raphiaura/KytyPS5
1
0
Fork 0
mirror of https://github.com/KytyPS5/KytyPS5.git synced 2026-10-20 10:56:58 +02:00
Code Issues Packages Projects Releases Wiki Activity

renderer: reduce synchronization and allocation overhead

Browse Source 
Performance changes from #1495:

- Sync changed BDA ranges instead of scanning all mappings.
- Let VMA suballocate BDA buffers.
- Remove the renderer lock from presentation.
- Keep compatible pipeline caches across builds.
- Fold masked shifts that always produce zero.
This commit is contained in:
shadowbeat070
2026-10-16 16:58:07 +02:00
committed by nmzik
parent 4c6be6cd1d
commit d58aefde49
12 changed files with 277 additions and 38 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/graphics/host_gpu/renderer/cache/streamBuffer.cpp
Vendored
+1 -3
View File
@@ -67,11 +67,9 @@ Buffer::Buffer(GraphicContext& graphics, CommandScheduler& scheduler, MemoryUsag
buffer_info.usage = flags;
const bool with_bda = bool(flags & vk::BufferUsageFlagBits::eShaderDeviceAddress);
const VmaAllocationCreateFlags bda_flag =
with_bda ? VMA_ALLOCATION_CREATE_DEDICATED_MEMORY_BIT : 0;
VmaAllocationCreateInfo allocation_info {};
allocation_info.flags =
VMA_ALLOCATION_CREATE_WITHIN_BUDGET_BIT | bda_flag | AllocationFlags(usage);
VMA_ALLOCATION_CREATE_WITHIN_BUDGET_BIT | AllocationFlags(usage);
allocation_info.usage = AllocationUsage(usage);
allocation_info.preferredFlags = usage == MemoryUsage::DeviceLocal
? VkMemoryPropertyFlags {}