fill_tile
-
void ckernel::fill_tile_init()
-
Please refer to documentation for any_init.
-
void ckernel::fill_tile(uint32_t idst, float param0)
-
Performs element-wise fill operation. The value to be filled in the tile is provided as const param0. The DST register buffer must be in acquired state via acquire_dst call. This call is blocking and is only available on the compute engine.
Return value: None
Argument
Description
Type
Valid Range
Required
idst
The index of the tile in DST register buffer to perform the computation on
uint32_t
Must be less than the size of the DST register buffer
True
param0
The value the output is if the input is greater than 0
float
True
-
void ckernel::fill_tile_int(uint32_t idst, uint param0)
-
void ckernel::fill_tile_bitcast(uint32_t idst, uint32_t param0)
-
Performs element-wise fill operation. The value to be filled in the tile is provided as const param0, which is interpreted as a bit-cast representation of a floating-point value. The DST register buffer must be in acquired state via acquire_dst call. This call is blocking and is only available on the compute engine.
Return value: None
Argument
Description
Type
Valid Range
Required
idst
The index of the tile in DST register buffer to perform the computation on
uint32_t
Must be less than the size of the DST register buffer
True
param0
The bit-cast representation of a floating-point value to be used as output
uint32_t
Must represent a valid bit-cast float value
True