Wait for the previous import operation to complete before beginning an import of the next tile. More...
#include <TTL_double_scheme.h>
Public Member Functions | |
TTL_import_double_buffering (TTL_local(TENSORTYPE *) int_base1, TTL_local(TENSORTYPE *) int_base2, TTL_tensor< TENSORTYPE > ext_tensor, TTL_event *event, TTL_tile first_tile) | |
Create a TTL_import_double_buffering and begin the buffering process. | |
TTL_sub_tensor< TENSORTYPE > | step_buffering (const TTL_tile next_tile) |
void | finish_buffering () |
Complete any transfers required to finish the buffering process. | |
Data Fields | |
TTL_event * | m_event |
TTL_tile | m_prev_tile |
Store of the previous imported tile */. | |
TTL_common_buffering< TENSORTYPE, 2 > | m_common |
The information that is m_common to all pipeline schemes. | |
Wait for the previous import operation to complete before beginning an import of the next tile.
db | TTL_import_double_buffering describing the attributes of the transfer |
next_tile | A description of the tile to begin importing. |
Definition at line 37 of file p/pipelines/TTL_double_scheme.h.
|
inline |
Create a TTL_import_double_buffering and begin the buffering process.
int_base1 | A pointer to the 1st local buffer |
int_base2 | A pointer to the 2nd local buffer |
ext_tensor | A tensor describing the input in global memory |
event | A pointer to the event to use for the inward (external to internal) transfer completion |
first_tile | The first tile to fetch for the scheme |
Example:
This can be optimized and standardized using the TTL_step_buffering call.
Definition at line 69 of file p/pipelines/TTL_double_scheme.h.
|
inline |
Complete any transfers required to finish the buffering process.
Any transfers that are still in progress will be completed and any transfers that need to be started and completed before finish_buffering returns
Definition at line 122 of file p/pipelines/TTL_double_scheme.h.
|
inline |
Definition at line 83 of file p/pipelines/TTL_double_scheme.h.
TTL_common_buffering<TENSORTYPE, 2> TTL_import_double_buffering< TENSORTYPE >::m_common |
The information that is m_common to all pipeline schemes.
Definition at line 130 of file p/pipelines/TTL_double_scheme.h.
TTL_event* TTL_import_double_buffering< TENSORTYPE >::m_event |
A pointer to the event that is used to track the progress of the transfer
Definition at line 126 of file p/pipelines/TTL_double_scheme.h.
TTL_tile TTL_import_double_buffering< TENSORTYPE >::m_prev_tile |
Store of the previous imported tile */.
Definition at line 128 of file p/pipelines/TTL_double_scheme.h.