intel/compute-runtime
1
0
Fork 0
mirror of https://github.com/intel/compute-runtime.git synced 2026-01-05 18:06:32 +08:00
Code Issues Packages Projects Releases Wiki Activity

graphic memory allocation with alignment in limited Range heap

Browse Source 
Change-Id: Iccfb0fdc2f161e30bfdd26154110185277f176f5
Signed-off-by: dongwonk <dongwon.kim@intel.com>
This commit is contained in:
dongwonk
2019-02-07 11:29:30 -08:00
parent 4ec5be0c99
commit 56972935ad
4 changed files with 131 additions and 37 deletions
Download Patch File Download Diff File Expand all files Collapse all files

61
runtime/os_interface/linux/drm_memory_manager.cpp
View File

@@ -237,10 +237,29 @@ DrmAllocation *DrmMemoryManager::allocateGraphicsMemoryWithAlignment(const Alloc
}
bo->isAllocated = true;
// if limitedRangeAlloction is enabled, memory allocation for bo in the limited Range heap is required
if (limitedGpuAddressRangeAllocator) {
StorageAllocatorType allocType;
bo->gpuAddress = acquireGpuRange(cSize, allocType, false);
if (!bo->gpuAddress) {
bo->close();
delete bo;
alignedFreeWrapper(res);
return nullptr;
}
bo->setUnmapSize(cSize);
bo->setAllocationType(allocType);
}
if (forcePinEnabled && pinBB != nullptr && allocationData.flags.forcePin && allocationData.size >= this->pinThreshold) {
pinBB->pin(&bo, 1, getDefaultCommandStreamReceiver(0)->getOsContext().get()->getDrmContextId());
}
return new DrmAllocation(bo, res, castToUint64(res), cSize, MemoryPool::System4KBPages, allocationData.flags.multiOsContextCapable);
auto allocation = new DrmAllocation(bo, res, bo->gpuAddress, cSize, MemoryPool::System4KBPages, allocationData.flags.multiOsContextCapable);
allocation->driverAllocatedCpuPointer = limitedGpuAddressRangeAllocator ? res : nullptr;
return allocation;
}
DrmAllocation *DrmMemoryManager::allocateGraphicsMemoryWithHostPtr(const AllocationData &allocationData) {
@@ -355,11 +374,11 @@ DrmAllocation *DrmMemoryManager::allocate32BitGraphicsMemoryImpl(const Allocatio
bo->setUnmapSize(realAllocationSize);
bo->gpuAddress = gpuVirtualAddress;
bo->setAllocationType(allocatorType);
auto drmAllocation = new DrmAllocation(bo, const_cast<void *>(allocationData.hostPtr), static_cast<uint64_t>(ptrOffset(gpuVirtualAddress, inputPointerOffset)),
allocationSize, MemoryPool::System4KBPagesWith32BitGpuAddressing, false);
drmAllocation->is32BitAllocation = true;
drmAllocation->gpuBaseAddress = allocatorToUse->getBase();
return drmAllocation;
auto allocation = new DrmAllocation(bo, const_cast<void *>(allocationData.hostPtr), static_cast<uint64_t>(ptrOffset(gpuVirtualAddress, inputPointerOffset)),
allocationSize, MemoryPool::System4KBPagesWith32BitGpuAddressing, false);
allocation->is32BitAllocation = true;
allocation->gpuBaseAddress = allocatorToUse->getBase();
return allocation;
}
size_t alignedAllocationSize = alignUp(allocationData.size, MemoryConstants::pageSize);
@@ -371,9 +390,9 @@ DrmAllocation *DrmMemoryManager::allocate32BitGraphicsMemoryImpl(const Allocatio
}
auto limitedRangeAllocation = false;
void *ptrAlloc = nullptr;
void *ptrAlloc = reinterpret_cast<void *>(res);
if (limitedGpuAddressRangeAllocator.get() && allocatorType == BIT32_ALLOCATOR_INTERNAL) {
if (limitedGpuAddressRangeAllocator && allocatorType == BIT32_ALLOCATOR_INTERNAL) {
limitedRangeAllocation = true;
ptrAlloc = alignedMallocWrapper(alignedAllocationSize, MemoryConstants::allocationAlignment);
@@ -383,10 +402,10 @@ DrmAllocation *DrmMemoryManager::allocate32BitGraphicsMemoryImpl(const Allocatio
}
}
BufferObject *bo = allocUserptr(limitedRangeAllocation ? reinterpret_cast<uintptr_t>(ptrAlloc) : res, alignedAllocationSize, 0, true);
BufferObject *bo = allocUserptr(reinterpret_cast<uintptr_t>(ptrAlloc), alignedAllocationSize, 0, true);
if (!bo) {
if (ptrAlloc != nullptr) {
if (limitedRangeAllocation) {
alignedFreeWrapper(ptrAlloc);
}
allocatorToUse->free(res, allocationSize);
@@ -397,22 +416,16 @@ DrmAllocation *DrmMemoryManager::allocate32BitGraphicsMemoryImpl(const Allocatio
bo->setUnmapSize(allocationSize);
bo->setAllocationType(allocatorType);
bo->gpuAddress = res;
DrmAllocation *drmAllocation = nullptr;
if (limitedRangeAllocation) {
// softpin to the GPU address, res if it uses limitedRangeAllocation
bo->gpuAddress = res;
drmAllocation = new DrmAllocation(bo, ptrAlloc, res, alignedAllocationSize,
MemoryPool::System4KBPagesWith32BitGpuAddressing, false);
} else {
drmAllocation = new DrmAllocation(bo, reinterpret_cast<void *>(res), res, alignedAllocationSize,
MemoryPool::System4KBPagesWith32BitGpuAddressing, false);
}
// softpin to the GPU address, res if it uses limitedRange Allocation
auto allocation = new DrmAllocation(bo, ptrAlloc, res, alignedAllocationSize,
MemoryPool::System4KBPagesWith32BitGpuAddressing, false);
drmAllocation->is32BitAllocation = true;
drmAllocation->gpuBaseAddress = allocatorToUse->getBase();
drmAllocation->driverAllocatedCpuPointer = ptrAlloc;
return drmAllocation;
allocation->is32BitAllocation = true;
allocation->gpuBaseAddress = allocatorToUse->getBase();
allocation->driverAllocatedCpuPointer = limitedRangeAllocation ? ptrAlloc : nullptr;
return allocation;
}
BufferObject *DrmMemoryManager::findAndReferenceSharedBufferObject(int boHandle) {