[Mesa-dev] [PATCH] radv: add a write-combining host-local memory type

[Fredrik Höglund]

Add the new memory type between the two device-local types. This makes
the list of supported memory types look like this:

1) DEVICE_LOCAL |              |               |
2)              | HOST_VISIBLE | HOST_COHERENT |
3) DEVICE_LOCAL | HOST_VISIBLE | HOST_COHERENT |
4)              | HOST_VISIBLE | HOST_COHERENT | HOST_CACHED

With this order a client that searches for a HOST_VISIBLE and
HOST_COHERENT memory type using the algorithm described in section
10.2 of the Vulkan specification (revision 32) will find the host-
local memory type first.

A client that requires the memory type to be HOST_VISIBLE and
HOST_COHERENT, but not DEVICE_LOCAL is most likely searching for
a memory type suitable for staging buffers / images.
---
 src/amd/vulkan/radv_device.c | 17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

diff --git a/src/amd/vulkan/radv_device.c b/src/amd/vulkan/radv_device.c
index af5a151..897e5dd 100644
--- a/src/amd/vulkan/radv_device.c
+++ b/src/amd/vulkan/radv_device.c
@@ -528,19 +528,24 @@ void radv_GetPhysicalDeviceMemoryProperties(
 {
 	RADV_FROM_HANDLE(radv_physical_device, physical_device, physicalDevice);
 
-	pMemoryProperties->memoryTypeCount = 3;
+	pMemoryProperties->memoryTypeCount = 4;
 	pMemoryProperties->memoryTypes[0] = (VkMemoryType) {
 		.propertyFlags = VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT,
 		.heapIndex = 0,
 	};
 	pMemoryProperties->memoryTypes[1] = (VkMemoryType) {
+		.propertyFlags = VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT |
+		VK_MEMORY_PROPERTY_HOST_COHERENT_BIT,
+		.heapIndex = 1,
+	};
+	pMemoryProperties->memoryTypes[2] = (VkMemoryType) {
 		.propertyFlags = VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT |
 		VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT |
 		VK_MEMORY_PROPERTY_HOST_COHERENT_BIT,
 		.heapIndex = 0,
 	};
-	pMemoryProperties->memoryTypes[2] = (VkMemoryType) {
-		.propertyFlags = VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT|
+	pMemoryProperties->memoryTypes[3] = (VkMemoryType) {
+		.propertyFlags = VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT |
 		VK_MEMORY_PROPERTY_HOST_COHERENT_BIT |
 		VK_MEMORY_PROPERTY_HOST_CACHED_BIT,
 		.heapIndex = 1,
@@ -860,7 +865,7 @@ VkResult radv_AllocateMemory(
 		return vk_error(VK_ERROR_OUT_OF_HOST_MEMORY);
 
 	uint64_t alloc_size = align_u64(pAllocateInfo->allocationSize, 4096);
-	if (pAllocateInfo->memoryTypeIndex == 2)
+	if (pAllocateInfo->memoryTypeIndex == 1 || pAllocateInfo->memoryTypeIndex == 3)
 		domain = RADEON_DOMAIN_GTT;
 	else
 		domain = RADEON_DOMAIN_VRAM;
@@ -869,6 +874,10 @@ VkResult radv_AllocateMemory(
 		flags |= RADEON_FLAG_NO_CPU_ACCESS;
 	else
 		flags |= RADEON_FLAG_CPU_ACCESS;
+
+	if (pAllocateInfo->memoryTypeIndex == 1)
+		flags |= RADEON_FLAG_GTT_WC;
+
 	mem->bo = device->ws->buffer_create(device->ws, alloc_size, 32768,
 					       domain, flags);
 
-- 
2.1.4