(CallInst *CI, IRBuilderBase &B) {
     Type *SizeTTy = IntegerType::get(CI->getContext(), SizeTBits);
     return copyFlags(*CI,
                      emitMemChr(SrcStr, CharVal, // include nul.
-                                ConstantInt::get(SizeTTy, Len), B,
-                                DL, TLI));
+                                ConstantInt::get(SizeTTy, Len), B, DL, TLI));
   }
 
   if (CharC->isZero()) {
@@ -523,22 +522,56 @@ Value *LibCallSimplifier::optimizeStrRChr(CallInst *CI, IRBuilderBase &B) {
   ConstantInt *CharC = dyn_cast<ConstantInt>(CharVal);
   annotateNonNullNoUndefBasedOnAccess(CI, 0);
 
+  if (!CharC) {
+    uint64_t Len = GetStringLength(SrcStr);
+    if (Len)
+      annotateDereferenceableBytes(CI, 0, Len);
+    else
+      return nullptr;
+
+    Function *Callee = CI->getCalledFunction();
+    FunctionType *FT = Callee->getFunctionType();
+    unsigned IntBits = TLI->getIntSize();
+    if (!FT->getParamType(1)->isIntegerTy(IntBits)) // memrchr needs 'int'.
+      return nullptr;
+
+    unsigned SizeTBits = TLI->getSizeTSize(*CI->getModule());
+    Type *SizeTTy = IntegerType::get(CI->getContext(), SizeTBits);
+
+    // Try to expand strrchr to the memrchr nonstandard extension if it's
+    // available, or simply fail otherwise.
+    return copyFlags(*CI,
+                     emitMemRChr(SrcStr, CharVal, // include nul.
+                                 ConstantInt::get(SizeTTy, Len), B, DL, TLI));
+  }
+
+  if (CharC->isZero()) {
+    Value *NullPtr = Constant::getNullValue(CI->getType());
+    if (isOnlyUsedInEqualityComparison(CI, NullPtr))
+      // Pre-empt the transformation to strlen below and fold
+      // strrchr(A, '\0') == null to false.
+      return B.CreateIntToPtr(B.getTrue(), CI->getType());
+  }
+
+  // Otherwise, the character is a constant, see if the first argument is
+  // a string literal.  If so, we can constant fold.
   StringRef Str;
   if (!getConstantStringInfo(SrcStr, Str)) {
-    // strrchr(s, 0) -> strchr(s, 0)
-    if (CharC && CharC->isZero())
+    if (CharC->isZero()) // strrchr(p, 0) -> strchr(p, 0)
       return copyFlags(*CI, emitStrChr(SrcStr, '\0', B, TLI));
     return nullptr;
   }
 
-  unsigned SizeTBits = TLI->getSizeTSize(*CI->getModule());
-  Type *SizeTTy = IntegerType::get(CI->getContext(), SizeTBits);
+  // Compute the offset, make sure to handle the case when we're searching for
+  // zero (a weird way to spell strlen).
+  size_t I = (0xFF & CharC->getSExtValue()) == 0
+                 ? Str.size()
+                 : Str.rfind(CharC->getSExtValue());
+  if (I == StringRef::npos) // Didn't find the char.  strrchr returns null.
+    return Constant::getNullValue(CI->getType());
 
-  // Try to expand strrchr to the memrchr nonstandard extension if it's
-  // available, or simply fail otherwise.
-  uint64_t NBytes = Str.size() + 1;   // Include the terminating nul.
-  Value *Size = ConstantInt::get(SizeTTy, NBytes);
-  return copyFlags(*CI, emitMemRChr(SrcStr, CharVal, Size, B, DL, TLI));
+  // strrchr(s+n,c)  -> gep(s+n+i,c)
+  return B.CreateInBoundsGEP(B.getInt8Ty(), SrcStr, B.getInt64(I), "strrchr");
 }
 
 Value *LibCallSimplifier::optimizeStrCmp(CallInst *CI, IRBuilderBase &B) {

>From 2170e6c0f9e7a25120a4dee84f46b8d8dd228f53 Mon Sep 17 00:00:00 2001
From: Rose <83477269+AtariDreams at users.noreply.github.com>
Date: Thu, 11 Jan 2024 10:23:13 -0500
Subject: [PATCH 2/2] test

---
 llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

diff --git a/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp b/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp
index d0da4771f698e6..39ea0ff03396c2 100644
--- a/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp
+++ b/llvm/lib/Transforms/Utils/SimplifyLibCalls.cpp
@@ -557,18 +557,19 @@ Value *LibCallSimplifier::optimizeStrRChr(CallInst *CI, IRBuilderBase &B) {
   // a string literal.  If so, we can constant fold.
   StringRef Str;
   if (!getConstantStringInfo(SrcStr, Str)) {
-    if (CharC->isZero()) // strrchr(p, 0) -> strchr(p, 0)
-      return copyFlags(*CI, emitStrChr(SrcStr, '\0', B, TLI));
+    if (CharC->isZero()) // strrchr(p, 0) -> p + strlen(p)
+      if (Value *StrLen = emitStrLen(SrcStr, B, DL, TLI))
+        return B.CreateInBoundsGEP(B.getInt8Ty(), SrcStr, StrLen, "strrchr");
     return nullptr;
   }
 
   // Compute the offset, make sure to handle the case when we're searching for
   // zero (a weird way to spell strlen).
-  size_t I = (0xFF & CharC->getSExtValue()) == 0
-                 ? Str.size()
-                 : Str.rfind(CharC->getSExtValue());
-  if (I == StringRef::npos) // Didn't find the char.  strrchr returns null.
-    return Constant::getNullValue(CI->getType());
+  // size_t I = (0xFF & CharC->getSExtValue()) == 0
+  //                ? Str.size()
+  //                : Str.rfind(CharC->getSExtValue());
+  // if (I == StringRef::npos) // Didn't find the char.  strrchr returns null.
+  //   return Constant::getNullValue(CI->getType());
 
   // strrchr(s+n,c)  -> gep(s+n+i,c)
   return B.CreateInBoundsGEP(B.getInt8Ty(), SrcStr, B.getInt64(I), "strrchr");