[clang] add sqrt{pd|ps}256

Signed-off-by: Shreeyash Pandey <[email protected]>

Author: bojle

clang/include/clang/Basic/BuiltinsX86.td

@@ -504,13 +504,13 @@ let Features = "avx", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWid
   def vinsertf128_pd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Vector<2, double>, _Constant int)">;
   def vinsertf128_ps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Vector<4, float>, _Constant int)">;
   def vinsertf128_si256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<4, int>, _Constant int)">;
+  def sqrtpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>)">;
+  def sqrtps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>)">;
 }
 
 let Features = "avx", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in {
   def vpermilpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int)">;
   def vpermilps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>, _Constant int)">;
-  def sqrtpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>)">;
-  def sqrtps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>)">;
   def rsqrtps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>)">;
   def rcpps256 : X86Builtin<"_Vector<8, float>(_Vector<8, float>)">;
   def roundpd256 : X86Builtin<"_Vector<4, double>(_Vector<4, double>, _Constant int)">;

clang/lib/AST/ByteCode/InterpBuiltin.cpp

@@ -3829,6 +3829,8 @@ bool InterpretBuiltin(InterpState &S, CodePtr OpPC, const CallExpr *Call,
     return interp__builtin_x86_insert_subvector(S, OpPC, Call, BuiltinID);
   case X86::BI__builtin_ia32_sqrtpd:
   case X86::BI__builtin_ia32_sqrtps:
+  case X86::BI__builtin_ia32_sqrtpd256:
+  case X86::BI__builtin_ia32_sqrtps256:
     return interp__builtin_x86_sqrt(S, OpPC, Call, BuiltinID);
 
   default:

clang/lib/AST/ExprConstant.cpp

@@ -12237,7 +12237,9 @@ bool VectorExprEvaluator::VisitCallExpr(const CallExpr *E) {
     return Success(APValue(ResultElements.data(), ResultElements.size()), E);
   }
   case X86::BI__builtin_ia32_sqrtpd: 
-  case X86::BI__builtin_ia32_sqrtps: {
+  case X86::BI__builtin_ia32_sqrtps: 
+  case X86::BI__builtin_ia32_sqrtpd256:
+  case X86::BI__builtin_ia32_sqrtps256: {
     llvm::errs() << "We are inside sqrtpd/sqrtps\n";
     APValue Source;
     if (!EvaluateAsRValue(Info, E->getArg(0), Source))

clang/lib/Headers/avxintrin.h

@@ -335,7 +335,7 @@ static __inline __m256 __DEFAULT_FN_ATTRS_CONSTEXPR _mm256_mul_ps(__m256 __a,
 ///    A 256-bit vector of [4 x double].
 /// \returns A 256-bit vector of [4 x double] containing the square roots of the
 ///    values in the operand.
-static __inline __m256d __DEFAULT_FN_ATTRS
+static __inline __m256d __DEFAULT_FN_ATTRS_CONSTEXPR
 _mm256_sqrt_pd(__m256d __a)
 {
   return (__m256d)__builtin_ia32_sqrtpd256((__v4df)__a);
@@ -352,7 +352,7 @@ _mm256_sqrt_pd(__m256d __a)
 ///    A 256-bit vector of [8 x float].
 /// \returns A 256-bit vector of [8 x float] containing the square roots of the
 ///    values in the operand.
-static __inline __m256 __DEFAULT_FN_ATTRS
+static __inline __m256 __DEFAULT_FN_ATTRS_CONSTEXPR
 _mm256_sqrt_ps(__m256 __a)
 {
   return (__m256)__builtin_ia32_sqrtps256((__v8sf)__a);

clang/test/CodeGen/X86/avx-builtins.c

@@ -1871,12 +1871,16 @@ __m256d test_mm256_sqrt_pd(__m256d A) {
   return _mm256_sqrt_pd(A);
 }
 
+TEST_CONSTEXPR(match_m256d(_mm256_sqrt_pd(_mm256_set_pd(16.0, 9.0, 4.0, 1.0)), 1.0, 2.0, 3.0, 4.0));
+
 __m256 test_mm256_sqrt_ps(__m256 A) {
   // CHECK-LABEL: test_mm256_sqrt_ps
   // CHECK: call {{.*}}<8 x float> @llvm.sqrt.v8f32(<8 x float> %{{.*}})
   return _mm256_sqrt_ps(A);
 }
 
+TEST_CONSTEXPR(match_m256(_mm256_sqrt_ps(_mm256_set_ps(64.0f, 49.0f, 36.0f, 25.0f, 16.0f, 9.0f, 4.0f, 1.0f)), 1.0f, 2.0f, 3.0f, 4.0f, 5.0f, 6.0f, 7.0f, 8.0f));
+
 void test_mm256_store_pd(double* A, __m256d B) {
   // CHECK-LABEL: test_mm256_store_pd
   // CHECK: store <4 x double> %{{.*}}, ptr %{{.*}}, align 32