[BOLT] Keep function which is directly called or has prologue

[yubingex007-a11y]

No description provided.

[github-actions]

⚠️ C/C++ code formatter, clang-format found issues in your code. ⚠️

You can test this locally with the following command:

git-clang-format --diff da5a86b53e7d6e7ff7407b16c2c869894493ee99 a846f4f3f4089c516a41db3b64d4cf750bd505ff -- bolt/include/bolt/Core/BinaryFunction.h bolt/include/bolt/Rewrite/RewriteInstance.h bolt/lib/Core/BinaryFunction.cpp bolt/lib/Rewrite/RewriteInstance.cpp

View the diff from clang-format here.

diff --git a/bolt/include/bolt/Core/BinaryFunction.h b/bolt/include/bolt/Core/BinaryFunction.h
index cd6ded17e4..bd1460580e 100644
--- a/bolt/include/bolt/Core/BinaryFunction.h
+++ b/bolt/include/bolt/Core/BinaryFunction.h
@@ -208,7 +208,7 @@ public:
   /// Mark injected functions
   bool IsInjected = false;
 
-  bool IsDirectCalled =false;
+  bool IsDirectCalled = false;
 
   using LSDATypeTableTy = SmallVector<uint64_t, 0>;
 
diff --git a/bolt/lib/Core/BinaryFunction.cpp b/bolt/lib/Core/BinaryFunction.cpp
index 2f7e6a8d36..0e7a19da93 100644
--- a/bolt/lib/Core/BinaryFunction.cpp
+++ b/bolt/lib/Core/BinaryFunction.cpp
@@ -1268,11 +1268,14 @@ Error BinaryFunction::disassemble() {
         bool IsCall = MIB->isCall(Instruction);
         const bool IsCondBranch = MIB->isConditionalBranch(Instruction);
         MCSymbol *TargetSymbol = nullptr;
-        if(IsCall) {
-          if (BinaryFunction *TargetFunc = BC.getBinaryFunctionContainingAddress(TargetAddress)) {
+        if (IsCall) {
+          if (BinaryFunction *TargetFunc =
+                  BC.getBinaryFunctionContainingAddress(TargetAddress)) {
             // direct call here.
             TargetFunc->IsDirectCalled = true;
-            LLVM_DEBUG(dbgs() << "[Disasm] Find function which is directly called: 0x" << Twine::utohexstr(TargetFunc->getAddress()) << "\n");
+            LLVM_DEBUG(dbgs()
+                       << "[Disasm] Find function which is directly called: 0x"
+                       << Twine::utohexstr(TargetFunc->getAddress()) << "\n");
           }
         }
         if (BC.MIB->isUnsupportedBranch(Instruction)) {
diff --git a/bolt/lib/Rewrite/RewriteInstance.cpp b/bolt/lib/Rewrite/RewriteInstance.cpp
index a0768ba870..7168804f8b 100644
--- a/bolt/lib/Rewrite/RewriteInstance.cpp
+++ b/bolt/lib/Rewrite/RewriteInstance.cpp
@@ -811,52 +811,65 @@ Error RewriteInstance::run() {
 }
 
 void RewriteInstance::keepDCPFunction() {
- BC->outs() << "keepDCPFunction!"<<"\n";
-  for(auto &BFI : BC->getBinaryFunctions()) {
+  BC->outs() << "keepDCPFunction!" << "\n";
+  for (auto &BFI : BC->getBinaryFunctions()) {
     BinaryFunction &BF = BFI.second;
-    bool hasPrologue =false;
+    bool hasPrologue = false;
     ErrorOr<ArrayRef<uint8_t>> ErrorOrFunctionData = BF.getData();
     assert(ErrorOrFunctionData && "function data is not available");
     ArrayRef<uint8_t> IData = *ErrorOrFunctionData;
-    // BC->outs()  << "Potential Function Entry Point: 0x" << Twine::utohexstr(IData[0]) << "\n";
-    // Check for common function prologue patterns
-    // push %rbp
-    // mov %rsp %rbp
-    if (BF.getSize()>=4 && (char)IData[0] == '\x55' && (char)IData[1] == '\x48' && (char)IData[2] == '\x89' && (char)IData[3] == '\xe5') {
-        LLVM_DEBUG(dbgs() << "[keepDCPFunction] Find function with prologue in: 0x" << Twine::utohexstr(BF.getAddress()) << "\n");
-        hasPrologue =true;
-    } else if (BF.getSize()>=4 && (char)IData[0] == '\xf3' && (char)IData[1] == '\x0f' && (char)IData[2] == '\x1e' && (char)IData[3] == '\xfa') {
-       LLVM_DEBUG(dbgs() <<  "[keepDCPFunction] Find function with prologue in: 0x" << Twine::utohexstr(BF.getAddress()) << "\n");
-        hasPrologue= true;
+    // BC->outs()  << "Potential Function Entry Point: 0x" <<
+    // Twine::utohexstr(IData[0]) << "\n"; Check for common function prologue
+    // patterns push %rbp mov %rsp %rbp
+    if (BF.getSize() >= 4 && (char)IData[0] == '\x55' &&
+        (char)IData[1] == '\x48' && (char)IData[2] == '\x89' &&
+        (char)IData[3] == '\xe5') {
+      LLVM_DEBUG(
+          dbgs() << "[keepDCPFunction] Find function with prologue in: 0x"
+                 << Twine::utohexstr(BF.getAddress()) << "\n");
+      hasPrologue = true;
+    } else if (BF.getSize() >= 4 && (char)IData[0] == '\xf3' &&
+               (char)IData[1] == '\x0f' && (char)IData[2] == '\x1e' &&
+               (char)IData[3] == '\xfa') {
+      LLVM_DEBUG(
+          dbgs() << "[keepDCPFunction] Find function with prologue in: 0x"
+                 << Twine::utohexstr(BF.getAddress()) << "\n");
+      hasPrologue = true;
     } else {
       hasPrologue = false;
     }
-   LLVM_DEBUG(
-    if(BF.IsDirectCalled) dbgs() << "[keepDCPFunction] Find function which is directly called: 0x" << Twine::utohexstr(BF.getAddress()) << "\n"
-    );
-    if(!BF.IsDirectCalled && !hasPrologue )
+    LLVM_DEBUG(if (BF.IsDirectCalled) dbgs()
+               << "[keepDCPFunction] Find function which is directly called: 0x"
+               << Twine::utohexstr(BF.getAddress()) << "\n");
+    if (!BF.IsDirectCalled && !hasPrologue)
       BF.setIgnored();
   }
 }
 void RewriteInstance::keepPrologueFunction() {
- BC->outs() << "keepPrologueFunction!"<<"\n";
-  for(auto &BFI : BC->getBinaryFunctions()) {
+  BC->outs() << "keepPrologueFunction!" << "\n";
+  for (auto &BFI : BC->getBinaryFunctions()) {
     BinaryFunction &BF = BFI.second;
     ErrorOr<ArrayRef<uint8_t>> ErrorOrFunctionData = BF.getData();
     assert(ErrorOrFunctionData && "function data is not available");
     ArrayRef<uint8_t> IData = *ErrorOrFunctionData;
-    // BC->outs()  << "Potential Function Entry Point: 0x" << Twine::utohexstr(IData[0]) << "\n";
-    // Check for common function prologue patterns
-    // push %rbp
-    // mov %rsp %rbp
-    if (BF.getSize()>=4 && (char)IData[0] == '\x55' && (char)IData[1] == '\x48' && (char)IData[2] == '\x89' && (char)IData[3] == '\xe5') {
-        LLVM_DEBUG(dbgs() << "[keepPrologueFunction] Find function with prologue in: 0x" << Twine::utohexstr(BF.getAddress()) << "\n");
-    } else if (BF.getSize()>=4 && (char)IData[0] == '\xf3' && (char)IData[1] == '\x0f' && (char)IData[2] == '\x1e' && (char)IData[3] == '\xfa') {
-        LLVM_DEBUG(dbgs() <<  "[keepPrologueFunction] Find function with prologue in: 0x" << Twine::utohexstr(BF.getAddress()) << "\n");
+    // BC->outs()  << "Potential Function Entry Point: 0x" <<
+    // Twine::utohexstr(IData[0]) << "\n"; Check for common function prologue
+    // patterns push %rbp mov %rsp %rbp
+    if (BF.getSize() >= 4 && (char)IData[0] == '\x55' &&
+        (char)IData[1] == '\x48' && (char)IData[2] == '\x89' &&
+        (char)IData[3] == '\xe5') {
+      LLVM_DEBUG(
+          dbgs() << "[keepPrologueFunction] Find function with prologue in: 0x"
+                 << Twine::utohexstr(BF.getAddress()) << "\n");
+    } else if (BF.getSize() >= 4 && (char)IData[0] == '\xf3' &&
+               (char)IData[1] == '\x0f' && (char)IData[2] == '\x1e' &&
+               (char)IData[3] == '\xfa') {
+      LLVM_DEBUG(
+          dbgs() << "[keepPrologueFunction] Find function with prologue in: 0x"
+                 << Twine::utohexstr(BF.getAddress()) << "\n");
     } else {
       BF.setIgnored();
     }
-
   }
 }