replica
/
hanchenye-llvm-project
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Define __SSE4A__ when targeting new AMD CPUs. 

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

llvm-svn: 157630
This commit is contained in:
Benjamin Kramer 2012-05-29 17:48:39 +00:00
parent 64fe1c559e
commit 8ac9c22391
2 changed files with 53 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
clang/lib/Basic/Targets.cpp
View File

@ -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)

42
clang/test/Preprocessor/predefined-arch-macros.c
View File

@ -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 ------------------