[OpenCL] Add cl_khr_extended_bit_ops

[svenvh]

Summary

See https://www.khronos.org/registry/OpenCL/specs/3.0-unified/html/OpenCL_Ext.html#cl_khr_extended_bit_ops

The opencl-c.h additions were generated using D104040.

[stuart]

This looks good to me.

clang/lib/Headers/opencl-c-base.h

 #if defined(__SPIR__)
 #define cl_khr_subgroup_extended_types 1
 #define cl_khr_subgroup_non_uniform_vote 1
 #define cl_khr_subgroup_ballot 1
 #define cl_khr_subgroup_non_uniform_arithmetic 1
 #define cl_khr_subgroup_shuffle 1
 #define cl_khr_subgroup_shuffle_relative 1
 #define cl_khr_subgroup_clustered_reduce 1
+#define cl_khr_extended_bit_ops 1
 #endif // defined(__SPIR__)
 #endif // (defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ >= 200)
 
 // Define feature macros for OpenCL C 2.0
 #if (defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ == 200)
 #define __opencl_c_pipes 1
 #define __opencl_c_generic_address_space 1
 #define __opencl_c_work_group_collective_functions 1

clang/lib/Headers/opencl-c.h

 double  __ovld sub_group_clustered_reduce_add( double value, uint clustersize );
 double  __ovld sub_group_clustered_reduce_mul( double value, uint clustersize );
 double  __ovld sub_group_clustered_reduce_min( double value, uint clustersize );
 double  __ovld sub_group_clustered_reduce_max( double value, uint clustersize );
 #endif // cl_khr_fp64
 
 #endif // cl_khr_subgroup_clustered_reduce
 
+#if defined(cl_khr_extended_bit_ops)
+char __ovld __cnfn bitfield_insert(char, char, uint, uint);
+uchar __ovld __cnfn bitfield_insert(uchar, uchar, uint, uint);
+short __ovld __cnfn bitfield_insert(short, short, uint, uint);
+ushort __ovld __cnfn bitfield_insert(ushort, ushort, uint, uint);
+int __ovld __cnfn bitfield_insert(int, int, uint, uint);
+uint __ovld __cnfn bitfield_insert(uint, uint, uint, uint);
+long __ovld __cnfn bitfield_insert(long, long, uint, uint);
+ulong __ovld __cnfn bitfield_insert(ulong, ulong, uint, uint);
+char2 __ovld __cnfn bitfield_insert(char2, char2, uint, uint);
+uchar2 __ovld __cnfn bitfield_insert(uchar2, uchar2, uint, uint);
+short2 __ovld __cnfn bitfield_insert(short2, short2, uint, uint);
+ushort2 __ovld __cnfn bitfield_insert(ushort2, ushort2, uint, uint);
+int2 __ovld __cnfn bitfield_insert(int2, int2, uint, uint);
+uint2 __ovld __cnfn bitfield_insert(uint2, uint2, uint, uint);
+long2 __ovld __cnfn bitfield_insert(long2, long2, uint, uint);
+ulong2 __ovld __cnfn bitfield_insert(ulong2, ulong2, uint, uint);
+char3 __ovld __cnfn bitfield_insert(char3, char3, uint, uint);
+uchar3 __ovld __cnfn bitfield_insert(uchar3, uchar3, uint, uint);
+short3 __ovld __cnfn bitfield_insert(short3, short3, uint, uint);
+ushort3 __ovld __cnfn bitfield_insert(ushort3, ushort3, uint, uint);
+int3 __ovld __cnfn bitfield_insert(int3, int3, uint, uint);
+uint3 __ovld __cnfn bitfield_insert(uint3, uint3, uint, uint);
+long3 __ovld __cnfn bitfield_insert(long3, long3, uint, uint);
+ulong3 __ovld __cnfn bitfield_insert(ulong3, ulong3, uint, uint);
+char4 __ovld __cnfn bitfield_insert(char4, char4, uint, uint);
+uchar4 __ovld __cnfn bitfield_insert(uchar4, uchar4, uint, uint);
+short4 __ovld __cnfn bitfield_insert(short4, short4, uint, uint);
+ushort4 __ovld __cnfn bitfield_insert(ushort4, ushort4, uint, uint);
+int4 __ovld __cnfn bitfield_insert(int4, int4, uint, uint);
+uint4 __ovld __cnfn bitfield_insert(uint4, uint4, uint, uint);
+long4 __ovld __cnfn bitfield_insert(long4, long4, uint, uint);
+ulong4 __ovld __cnfn bitfield_insert(ulong4, ulong4, uint, uint);
+char8 __ovld __cnfn bitfield_insert(char8, char8, uint, uint);
+uchar8 __ovld __cnfn bitfield_insert(uchar8, uchar8, uint, uint);
+short8 __ovld __cnfn bitfield_insert(short8, short8, uint, uint);
+ushort8 __ovld __cnfn bitfield_insert(ushort8, ushort8, uint, uint);
+int8 __ovld __cnfn bitfield_insert(int8, int8, uint, uint);
+uint8 __ovld __cnfn bitfield_insert(uint8, uint8, uint, uint);
+long8 __ovld __cnfn bitfield_insert(long8, long8, uint, uint);
+ulong8 __ovld __cnfn bitfield_insert(ulong8, ulong8, uint, uint);
+char16 __ovld __cnfn bitfield_insert(char16, char16, uint, uint);
+uchar16 __ovld __cnfn bitfield_insert(uchar16, uchar16, uint, uint);
+short16 __ovld __cnfn bitfield_insert(short16, short16, uint, uint);
+ushort16 __ovld __cnfn bitfield_insert(ushort16, ushort16, uint, uint);
+int16 __ovld __cnfn bitfield_insert(int16, int16, uint, uint);
+uint16 __ovld __cnfn bitfield_insert(uint16, uint16, uint, uint);
+long16 __ovld __cnfn bitfield_insert(long16, long16, uint, uint);
+ulong16 __ovld __cnfn bitfield_insert(ulong16, ulong16, uint, uint);
+
+char __ovld __cnfn bitfield_extract_signed(char, uint, uint);
+short __ovld __cnfn bitfield_extract_signed(short, uint, uint);
+int __ovld __cnfn bitfield_extract_signed(int, uint, uint);
+long __ovld __cnfn bitfield_extract_signed(long, uint, uint);
+char2 __ovld __cnfn bitfield_extract_signed(char2, uint, uint);
+short2 __ovld __cnfn bitfield_extract_signed(short2, uint, uint);
+int2 __ovld __cnfn bitfield_extract_signed(int2, uint, uint);
+long2 __ovld __cnfn bitfield_extract_signed(long2, uint, uint);
+char3 __ovld __cnfn bitfield_extract_signed(char3, uint, uint);
+short3 __ovld __cnfn bitfield_extract_signed(short3, uint, uint);
+int3 __ovld __cnfn bitfield_extract_signed(int3, uint, uint);
+long3 __ovld __cnfn bitfield_extract_signed(long3, uint, uint);
+char4 __ovld __cnfn bitfield_extract_signed(char4, uint, uint);
+short4 __ovld __cnfn bitfield_extract_signed(short4, uint, uint);
+int4 __ovld __cnfn bitfield_extract_signed(int4, uint, uint);
+long4 __ovld __cnfn bitfield_extract_signed(long4, uint, uint);
+char8 __ovld __cnfn bitfield_extract_signed(char8, uint, uint);
+short8 __ovld __cnfn bitfield_extract_signed(short8, uint, uint);
+int8 __ovld __cnfn bitfield_extract_signed(int8, uint, uint);
+long8 __ovld __cnfn bitfield_extract_signed(long8, uint, uint);
+char16 __ovld __cnfn bitfield_extract_signed(char16, uint, uint);
+short16 __ovld __cnfn bitfield_extract_signed(short16, uint, uint);
+int16 __ovld __cnfn bitfield_extract_signed(int16, uint, uint);
+long16 __ovld __cnfn bitfield_extract_signed(long16, uint, uint);
+
+char __ovld __cnfn bitfield_extract_signed(uchar, uint, uint);
+short __ovld __cnfn bitfield_extract_signed(ushort, uint, uint);
+int __ovld __cnfn bitfield_extract_signed(uint, uint, uint);
+long __ovld __cnfn bitfield_extract_signed(ulong, uint, uint);
+char2 __ovld __cnfn bitfield_extract_signed(uchar2, uint, uint);
+short2 __ovld __cnfn bitfield_extract_signed(ushort2, uint, uint);
+int2 __ovld __cnfn bitfield_extract_signed(uint2, uint, uint);
+long2 __ovld __cnfn bitfield_extract_signed(ulong2, uint, uint);
+char3 __ovld __cnfn bitfield_extract_signed(uchar3, uint, uint);
+short3 __ovld __cnfn bitfield_extract_signed(ushort3, uint, uint);
+int3 __ovld __cnfn bitfield_extract_signed(uint3, uint, uint);
+long3 __ovld __cnfn bitfield_extract_signed(ulong3, uint, uint);
+char4 __ovld __cnfn bitfield_extract_signed(uchar4, uint, uint);
+short4 __ovld __cnfn bitfield_extract_signed(ushort4, uint, uint);
+int4 __ovld __cnfn bitfield_extract_signed(uint4, uint, uint);
+long4 __ovld __cnfn bitfield_extract_signed(ulong4, uint, uint);
+char8 __ovld __cnfn bitfield_extract_signed(uchar8, uint, uint);
+short8 __ovld __cnfn bitfield_extract_signed(ushort8, uint, uint);
+int8 __ovld __cnfn bitfield_extract_signed(uint8, uint, uint);
+long8 __ovld __cnfn bitfield_extract_signed(ulong8, uint, uint);
+char16 __ovld __cnfn bitfield_extract_signed(uchar16, uint, uint);
+short16 __ovld __cnfn bitfield_extract_signed(ushort16, uint, uint);
+int16 __ovld __cnfn bitfield_extract_signed(uint16, uint, uint);
+long16 __ovld __cnfn bitfield_extract_signed(ulong16, uint, uint);
+
+uchar __ovld __cnfn bitfield_extract_unsigned(char, uint, uint);
+ushort __ovld __cnfn bitfield_extract_unsigned(short, uint, uint);
+uint __ovld __cnfn bitfield_extract_unsigned(int, uint, uint);
+ulong __ovld __cnfn bitfield_extract_unsigned(long, uint, uint);
+uchar2 __ovld __cnfn bitfield_extract_unsigned(char2, uint, uint);
+ushort2 __ovld __cnfn bitfield_extract_unsigned(short2, uint, uint);
+uint2 __ovld __cnfn bitfield_extract_unsigned(int2, uint, uint);
+ulong2 __ovld __cnfn bitfield_extract_unsigned(long2, uint, uint);
+uchar3 __ovld __cnfn bitfield_extract_unsigned(char3, uint, uint);
+ushort3 __ovld __cnfn bitfield_extract_unsigned(short3, uint, uint);
+uint3 __ovld __cnfn bitfield_extract_unsigned(int3, uint, uint);
+ulong3 __ovld __cnfn bitfield_extract_unsigned(long3, uint, uint);
+uchar4 __ovld __cnfn bitfield_extract_unsigned(char4, uint, uint);
+ushort4 __ovld __cnfn bitfield_extract_unsigned(short4, uint, uint);
+uint4 __ovld __cnfn bitfield_extract_unsigned(int4, uint, uint);
+ulong4 __ovld __cnfn bitfield_extract_unsigned(long4, uint, uint);
+uchar8 __ovld __cnfn bitfield_extract_unsigned(char8, uint, uint);
+ushort8 __ovld __cnfn bitfield_extract_unsigned(short8, uint, uint);
+uint8 __ovld __cnfn bitfield_extract_unsigned(int8, uint, uint);
+ulong8 __ovld __cnfn bitfield_extract_unsigned(long8, uint, uint);
+uchar16 __ovld __cnfn bitfield_extract_unsigned(char16, uint, uint);
+ushort16 __ovld __cnfn bitfield_extract_unsigned(short16, uint, uint);
+uint16 __ovld __cnfn bitfield_extract_unsigned(int16, uint, uint);
+ulong16 __ovld __cnfn bitfield_extract_unsigned(long16, uint, uint);
+
+uchar __ovld __cnfn bitfield_extract_unsigned(uchar, uint, uint);
+ushort __ovld __cnfn bitfield_extract_unsigned(ushort, uint, uint);
+uint __ovld __cnfn bitfield_extract_unsigned(uint, uint, uint);
+ulong __ovld __cnfn bitfield_extract_unsigned(ulong, uint, uint);
+uchar2 __ovld __cnfn bitfield_extract_unsigned(uchar2, uint, uint);
+ushort2 __ovld __cnfn bitfield_extract_unsigned(ushort2, uint, uint);
+uint2 __ovld __cnfn bitfield_extract_unsigned(uint2, uint, uint);
+ulong2 __ovld __cnfn bitfield_extract_unsigned(ulong2, uint, uint);
+uchar3 __ovld __cnfn bitfield_extract_unsigned(uchar3, uint, uint);
+ushort3 __ovld __cnfn bitfield_extract_unsigned(ushort3, uint, uint);
+uint3 __ovld __cnfn bitfield_extract_unsigned(uint3, uint, uint);
+ulong3 __ovld __cnfn bitfield_extract_unsigned(ulong3, uint, uint);
+uchar4 __ovld __cnfn bitfield_extract_unsigned(uchar4, uint, uint);
+ushort4 __ovld __cnfn bitfield_extract_unsigned(ushort4, uint, uint);
+uint4 __ovld __cnfn bitfield_extract_unsigned(uint4, uint, uint);
+ulong4 __ovld __cnfn bitfield_extract_unsigned(ulong4, uint, uint);
+uchar8 __ovld __cnfn bitfield_extract_unsigned(uchar8, uint, uint);
+ushort8 __ovld __cnfn bitfield_extract_unsigned(ushort8, uint, uint);
+uint8 __ovld __cnfn bitfield_extract_unsigned(uint8, uint, uint);
+ulong8 __ovld __cnfn bitfield_extract_unsigned(ulong8, uint, uint);
+uchar16 __ovld __cnfn bitfield_extract_unsigned(uchar16, uint, uint);
+ushort16 __ovld __cnfn bitfield_extract_unsigned(ushort16, uint, uint);
+uint16 __ovld __cnfn bitfield_extract_unsigned(uint16, uint, uint);
+ulong16 __ovld __cnfn bitfield_extract_unsigned(ulong16, uint, uint);
+
+char __ovld __cnfn bit_reverse(char);
+uchar __ovld __cnfn bit_reverse(uchar);
+short __ovld __cnfn bit_reverse(short);
+ushort __ovld __cnfn bit_reverse(ushort);
+int __ovld __cnfn bit_reverse(int);
+uint __ovld __cnfn bit_reverse(uint);
+long __ovld __cnfn bit_reverse(long);
+ulong __ovld __cnfn bit_reverse(ulong);
+char2 __ovld __cnfn bit_reverse(char2);
+uchar2 __ovld __cnfn bit_reverse(uchar2);
+short2 __ovld __cnfn bit_reverse(short2);
+ushort2 __ovld __cnfn bit_reverse(ushort2);
+int2 __ovld __cnfn bit_reverse(int2);
+uint2 __ovld __cnfn bit_reverse(uint2);
+long2 __ovld __cnfn bit_reverse(long2);
+ulong2 __ovld __cnfn bit_reverse(ulong2);
+char3 __ovld __cnfn bit_reverse(char3);
+uchar3 __ovld __cnfn bit_reverse(uchar3);
+short3 __ovld __cnfn bit_reverse(short3);
+ushort3 __ovld __cnfn bit_reverse(ushort3);
+int3 __ovld __cnfn bit_reverse(int3);
+uint3 __ovld __cnfn bit_reverse(uint3);
+long3 __ovld __cnfn bit_reverse(long3);
+ulong3 __ovld __cnfn bit_reverse(ulong3);
+char4 __ovld __cnfn bit_reverse(char4);
+uchar4 __ovld __cnfn bit_reverse(uchar4);
+short4 __ovld __cnfn bit_reverse(short4);
+ushort4 __ovld __cnfn bit_reverse(ushort4);
+int4 __ovld __cnfn bit_reverse(int4);
+uint4 __ovld __cnfn bit_reverse(uint4);
+long4 __ovld __cnfn bit_reverse(long4);
+ulong4 __ovld __cnfn bit_reverse(ulong4);
+char8 __ovld __cnfn bit_reverse(char8);
+uchar8 __ovld __cnfn bit_reverse(uchar8);
+short8 __ovld __cnfn bit_reverse(short8);
+ushort8 __ovld __cnfn bit_reverse(ushort8);
+int8 __ovld __cnfn bit_reverse(int8);
+uint8 __ovld __cnfn bit_reverse(uint8);
+long8 __ovld __cnfn bit_reverse(long8);
+ulong8 __ovld __cnfn bit_reverse(ulong8);
+char16 __ovld __cnfn bit_reverse(char16);
+uchar16 __ovld __cnfn bit_reverse(uchar16);
+short16 __ovld __cnfn bit_reverse(short16);
+ushort16 __ovld __cnfn bit_reverse(ushort16);
+int16 __ovld __cnfn bit_reverse(int16);
+uint16 __ovld __cnfn bit_reverse(uint16);
+long16 __ovld __cnfn bit_reverse(long16);
+ulong16 __ovld __cnfn bit_reverse(ulong16);
+#endif // cl_khr_extended_bit_ops
+
 #if defined(cl_intel_subgroups)
 // Intel-Specific Sub Group Functions
 float   __ovld __conv intel_sub_group_shuffle( float  x, uint c );
 float2  __ovld __conv intel_sub_group_shuffle( float2 x, uint c );
 float3  __ovld __conv intel_sub_group_shuffle( float3 x, uint c );
 float4  __ovld __conv intel_sub_group_shuffle( float4 x, uint c );
 float8  __ovld __conv intel_sub_group_shuffle( float8 x, uint c );
 float16 __ovld __conv intel_sub_group_shuffle( float16 x, uint c );

clang/lib/Sema/OpenCLBuiltins.td

 def FuncExtKhrSubgroups                  : FunctionExtension<"cl_khr_subgroups">;
 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">;
let Extension = FuncExtKhrSubgroupClusteredReduce in {
   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>;
+}
+
 //--------------------------------------------------------------------
 // Arm extensions.
 let Extension = ArmIntegerDotProductInt8 in {
   foreach name = ["arm_dot"] in {
     def : Builtin<name, [UInt, VectorType<UChar, 4>, VectorType<UChar, 4>]>;
     def : Builtin<name, [Int, VectorType<Char, 4>, VectorType<Char, 4>]>;
   }
 }

clang/test/Headers/opencl-c-header.cl

 #error "Incorrectly defined cl_khr_subgroup_shuffle"
 #endif
 #if cl_khr_subgroup_shuffle_relative != 1
 #error "Incorrectly defined cl_khr_subgroup_shuffle_relative"
 #endif
 #if cl_khr_subgroup_clustered_reduce != 1
 #error "Incorrectly defined cl_khr_subgroup_clustered_reduce"
 #endif
+#if cl_khr_extended_bit_ops != 1
+#error "Incorrectly defined cl_khr_extended_bit_ops"
+#endif
 
 #else
 
 #ifdef cl_khr_subgroup_extended_types
 #error "Incorrect cl_khr_subgroup_extended_types define"
 #endif
 #ifdef cl_khr_subgroup_non_uniform_vote
 #error "Incorrect cl_khr_subgroup_non_uniform_vote define"
 #endif
 #ifdef cl_khr_subgroup_ballot
 #error "Incorrect cl_khr_subgroup_ballot define"
 #endif
 #ifdef cl_khr_subgroup_non_uniform_arithmetic
 #error "Incorrect cl_khr_subgroup_non_uniform_arithmetic define"
 #endif
 #ifdef cl_khr_subgroup_shuffle
 #error "Incorrect cl_khr_subgroup_shuffle define"
 #endif
 #ifdef cl_khr_subgroup_shuffle_relative
 #error "Incorrect cl_khr_subgroup_shuffle_relative define"
 #endif
 #ifdef cl_khr_subgroup_clustered_reduce
 #error "Incorrect cl_khr_subgroup_clustered_reduce define"
 #endif
+#ifdef cl_khr_extended_bit_ops
+#error "Incorrect cl_khr_extended_bit_ops define"
+#endif
 
 #endif //(defined(__OPENCL_CPP_VERSION__) || __OPENCL_C_VERSION__ >= 200)
 
 // OpenCL C features.
 #if (__OPENCL_C_VERSION__ == 300)
 
 #if __opencl_c_atomic_scope_all_devices != 1
 #error "Incorrectly defined feature macro __opencl_c_atomic_scope_all_devices"