From: Benjamin Kramer <benny.kra@googlemail.com>
Date: Tue, 29 May 2012 17:48:39 +0000 (+0000)
Subject: Define __SSE4A__ when targeting new AMD CPUs.
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=4dfa5ad7b1f70520f02c754750209138e94b6eb5;p=clang

Define __SSE4A__ when targeting new AMD CPUs.

This doesn't really fit the existing SSELevel so it gets an extra flag.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@157630 91177308-0d34-0410-b5e6-96231b3b80d8
---

diff --git a/lib/Basic/Targets.cpp b/lib/Basic/Targets.cpp
index e5cbc3c0f4..21992363da 100644
--- a/lib/Basic/Targets.cpp
+++ b/lib/Basic/Targets.cpp
@@ -1242,6 +1242,7 @@ class X86TargetInfo : public TargetInfo {
   bool HasBMI;
   bool HasBMI2;
   bool HasPOPCNT;
+  bool HasSSE4a;
   bool HasFMA4;
 
   /// \brief Enumeration of all of the X86 CPUs supported by Clang.
@@ -1388,7 +1389,7 @@ public:
   X86TargetInfo(const std::string& triple)
     : TargetInfo(triple), SSELevel(NoSSE), MMX3DNowLevel(NoMMX3DNow),
       HasAES(false), HasLZCNT(false), HasBMI(false), HasBMI2(false),
-      HasPOPCNT(false), HasFMA4(false), CPU(CK_Generic) {
+      HasPOPCNT(false), HasSSE4a(false), HasFMA4(false), CPU(CK_Generic) {
     BigEndian = false;
     LongDoubleFormat = &llvm::APFloat::x87DoubleExtended;
   }
@@ -1843,6 +1844,11 @@ void X86TargetInfo::HandleTargetFeatures(std::vector<std::string> &Features) {
       continue;
     }
 
+    if (Feature == "sse4a") {
+      HasSSE4a = true;
+      continue;
+    }
+
     if (Feature == "fma4") {
       HasFMA4 = true;
       continue;
@@ -2042,6 +2048,9 @@ void X86TargetInfo::getTargetDefines(const LangOptions &Opts,
   if (HasPOPCNT)
     Builder.defineMacro("__POPCNT__");
 
+  if (HasSSE4a)
+    Builder.defineMacro("__SSE4A__");
+
   if (HasFMA4)
     Builder.defineMacro("__FMA4__");
 
@@ -2108,6 +2117,7 @@ bool X86TargetInfo::hasFeature(StringRef Feature) const {
       .Case("avx2", SSELevel >= AVX2)
       .Case("bmi", HasBMI)
       .Case("bmi2", HasBMI2)
+      .Case("sse4a", HasSSE4a)
       .Case("fma4", HasFMA4)
       .Case("lzcnt", HasLZCNT)
       .Case("mm3dnow", MMX3DNowLevel >= AMD3DNow)
diff --git a/test/Preprocessor/predefined-arch-macros.c b/test/Preprocessor/predefined-arch-macros.c
index 0b9d343b24..f4def04378 100644
--- a/test/Preprocessor/predefined-arch-macros.c
+++ b/test/Preprocessor/predefined-arch-macros.c
@@ -940,5 +940,47 @@
 // CHECK_ATHLON_FX_M64: #define __tune_k8__ 1
 // CHECK_ATHLON_FX_M64: #define __x86_64 1
 // CHECK_ATHLON_FX_M64: #define __x86_64__ 1
+// RUN: %clang -march=amdfam10 -m64 -E -dM %s -o - 2>&1 \
+// RUN:     -target i386-unknown-linux \
+// RUN:   | FileCheck %s -check-prefix=CHECK_AMDFAM10_M64
+// CHECK_AMDFAM10_M64: #define __3dNOW_A__ 1
+// CHECK_AMDFAM10_M64: #define __3dNOW__ 1
+// CHECK_AMDFAM10_M64: #define __MMX__ 1
+// CHECK_AMDFAM10_M64: #define __SSE2_MATH__ 1
+// CHECK_AMDFAM10_M64: #define __SSE2__ 1
+// CHECK_AMDFAM10_M64: #define __SSE3__ 1
+// CHECK_AMDFAM10_M64: #define __SSE4A__ 1
+// CHECK_AMDFAM10_M64: #define __SSE_MATH__ 1
+// CHECK_AMDFAM10_M64: #define __SSE__ 1
+// CHECK_AMDFAM10_M64: #define __amd64 1
+// CHECK_AMDFAM10_M64: #define __amd64__ 1
+// CHECK_AMDFAM10_M64: #define __amdfam10 1
+// CHECK_AMDFAM10_M64: #define __amdfam10__ 1
+// CHECK_AMDFAM10_M64: #define __tune_amdfam10__ 1
+// CHECK_AMDFAM10_M64: #define __x86_64 1
+// CHECK_AMDFAM10_M64: #define __x86_64__ 1
+// RUN: %clang -march=bdver1 -m64 -E -dM %s -o - 2>&1 \
+// RUN:     -target i386-unknown-linux \
+// RUN:   | FileCheck %s -check-prefix=CHECK_BDVER1_M64
+// CHECK_BDVER1_M64: #define __AVX__ 1
+// CHECK_BDVER1_M64-NOT: #define __3dNOW_A__ 1
+// CHECK_BDVER1_M64-NOT: #define __3dNOW__ 1
+// CHECK_BDVER1_M64: #define __MMX__ 1
+// CHECK_BDVER1_M64: #define __SSE2_MATH__ 1
+// CHECK_BDVER1_M64: #define __SSE2__ 1
+// CHECK_BDVER1_M64: #define __SSE3__ 1
+// CHECK_BDVER1_M64: #define __SSE4A__ 1
+// CHECK_BDVER1_M64: #define __SSE4_1__ 1
+// CHECK_BDVER1_M64: #define __SSE4_2__ 1
+// CHECK_BDVER1_M64: #define __SSE_MATH__ 1
+// CHECK_BDVER1_M64: #define __SSE__ 1
+// CHECK_BDVER1_M64: #define __SSSE3__ 1
+// CHECK_BDVER1_M64: #define __amd64 1
+// CHECK_BDVER1_M64: #define __amd64__ 1
+// CHECK_BDVER1_M64: #define __bdver1 1
+// CHECK_BDVER1_M64: #define __bdver1__ 1
+// CHECK_BDVER1_M64: #define __tune_bdver1__ 1
+// CHECK_BDVER1_M64: #define __x86_64 1
+// CHECK_BDVER1_M64: #define __x86_64__ 1
 //
 // End X86/GCC/Linux tests ------------------