intel/compute-runtime
1
0
Fork 0
mirror of https://github.com/intel/compute-runtime.git synced 2026-01-04 15:53:45 +08:00
Code Issues Packages Projects Releases Wiki Activity

[2/n] Optimize indirect calls.

Browse Source 
Migrate shared allocation when command list sets indirect flags.

Signed-off-by: Michal Mrozek <michal.mrozek@intel.com>
This commit is contained in:
Michal Mrozek
2022-01-25 10:26:22 +00:00
committed by Compute-Runtime-Automation
parent 3d9e1ea3a5
commit a12f9cb377
3 changed files with 59 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
level_zero/core/source/cmdlist/cmdlist.cpp
View File

@@ -160,6 +160,10 @@ void CommandList::migrateSharedAllocations() {
for (auto alloc : driverHandleImp->sharedMakeResidentAllocations) {
pageFaultManager->moveAllocationToGpuDomain(reinterpret_cast<void *>(alloc.second->getGpuAddress()));
}
if (this->unifiedMemoryControls.indirectSharedAllocationsAllowed) {
auto pageFaultManager = device->getDriverHandle()->getMemoryManager()->getPageFaultManager();
pageFaultManager->moveAllocationsWithinUMAllocsManagerToGpuDomain(this->device->getDriverHandle()->getSvmAllocsManager());
}
}
} // namespace L0