AuroraMiddleware/VulkanMemoryAllocator
1
0
Fork 0
mirror of https://github.com/GPUOpen-LibrariesAndSDKs/VulkanMemoryAllocator synced 2024-11-05 04:10:06 +00:00
Code Issues Projects Releases Wiki Activity

VMA_MEMORY_USAGE_AUTO* flags: Added support for rare case when a resource used on the device, not used on the host, may still be preferred in host memory

Browse Source 
An update in the documentation.
This commit is contained in:
Adam Sawicki 2022-02-26 11:49:51 +01:00
parent 75446de4cc
commit 86c124c772
1 changed files with 6 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
include/vk_mem_alloc.h
View File

@ -600,7 +600,7 @@ typedef enum VmaAllocationCreateFlagBits
This includes allocations created in \ref custom_memory_pools.
Declares that mapped memory can be read, written, and accessed in random order,
so a `HOST_CACHED` memory type is preferred.
so a `HOST_CACHED` memory type is required.
*/
VMA_ALLOCATION_CREATE_HOST_ACCESS_RANDOM_BIT = 0x00000800,
/**
@ -3639,7 +3639,11 @@ static bool FindMemoryPreferences(
// GPU access, no CPU access (e.g. a color attachment image) - prefer GPU memory
if(deviceAccess)
{
outPreferredFlags |= VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT;
// ...unless there is a clear preference from the user not to do so.
if(preferHost)
outNotPreferredFlags |= VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT;
else
outPreferredFlags |= VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT;
}
// No direct GPU access, no CPU access, just transfers.
// It may be staging copy intended for e.g. preserving image for next frame (then better GPU memory) or