Fix the required args count for variadic blocks.

We were emitting calls to blocks as if all arguments were
required --- i.e. with signature (A,B,C,D,...) rather than
(A,B,...).  This patch fixes that and accounts for the
implicit block-context argument as a required argument.
In addition, this patch changes the function type under which
we call unprototyped functions on platforms like x86-64 that
guarantee compatibility of variadic functions with unprototyped
function types;  previously we would always call such functions
under the LLVM type T (...)*, but now we will call them under
the type T (A,B,C,D,...)*.  This last change should have no
material effect except for making the type conventions more
explicit;  it was a side-effect of the most convenient implementation.

llvm-svn: 169588

clang/lib/CodeGen/CGBlocks.cpp

@@ -905,7 +905,7 @@ RValue CodeGenFunction::EmitBlockCallExpr(const CallExpr* E,
 
   const FunctionType *FuncTy = FnType->castAs<FunctionType>();
   const CGFunctionInfo &FnInfo =
-    CGM.getTypes().arrangeFreeFunctionCall(Args, FuncTy);
+    CGM.getTypes().arrangeBlockFunctionCall(Args, FuncTy);
 
   // Cast the function pointer to the right type.
   llvm::Type *BlockFTy = CGM.getTypes().GetFunctionType(FnInfo);

clang/lib/CodeGen/CGCall.cpp

@@ -316,6 +316,37 @@ CodeGenTypes::arrangeGlobalDeclaration(GlobalDecl GD) {
   return arrangeFunctionDeclaration(FD);
 }
 
+/// Arrange a call as unto a free function, except possibly with an
+/// additional number of formal parameters considered required.
+static const CGFunctionInfo &
+arrangeFreeFunctionLikeCall(CodeGenTypes &CGT,
+                            const CallArgList &args,
+                            const FunctionType *fnType,
+                            unsigned numExtraRequiredArgs) {
+  assert(args.size() >= numExtraRequiredArgs);
+
+  // In most cases, there are no optional arguments.
+  RequiredArgs required = RequiredArgs::All;
+
+  // If we have a variadic prototype, the required arguments are the
+  // extra prefix plus the arguments in the prototype.
+  if (const FunctionProtoType *proto = dyn_cast<FunctionProtoType>(fnType)) {
+    if (proto->isVariadic())
+      required = RequiredArgs(proto->getNumArgs() + numExtraRequiredArgs);
+
+  // If we don't have a prototype at all, but we're supposed to
+  // explicitly use the variadic convention for unprototyped calls,
+  // treat all of the arguments as required but preserve the nominal
+  // possibility of variadics.
+  } else if (CGT.CGM.getTargetCodeGenInfo()
+               .isNoProtoCallVariadic(args, cast<FunctionNoProtoType>(fnType))) {
+    required = RequiredArgs(args.size());
+  }
+
+  return CGT.arrangeFreeFunctionCall(fnType->getResultType(), args,
+                                     fnType->getExtInfo(), required);
+}
+
 /// Figure out the rules for calling a function with the given formal
 /// type using the given arguments.  The arguments are necessary
 /// because the function might be unprototyped, in which case it's
@@ -323,17 +354,15 @@ CodeGenTypes::arrangeGlobalDeclaration(GlobalDecl GD) {
 const CGFunctionInfo &
 CodeGenTypes::arrangeFreeFunctionCall(const CallArgList &args,
                                       const FunctionType *fnType) {
-  RequiredArgs required = RequiredArgs::All;
-  if (const FunctionProtoType *proto = dyn_cast<FunctionProtoType>(fnType)) {
-    if (proto->isVariadic())
-      required = RequiredArgs(proto->getNumArgs());
-  } else if (CGM.getTargetCodeGenInfo()
-               .isNoProtoCallVariadic(args, cast<FunctionNoProtoType>(fnType))) {
-    required = RequiredArgs(0);
-  }
+  return arrangeFreeFunctionLikeCall(*this, args, fnType, 0);
+}
 
-  return arrangeFreeFunctionCall(fnType->getResultType(), args,
-                                 fnType->getExtInfo(), required);
+/// A block function call is essentially a free-function call with an
+/// extra implicit argument.
+const CGFunctionInfo &
+CodeGenTypes::arrangeBlockFunctionCall(const CallArgList &args,
+                                       const FunctionType *fnType) {
+  return arrangeFreeFunctionLikeCall(*this, args, fnType, 1);
 }
 
 const CGFunctionInfo &
@@ -865,8 +894,14 @@ CodeGenTypes::GetFunctionType(const CGFunctionInfo &FI) {
     break;
   }
 
-  for (CGFunctionInfo::const_arg_iterator it = FI.arg_begin(),
-         ie = FI.arg_end(); it != ie; ++it) {
+  // Add in all of the required arguments.
+  CGFunctionInfo::const_arg_iterator it = FI.arg_begin(), ie;
+  if (FI.isVariadic()) {
+    ie = it + FI.getRequiredArgs().getNumRequiredArgs();
+  } else {
+    ie = FI.arg_end();
+  }
+  for (; it != ie; ++it) {
     const ABIArgInfo &argAI = it->info;
 
     // Insert a padding type to ensure proper alignment.

clang/lib/CodeGen/CGCall.h

@@ -134,7 +134,7 @@ namespace CodeGen {
     }
 
     bool allowsOptionalArgs() const { return NumRequired != ~0U; }
-    bool getNumRequiredArgs() const {
+    unsigned getNumRequiredArgs() const {
       assert(allowsOptionalArgs());
       return NumRequired;
     }

clang/lib/CodeGen/CGExpr.cpp

@@ -2930,7 +2930,7 @@ RValue CodeGenFunction::EmitCall(QualType CalleeType, llvm::Value *Callee,
   // through an unprototyped function type works like a *non-variadic*
   // call.  The way we make this work is to cast to the exact type
   // of the promoted arguments.
-  if (isa<FunctionNoProtoType>(FnType) && !FnInfo.isVariadic()) {
+  if (isa<FunctionNoProtoType>(FnType)) {
     llvm::Type *CalleeTy = getTypes().GetFunctionType(FnInfo);
     CalleeTy = CalleeTy->getPointerTo();
     Callee = Builder.CreateBitCast(Callee, CalleeTy, "callee.knr.cast");

clang/lib/CodeGen/CodeGenTypes.h

@@ -58,6 +58,7 @@ namespace CodeGen {
 /// CodeGenTypes - This class organizes the cross-module state that is used
 /// while lowering AST types to LLVM types.
 class CodeGenTypes {
+public:
   // Some of this stuff should probably be left on the CGM.
   ASTContext &Context;
   const TargetInfo &Target;
@@ -68,6 +69,7 @@ class CodeGenTypes {
   const CodeGenOptions &CodeGenOpts;
   CodeGenModule &CGM;
 
+private:
   /// The opaque type map for Objective-C interfaces. All direct
   /// manipulation is done by the runtime interfaces, which are
   /// responsible for coercing to the appropriate type; these opaque
@@ -195,6 +197,8 @@ public:
                                                 const CallArgList &args,
                                                 FunctionType::ExtInfo info,
                                                 RequiredArgs required);
+  const CGFunctionInfo &arrangeBlockFunctionCall(const CallArgList &args,
+                                                 const FunctionType *type);
 
   const CGFunctionInfo &arrangeCXXMethodCall(const CallArgList &args,
                                              const FunctionProtoType *type,

clang/lib/CodeGen/TargetInfo.h

@@ -158,10 +158,13 @@ namespace clang {
     ///   - the conventions are substantively different in how they pass
     ///     arguments, because in this case using the variadic convention
     ///     will lead to C99 violations.
-    /// It is not necessarily correct when arguments are passed in the
-    /// same way and some out-of-band information is passed for the
-    /// benefit of variadic callees, as is the case for x86-64.
-    /// In this case the ABI should be consulted.
+    ///
+    /// However, some platforms make the conventions identical except
+    /// for passing additional out-of-band information to a variadic
+    /// function: for example, x86-64 passes the number of SSE
+    /// arguments in %al.  On these platforms, it is desireable to
+    /// call unprototyped functions using the variadic convention so
+    /// that unprototyped calls to varargs functions still succeed.
     virtual bool isNoProtoCallVariadic(const CodeGen::CallArgList &args,
                                        const FunctionNoProtoType *fnType) const;
   };

clang/test/CodeGen/x86_64-arguments.c

@@ -374,3 +374,21 @@ T1 test48(void) {
 // CHECK: memcpy
   return T1_retval;
 }
+
+void test49_helper(double, ...);
+void test49(double d, double e) {
+  test49_helper(d, e);
+}
+// CHECK:    define void @test49(
+// CHECK:      [[T0:%.*]] = load double*
+// CHECK-NEXT: [[T1:%.*]] = load double*
+// CHECK-NEXT: call void (double, ...)* @test49_helper(double [[T0]], double [[T1]])
+
+void test50_helper();
+void test50(double d, double e) {
+  test50_helper(d, e);
+}
+// CHECK:    define void @test50(
+// CHECK:      [[T0:%.*]] = load double*
+// CHECK-NEXT: [[T1:%.*]] = load double*
+// CHECK-NEXT: call void (double, double, ...)* bitcast (void (...)* @test50_helper to void (double, double, ...)*)(double [[T0]], double [[T1]])

clang/test/CodeGenObjC/blocks.m

@@ -100,3 +100,35 @@ void test2(Test2 *x) {
 // CHECK-NEXT: [[T4:%.*]] = load [[WEAK_T]]{{.*}}** [[T3]]
 // CHECK-NEXT: [[WEAKX:%.*]] = getelementptr inbounds [[WEAK_T]]{{.*}}* [[T4]], i32 0, i32 6
 // CHECK-NEXT: [[T0:%.*]] = load [[TEST2]]** [[WEAKX]], align 4
+
+// rdar://problem/12722954
+// Make sure that ... is appropriately positioned in a block call.
+void test3(void (^block)(int, ...)) {
+  block(0, 1, 2, 3);
+}
+// CHECK:    define void @test3(
+// CHECK:      [[BLOCK:%.*]] = alloca void (i32, ...)*, align 4
+// CHECK-NEXT: store void (i32, ...)*
+// CHECK-NEXT: [[T0:%.*]] = load void (i32, ...)** [[BLOCK]], align 4
+// CHECK-NEXT: [[T1:%.*]] = bitcast void (i32, ...)* [[T0]] to [[BLOCK_T:%.*]]*
+// CHECK-NEXT: [[T2:%.*]] = getelementptr inbounds [[BLOCK_T]]* [[T1]], i32 0, i32 3
+// CHECK-NEXT: [[T3:%.*]] = bitcast [[BLOCK_T]]* [[T1]] to i8*
+// CHECK-NEXT: [[T4:%.*]] = load i8** [[T2]]
+// CHECK-NEXT: [[T5:%.*]] = bitcast i8* [[T4]] to void (i8*, i32, ...)*
+// CHECK-NEXT: call void (i8*, i32, ...)* [[T5]](i8* [[T3]], i32 0, i32 1, i32 2, i32 3)
+// CHECK-NEXT: ret void
+
+void test4(void (^block)()) {
+  block(0, 1, 2, 3);
+}
+// CHECK:    define void @test4(
+// CHECK:      [[BLOCK:%.*]] = alloca void (...)*, align 4
+// CHECK-NEXT: store void (...)*
+// CHECK-NEXT: [[T0:%.*]] = load void (...)** [[BLOCK]], align 4
+// CHECK-NEXT: [[T1:%.*]] = bitcast void (...)* [[T0]] to [[BLOCK_T:%.*]]*
+// CHECK-NEXT: [[T2:%.*]] = getelementptr inbounds [[BLOCK_T]]* [[T1]], i32 0, i32 3
+// CHECK-NEXT: [[T3:%.*]] = bitcast [[BLOCK_T]]* [[T1]] to i8*
+// CHECK-NEXT: [[T4:%.*]] = load i8** [[T2]]
+// CHECK-NEXT: [[T5:%.*]] = bitcast i8* [[T4]] to void (i8*, i32, i32, i32, i32)*
+// CHECK-NEXT: call void [[T5]](i8* [[T3]], i32 0, i32 1, i32 2, i32 3)
+// CHECK-NEXT: ret void