//==--- OpenCLBuiltins.td - OpenCL builtin declarations -------------------===// // // The LLVM Compiler Infrastructure // // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. // See https://llvm.org/LICENSE.txt for license information. // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception // //===----------------------------------------------------------------------===// // // This file contains TableGen definitions for OpenCL builtin function // declarations. In case of an unresolved function name in OpenCL, Clang will // check for a function described in this file when -fdeclare-opencl-builtins // is specified. // //===----------------------------------------------------------------------===// //===----------------------------------------------------------------------===// // Definitions of miscellaneous basic entities. //===----------------------------------------------------------------------===// // Versions of OpenCL class Version { int ID = _Version; } def CLAll : Version< 0>; def CL10 : Version<100>; def CL11 : Version<110>; def CL12 : Version<120>; def CL20 : Version<200>; // Address spaces // Pointer types need to be assigned an address space. class AddressSpace { string Name = _AS; } def DefaultAS : AddressSpace<"clang::LangAS::Default">; def PrivateAS : AddressSpace<"clang::LangAS::opencl_private">; def GlobalAS : AddressSpace<"clang::LangAS::opencl_global">; def ConstantAS : AddressSpace<"clang::LangAS::opencl_constant">; def LocalAS : AddressSpace<"clang::LangAS::opencl_local">; def GenericAS : AddressSpace<"clang::LangAS::opencl_generic">; // OpenCL language extension. class AbstractExtension { // One or more OpenCL extensions, space separated. Each extension must be // a valid extension name for the opencl extension pragma. string ExtName = _Ext; } // Extension associated to a builtin function. class FunctionExtension : AbstractExtension<_Ext>; // Extension associated to a type. This enables implicit conditionalization of // builtin function overloads containing a type that depends on an extension. // During overload resolution, when a builtin function overload contains a type // with a TypeExtension, those overloads are skipped when the extension is // disabled. class TypeExtension : AbstractExtension<_Ext>; // Concatenate zero or more space-separated extensions in NewExts to Base and // return the resulting FunctionExtension in ret. class concatExtension { FunctionExtension ret = FunctionExtension< !cond( // Return Base extension if NewExts is empty, !empty(NewExts) : Base.ExtName, // otherwise, return NewExts if Base extension is empty, !empty(Base.ExtName) : NewExts, // otherwise, concatenate NewExts to Base. true : Base.ExtName # " " # NewExts ) >; } // TypeExtension definitions. def NoTypeExt : TypeExtension<"">; def Fp16TypeExt : TypeExtension<"cl_khr_fp16">; def Fp64TypeExt : TypeExtension<"cl_khr_fp64">; def Atomic64TypeExt : TypeExtension<"cl_khr_int64_base_atomics cl_khr_int64_extended_atomics">; def AtomicFp64TypeExt : TypeExtension<"cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_fp64">; // FunctionExtension definitions. def FuncExtNone : FunctionExtension<"">; def FuncExtKhrSubgroups : FunctionExtension<"__opencl_subgroup_builtins">; def FuncExtKhrSubgroupExtendedTypes : FunctionExtension<"cl_khr_subgroup_extended_types">; def FuncExtKhrSubgroupNonUniformVote : FunctionExtension<"cl_khr_subgroup_non_uniform_vote">; def FuncExtKhrSubgroupBallot : FunctionExtension<"cl_khr_subgroup_ballot">; def FuncExtKhrSubgroupNonUniformArithmetic: FunctionExtension<"cl_khr_subgroup_non_uniform_arithmetic">; def FuncExtKhrSubgroupShuffle : FunctionExtension<"cl_khr_subgroup_shuffle">; def FuncExtKhrSubgroupShuffleRelative : FunctionExtension<"cl_khr_subgroup_shuffle_relative">; def FuncExtKhrSubgroupClusteredReduce : FunctionExtension<"cl_khr_subgroup_clustered_reduce">; def FuncExtKhrExtendedBitOps : FunctionExtension<"cl_khr_extended_bit_ops">; def FuncExtKhrGlobalInt32BaseAtomics : FunctionExtension<"cl_khr_global_int32_base_atomics">; def FuncExtKhrGlobalInt32ExtendedAtomics : FunctionExtension<"cl_khr_global_int32_extended_atomics">; def FuncExtKhrLocalInt32BaseAtomics : FunctionExtension<"cl_khr_local_int32_base_atomics">; def FuncExtKhrLocalInt32ExtendedAtomics : FunctionExtension<"cl_khr_local_int32_extended_atomics">; def FuncExtKhrInt64BaseAtomics : FunctionExtension<"cl_khr_int64_base_atomics">; def FuncExtKhrInt64ExtendedAtomics : FunctionExtension<"cl_khr_int64_extended_atomics">; def FuncExtKhrMipmapImage : FunctionExtension<"cl_khr_mipmap_image">; def FuncExtKhrMipmapImageWrites : FunctionExtension<"cl_khr_mipmap_image_writes">; def FuncExtKhrGlMsaaSharing : FunctionExtension<"cl_khr_gl_msaa_sharing">; def FuncExtOpenCLCDeviceEnqueue : FunctionExtension<"__opencl_c_device_enqueue">; def FuncExtOpenCLCGenericAddressSpace : FunctionExtension<"__opencl_c_generic_address_space">; def FuncExtOpenCLCNamedAddressSpaceBuiltins : FunctionExtension<"__opencl_c_named_address_space_builtins">; def FuncExtOpenCLCPipes : FunctionExtension<"__opencl_c_pipes">; def FuncExtOpenCLCWGCollectiveFunctions : FunctionExtension<"__opencl_c_work_group_collective_functions">; def FuncExtOpenCLCReadWriteImages : FunctionExtension<"__opencl_c_read_write_images">; def FuncExtFloatAtomicsFp16GlobalASLoadStore : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp16_global_atomic_load_store">; def FuncExtFloatAtomicsFp16LocalASLoadStore : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp16_local_atomic_load_store">; def FuncExtFloatAtomicsFp16GenericASLoadStore : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp16_global_atomic_load_store __opencl_c_ext_fp16_local_atomic_load_store">; def FuncExtFloatAtomicsFp16GlobalASAdd : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp16_global_atomic_add">; def FuncExtFloatAtomicsFp32GlobalASAdd : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp32_global_atomic_add">; def FuncExtFloatAtomicsFp64GlobalASAdd : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp64_global_atomic_add">; def FuncExtFloatAtomicsFp16LocalASAdd : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp16_local_atomic_add">; def FuncExtFloatAtomicsFp32LocalASAdd : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp32_local_atomic_add">; def FuncExtFloatAtomicsFp64LocalASAdd : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp64_local_atomic_add">; def FuncExtFloatAtomicsFp16GenericASAdd : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp16_local_atomic_add __opencl_c_ext_fp16_global_atomic_add">; def FuncExtFloatAtomicsFp32GenericASAdd : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp32_local_atomic_add __opencl_c_ext_fp32_global_atomic_add">; def FuncExtFloatAtomicsFp64GenericASAdd : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp64_local_atomic_add __opencl_c_ext_fp64_global_atomic_add">; def FuncExtFloatAtomicsFp16GlobalASMinMax : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp16_global_atomic_min_max">; def FuncExtFloatAtomicsFp32GlobalASMinMax : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp32_global_atomic_min_max">; def FuncExtFloatAtomicsFp64GlobalASMinMax : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp64_global_atomic_min_max">; def FuncExtFloatAtomicsFp16LocalASMinMax : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp16_local_atomic_min_max">; def FuncExtFloatAtomicsFp32LocalASMinMax : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp32_local_atomic_min_max">; def FuncExtFloatAtomicsFp64LocalASMinMax : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp64_local_atomic_min_max">; def FuncExtFloatAtomicsFp16GenericASMinMax : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp16_local_atomic_min_max __opencl_c_ext_fp16_global_atomic_min_max">; def FuncExtFloatAtomicsFp32GenericASMinMax : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp32_local_atomic_min_max __opencl_c_ext_fp32_global_atomic_min_max">; def FuncExtFloatAtomicsFp64GenericASMinMax : FunctionExtension<"cl_ext_float_atomics __opencl_c_ext_fp64_local_atomic_min_max __opencl_c_ext_fp64_global_atomic_min_max">; // Not a real extension, but a workaround to add C++ for OpenCL specific builtins. def FuncExtOpenCLCxx : FunctionExtension<"__cplusplus">; // Arm extensions. def ArmIntegerDotProductInt8 : FunctionExtension<"cl_arm_integer_dot_product_int8">; def ArmIntegerDotProductAccumulateInt8 : FunctionExtension<"cl_arm_integer_dot_product_accumulate_int8">; def ArmIntegerDotProductAccumulateInt16 : FunctionExtension<"cl_arm_integer_dot_product_accumulate_int16">; def ArmIntegerDotProductAccumulateSaturateInt8 : FunctionExtension<"cl_arm_integer_dot_product_accumulate_saturate_int8">; // Qualified Type. These map to ASTContext::QualType. class QualType { // Expression to obtain the QualType inside OCL2Qual. // E.g. TypeExpr="Context.IntTy" for the int type. string TypeExpr = _TypeExpr; // Some QualTypes in this file represent an abstract type for which there is // no corresponding AST QualType, e.g. a GenType or an `image2d_t` type // without access qualifiers. bit IsAbstract = _IsAbstract; } // List of integers. class IntList _List> { string Name = _Name; list List = _List; } //===----------------------------------------------------------------------===// // OpenCL C classes for types //===----------------------------------------------------------------------===// // OpenCL C basic data types (int, float, image2d_t, ...). // Its child classes can represent concrete types (e.g. VectorType) or // abstract types (e.g. GenType). class Type { // Name of the Type. string Name = _Name; // QualType associated with this type. QualType QTExpr = _QTExpr; // Size of the vector (if applicable). int VecWidth = 1; // Is a pointer. bit IsPointer = 0; // "const" qualifier. bit IsConst = 0; // "volatile" qualifier. bit IsVolatile = 0; // Access qualifier. Must be one of ("RO", "WO", "RW"). string AccessQualifier = ""; // Address space. string AddrSpace = DefaultAS.Name; // Extension that needs to be enabled to expose a builtin that uses this type. TypeExtension Extension = NoTypeExt; } // OpenCL vector types (e.g. int2, int3, int16, float8, ...). class VectorType : Type<_Ty.Name, _Ty.QTExpr> { let VecWidth = _VecWidth; let AccessQualifier = ""; // Inherited fields let IsPointer = _Ty.IsPointer; let IsConst = _Ty.IsConst; let IsVolatile = _Ty.IsVolatile; let AddrSpace = _Ty.AddrSpace; let Extension = _Ty.Extension; } // OpenCL pointer types (e.g. int*, float*, ...). class PointerType : Type<_Ty.Name, _Ty.QTExpr> { let AddrSpace = _AS.Name; // Inherited fields let VecWidth = _Ty.VecWidth; let IsPointer = 1; let IsConst = _Ty.IsConst; let IsVolatile = _Ty.IsVolatile; let AccessQualifier = _Ty.AccessQualifier; let Extension = _Ty.Extension; } // OpenCL const types (e.g. const int). class ConstType : Type<_Ty.Name, _Ty.QTExpr> { let IsConst = 1; // Inherited fields let VecWidth = _Ty.VecWidth; let IsPointer = _Ty.IsPointer; let IsVolatile = _Ty.IsVolatile; let AccessQualifier = _Ty.AccessQualifier; let AddrSpace = _Ty.AddrSpace; let Extension = _Ty.Extension; } // OpenCL volatile types (e.g. volatile int). class VolatileType : Type<_Ty.Name, _Ty.QTExpr> { let IsVolatile = 1; // Inherited fields let VecWidth = _Ty.VecWidth; let IsPointer = _Ty.IsPointer; let IsConst = _Ty.IsConst; let AccessQualifier = _Ty.AccessQualifier; let AddrSpace = _Ty.AddrSpace; let Extension = _Ty.Extension; } // OpenCL image types (e.g. image2d). class ImageType : Type<_Ty.Name, QualType<_Ty.QTExpr.TypeExpr # _AccessQualifier # "Ty", 0>> { let VecWidth = 0; let AccessQualifier = _AccessQualifier; // Inherited fields let IsPointer = _Ty.IsPointer; let IsConst = _Ty.IsConst; let IsVolatile = _Ty.IsVolatile; let AddrSpace = _Ty.AddrSpace; // Add TypeExtensions for writable "image3d_t" and "read_write" image types. let Extension = !cond( !and(!eq(_Ty.Name, "image3d_t"), !eq(_AccessQualifier, "WO")) : TypeExtension<"cl_khr_3d_image_writes">, !and(!eq(_Ty.Name, "image3d_t"), !eq(_AccessQualifier, "RW")) : TypeExtension<"cl_khr_3d_image_writes __opencl_c_read_write_images">, !or(!eq(_Ty.Name, "image2d_depth_t"), !eq(_Ty.Name, "image2d_array_depth_t")) : TypeExtension<"cl_khr_depth_images">, !eq(_AccessQualifier, "RW") : TypeExtension<"__opencl_c_read_write_images">, true : _Ty.Extension); } // OpenCL enum type (e.g. memory_scope). class EnumType : Type<_Name, QualType<"getOpenCLEnumType(S, \"" # _Name # "\")", 0>> { } // OpenCL typedef type (e.g. cl_mem_fence_flags). class TypedefType : Type<_Name, QualType<"getOpenCLTypedefType(S, \"" # _Name # "\")", 0>> { } // List of Types. class TypeList _Type> { list List = _Type; } // A GenericType is an abstract type that defines a set of types as a // combination of Types and vector sizes. // // For example, if TypeList = and VectorList = <1, 2, 4>, then it // represents . // // Some rules apply when using multiple GenericType arguments in a declaration: // 1. The number of vector sizes must be equal or 1 for all gentypes in a // declaration. // 2. The number of Types must be equal or 1 for all gentypes in a // declaration. // 3. Generic types are combined by iterating over all generic types at once. // For example, for the following GenericTypes // GenT1 = GenericType and // GenT2 = GenericType // A declaration f(GenT1, GenT2) results in the combinations // f(half, float), f(half2, float2), f(half, int), f(half2, int2) . // 4. "sgentype" from the OpenCL specification is supported by specifying // a single vector size. // For example, for the following GenericTypes // GenT = GenericType and // SGenT = GenericType // A declaration f(GenT, SGenT) results in the combinations // f(half, half), f(half2, half), f(int, int), f(int2, int) . class GenericType : Type<_Ty, QualType<"null", 1>> { // Possible element types of the generic type. TypeList TypeList = _TypeList; // Possible vector sizes of the types in the TypeList. IntList VectorList = _VectorList; // The VecWidth field is ignored for GenericTypes. Use VectorList instead. let VecWidth = 0; } // Builtin function attributes. def Attr { list None = [0, 0, 0]; list Pure = [1, 0, 0]; list Const = [0, 1, 0]; list Convergent = [0, 0, 1]; } //===----------------------------------------------------------------------===// // OpenCL C class for builtin functions //===----------------------------------------------------------------------===// class Builtin _Signature, list _Attributes = Attr.None> { // Name of the builtin function string Name = _Name; // List of types used by the function. The first one is the return type and // the following are the arguments. The list must have at least one element // (the return type). list Signature = _Signature; // Function attribute __attribute__((pure)) bit IsPure = _Attributes[0]; // Function attribute __attribute__((const)) bit IsConst = _Attributes[1]; // Function attribute __attribute__((convergent)) bit IsConv = _Attributes[2]; // OpenCL extensions to which the function belongs. FunctionExtension Extension = FuncExtNone; // Version of OpenCL from which the function is available (e.g.: CL10). // MinVersion is inclusive. Version MinVersion = CL10; // Version of OpenCL from which the function is not supported anymore. // MaxVersion is exclusive. // CLAll makes the function available for all versions. Version MaxVersion = CLAll; } //===----------------------------------------------------------------------===// // Definitions of OpenCL C types //===----------------------------------------------------------------------===// // OpenCL v1.0/1.2/2.0 s6.1.1: Built-in Scalar Data Types. def Bool : Type<"bool", QualType<"Context.BoolTy">>; def Char : Type<"char", QualType<"Context.CharTy">>; def UChar : Type<"uchar", QualType<"Context.UnsignedCharTy">>; def Short : Type<"short", QualType<"Context.ShortTy">>; def UShort : Type<"ushort", QualType<"Context.UnsignedShortTy">>; def Int : Type<"int", QualType<"Context.IntTy">>; def UInt : Type<"uint", QualType<"Context.UnsignedIntTy">>; def Long : Type<"long", QualType<"Context.LongTy">>; def ULong : Type<"ulong", QualType<"Context.UnsignedLongTy">>; def Float : Type<"float", QualType<"Context.FloatTy">>; let Extension = Fp64TypeExt in { def Double : Type<"double", QualType<"Context.DoubleTy">>; } // The half type for builtins that require the cl_khr_fp16 extension. let Extension = Fp16TypeExt in { def Half : Type<"half", QualType<"Context.HalfTy">>; } // Without the cl_khr_fp16 extension, the half type can only be used to declare // a pointer. Define const and non-const pointer types in all address spaces. // Use the "__half" alias to allow the TableGen emitter to distinguish the // (extensionless) pointee type of these pointer-to-half types from the "half" // type defined above that already carries the cl_khr_fp16 extension. foreach AS = [PrivateAS, GlobalAS, ConstantAS, LocalAS, GenericAS] in { def "HalfPtr" # AS : PointerType>, AS>; def "HalfPtrConst" # AS : PointerType>>, AS>; } def Size : Type<"size_t", QualType<"Context.getSizeType()">>; def PtrDiff : Type<"ptrdiff_t", QualType<"Context.getPointerDiffType()">>; def IntPtr : Type<"intptr_t", QualType<"Context.getIntPtrType()">>; def UIntPtr : Type<"uintptr_t", QualType<"Context.getUIntPtrType()">>; def Void : Type<"void", QualType<"Context.VoidTy">>; // OpenCL v1.0/1.2/2.0 s6.1.2: Built-in Vector Data Types. // Built-in vector data types are created by TableGen's OpenCLBuiltinEmitter. // OpenCL v1.0/1.2/2.0 s6.1.3: Other Built-in Data Types. // The image definitions are "abstract". They should not be used without // specifying an access qualifier (RO/WO/RW). def Image1d : Type<"image1d_t", QualType<"Context.OCLImage1d", 1>>; def Image2d : Type<"image2d_t", QualType<"Context.OCLImage2d", 1>>; def Image3d : Type<"image3d_t", QualType<"Context.OCLImage3d", 1>>; def Image1dArray : Type<"image1d_array_t", QualType<"Context.OCLImage1dArray", 1>>; def Image1dBuffer : Type<"image1d_buffer_t", QualType<"Context.OCLImage1dBuffer", 1>>; def Image2dArray : Type<"image2d_array_t", QualType<"Context.OCLImage2dArray", 1>>; def Image2dDepth : Type<"image2d_depth_t", QualType<"Context.OCLImage2dDepth", 1>>; def Image2dArrayDepth : Type<"image2d_array_depth_t", QualType<"Context.OCLImage2dArrayDepth", 1>>; def Image2dMsaa : Type<"image2d_msaa_t", QualType<"Context.OCLImage2dMSAA", 1>>; def Image2dArrayMsaa : Type<"image2d_array_msaa_t", QualType<"Context.OCLImage2dArrayMSAA", 1>>; def Image2dMsaaDepth : Type<"image2d_msaa_depth_t", QualType<"Context.OCLImage2dMSAADepth", 1>>; def Image2dArrayMsaaDepth : Type<"image2d_array_msaa_depth_t", QualType<"Context.OCLImage2dArrayMSAADepth", 1>>; def Sampler : Type<"sampler_t", QualType<"Context.OCLSamplerTy">>; def ClkEvent : Type<"clk_event_t", QualType<"Context.OCLClkEventTy">>; def Event : Type<"event_t", QualType<"Context.OCLEventTy">>; def Queue : Type<"queue_t", QualType<"Context.OCLQueueTy">>; def ReserveId : Type<"reserve_id_t", QualType<"Context.OCLReserveIDTy">>; def MemFenceFlags : TypedefType<"cl_mem_fence_flags">; def ClkProfilingInfo : TypedefType<"clk_profiling_info">; def NDRange : TypedefType<"ndrange_t">; // OpenCL v2.0 s6.13.11: Atomic integer and floating-point types. def AtomicInt : Type<"atomic_int", QualType<"Context.getAtomicType(Context.IntTy)">>; def AtomicUInt : Type<"atomic_uint", QualType<"Context.getAtomicType(Context.UnsignedIntTy)">>; let Extension = Atomic64TypeExt in { def AtomicLong : Type<"atomic_long", QualType<"Context.getAtomicType(Context.LongTy)">>; def AtomicULong : Type<"atomic_ulong", QualType<"Context.getAtomicType(Context.UnsignedLongTy)">>; } def AtomicFloat : Type<"atomic_float", QualType<"Context.getAtomicType(Context.FloatTy)">>; let Extension = AtomicFp64TypeExt in { def AtomicDouble : Type<"atomic_double", QualType<"Context.getAtomicType(Context.DoubleTy)">>; } def AtomicHalf : Type<"atomic_half", QualType<"Context.getAtomicType(Context.HalfTy)">>; def AtomicIntPtr : Type<"atomic_intptr_t", QualType<"Context.getAtomicType(Context.getIntPtrType())">>; def AtomicUIntPtr : Type<"atomic_uintptr_t", QualType<"Context.getAtomicType(Context.getUIntPtrType())">>; def AtomicSize : Type<"atomic_size_t", QualType<"Context.getAtomicType(Context.getSizeType())">>; def AtomicPtrDiff : Type<"atomic_ptrdiff_t", QualType<"Context.getAtomicType(Context.getPointerDiffType())">>; def AtomicFlag : TypedefType<"atomic_flag">; def MemoryOrder : EnumType<"memory_order">; def MemoryScope : EnumType<"memory_scope">; //===----------------------------------------------------------------------===// // Definitions of OpenCL gentype variants //===----------------------------------------------------------------------===// // The OpenCL specification often uses "gentype" in builtin function // declarations to indicate that a builtin function is available with various // argument and return types. The types represented by "gentype" vary between // different parts of the specification. The following definitions capture // the different type lists for gentypes in different parts of the // specification. // Vector width lists. def VecAndScalar: IntList<"VecAndScalar", [1, 2, 3, 4, 8, 16]>; def VecNoScalar : IntList<"VecNoScalar", [2, 3, 4, 8, 16]>; def Vec1 : IntList<"Vec1", [1]>; def Vec1234 : IntList<"Vec1234", [1, 2, 3, 4]>; // Type lists. def TLAll : TypeList<[Char, UChar, Short, UShort, Int, UInt, Long, ULong, Float, Double, Half]>; def TLFloat : TypeList<[Float, Double, Half]>; def TLSignedInts : TypeList<[Char, Short, Int, Long]>; def TLUnsignedInts : TypeList<[UChar, UShort, UInt, ULong]>; def TLIntLongFloats : TypeList<[Int, UInt, Long, ULong, Float, Double, Half]>; // All unsigned integer types twice, to facilitate unsigned return types for e.g. // uchar abs(char) and // uchar abs(uchar). def TLAllUIntsTwice : TypeList<[UChar, UChar, UShort, UShort, UInt, UInt, ULong, ULong]>; def TLAllInts : TypeList<[Char, UChar, Short, UShort, Int, UInt, Long, ULong]>; // GenType definitions for multiple base types (e.g. all floating point types, // or all integer types). // All types def AGenType1 : GenericType<"AGenType1", TLAll, Vec1>; def AGenTypeN : GenericType<"AGenTypeN", TLAll, VecAndScalar>; def AGenTypeNNoScalar : GenericType<"AGenTypeNNoScalar", TLAll, VecNoScalar>; // All integer def AIGenType1 : GenericType<"AIGenType1", TLAllInts, Vec1>; def AIGenTypeN : GenericType<"AIGenTypeN", TLAllInts, VecAndScalar>; def AIGenTypeNNoScalar : GenericType<"AIGenTypeNNoScalar", TLAllInts, VecNoScalar>; // All integer to unsigned def AI2UGenTypeN : GenericType<"AI2UGenTypeN", TLAllUIntsTwice, VecAndScalar>; // Signed integer def SGenTypeN : GenericType<"SGenTypeN", TLSignedInts, VecAndScalar>; // Unsigned integer def UGenTypeN : GenericType<"UGenTypeN", TLUnsignedInts, VecAndScalar>; // Float def FGenTypeN : GenericType<"FGenTypeN", TLFloat, VecAndScalar>; // (u)int, (u)long, and all floats def IntLongFloatGenType1 : GenericType<"IntLongFloatGenType1", TLIntLongFloats, Vec1>; // (u)char and (u)short def CharShortGenType1 : GenericType<"CharShortGenType1", TypeList<[Char, UChar, Short, UShort]>, Vec1>; // GenType definitions for every single base type (e.g. fp32 only). // Names are like: GenTypeFloatVecAndScalar. foreach Type = [Char, UChar, Short, UShort, Int, UInt, Long, ULong, Float, Double, Half] in { foreach VecSizes = [VecAndScalar, VecNoScalar] in { def "GenType" # Type # VecSizes : GenericType<"GenType" # Type # VecSizes, TypeList<[Type]>, VecSizes>; } } // GenType definitions for vec1234. foreach Type = [Float, Double, Half] in { def "GenType" # Type # Vec1234 : GenericType<"GenType" # Type # Vec1234, TypeList<[Type]>, Vec1234>; } //===----------------------------------------------------------------------===// // Definitions of OpenCL builtin functions //===----------------------------------------------------------------------===// //-------------------------------------------------------------------- // OpenCL v1.1/1.2/2.0 s6.2.3 - Explicit conversions. // OpenCL v2.0 Extensions s5.1.1 and s6.1.1 - Conversions. // Generate the convert_* builtins functions. foreach RType = [Float, Double, Half, Char, UChar, Short, UShort, Int, UInt, Long, ULong] in { foreach IType = [Float, Double, Half, Char, UChar, Short, UShort, Int, UInt, Long, ULong] in { // Conversions to integer type have a sat and non-sat variant. foreach sat = !cond(!eq(RType.Name, "float") : [""], !eq(RType.Name, "double") : [""], !eq(RType.Name, "half") : [""], 1 : ["", "_sat"]) in { foreach rnd = ["", "_rte", "_rtn", "_rtp", "_rtz"] in { def : Builtin<"convert_" # RType.Name # sat # rnd, [RType, IType], Attr.Const>; foreach v = [2, 3, 4, 8, 16] in { def : Builtin<"convert_" # RType.Name # v # sat # rnd, [VectorType, VectorType], Attr.Const>; } } } } } //-------------------------------------------------------------------- // OpenCL v1.1 s6.11.1, v1.2 s6.12.1, v2.0 s6.13.1 - Work-item Functions // --- Table 7 --- def : Builtin<"get_work_dim", [UInt], Attr.Const>; foreach name = ["get_global_size", "get_global_id", "get_local_size", "get_local_id", "get_num_groups", "get_group_id", "get_global_offset"] in { def : Builtin; } let MinVersion = CL20 in { def : Builtin<"get_enqueued_local_size", [Size, UInt]>; foreach name = ["get_global_linear_id", "get_local_linear_id"] in { def : Builtin; } } //-------------------------------------------------------------------- // OpenCL v1.1 s6.11.2, v1.2 s6.12.2, v2.0 s6.13.2 - Math functions // OpenCL Extension v2.0 s5.1.2 and s6.1.2 - Math Functions // --- Table 8 --- // --- 1 argument --- foreach name = ["acos", "acosh", "acospi", "asin", "asinh", "asinpi", "atan", "atanh", "atanpi", "cbrt", "ceil", "cos", "cosh", "cospi", "erfc", "erf", "exp", "exp2", "exp10", "expm1", "fabs", "floor", "log", "log2", "log10", "log1p", "logb", "rint", "round", "rsqrt", "sin", "sinh", "sinpi", "sqrt", "tan", "tanh", "tanpi", "tgamma", "trunc", "lgamma"] in { def : Builtin; } foreach name = ["nan"] in { def : Builtin; def : Builtin; def : Builtin; } // --- 2 arguments --- foreach name = ["atan2", "atan2pi", "copysign", "fdim", "fmod", "hypot", "maxmag", "minmag", "nextafter", "pow", "powr", "remainder"] in { def : Builtin; } foreach name = ["fmax", "fmin"] in { def : Builtin; def : Builtin; def : Builtin; def : Builtin; } foreach name = ["ilogb"] in { def : Builtin; def : Builtin; def : Builtin; } foreach name = ["ldexp"] in { def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; } foreach name = ["pown", "rootn"] in { def : Builtin; def : Builtin; def : Builtin; } // --- 3 arguments --- foreach name = ["fma", "mad"] in { def : Builtin; } // The following math builtins take pointer arguments. Which overloads are // available depends on whether the generic address space feature is enabled. multiclass MathWithPointer addrspaces> { foreach AS = addrspaces in { foreach name = ["fract", "modf", "sincos"] in { def : Builtin]>; } foreach name = ["frexp", "lgamma_r"] in { foreach Type = [GenTypeFloatVecAndScalar, GenTypeDoubleVecAndScalar, GenTypeHalfVecAndScalar] in { def : Builtin]>; } } foreach name = ["remquo"] in { foreach Type = [GenTypeFloatVecAndScalar, GenTypeDoubleVecAndScalar, GenTypeHalfVecAndScalar] in { def : Builtin]>; } } } } let Extension = FuncExtOpenCLCNamedAddressSpaceBuiltins in { defm : MathWithPointer<[GlobalAS, LocalAS, PrivateAS]>; } let Extension = FuncExtOpenCLCGenericAddressSpace in { defm : MathWithPointer<[GenericAS]>; } // --- Table 9 --- foreach name = ["half_cos", "half_exp", "half_exp2", "half_exp10", "half_log", "half_log2", "half_log10", "half_recip", "half_rsqrt", "half_sin", "half_sqrt", "half_tan", "native_cos", "native_exp", "native_exp2", "native_exp10", "native_log", "native_log2", "native_log10", "native_recip", "native_rsqrt", "native_sin", "native_sqrt", "native_tan"] in { def : Builtin; } foreach name = ["half_divide", "half_powr", "native_divide", "native_powr"] in { def : Builtin; } //-------------------------------------------------------------------- // OpenCL v1.1 s6.11.3, v1.2 s6.12.3, v2.0 s6.13.3 - Integer Functions // --- Table 10 --- // --- 1 argument --- foreach name = ["abs"] in { def : Builtin; } def : Builtin<"clz", [AIGenTypeN, AIGenTypeN], Attr.Const>; let MinVersion = CL12 in { def : Builtin<"popcount", [AIGenTypeN, AIGenTypeN], Attr.Const>; } let MinVersion = CL20 in { foreach name = ["ctz"] in { def : Builtin; } } // --- 2 arguments --- foreach name = ["abs_diff"] in { def : Builtin; } foreach name = ["add_sat", "hadd", "rhadd", "mul_hi", "rotate", "sub_sat"] in { def : Builtin; } foreach name = ["max", "min"] in { def : Builtin; def : Builtin; } foreach name = ["upsample"] in { def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; } // --- 3 arguments --- foreach name = ["clamp"] in { def : Builtin; def : Builtin; } foreach name = ["mad_hi", "mad_sat"] in { def : Builtin; } // --- Table 11 --- foreach name = ["mad24"] in { def : Builtin; def : Builtin; } foreach name = ["mul24"] in { def : Builtin; def : Builtin; } //-------------------------------------------------------------------- // OpenCL v1.1 s6.11.4, v1.2 s6.12.4, v2.0 s6.13.4 - Common Functions // OpenCL Extension v2.0 s5.1.3 and s6.1.3 - Common Functions // --- Table 12 --- // --- 1 argument --- foreach name = ["degrees", "radians", "sign"] in { def : Builtin; } // --- 2 arguments --- foreach name = ["max", "min"] in { def : Builtin; def : Builtin; def : Builtin; def : Builtin; } foreach name = ["step"] in { def : Builtin; def : Builtin; def : Builtin; def : Builtin; } // --- 3 arguments --- foreach name = ["clamp"] in { def : Builtin; def : Builtin; def : Builtin; def : Builtin; } foreach name = ["mix"] in { def : Builtin; def : Builtin; def : Builtin; def : Builtin; } foreach name = ["smoothstep"] in { def : Builtin; def : Builtin; def : Builtin; def : Builtin; } //-------------------------------------------------------------------- // OpenCL v1.1 s6.11.5, v1.2 s6.12.5, v2.0 s6.13.5 - Geometric Functions // OpenCL Extension v2.0 s5.1.4 and s6.1.4 - Geometric Functions // --- Table 13 --- // --- 1 argument --- foreach name = ["length"] in { def : Builtin; def : Builtin; def : Builtin; } foreach name = ["normalize"] in { def : Builtin; def : Builtin; def : Builtin; } foreach name = ["fast_length"] in { def : Builtin; } foreach name = ["fast_normalize"] in { def : Builtin; } // --- 2 arguments --- foreach name = ["cross"] in { foreach VSize = [3, 4] in { def : Builtin, VectorType, VectorType], Attr.Const>; def : Builtin, VectorType, VectorType], Attr.Const>; def : Builtin, VectorType, VectorType], Attr.Const>; } } foreach name = ["dot", "distance"] in { def : Builtin; def : Builtin; def : Builtin; } foreach name = ["fast_distance"] in { def : Builtin; } //-------------------------------------------------------------------- // OpenCL v1.1 s6.11.6, v1.2 s6.12.6, v2.0 s6.13.6 - Relational Functions // OpenCL Extension v2.0 s5.1.5 and s6.1.5 - Relational Functions // --- Table 14 --- // --- 1 argument --- foreach name = ["isfinite", "isinf", "isnan", "isnormal", "signbit"] in { def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; } foreach name = ["any", "all"] in { def : Builtin; } // --- 2 arguments --- foreach name = ["isequal", "isnotequal", "isgreater", "isgreaterequal", "isless", "islessequal", "islessgreater", "isordered", "isunordered"] in { def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; } // --- 3 arguments --- foreach name = ["bitselect"] in { def : Builtin; } foreach name = ["select"] in { def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; def : Builtin; } //-------------------------------------------------------------------- // OpenCL v1.1 s6.11.7, v1.2 s6.12.7, v2.0 s6.13.7 - Vector Data Load and Store Functions // OpenCL Extension v1.1 s9.3.6 and s9.6.6, v1.2 s9.5.6, v2.0 s5.1.6 and s6.1.6 - Vector Data Load and Store Functions // --- Table 15 --- multiclass VloadVstore addrspaces, bit defStores> { foreach AS = addrspaces in { foreach VSize = [2, 3, 4, 8, 16] in { foreach name = ["vload" # VSize] in { def : Builtin, Size, PointerType, AS>], Attr.Pure>; def : Builtin, Size, PointerType, AS>], Attr.Pure>; def : Builtin, Size, PointerType, AS>], Attr.Pure>; def : Builtin, Size, PointerType, AS>], Attr.Pure>; def : Builtin, Size, PointerType, AS>], Attr.Pure>; def : Builtin, Size, PointerType, AS>], Attr.Pure>; def : Builtin, Size, PointerType, AS>], Attr.Pure>; def : Builtin, Size, PointerType, AS>], Attr.Pure>; def : Builtin, Size, PointerType, AS>], Attr.Pure>; def : Builtin, Size, PointerType, AS>], Attr.Pure>; def : Builtin, Size, PointerType, AS>], Attr.Pure>; } if defStores then { foreach name = ["vstore" # VSize] in { def : Builtin, Size, PointerType]>; def : Builtin, Size, PointerType]>; def : Builtin, Size, PointerType]>; def : Builtin, Size, PointerType]>; def : Builtin, Size, PointerType]>; def : Builtin, Size, PointerType]>; def : Builtin, Size, PointerType]>; def : Builtin, Size, PointerType]>; def : Builtin, Size, PointerType]>; def : Builtin, Size, PointerType]>; def : Builtin, Size, PointerType]>; } } } } } let Extension = FuncExtOpenCLCNamedAddressSpaceBuiltins in { defm : VloadVstore<[GlobalAS, LocalAS, PrivateAS], 1>; } let Extension = FuncExtOpenCLCGenericAddressSpace in { defm : VloadVstore<[GenericAS], 1>; } // vload with constant address space is available regardless of version. defm : VloadVstore<[ConstantAS], 0>; multiclass VloadVstoreHalf addrspaces, bit defStores> { foreach AS = addrspaces in { def : Builtin<"vload_half", [Float, Size, !cast("HalfPtrConst" # AS)], Attr.Pure>; foreach VSize = [2, 3, 4, 8, 16] in { foreach name = ["vload_half" # VSize, "vloada_half" # VSize] in { def : Builtin, Size, !cast("HalfPtrConst" # AS)], Attr.Pure>; } } if defStores then { foreach rnd = ["", "_rte", "_rtz", "_rtp", "_rtn"] in { foreach name = ["vstore_half" # rnd] in { def : Builtin("HalfPtr" # AS)]>; def : Builtin("HalfPtr" # AS)]>; } foreach VSize = [2, 3, 4, 8, 16] in { foreach name = ["vstore_half" # VSize # rnd, "vstorea_half" # VSize # rnd] in { def : Builtin, Size, !cast("HalfPtr" # AS)]>; def : Builtin, Size, !cast("HalfPtr" # AS)]>; } } } } } } let Extension = FuncExtOpenCLCNamedAddressSpaceBuiltins in { defm : VloadVstoreHalf<[GlobalAS, LocalAS, PrivateAS], 1>; } let Extension = FuncExtOpenCLCGenericAddressSpace in { defm : VloadVstoreHalf<[GenericAS], 1>; } // vload_half and vloada_half with constant address space are available regardless of version. defm : VloadVstoreHalf<[ConstantAS], 0>; // OpenCL v3.0 s6.15.8 - Synchronization Functions. def : Builtin<"barrier", [Void, MemFenceFlags], Attr.Convergent>; let MinVersion = CL20 in { def : Builtin<"work_group_barrier", [Void, MemFenceFlags], Attr.Convergent>; def : Builtin<"work_group_barrier", [Void, MemFenceFlags, MemoryScope], Attr.Convergent>; } // OpenCL v3.0 s6.15.9 - Legacy Explicit Memory Fence Functions. def : Builtin<"mem_fence", [Void, MemFenceFlags]>; def : Builtin<"read_mem_fence", [Void, MemFenceFlags]>; def : Builtin<"write_mem_fence", [Void, MemFenceFlags]>; // OpenCL v3.0 s6.15.10 - Address Space Qualifier Functions. // to_global, to_local, to_private are declared in Builtins.def. let Extension = FuncExtOpenCLCGenericAddressSpace in { // The OpenCL 3.0 specification defines these with a "gentype" argument indicating any builtin // type or user-defined type, which cannot be represented currently. Hence we slightly diverge // by providing only the following overloads with a void pointer. def : Builtin<"get_fence", [MemFenceFlags, PointerType]>; def : Builtin<"get_fence", [MemFenceFlags, PointerType, GenericAS>]>; } //-------------------------------------------------------------------- // OpenCL v1.1 s6.11.10, v1.2 s6.12.10, v2.0 s6.13.10: Async Copies from Global to Local Memory, Local to Global Memory, and Prefetch // OpenCL Extension v2.0 s5.1.7 and s6.1.7: Async Copies from Global to Local Memory, Local to Global Memory, and Prefetch // --- Table 18 --- foreach name = ["async_work_group_copy"] in { def : Builtin, PointerType, GlobalAS>, Size, Event]>; def : Builtin, PointerType, LocalAS>, Size, Event]>; } foreach name = ["async_work_group_strided_copy"] in { def : Builtin, PointerType, GlobalAS>, Size, Size, Event]>; def : Builtin, PointerType, LocalAS>, Size, Size, Event]>; } foreach name = ["wait_group_events"] in { def : Builtin]>; } foreach name = ["prefetch"] in { def : Builtin, GlobalAS>, Size]>; } //-------------------------------------------------------------------- // OpenCL v2.0 s6.13.11 - Atomics Functions. // Functions that use memory_order and cl_mem_fence_flags enums are not // declared here as the TableGen backend does not handle enums. // OpenCL v1.0 s9.5, s9.6, s9.7 - Atomic Functions for 32-bit integers // --- Table 9.1 --- let Extension = FuncExtKhrGlobalInt32BaseAtomics in { foreach Type = [Int, UInt] in { foreach name = ["atom_add", "atom_sub", "atom_xchg"] in { def : Builtin, GlobalAS>, Type]>; } foreach name = ["atom_inc", "atom_dec"] in { def : Builtin, GlobalAS>]>; } foreach name = ["atom_cmpxchg"] in { def : Builtin, GlobalAS>, Type, Type]>; } } } // --- Table 9.3 --- let Extension = FuncExtKhrLocalInt32BaseAtomics in { foreach Type = [Int, UInt] in { foreach name = ["atom_add", "atom_sub", "atom_xchg"] in { def : Builtin, LocalAS>, Type]>; } foreach name = ["atom_inc", "atom_dec"] in { def : Builtin, LocalAS>]>; } foreach name = ["atom_cmpxchg"] in { def : Builtin, LocalAS>, Type, Type]>; } } } // --- Table 9.5 --- let Extension = FuncExtKhrInt64BaseAtomics in { foreach AS = [GlobalAS, LocalAS] in { foreach Type = [Long, ULong] in { foreach name = ["atom_add", "atom_sub", "atom_xchg"] in { def : Builtin, AS>, Type]>; } foreach name = ["atom_inc", "atom_dec"] in { def : Builtin, AS>]>; } foreach name = ["atom_cmpxchg"] in { def : Builtin, AS>, Type, Type]>; } } } } // --- Table 9.2 --- let Extension = FuncExtKhrGlobalInt32ExtendedAtomics in { foreach Type = [Int, UInt] in { foreach name = ["atom_min", "atom_max", "atom_and", "atom_or", "atom_xor"] in { def : Builtin, GlobalAS>, Type]>; } } } // --- Table 9.4 --- let Extension = FuncExtKhrLocalInt32ExtendedAtomics in { foreach Type = [Int, UInt] in { foreach name = ["atom_min", "atom_max", "atom_and", "atom_or", "atom_xor"] in { def : Builtin, LocalAS>, Type]>; } } } // --- Table 9.6 --- let Extension = FuncExtKhrInt64ExtendedAtomics in { foreach AS = [GlobalAS, LocalAS] in { foreach Type = [Long, ULong] in { foreach name = ["atom_min", "atom_max", "atom_and", "atom_or", "atom_xor"] in { def : Builtin, AS>, Type]>; } } } } // OpenCL v1.1 s6.11.1, v1.2 s6.12.11 - Atomic Functions foreach AS = [GlobalAS, LocalAS] in { def : Builtin<"atomic_xchg", [Float, PointerType, AS>, Float]>; foreach Type = [Int, UInt] in { foreach name = ["atomic_add", "atomic_sub", "atomic_xchg", "atomic_min", "atomic_max", "atomic_and", "atomic_or", "atomic_xor"] in { def : Builtin, AS>, Type]>; } foreach name = ["atomic_inc", "atomic_dec"] in { def : Builtin, AS>]>; } foreach name = ["atomic_cmpxchg"] in { def : Builtin, AS>, Type, Type]>; } } } let Extension = FuncExtOpenCLCxx in { foreach Type = [Int, UInt] in { foreach name = ["atomic_add", "atomic_sub", "atomic_xchg", "atomic_min", "atomic_max", "atomic_and", "atomic_or", "atomic_xor"] in { def : Builtin, GenericAS>, Type]>; } foreach name = ["atomic_inc", "atomic_dec"] in { def : Builtin, GenericAS>]>; } foreach name = ["atomic_cmpxchg"] in { def : Builtin, GenericAS>, Type, Type]>; } } } // OpenCL v2.0 s6.13.11 - Atomic Functions. // An atomic builtin with 2 additional _explicit variants. multiclass BuiltinAtomicExplicit Types, FunctionExtension BaseExt> { // Without explicit MemoryOrder or MemoryScope. let Extension = concatExtension.ret in { def : Builtin; } // With an explicit MemoryOrder argument. let Extension = concatExtension.ret in { def : Builtin; } // With explicit MemoryOrder and MemoryScope arguments. let Extension = BaseExt in { def : Builtin; } } // OpenCL 2.0 atomic functions that have a pointer argument in a given address space. multiclass OpenCL2Atomics { foreach TypePair = [[AtomicInt, Int], [AtomicUInt, UInt], [AtomicLong, Long], [AtomicULong, ULong], [AtomicFloat, Float], [AtomicDouble, Double]] in { let Extension = BaseExt in { def : Builtin<"atomic_init", [Void, PointerType, addrspace>, TypePair[1]]>; } defm : BuiltinAtomicExplicit<"atomic_store", [Void, PointerType, addrspace>, TypePair[1]], BaseExt>; defm : BuiltinAtomicExplicit<"atomic_load", [TypePair[1], PointerType, addrspace>], BaseExt>; defm : BuiltinAtomicExplicit<"atomic_exchange", [TypePair[1], PointerType, addrspace>, TypePair[1]], BaseExt>; foreach Variant = ["weak", "strong"] in { foreach exp_ptr_addrspace = !cond( !eq(BaseExt, FuncExtOpenCLCGenericAddressSpace): [GenericAS], !eq(BaseExt, FuncExtOpenCLCNamedAddressSpaceBuiltins): [GlobalAS, LocalAS, PrivateAS]) in { let Extension = concatExtension.ret in { def : Builtin<"atomic_compare_exchange_" # Variant, [Bool, PointerType, addrspace>, PointerType, TypePair[1]]>; } let Extension = concatExtension.ret in { def : Builtin<"atomic_compare_exchange_" # Variant # "_explicit", [Bool, PointerType, addrspace>, PointerType, TypePair[1], MemoryOrder, MemoryOrder]>; } let Extension = BaseExt in { def : Builtin<"atomic_compare_exchange_" # Variant # "_explicit", [Bool, PointerType, addrspace>, PointerType, TypePair[1], MemoryOrder, MemoryOrder, MemoryScope]>; } } } } foreach TypePair = [[AtomicInt, Int, Int], [AtomicUInt, UInt, UInt], [AtomicLong, Long, Long], [AtomicULong, ULong, ULong], [AtomicUIntPtr, UIntPtr, PtrDiff]] in { foreach ModOp = ["add", "sub"] in { defm : BuiltinAtomicExplicit<"atomic_fetch_" # ModOp, [TypePair[1], PointerType, addrspace>, TypePair[2]], BaseExt>; } } foreach TypePair = [[AtomicInt, Int, Int], [AtomicUInt, UInt, UInt], [AtomicLong, Long, Long], [AtomicULong, ULong, ULong]] in { foreach ModOp = ["or", "xor", "and", "min", "max"] in { defm : BuiltinAtomicExplicit<"atomic_fetch_" # ModOp, [TypePair[1], PointerType, addrspace>, TypePair[2]], BaseExt>; } } defm : BuiltinAtomicExplicit<"atomic_flag_clear", [Void, PointerType, addrspace>], BaseExt>; defm : BuiltinAtomicExplicit<"atomic_flag_test_and_set", [Bool, PointerType, addrspace>], BaseExt>; } let MinVersion = CL20 in { def : Builtin<"atomic_work_item_fence", [Void, MemFenceFlags, MemoryOrder, MemoryScope]>; defm : OpenCL2Atomics; defm : OpenCL2Atomics; defm : OpenCL2Atomics; } // The functionality added by cl_ext_float_atomics extension let MinVersion = CL20 in { foreach addrspace = [GlobalAS, LocalAS, GenericAS] in { defvar extension_fp16 = !cast("FuncExtFloatAtomicsFp16" # addrspace # "LoadStore"); defm : BuiltinAtomicExplicit<"atomic_store", [Void, PointerType, addrspace>, AtomicHalf], extension_fp16>; defm : BuiltinAtomicExplicit<"atomic_load", [Half, PointerType, addrspace>], extension_fp16>; defm : BuiltinAtomicExplicit<"atomic_exchange", [Half, PointerType, addrspace>, Half], extension_fp16>; foreach ModOp = ["add", "sub"] in { defvar extension_fp16 = !cast("FuncExtFloatAtomicsFp16" # addrspace # "Add"); defvar extension_fp32 = !cast("FuncExtFloatAtomicsFp32" # addrspace # "Add"); defvar extension_fp64 = !cast("FuncExtFloatAtomicsFp64" # addrspace # "Add"); defm : BuiltinAtomicExplicit<"atomic_fetch_" # ModOp, [Half, PointerType, addrspace>, Half], extension_fp16>; defm : BuiltinAtomicExplicit<"atomic_fetch_" # ModOp, [Float, PointerType, addrspace>, Float], extension_fp32>; defm : BuiltinAtomicExplicit<"atomic_fetch_" # ModOp, [Double, PointerType, addrspace>, Double], extension_fp64>; } foreach ModOp = ["min", "max"] in { defvar extension_fp16 = !cast("FuncExtFloatAtomicsFp16" # addrspace # "MinMax"); defvar extension_fp32 = !cast("FuncExtFloatAtomicsFp32" # addrspace # "MinMax"); defvar extension_fp64 = !cast("FuncExtFloatAtomicsFp64" # addrspace # "MinMax"); defm : BuiltinAtomicExplicit<"atomic_fetch_" # ModOp, [Half, PointerType, addrspace>, Half], extension_fp16>; defm : BuiltinAtomicExplicit<"atomic_fetch_" # ModOp, [Float, PointerType, addrspace>, Float], extension_fp32>; defm : BuiltinAtomicExplicit<"atomic_fetch_" # ModOp, [Double, PointerType, addrspace>, Double], extension_fp64>; } } } //-------------------------------------------------------------------- // OpenCL v1.1 s6.11.12, v1.2 s6.12.12, v2.0 s6.13.12 - Miscellaneous Vector Functions // --- Table 19 --- foreach VSize1 = [2, 4, 8, 16] in { foreach VSize2 = [2, 4, 8, 16] in { foreach VecAndMaskType = [[Char, UChar], [UChar, UChar], [Short, UShort], [UShort, UShort], [Int, UInt], [UInt, UInt], [Long, ULong], [ULong, ULong], [Float, UInt], [Double, ULong], [Half, UShort]] in { def : Builtin<"shuffle", [VectorType, VectorType, VectorType], Attr.Const>; } } } foreach VSize1 = [2, 4, 8, 16] in { foreach VSize2 = [2, 4, 8, 16] in { foreach VecAndMaskType = [[Char, UChar], [UChar, UChar], [Short, UShort], [UShort, UShort], [Int, UInt], [UInt, UInt], [Long, ULong], [ULong, ULong], [Float, UInt], [Double, ULong], [Half, UShort]] in { def : Builtin<"shuffle2", [VectorType, VectorType, VectorType, VectorType], Attr.Const>; } } } //-------------------------------------------------------------------- // OpenCL v1.1 s6.11.3, v1.2 s6.12.14, v2.0 s6.13.14: Image Read and Write Functions // OpenCL Extension v2.0 s5.1.8 and s6.1.8: Image Read and Write Functions // --- Table 22: Image Read Functions with Samplers --- foreach imgTy = [Image1d] in { foreach coordTy = [Int, Float] in { def : Builtin<"read_imagef", [VectorType, ImageType, Sampler, coordTy], Attr.Pure>; def : Builtin<"read_imagei", [VectorType, ImageType, Sampler, coordTy], Attr.Pure>; def : Builtin<"read_imageui", [VectorType, ImageType, Sampler, coordTy], Attr.Pure>; } } foreach imgTy = [Image2d, Image1dArray] in { foreach coordTy = [Int, Float] in { def : Builtin<"read_imagef", [VectorType, ImageType, Sampler, VectorType], Attr.Pure>; def : Builtin<"read_imagei", [VectorType, ImageType, Sampler, VectorType], Attr.Pure>; def : Builtin<"read_imageui", [VectorType, ImageType, Sampler, VectorType], Attr.Pure>; } } foreach imgTy = [Image3d, Image2dArray] in { foreach coordTy = [Int, Float] in { def : Builtin<"read_imagef", [VectorType, ImageType, Sampler, VectorType], Attr.Pure>; def : Builtin<"read_imagei", [VectorType, ImageType, Sampler, VectorType], Attr.Pure>; def : Builtin<"read_imageui", [VectorType, ImageType, Sampler, VectorType], Attr.Pure>; } } foreach coordTy = [Int, Float] in { def : Builtin<"read_imagef", [Float, ImageType, Sampler, VectorType], Attr.Pure>; def : Builtin<"read_imagef", [Float, ImageType, Sampler, VectorType], Attr.Pure>; } // --- Table 23: Sampler-less Read Functions --- multiclass ImageReadSamplerless { foreach imgTy = [Image2d, Image1dArray] in { def : Builtin<"read_imagef", [VectorType, ImageType, VectorType], Attr.Pure>; def : Builtin<"read_imagei", [VectorType, ImageType, VectorType], Attr.Pure>; def : Builtin<"read_imageui", [VectorType, ImageType, VectorType], Attr.Pure>; } foreach imgTy = [Image3d, Image2dArray] in { def : Builtin<"read_imagef", [VectorType, ImageType, VectorType], Attr.Pure>; def : Builtin<"read_imagei", [VectorType, ImageType, VectorType], Attr.Pure>; def : Builtin<"read_imageui", [VectorType, ImageType, VectorType], Attr.Pure>; } foreach imgTy = [Image1d, Image1dBuffer] in { def : Builtin<"read_imagef", [VectorType, ImageType, Int], Attr.Pure>; def : Builtin<"read_imagei", [VectorType, ImageType, Int], Attr.Pure>; def : Builtin<"read_imageui", [VectorType, ImageType, Int], Attr.Pure>; } def : Builtin<"read_imagef", [Float, ImageType, VectorType], Attr.Pure>; def : Builtin<"read_imagef", [Float, ImageType, VectorType], Attr.Pure>; } let MinVersion = CL12 in { defm : ImageReadSamplerless<"RO">; let Extension = FuncExtOpenCLCReadWriteImages in { defm : ImageReadSamplerless<"RW">; } } // --- Table 24: Image Write Functions --- multiclass ImageWrite { foreach imgTy = [Image2d] in { def : Builtin<"write_imagef", [Void, ImageType, VectorType, VectorType]>; def : Builtin<"write_imagei", [Void, ImageType, VectorType, VectorType]>; def : Builtin<"write_imageui", [Void, ImageType, VectorType, VectorType]>; } foreach imgTy = [Image2dArray] in { def : Builtin<"write_imagef", [Void, ImageType, VectorType, VectorType]>; def : Builtin<"write_imagei", [Void, ImageType, VectorType, VectorType]>; def : Builtin<"write_imageui", [Void, ImageType, VectorType, VectorType]>; } foreach imgTy = [Image1d, Image1dBuffer] in { def : Builtin<"write_imagef", [Void, ImageType, Int, VectorType]>; def : Builtin<"write_imagei", [Void, ImageType, Int, VectorType]>; def : Builtin<"write_imageui", [Void, ImageType, Int, VectorType]>; } foreach imgTy = [Image1dArray] in { def : Builtin<"write_imagef", [Void, ImageType, VectorType, VectorType]>; def : Builtin<"write_imagei", [Void, ImageType, VectorType, VectorType]>; def : Builtin<"write_imageui", [Void, ImageType, VectorType, VectorType]>; } foreach imgTy = [Image3d] in { def : Builtin<"write_imagef", [Void, ImageType, VectorType, VectorType]>; def : Builtin<"write_imagei", [Void, ImageType, VectorType, VectorType]>; def : Builtin<"write_imageui", [Void, ImageType, VectorType, VectorType]>; } def : Builtin<"write_imagef", [Void, ImageType, VectorType, Float]>; def : Builtin<"write_imagef", [Void, ImageType, VectorType, Float]>; } defm : ImageWrite<"WO">; let Extension = FuncExtOpenCLCReadWriteImages in { defm : ImageWrite<"RW">; } // --- Table 25: Image Query Functions --- multiclass ImageQuery { foreach imgTy = [Image1d, Image1dBuffer, Image2d, Image3d, Image1dArray, Image2dArray, Image2dDepth, Image2dArrayDepth] in { foreach name = ["get_image_width", "get_image_channel_data_type", "get_image_channel_order"] in { def : Builtin], Attr.Const>; } } foreach imgTy = [Image2d, Image3d, Image2dArray, Image2dDepth, Image2dArrayDepth] in { def : Builtin<"get_image_height", [Int, ImageType], Attr.Const>; } def : Builtin<"get_image_depth", [Int, ImageType], Attr.Const>; foreach imgTy = [Image2d, Image2dArray, Image2dDepth, Image2dArrayDepth] in { def : Builtin<"get_image_dim", [VectorType, ImageType], Attr.Const>; } def : Builtin<"get_image_dim", [VectorType, ImageType], Attr.Const>; foreach imgTy = [Image1dArray, Image2dArray, Image2dArrayDepth] in { def : Builtin<"get_image_array_size", [Size, ImageType], Attr.Const>; } } defm : ImageQuery<"RO">; defm : ImageQuery<"WO">; let Extension = FuncExtOpenCLCReadWriteImages in { defm : ImageQuery<"RW">; } // OpenCL extension v2.0 s5.1.9: Built-in Image Read Functions // --- Table 8 --- foreach aQual = ["RO"] in { foreach name = ["read_imageh"] in { foreach coordTy = [Int, Float] in { foreach imgTy = [Image2d, Image1dArray] in { def : Builtin, ImageType, Sampler, VectorType], Attr.Pure>; } foreach imgTy = [Image3d, Image2dArray] in { def : Builtin, ImageType, Sampler, VectorType], Attr.Pure>; } foreach imgTy = [Image1d] in { def : Builtin, ImageType, Sampler, coordTy], Attr.Pure>; } } } } // OpenCL extension v2.0 s5.1.10: Built-in Image Sampler-less Read Functions // --- Table 9 --- let MinVersion = CL12 in { multiclass ImageReadHalf { foreach name = ["read_imageh"] in { foreach imgTy = [Image2d, Image1dArray] in { def : Builtin, ImageType, VectorType], Attr.Pure>; } foreach imgTy = [Image3d, Image2dArray] in { def : Builtin, ImageType, VectorType], Attr.Pure>; } foreach imgTy = [Image1d, Image1dBuffer] in { def : Builtin, ImageType, Int], Attr.Pure>; } } } defm : ImageReadHalf<"RO">; let Extension = FuncExtOpenCLCReadWriteImages in { defm : ImageReadHalf<"RW">; } } // OpenCL extension v2.0 s5.1.11: Built-in Image Write Functions // --- Table 10 --- multiclass ImageWriteHalf { foreach name = ["write_imageh"] in { def : Builtin, VectorType, VectorType]>; def : Builtin, VectorType, VectorType]>; def : Builtin, Int, VectorType]>; def : Builtin, Int, VectorType]>; def : Builtin, VectorType, VectorType]>; def : Builtin, VectorType, VectorType]>; } } defm : ImageWriteHalf<"WO">; let Extension = FuncExtOpenCLCReadWriteImages in { defm : ImageWriteHalf<"RW">; } //-------------------------------------------------------------------- // OpenCL v2.0 s6.13.15 - Work-group Functions // --- Table 26 --- let Extension = FuncExtOpenCLCWGCollectiveFunctions in { foreach name = ["work_group_all", "work_group_any"] in { def : Builtin; } foreach name = ["work_group_broadcast"] in { def : Builtin; def : Builtin; def : Builtin; } foreach op = ["add", "min", "max"] in { foreach name = ["work_group_reduce_", "work_group_scan_exclusive_", "work_group_scan_inclusive_"] in { def : Builtin; } } } //-------------------------------------------------------------------- // OpenCL2.0 : 6.13.16 : Pipe Functions // --- Table 27 --- // Defined in Builtins.def // --- Table 28 --- // Builtins taking pipe arguments are defined in Builtins.def let Extension = FuncExtOpenCLCPipes in { def : Builtin<"is_valid_reserve_id", [Bool, ReserveId]>; } // --- Table 29 --- // Defined in Builtins.def //-------------------------------------------------------------------- // OpenCL2.0 : 6.13.17 : Enqueuing Kernels // --- Table 30 --- // Defined in Builtins.def // --- Table 32 --- // Defined in Builtins.def // --- Table 33 --- let Extension = FuncExtOpenCLCDeviceEnqueue in { def : Builtin<"enqueue_marker", [Int, Queue, UInt, PointerType, GenericAS>, PointerType]>; // --- Table 34 --- def : Builtin<"retain_event", [Void, ClkEvent]>; def : Builtin<"release_event", [Void, ClkEvent]>; def : Builtin<"create_user_event", [ClkEvent]>; def : Builtin<"is_valid_event", [Bool, ClkEvent]>; def : Builtin<"set_user_event_status", [Void, ClkEvent, Int]>; def : Builtin<"capture_event_profiling_info", [Void, ClkEvent, ClkProfilingInfo, PointerType]>; // --- Table 35 --- def : Builtin<"get_default_queue", [Queue]>; def : Builtin<"ndrange_1D", [NDRange, Size]>; def : Builtin<"ndrange_1D", [NDRange, Size, Size]>; def : Builtin<"ndrange_1D", [NDRange, Size, Size, Size]>; def : Builtin<"ndrange_2D", [NDRange, PointerType, PrivateAS>]>; def : Builtin<"ndrange_2D", [NDRange, PointerType, PrivateAS>, PointerType, PrivateAS>]>; def : Builtin<"ndrange_2D", [NDRange, PointerType, PrivateAS>, PointerType, PrivateAS>, PointerType, PrivateAS>]>; def : Builtin<"ndrange_3D", [NDRange, PointerType, PrivateAS>]>; def : Builtin<"ndrange_3D", [NDRange, PointerType, PrivateAS>, PointerType, PrivateAS>]>; def : Builtin<"ndrange_3D", [NDRange, PointerType, PrivateAS>, PointerType, PrivateAS>, PointerType, PrivateAS>]>; } //-------------------------------------------------------------------- // End of the builtin functions defined in the OpenCL C specification. // Builtin functions defined in the OpenCL C Extension are below. //-------------------------------------------------------------------- // OpenCL Extension v2.0 s9.18 - Mipmaps let Extension = FuncExtKhrMipmapImage in { // Added to section 6.13.14.2. foreach aQual = ["RO"] in { foreach imgTy = [Image2d] in { foreach name = ["read_imagef"] in { def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, VectorType, VectorType], Attr.Pure>; } foreach name = ["read_imagei"] in { def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, VectorType, VectorType], Attr.Pure>; } foreach name = ["read_imageui"] in { def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, VectorType, VectorType], Attr.Pure>; } } foreach imgTy = [Image2dDepth] in { foreach name = ["read_imagef"] in { def : Builtin, Sampler, VectorType, Float], Attr.Pure>; def : Builtin, Sampler, VectorType, VectorType, VectorType], Attr.Pure>; } } foreach imgTy = [Image1d] in { foreach name = ["read_imagef"] in { def : Builtin, ImageType, Sampler, Float, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, Float, Float, Float], Attr.Pure>; } foreach name = ["read_imagei"] in { def : Builtin, ImageType, Sampler, Float, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, Float, Float, Float], Attr.Pure>; } foreach name = ["read_imageui"] in { def : Builtin, ImageType, Sampler, Float, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, Float, Float, Float], Attr.Pure>; } } foreach imgTy = [Image3d] in { foreach name = ["read_imagef"] in { def : Builtin, ImageType, Sampler, VectorType, VectorType, VectorType], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; } foreach name = ["read_imagei"] in { def : Builtin, ImageType, Sampler, VectorType, VectorType, VectorType], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; } foreach name = ["read_imageui"] in { def : Builtin, ImageType, Sampler, VectorType, VectorType, VectorType], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; } } foreach imgTy = [Image1dArray] in { foreach name = ["read_imagef"] in { def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, Float, Float], Attr.Pure>; } foreach name = ["read_imagei"] in { def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, Float, Float], Attr.Pure>; } foreach name = ["read_imageui"] in { def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, Float, Float], Attr.Pure>; } } foreach imgTy = [Image2dArray] in { foreach name = ["read_imagef"] in { def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, VectorType, VectorType], Attr.Pure>; } foreach name = ["read_imagei"] in { def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, VectorType, VectorType], Attr.Pure>; } foreach name = ["read_imageui"] in { def : Builtin, ImageType, Sampler, VectorType, Float], Attr.Pure>; def : Builtin, ImageType, Sampler, VectorType, VectorType, VectorType], Attr.Pure>; } } foreach imgTy = [Image2dArrayDepth] in { foreach name = ["read_imagef"] in { def : Builtin, Sampler, VectorType, Float], Attr.Pure>; def : Builtin, Sampler, VectorType, VectorType, VectorType], Attr.Pure>; } } } } // Added to section 6.13.14.5 multiclass ImageQueryNumMipLevels { foreach imgTy = [Image1d, Image2d, Image3d, Image1dArray, Image2dArray, Image2dDepth, Image2dArrayDepth] in { def : Builtin<"get_image_num_mip_levels", [Int, ImageType]>; } } let Extension = FuncExtKhrMipmapImage in { defm : ImageQueryNumMipLevels<"RO">; defm : ImageQueryNumMipLevels<"WO">; defm : ImageQueryNumMipLevels<"RW">; } // Write functions are enabled using a separate extension. let Extension = FuncExtKhrMipmapImageWrites in { // Added to section 6.13.14.4. foreach aQual = ["WO"] in { foreach imgTy = [Image2d] in { def : Builtin<"write_imagef", [Void, ImageType, VectorType, Int, VectorType]>; def : Builtin<"write_imagei", [Void, ImageType, VectorType, Int, VectorType]>; def : Builtin<"write_imageui", [Void, ImageType, VectorType, Int, VectorType]>; } def : Builtin<"write_imagef", [Void, ImageType, VectorType, Int, Float]>; foreach imgTy = [Image1d] in { def : Builtin<"write_imagef", [Void, ImageType, Int, Int, VectorType]>; def : Builtin<"write_imagei", [Void, ImageType, Int, Int, VectorType]>; def : Builtin<"write_imageui", [Void, ImageType, Int, Int, VectorType]>; } foreach imgTy = [Image1dArray] in { def : Builtin<"write_imagef", [Void, ImageType, VectorType, Int, VectorType]>; def : Builtin<"write_imagei", [Void, ImageType, VectorType, Int, VectorType]>; def : Builtin<"write_imageui", [Void, ImageType, VectorType, Int, VectorType]>; } foreach imgTy = [Image2dArray] in { def : Builtin<"write_imagef", [Void, ImageType, VectorType, Int, VectorType]>; def : Builtin<"write_imagei", [Void, ImageType, VectorType, Int, VectorType]>; def : Builtin<"write_imageui", [Void, ImageType, VectorType, Int, VectorType]>; } def : Builtin<"write_imagef", [Void, ImageType, VectorType, Int, Float]>; foreach imgTy = [Image3d] in { def : Builtin<"write_imagef", [Void, ImageType, VectorType, Int, VectorType]>; def : Builtin<"write_imagei", [Void, ImageType, VectorType, Int, VectorType]>; def : Builtin<"write_imageui", [Void, ImageType, VectorType, Int, VectorType]>; } } } //-------------------------------------------------------------------- // OpenCL Extension v2.0 s18.3 - Creating OpenCL Memory Objects from OpenGL MSAA Textures // --- Table 6.13.14.3 --- multiclass ImageReadMsaa { foreach imgTy = [Image2dMsaa] in { def : Builtin<"read_imagef", [VectorType, ImageType, VectorType, Int], Attr.Pure>; def : Builtin<"read_imagei", [VectorType, ImageType, VectorType, Int], Attr.Pure>; def : Builtin<"read_imageui", [VectorType, ImageType, VectorType, Int], Attr.Pure>; } foreach imgTy = [Image2dArrayMsaa] in { def : Builtin<"read_imagef", [VectorType, ImageType, VectorType, Int], Attr.Pure>; def : Builtin<"read_imagei", [VectorType, ImageType, VectorType, Int], Attr.Pure>; def : Builtin<"read_imageui", [VectorType, ImageType, VectorType, Int], Attr.Pure>; } foreach name = ["read_imagef"] in { def : Builtin, VectorType, Int], Attr.Pure>; def : Builtin, VectorType, Int], Attr.Pure>; } } // --- Table 6.13.14.5 --- multiclass ImageQueryMsaa { foreach imgTy = [Image2dMsaa, Image2dArrayMsaa, Image2dMsaaDepth, Image2dArrayMsaaDepth] in { foreach name = ["get_image_width", "get_image_height", "get_image_channel_data_type", "get_image_channel_order", "get_image_num_samples"] in { def : Builtin], Attr.Const>; } def : Builtin<"get_image_dim", [VectorType, ImageType], Attr.Const>; } foreach imgTy = [Image2dArrayMsaa, Image2dArrayMsaaDepth] in { def : Builtin<"get_image_array_size", [Size, ImageType], Attr.Const>; } } let Extension = FuncExtKhrGlMsaaSharing in { defm : ImageReadMsaa<"RO">; defm : ImageQueryMsaa<"RO">; defm : ImageQueryMsaa<"WO">; defm : ImageReadMsaa<"RW">; defm : ImageQueryMsaa<"RW">; } //-------------------------------------------------------------------- // OpenCL Extension v2.0 s28 - Subgroups // --- Table 28.2.1 --- let Extension = FuncExtKhrSubgroups in { foreach name = ["get_sub_group_size", "get_max_sub_group_size", "get_num_sub_groups", "get_sub_group_id", "get_sub_group_local_id"] in { def : Builtin; } let MinVersion = CL20 in { foreach name = ["get_enqueued_num_sub_groups"] in { def : Builtin; } } } // --- Table 28.2.2 --- let Extension = FuncExtKhrSubgroups in { def : Builtin<"sub_group_barrier", [Void, MemFenceFlags], Attr.Convergent>; let MinVersion = CL20 in { def : Builtin<"sub_group_barrier", [Void, MemFenceFlags, MemoryScope], Attr.Convergent>; } } // --- Table 28.2.4 --- let Extension = FuncExtKhrSubgroups in { foreach name = ["sub_group_all", "sub_group_any"] in { def : Builtin; } foreach name = ["sub_group_broadcast"] in { def : Builtin; } foreach name = ["sub_group_reduce_", "sub_group_scan_exclusive_", "sub_group_scan_inclusive_"] in { foreach op = ["add", "min", "max"] in { def : Builtin; } } } // OpenCL Extension v3.0 s38 - Extended Subgroup Functions // Section 38.4.1 - cl_khr_subgroup_extended_types let Extension = FuncExtKhrSubgroupExtendedTypes in { // For sub_group_broadcast, add scalar char, uchar, short, and ushort support, def : Builtin<"sub_group_broadcast", [CharShortGenType1, CharShortGenType1, UInt], Attr.Convergent>; // gentype may additionally be one of the supported built-in vector data types. def : Builtin<"sub_group_broadcast", [AGenTypeNNoScalar, AGenTypeNNoScalar, UInt], Attr.Convergent>; foreach name = ["sub_group_reduce_", "sub_group_scan_exclusive_", "sub_group_scan_inclusive_"] in { foreach op = ["add", "min", "max"] in { def : Builtin; } } } // Section 38.5.1 - cl_khr_subgroup_non_uniform_vote let Extension = FuncExtKhrSubgroupNonUniformVote in { def : Builtin<"sub_group_elect", [Int]>; def : Builtin<"sub_group_non_uniform_all", [Int, Int]>; def : Builtin<"sub_group_non_uniform_any", [Int, Int]>; def : Builtin<"sub_group_non_uniform_all_equal", [Int, AGenType1]>; } // Section 38.6.1 - cl_khr_subgroup_ballot let Extension = FuncExtKhrSubgroupBallot in { def : Builtin<"sub_group_non_uniform_broadcast", [AGenTypeN, AGenTypeN, UInt]>; def : Builtin<"sub_group_broadcast_first", [AGenType1, AGenType1]>; def : Builtin<"sub_group_ballot", [VectorType, Int]>; def : Builtin<"sub_group_inverse_ballot", [Int, VectorType], Attr.Const>; def : Builtin<"sub_group_ballot_bit_extract", [Int, VectorType, UInt], Attr.Const>; def : Builtin<"sub_group_ballot_bit_count", [UInt, VectorType], Attr.Const>; def : Builtin<"sub_group_ballot_inclusive_scan", [UInt, VectorType]>; def : Builtin<"sub_group_ballot_exclusive_scan", [UInt, VectorType]>; def : Builtin<"sub_group_ballot_find_lsb", [UInt, VectorType]>; def : Builtin<"sub_group_ballot_find_msb", [UInt, VectorType]>; foreach op = ["eq", "ge", "gt", "le", "lt"] in { def : Builtin<"get_sub_group_" # op # "_mask", [VectorType], Attr.Const>; } } // Section 38.7.1 - cl_khr_subgroup_non_uniform_arithmetic let Extension = FuncExtKhrSubgroupNonUniformArithmetic in { foreach name = ["reduce_", "scan_exclusive_", "scan_inclusive_"] in { foreach op = ["add", "min", "max", "mul"] in { def : Builtin<"sub_group_non_uniform_" # name # op, [AGenType1, AGenType1]>; } foreach op = ["and", "or", "xor"] in { def : Builtin<"sub_group_non_uniform_" # name # op, [AIGenType1, AIGenType1]>; } foreach op = ["and", "or", "xor"] in { def : Builtin<"sub_group_non_uniform_" # name # "logical_" # op, [Int, Int]>; } } } // Section 38.8.1 - cl_khr_subgroup_shuffle let Extension = FuncExtKhrSubgroupShuffle in { def : Builtin<"sub_group_shuffle", [AGenType1, AGenType1, UInt]>; def : Builtin<"sub_group_shuffle_xor", [AGenType1, AGenType1, UInt]>; } // Section 38.9.1 - cl_khr_subgroup_shuffle_relative let Extension = FuncExtKhrSubgroupShuffleRelative in { def : Builtin<"sub_group_shuffle_up", [AGenType1, AGenType1, UInt]>; def : Builtin<"sub_group_shuffle_down", [AGenType1, AGenType1, UInt]>; } // Section 38.10.1 - cl_khr_subgroup_clustered_reduce let Extension = FuncExtKhrSubgroupClusteredReduce in { foreach op = ["add", "min", "max", "mul"] in { def : Builtin<"sub_group_clustered_reduce_" # op, [AGenType1, AGenType1, UInt]>; } foreach op = ["and", "or", "xor"] in { def : Builtin<"sub_group_clustered_reduce_" # op, [AIGenType1, AIGenType1, UInt]>; } foreach op = ["and", "or", "xor"] in { def : Builtin<"sub_group_clustered_reduce_logical_" # op, [Int, Int, UInt]>; } } // Section 40.3.1 - cl_khr_extended_bit_ops let Extension = FuncExtKhrExtendedBitOps in { def : Builtin<"bitfield_insert", [AIGenTypeN, AIGenTypeN, AIGenTypeN, UInt, UInt], Attr.Const>; def : Builtin<"bitfield_extract_signed", [SGenTypeN, SGenTypeN, UInt, UInt], Attr.Const>; def : Builtin<"bitfield_extract_signed", [SGenTypeN, UGenTypeN, UInt, UInt], Attr.Const>; def : Builtin<"bitfield_extract_unsigned", [UGenTypeN, SGenTypeN, UInt, UInt], Attr.Const>; def : Builtin<"bitfield_extract_unsigned", [UGenTypeN, UGenTypeN, UInt, UInt], Attr.Const>; def : Builtin<"bit_reverse", [AIGenTypeN, AIGenTypeN], Attr.Const>; } // Section 42.3 - cl_khr_integer_dot_product let Extension = FunctionExtension<"__opencl_c_integer_dot_product_input_4x8bit"> in { def : Builtin<"dot", [UInt, VectorType, VectorType], Attr.Const>; def : Builtin<"dot", [Int, VectorType, VectorType], Attr.Const>; def : Builtin<"dot", [Int, VectorType, VectorType], Attr.Const>; def : Builtin<"dot", [Int, VectorType, VectorType], Attr.Const>; def : Builtin<"dot_acc_sat", [UInt, VectorType, VectorType, UInt], Attr.Const>; def : Builtin<"dot_acc_sat", [Int, VectorType, VectorType, Int], Attr.Const>; def : Builtin<"dot_acc_sat", [Int, VectorType, VectorType, Int], Attr.Const>; def : Builtin<"dot_acc_sat", [Int, VectorType, VectorType, Int], Attr.Const>; } let Extension = FunctionExtension<"__opencl_c_integer_dot_product_input_4x8bit_packed"> in { def : Builtin<"dot_4x8packed_uu_uint", [UInt, UInt, UInt], Attr.Const>; def : Builtin<"dot_4x8packed_ss_int", [Int, UInt, UInt], Attr.Const>; def : Builtin<"dot_4x8packed_us_int", [Int, UInt, UInt], Attr.Const>; def : Builtin<"dot_4x8packed_su_int", [Int, UInt, UInt], Attr.Const>; def : Builtin<"dot_acc_sat_4x8packed_uu_uint", [UInt, UInt, UInt, UInt], Attr.Const>; def : Builtin<"dot_acc_sat_4x8packed_ss_int", [Int, UInt, UInt, Int], Attr.Const>; def : Builtin<"dot_acc_sat_4x8packed_us_int", [Int, UInt, UInt, Int], Attr.Const>; def : Builtin<"dot_acc_sat_4x8packed_su_int", [Int, UInt, UInt, Int], Attr.Const>; } // Section 48.3 - cl_khr_subgroup_rotate let Extension = FunctionExtension<"cl_khr_subgroup_rotate"> in { def : Builtin<"sub_group_rotate", [AGenType1, AGenType1, Int], Attr.Convergent>; def : Builtin<"sub_group_clustered_rotate", [AGenType1, AGenType1, Int, UInt], Attr.Convergent>; } //-------------------------------------------------------------------- // Arm extensions. let Extension = ArmIntegerDotProductInt8 in { foreach name = ["arm_dot"] in { def : Builtin, VectorType]>; def : Builtin, VectorType]>; } } let Extension = ArmIntegerDotProductAccumulateInt8 in { foreach name = ["arm_dot_acc"] in { def : Builtin, VectorType, UInt]>; def : Builtin, VectorType, Int]>; } } let Extension = ArmIntegerDotProductAccumulateInt16 in { foreach name = ["arm_dot_acc"] in { def : Builtin, VectorType, UInt]>; def : Builtin, VectorType, Int]>; } } let Extension = ArmIntegerDotProductAccumulateSaturateInt8 in { foreach name = ["arm_dot_acc_sat"] in { def : Builtin, VectorType, UInt]>; def : Builtin, VectorType, Int]>; } }