Enum CUjit_option

Namespace

DotCompute.Backends.CUDA.Types.Native

Assembly

DotCompute.Backends.CUDA.dll

JIT compiler options

public enum CUjit_option

Fields

CU_JIT_CACHE_MODE = 14

Specifies whether to enable caching explicitly

CU_JIT_ERROR_LOG_BUFFER = 5

Pointer to a buffer in which to print any error log messages

CU_JIT_ERROR_LOG_BUFFER_SIZE_BYTES = 6

Size in bytes of the buffer

CU_JIT_FALLBACK_STRATEGY = 10

Specifies choice of fallback strategy if matching cubin is not found

CU_JIT_FAST_COMPILE = 16

Use fast math operations

CU_JIT_FMA = 24

Enable floating-point multiplication and addition/subtraction operations fusion

CU_JIT_FTZ = 21

Enable position-independent code generation

CU_JIT_GENERATE_DEBUG_INFO = 11

Specifies whether to create debug information in output

CU_JIT_GENERATE_LINE_INFO = 13

Generate line number information

CU_JIT_GLOBAL_SYMBOL_ADDRESSES = 18

Array of host addresses to be used for relocated device symbols

CU_JIT_GLOBAL_SYMBOL_COUNT = 19

Number of entries in arrays for CU_JIT_GLOBAL_SYMBOL_NAMES and CU_JIT_GLOBAL_SYMBOL_ADDRESSES

CU_JIT_GLOBAL_SYMBOL_NAMES = 17

Array of device symbol names to be relocated

CU_JIT_INFO_LOG_BUFFER = 3

Pointer to a buffer in which to print any info/error log messages

CU_JIT_INFO_LOG_BUFFER_SIZE_BYTES = 4

Size in bytes of the buffer

CU_JIT_LOG_VERBOSE = 12

Generate verbose log messages

CU_JIT_LTO = 20

Enable Link Time Optimization for device code

CU_JIT_MAX_REGISTERS = 0

Max number of registers that a thread may use

CU_JIT_NEW_SM3X_OPT = 15

Device to use for compilation when using CU_JIT_LTO

CU_JIT_OPTIMIZATION_LEVEL = 7

Level of optimization to apply to generated code (0 - 4), with 4 being the default

CU_JIT_PREC_DIV = 22

Control single-precision floating-point division and reciprocals

CU_JIT_PREC_SQRT = 23

Control single-precision floating-point square root

CU_JIT_TARGET = 9

Target is chosen based on supplied CUjit_target

CU_JIT_TARGET_FROM_CUCONTEXT = 8

No option value required. Determines the target based on the current attached context

CU_JIT_THREADS_PER_BLOCK = 1

Specifies minimum number of threads per block to target compilation for

CU_JIT_WALL_TIME = 2

Overwrites the option value with the total wall clock time