Add __builtin_dcbf support for PPC

Summary:
This patch adds support for __builtin_dcbf for PPC.

__builtin_dcbf copies the contents of a modified block from the data cache
to main memory and flushes the copy from the data cache.

Differential revision: https://reviews.llvm.org/D59843

llvm-svn: 359517

clang/docs/LanguageExtensions.rst

@@ -2448,6 +2448,31 @@ Note that the mode argument will modulo 4, so if the integer argument is greater
 than 3, it will only use the least significant two bits of the mode. 
 Namely, ``__builtin_setrnd(102))`` is equal to ``__builtin_setrnd(2)``.
 
+PowerPC cache builtins
+^^^^^^^^^^^^^^^^^^^^^^
+
+The PowerPC architecture specifies instructions implementing cache operations.
+Clang provides builtins that give direct programmer access to these cache
+instructions.
+
+Currently the following builtins are implemented in clang:
+
+``__builtin_dcbf`` copies the contents of a modified block from the data cache
+to main memory and flushes the copy from the data cache.
+
+**Syntax**:
+
+.. code-block:: c
+
+  void __dcbf(const void* addr); /* Data Cache Block Flush */
+
+**Example of Use**:
+
+.. code-block:: c
+
+  int a = 1;
+  __builtin_dcbf (&a);
+
 Extensions for Static Analysis
 ==============================
 

clang/include/clang/Basic/BuiltinsPPC.def

@@ -478,6 +478,9 @@ BUILTIN(__builtin_pack_vector_int128, "V1LLLiULLiULLi", "")
 // Set the floating point rounding mode
 BUILTIN(__builtin_setrnd, "di", "")
 
+// Cache built-ins
+BUILTIN(__builtin_dcbf, "vvC*", "")
+
 // FIXME: Obviously incomplete.
 
 #undef BUILTIN

clang/test/CodeGen/builtins-ppc-cache.c

@@ -0,0 +1,47 @@
+// RUN: %clang_cc1 -triple powerpc64le-unknown-linux-gnu -emit-llvm \
+// RUN:   -o - %s | FileCheck %s
+
+int A;
+int B[5];
+float C;
+float D[5];
+double E;
+double F[5];
+
+void func(int a, int b[], float c, float d[], double e, double f[]) {
+  __builtin_dcbf (&a);
+  // CHECK: @llvm.ppc.dcbf(i8*
+
+  __builtin_dcbf (&A);
+  // CHECK: @llvm.ppc.dcbf(i8*
+
+  __builtin_dcbf (&b[2]);
+  // CHECK: @llvm.ppc.dcbf(i8*
+
+  __builtin_dcbf (&B[2]);
+  // CHECK: @llvm.ppc.dcbf(i8*
+
+  __builtin_dcbf (&c);
+  // CHECK: @llvm.ppc.dcbf(i8*
+
+  __builtin_dcbf (&C);
+  // CHECK: @llvm.ppc.dcbf(i8*
+
+  __builtin_dcbf (&d[2]);
+  // CHECK: @llvm.ppc.dcbf(i8*
+
+  __builtin_dcbf (&D[2]);
+  // CHECK: @llvm.ppc.dcbf(i8*
+
+  __builtin_dcbf (&e);
+  // CHECK: @llvm.ppc.dcbf(i8*
+
+  __builtin_dcbf (&E);
+  // CHECK: @llvm.ppc.dcbf(i8*
+
+  __builtin_dcbf (&f[0]);
+  // CHECK: @llvm.ppc.dcbf(i8*
+
+  __builtin_dcbf (&F[0]);
+  // CHECK: @llvm.ppc.dcbf(i8*
+}

llvm/include/llvm/IR/IntrinsicsPowerPC.td

@@ -18,7 +18,8 @@
 let TargetPrefix = "ppc" in {  // All intrinsics start with "llvm.ppc.".
   // dcba/dcbf/dcbi/dcbst/dcbt/dcbz/dcbzl(PPC970) instructions.
   def int_ppc_dcba  : Intrinsic<[], [llvm_ptr_ty], []>;
-  def int_ppc_dcbf  : Intrinsic<[], [llvm_ptr_ty], []>;
+  def int_ppc_dcbf  : GCCBuiltin<"__builtin_dcbf">,
+                      Intrinsic<[], [llvm_ptr_ty], []>;
   def int_ppc_dcbi  : Intrinsic<[], [llvm_ptr_ty], []>;
   def int_ppc_dcbst : Intrinsic<[], [llvm_ptr_ty], []>;
   def int_ppc_dcbt  : Intrinsic<[], [llvm_ptr_ty],

llvm/test/CodeGen/PowerPC/dcbf.ll

@@ -0,0 +1,15 @@
+; RUN: llc -mtriple=powerpc64le-unknown-linux-gnu < %s \
+; RUN:     -verify-machineinstrs -ppc-asm-full-reg-names \
+; RUN:     -ppc-vsr-nums-as-vr | FileCheck %s
+
+; Function Attrs: nounwind
+define void @dcbf_test(i8* %a) {
+entry:
+  tail call void @llvm.ppc.dcbf(i8* %a)
+; CHECK-LABEL: @dcbf_test
+; CHECK: dcbf 0, r3
+; CHECK-NEXT: blr
+ret void
+}
+
+declare void @llvm.ppc.dcbf(i8*)