intel/compute-runtime
1
0
Fork 0
mirror of https://github.com/intel/compute-runtime.git synced 2025-12-24 12:23:05 +08:00
Code Issues Packages Projects Releases Wiki Activity

feature: region group barrier allocation support

Browse Source 
Related-To: NEO-11031

Signed-off-by: Bartosz Dunajski <bartosz.dunajski@intel.com>
This commit is contained in:
Bartosz Dunajski
2024-06-03 14:19:36 +00:00
committed by Compute-Runtime-Automation
parent 844b29d17c
commit 692def2c79
13 changed files with 112 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
shared/source/kernel/kernel_descriptor.h
View File

@@ -124,7 +124,8 @@ struct KernelDescriptor {
bool hasSample : 1;
// 3
bool usesAssert : 1;
bool reserved : 7;
bool usesRegionGroupBarrier : 1;
bool reserved : 6;
};
std::array<bool, 4> packed;
} flags = {};
@@ -158,7 +159,6 @@ struct KernelDescriptor {
CrossThreadDataOffset regionGroupSize[3] = {undefined<CrossThreadDataOffset>, undefined<CrossThreadDataOffset>, undefined<CrossThreadDataOffset>};
CrossThreadDataOffset regionGroupDimension = undefined<CrossThreadDataOffset>;
CrossThreadDataOffset regionGroupWgCount = undefined<CrossThreadDataOffset>;
CrossThreadDataOffset regionGroupBarrierBuffer = undefined<CrossThreadDataOffset>;
} dispatchTraits;
struct {
@@ -185,6 +185,7 @@ struct KernelDescriptor {
ArgDescPointer syncBufferAddress;
ArgDescPointer rtDispatchGlobals;
ArgDescPointer assertBufferAddress;
ArgDescPointer regionGroupBarrierBuffer;
CrossThreadDataOffset privateMemorySize = undefined<CrossThreadDataOffset>;
CrossThreadDataOffset maxWorkGroupSize = undefined<CrossThreadDataOffset>;
CrossThreadDataOffset simdSize = undefined<CrossThreadDataOffset>;