Skip to content

w2r: release copy immediately if no host copy is available #684

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
devreal wants to merge 2 commits into from
Closed

w2r: release copy immediately if no host copy is available #684

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
bb76079
w2r: skip transfer if no host data is available
devreal Oct 18, 2024
3996a9d
Bookkeeping for w2r tasks
devreal Oct 29, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions parsec/mca/device/device_gpu.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1227,6 +1227,7 @@ parsec_default_gpu_stage_out(parsec_gpu_task_t *gtask,
if(flow_mask & (1U << i)){
source = task->data[i].data_out;
dest = source->original->device_copies[0];
if (NULL == dest) continue;
dst_dev = (parsec_device_gpu_module_t*)parsec_mca_device_get(dest->device_index);
src_dev = (parsec_device_gpu_module_t*)parsec_mca_device_get(source->device_index);

Expand Down
40 changes: 32 additions & 8 deletions parsec/mca/device/transfer_gpu.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -241,6 +241,25 @@ parsec_gpu_create_w2r_task(parsec_device_gpu_module_t *gpu_device,
parsec_atomic_lock( &gpu_copy->original->lock );
/* get the next item before altering the next pointer */
item = (parsec_list_item_t*)item->list_next; /* conversion needed for volatile */

if (gpu_copy->original->device_copies[0] == NULL) {
/* no host-side copy available, release immediately if no readers */
if (gpu_copy->readers == 0) {
PARSEC_DEBUG_VERBOSE(10, parsec_gpu_output_stream,
"D2H[%d:%s] CPU data copy discarded, GPU data copy %p [%p] now available",
gpu_device->super.device_index, gpu_device->super.name, gpu_copy, gpu_copy->original);

parsec_list_item_ring_chop((parsec_list_item_t*)gpu_copy);
PARSEC_LIST_ITEM_SINGLETON(gpu_copy);
gpu_copy->coherency_state = PARSEC_DATA_COHERENCY_SHARED;
gpu_copy->data_transfer_status = PARSEC_DATA_STATUS_COMPLETE_TRANSFER;
parsec_list_push_back(&gpu_device->gpu_mem_lru, (parsec_list_item_t*)gpu_copy);
gpu_device->data_avail_epoch++;
}
parsec_atomic_unlock( &gpu_copy->original->lock );
continue;
}

if( 0 == gpu_copy->readers ) {
if( PARSEC_UNLIKELY(NULL == d2h_task) ) { /* allocate on-demand */
d2h_task = (parsec_gpu_d2h_task_t*)parsec_thread_mempool_allocate(es->context_mempool);
Expand Down Expand Up @@ -284,6 +303,8 @@ parsec_gpu_create_w2r_task(parsec_device_gpu_module_t *gpu_device,
w2r_task->stage_out = &parsec_default_gpu_stage_out;
w2r_task->complete_stage = NULL;

parsec_atomic_fetch_inc_int32( &(gpu_device->mutex) );

(void)es;
return w2r_task;
}
Expand Down Expand Up @@ -314,20 +335,22 @@ int parsec_gpu_complete_w2r_task(parsec_device_gpu_module_t *gpu_device,

cpu_copy = original->device_copies[0];

if( cpu_copy->version < gpu_copy->version ) {
if( NULL != cpu_copy && cpu_copy->version < gpu_copy->version ) {
/* the GPU version has been acquired by a new task that is waiting for submission */
PARSEC_DEBUG_VERBOSE(10, parsec_gpu_output_stream,
"D2H[%d:%s] task %p:%i GPU data copy %p [%p] has a backup in memory",
gpu_device->super.device_index, gpu_device->super.name, (void*)task, i, gpu_copy, gpu_copy->original);
} else {
gpu_copy->coherency_state = PARSEC_DATA_COHERENCY_SHARED;
cpu_copy->coherency_state = PARSEC_DATA_COHERENCY_SHARED;
cpu_copy->version = gpu_copy->version;
PARSEC_DEBUG_VERBOSE(10, parsec_gpu_output_stream,
"D2H[%d:%s]: CPU copy %p gets the same version %d as GPU copy %p at %s:%d",
gpu_device->super.device_index, gpu_device->super.name,
cpu_copy, cpu_copy->version, gpu_copy,
__FILE__, __LINE__);
if (NULL != cpu_copy) {
cpu_copy->coherency_state = PARSEC_DATA_COHERENCY_SHARED;
cpu_copy->version = gpu_copy->version;
PARSEC_DEBUG_VERBOSE(10, parsec_gpu_output_stream,
"D2H[%d:%s]: CPU copy %p gets the same version %d as GPU copy %p at %s:%d",
gpu_device->super.device_index, gpu_device->super.name,
cpu_copy, cpu_copy->version, gpu_copy,
__FILE__, __LINE__);
}
PARSEC_DEBUG_VERBOSE(10, parsec_gpu_output_stream,
"D2H[%d:%s] task %p:%i GPU data copy %p [%p] now available",
gpu_device->super.device_index, gpu_device->super.name, (void*)task, i, gpu_copy, gpu_copy->original);
Expand All @@ -338,5 +361,6 @@ int parsec_gpu_complete_w2r_task(parsec_device_gpu_module_t *gpu_device,
parsec_thread_mempool_free(es->context_mempool, task);
free(gpu_task);
gpu_device->data_avail_epoch++;
parsec_atomic_fetch_dec_int32( &(gpu_device->mutex) );
return 0;
}