clang 22.0.0git
__clang_cuda_intrinsics.h File Reference

Go to the source code of this file.

Macros

#define __SM_30_INTRINSICS_H__
#define __SM_30_INTRINSICS_HPP__
#define __MAKE_SHUFFLES(__FnName, __IntIntrinsic, __FloatIntrinsic, __Mask, __Type)
#define __SM_32_INTRINSICS_H__
#define __SM_32_INTRINSICS_HPP__

Functions

 __MAKE_SHUFFLES (__shfl, __nvvm_shfl_idx_i32, __nvvm_shfl_idx_f32, 0x1f, int)
 __MAKE_SHUFFLES (__shfl_up, __nvvm_shfl_up_i32, __nvvm_shfl_up_f32, 0, unsigned int)
 __MAKE_SHUFFLES (__shfl_down, __nvvm_shfl_down_i32, __nvvm_shfl_down_f32, 0x1f, unsigned int)
 __MAKE_SHUFFLES (__shfl_xor, __nvvm_shfl_bfly_i32, __nvvm_shfl_bfly_f32, 0x1f, int)
__device__ char __ldg (const char *ptr)
__device__ short __ldg (const short *ptr)
__device__ int __ldg (const int *ptr)
__device__ long __ldg (const long *ptr)
__device__ long long __ldg (const long long *ptr)
__device__ unsigned char __ldg (const unsigned char *ptr)
__device__ signed char __ldg (const signed char *ptr)
__device__ unsigned short __ldg (const unsigned short *ptr)
__device__ unsigned int __ldg (const unsigned int *ptr)
__device__ unsigned long __ldg (const unsigned long *ptr)
__device__ unsigned long long __ldg (const unsigned long long *ptr)
__device__ float __ldg (const float *ptr)
__device__ double __ldg (const double *ptr)
__device__ char2 __ldg (const char2 *ptr)
__device__ char4 __ldg (const char4 *ptr)
__device__ short2 __ldg (const short2 *ptr)
__device__ short4 __ldg (const short4 *ptr)
__device__ int2 __ldg (const int2 *ptr)
__device__ int4 __ldg (const int4 *ptr)
__device__ longlong2 __ldg (const longlong2 *ptr)
__device__ uchar2 __ldg (const uchar2 *ptr)
__device__ uchar4 __ldg (const uchar4 *ptr)
__device__ ushort2 __ldg (const ushort2 *ptr)
__device__ ushort4 __ldg (const ushort4 *ptr)
__device__ uint2 __ldg (const uint2 *ptr)
__device__ uint4 __ldg (const uint4 *ptr)
__device__ ulonglong2 __ldg (const ulonglong2 *ptr)
__device__ float2 __ldg (const float2 *ptr)
__device__ float4 __ldg (const float4 *ptr)
__device__ double2 __ldg (const double2 *ptr)
__device__ unsigned __funnelshift_l (unsigned low32, unsigned high32, unsigned shiftWidth)
__device__ unsigned __funnelshift_lc (unsigned low32, unsigned high32, unsigned shiftWidth)
__device__ unsigned __funnelshift_r (unsigned low32, unsigned high32, unsigned shiftWidth)
__device__ unsigned __funnelshift_rc (unsigned low32, unsigned high32, unsigned shiftWidth)

Macro Definition Documentation

◆ __MAKE_SHUFFLES

#define __MAKE_SHUFFLES ( __FnName,
__IntIntrinsic,
__FloatIntrinsic,
__Mask,
__Type )

Definition at line 23 of file __clang_cuda_intrinsics.h.

◆ __SM_30_INTRINSICS_H__

#define __SM_30_INTRINSICS_H__

Definition at line 17 of file __clang_cuda_intrinsics.h.

◆ __SM_30_INTRINSICS_HPP__

#define __SM_30_INTRINSICS_HPP__

Definition at line 18 of file __clang_cuda_intrinsics.h.

◆ __SM_32_INTRINSICS_H__

#define __SM_32_INTRINSICS_H__

Definition at line 257 of file __clang_cuda_intrinsics.h.

◆ __SM_32_INTRINSICS_HPP__

#define __SM_32_INTRINSICS_HPP__

Definition at line 258 of file __clang_cuda_intrinsics.h.

Function Documentation

◆ __funnelshift_l()

__device__ unsigned __funnelshift_l ( unsigned low32,
unsigned high32,
unsigned shiftWidth )
inline

Definition at line 449 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __funnelshift_lc()

__device__ unsigned __funnelshift_lc ( unsigned low32,
unsigned high32,
unsigned shiftWidth )
inline

Definition at line 457 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __funnelshift_r()

__device__ unsigned __funnelshift_r ( unsigned low32,
unsigned high32,
unsigned shiftWidth )
inline

Definition at line 465 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __funnelshift_rc()

__device__ unsigned __funnelshift_rc ( unsigned low32,
unsigned high32,
unsigned shiftWidth )
inline

Definition at line 473 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [1/30]

__device__ char __ldg ( const char * ptr)
inline

Definition at line 262 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [2/30]

__device__ char2 __ldg ( const char2 * ptr)
inline

Definition at line 290 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [3/30]

__device__ char4 __ldg ( const char4 * ptr)
inline

Definition at line 301 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [4/30]

__device__ double __ldg ( const double * ptr)
inline

Definition at line 288 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [5/30]

__device__ double2 __ldg ( const double2 * ptr)
inline

Definition at line 437 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [6/30]

__device__ float __ldg ( const float * ptr)
inline

Definition at line 287 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [7/30]

__device__ float2 __ldg ( const float2 * ptr)
inline

Definition at line 419 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [8/30]

__device__ float4 __ldg ( const float4 * ptr)
inline

Definition at line 427 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [9/30]

__device__ int __ldg ( const int * ptr)
inline

Definition at line 264 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [10/30]

__device__ int2 __ldg ( const int2 * ptr)
inline

Definition at line 329 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [11/30]

__device__ int4 __ldg ( const int4 * ptr)
inline

Definition at line 337 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [12/30]

__device__ long __ldg ( const long * ptr)
inline

Definition at line 265 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [13/30]

__device__ long long __ldg ( const long long * ptr)
inline

Definition at line 266 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [14/30]

__device__ longlong2 __ldg ( const longlong2 * ptr)
inline

Definition at line 347 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [15/30]

__device__ short __ldg ( const short * ptr)
inline

Definition at line 263 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [16/30]

__device__ short2 __ldg ( const short2 * ptr)
inline

Definition at line 311 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [17/30]

__device__ short4 __ldg ( const short4 * ptr)
inline

Definition at line 319 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [18/30]

__device__ signed char __ldg ( const signed char * ptr)
inline

Definition at line 272 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [19/30]

__device__ uchar2 __ldg ( const uchar2 * ptr)
inline

Definition at line 356 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [20/30]

__device__ uchar4 __ldg ( const uchar4 * ptr)
inline

Definition at line 364 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [21/30]

__device__ uint2 __ldg ( const uint2 * ptr)
inline

Definition at line 392 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [22/30]

__device__ uint4 __ldg ( const uint4 * ptr)
inline

Definition at line 400 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [23/30]

__device__ ulonglong2 __ldg ( const ulonglong2 * ptr)
inline

Definition at line 410 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [24/30]

__device__ unsigned char __ldg ( const unsigned char * ptr)
inline

Definition at line 269 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [25/30]

__device__ unsigned int __ldg ( const unsigned int * ptr)
inline

Definition at line 278 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [26/30]

__device__ unsigned long __ldg ( const unsigned long * ptr)
inline

Definition at line 281 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [27/30]

__device__ unsigned long long __ldg ( const unsigned long long * ptr)
inline

Definition at line 284 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [28/30]

__device__ unsigned short __ldg ( const unsigned short * ptr)
inline

Definition at line 275 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [29/30]

__device__ ushort2 __ldg ( const ushort2 * ptr)
inline

Definition at line 374 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __ldg() [30/30]

__device__ ushort4 __ldg ( const ushort4 * ptr)
inline

Definition at line 382 of file __clang_cuda_intrinsics.h.

References __device__.

◆ __MAKE_SHUFFLES() [1/4]

__MAKE_SHUFFLES ( __shfl ,
__nvvm_shfl_idx_i32 ,
__nvvm_shfl_idx_f32 ,
0x1f ,
int  )

◆ __MAKE_SHUFFLES() [2/4]

__MAKE_SHUFFLES ( __shfl_down ,
__nvvm_shfl_down_i32 ,
__nvvm_shfl_down_f32 ,
0x1f ,
unsigned int  )

◆ __MAKE_SHUFFLES() [3/4]

__MAKE_SHUFFLES ( __shfl_up ,
__nvvm_shfl_up_i32 ,
__nvvm_shfl_up_f32 ,
0 ,
unsigned int  )

◆ __MAKE_SHUFFLES() [4/4]

__MAKE_SHUFFLES ( __shfl_xor ,
__nvvm_shfl_bfly_i32 ,
__nvvm_shfl_bfly_f32 ,
0x1f ,
int  )

References __device__.