intel/compute-runtime
1
0
Fork 0
mirror of https://github.com/intel/compute-runtime.git synced 2025-12-30 09:58:55 +08:00
Code Issues Packages Projects Releases Wiki Activity

Implement zeCommandListAppendLaunchCooperativeKernel

Browse Source 
Resolves: NEO-4725


Signed-off-by: Filip Hazubski <filip.hazubski@intel.com>
This commit is contained in:
Filip Hazubski
2021-03-19 23:14:09 +00:00
committed by Compute-Runtime-Automation
parent 3dc3ad36f8
commit 8d55bfe21d
26 changed files with 216 additions and 82 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
opencl/source/command_queue/enqueue_common.h
View File

@@ -15,6 +15,7 @@
#include "shared/source/memory_manager/surface.h"
#include "shared/source/os_interface/os_context.h"
#include "shared/source/program/sync_buffer_handler.h"
#include "shared/source/program/sync_buffer_handler.inl"
#include "shared/source/utilities/range.h"
#include "shared/source/utilities/tag_allocator.h"
@@ -404,7 +405,7 @@ void CommandQueueHw<GfxFamily>::processDispatchForKernels(const MultiDispatchInf
auto &lws = multiDispatchInfo.begin()->getLocalWorkgroupSize();
size_t workGroupsCount = (gws.x * gws.y * gws.z) /
(lws.x * lws.y * lws.z);
device->syncBufferHandler->prepareForEnqueue(workGroupsCount, *multiDispatchInfo.peekMainKernel());
device->getDevice().syncBufferHandler->prepareForEnqueue(workGroupsCount, *multiDispatchInfo.peekMainKernel());
}
if (commandType == CL_COMMAND_NDRANGE_KERNEL) {
@@ -685,7 +686,7 @@ CompletionStamp CommandQueueHw<GfxFamily>::enqueueNonBlocked(
auto rootDeviceIndex = device->getRootDeviceIndex();
if (multiDispatchInfo.peekMainKernel()->usesSyncBuffer(rootDeviceIndex)) {
device->syncBufferHandler->makeResident(getGpgpuCommandStreamReceiver());
device->getDevice().syncBufferHandler->makeResident(getGpgpuCommandStreamReceiver());
}
if (timestampPacketContainer) {