vulkan: keep the stream buffer within a small iGPU heap

The stream buffer prefers device-local host-visible memory. On a Radeon
780M that is a 256 MiB heap with a 243 MiB budget, and the 256 MiB
buffer alone kept it over budget for the whole session ("VMA heap 2:
usage 256 MiB of budget 243 MiB"). On integrated GPUs whose
device-local host-visible heap is no bigger than the stream buffer, use
three quarters of the heap instead (192 MiB there).

Radeon 780M, Mario Kart World title screen: heap 2 at 192 of 243 MiB,
picture and frame rate unchanged.

2 files changed, 18 insertions, 0 deletions

diff --git a/src/video_core/renderer_vulkan/vk_staging_buffer_pool.cpp b/src/video_core/renderer_vulkan/vk_staging_buffer_pool.cpp
index 7b6571d33d..95fa67c92a 100644
--- a/src/video_core/renderer_vulkan/vk_staging_buffer_pool.cpp
+++ b/src/video_core/renderer_vulkan/vk_staging_buffer_pool.cpp
@@ -41,6 +41,19 @@ constexpr VkDeviceSize MAX_STREAM_BUFFER_SIZE = 256_MiB;
 
 size_t GetStreamBufferSize(const Device& device) {
     if (!device.HasDebuggingToolAttached()) {
+        if (device.IsIntegrated()) {
+            // An integrated GPU's device-local host-visible heap can be no bigger than the stream
+            // buffer (Radeon 780M: 256 MiB, 243 MiB budget), which then sat over that budget.
+            // Leave a quarter of it free.
+            VkDeviceSize heap_size = 0;
+            ForEachDeviceLocalHostVisibleHeap(
+                device, [&heap_size](size_t, VkMemoryHeap& heap) {
+                    heap_size = (std::max)(heap_size, heap.size);
+                });
+            if (heap_size != 0 && heap_size <= MAX_STREAM_BUFFER_SIZE) {
+                return Common::AlignUp(heap_size * 3 / 4, MAX_ALIGNMENT);
+            }
+        }
         return MAX_STREAM_BUFFER_SIZE;
     }
 
diff --git a/src/video_core/vulkan_common/vulkan_device.h b/src/video_core/vulkan_common/vulkan_device.h
index d8524f6575..eb7cb15dc8 100644
--- a/src/video_core/vulkan_common/vulkan_device.h
+++ b/src/video_core/vulkan_common/vulkan_device.h
@@ -928,6 +928,11 @@ FN_MAX_LIMIT_LIST
         return 0x00010300U;
     }
 
+    /// Returns true for an integrated GPU (its device-local memory is system memory).
+    bool IsIntegrated() const {
+        return is_integrated;
+    }
+
     /// Returns true when a known debugging tool is attached.
     bool HasDebuggingToolAttached() const {
         return has_renderdoc || has_nsight_graphics || has_radeon_gpu_profiler;