From d4ee7fcd5129e7ee32ff38a5aef9b449d5a756de Mon Sep 17 00:00:00 2001
From: Takoroka <137927095+Takoroka@users.noreply.github.com>
Date: Fri, 30 Oct 2026 05:28:53 -0400
Subject: vulkan: bind memory to sparse terrain arrays layer by layer

The written part of Mario Kart World's sparse terrain arrays is small:
13-26% of their blocks. The two 1.45 GB ASTC arrays were backed in full
on the host, and the two 5.8 GB BC6H arrays, too large to back, stayed
unbacked, so what the game streamed into them never showed.

A color array of 32 MiB or more lying in one sparse region, in a format
uploaded layer by layer (huge ASTC arrays decoded on the GPU, or formats
sampled as they are), is now a sparse resident image (SparseLayers).
When it is refreshed, the layers written since (or, for a new image,
the layers any of whose blocks was written) get device memory bound
with vkQueueBindSparse (under the submit lock, waited on, as the
multi-range buffer cache does) and only those layers are uploaded.
Unwritten layers read zero. One image binds at most 768 MiB (384 MiB on
GPUs with less than 7 GiB); rendering into one binds all its layers.

Mario Kart World, 180 s: the ASTC arrays hold 1.4 MiB per written layer,
four BC6H layers (341 MiB) now have their data, device memory in use
4.8-5.0 GB instead of 6.0-7.1 GB with the texture cache below its
collection threshold, 16.8-17.2 ms average frames, peak working set
13.4 GB.
---
 src/video_core/vulkan_common/vulkan_wrapper.cpp | 11 +++++++++++
 1 file changed, 11 insertions(+)

(limited to 'src/video_core/vulkan_common/vulkan_wrapper.cpp')

diff --git a/src/video_core/vulkan_common/vulkan_wrapper.cpp b/src/video_core/vulkan_common/vulkan_wrapper.cpp
index bb43fe399f..dbaaf450a0 100644
--- a/src/video_core/vulkan_common/vulkan_wrapper.cpp
+++ b/src/video_core/vulkan_common/vulkan_wrapper.cpp
@@ -219,6 +219,7 @@ void Load(VkDevice device, DeviceDispatch& dld) noexcept {
     X(vkGetEventStatus);
     X(vkGetFenceStatus);
     X(vkGetImageMemoryRequirements);
+    X(vkGetImageSparseMemoryRequirements);
     X(vkGetPipelineCacheData);
     X(vkGetMemoryFdKHR);
 #ifdef _WIN32
@@ -866,6 +867,16 @@ VkMemoryRequirements Device::GetImageMemoryRequirements(VkImage image) const noe
     return requirements;
 }
 
+std::vector<VkSparseImageMemoryRequirements> Device::GetImageSparseMemoryRequirements(
+    VkImage image) const {
+    u32 count = 0;
+    dld->vkGetImageSparseMemoryRequirements(handle, image, &count, nullptr);
+    std::vector<VkSparseImageMemoryRequirements> requirements(count);
+    dld->vkGetImageSparseMemoryRequirements(handle, image, &count, requirements.data());
+    requirements.resize(count);
+    return requirements;
+}
+
 std::vector<VkPipelineExecutablePropertiesKHR> Device::GetPipelineExecutablePropertiesKHR(
     VkPipeline pipeline) const {
     const VkPipelineInfoKHR info{
-- 
cgit v1.2.3