From 940d44ca25628288f1b5f64c25c022b3a4ae5ce2 Mon Sep 17 00:00:00 2001 From: Takoroka <137927095+Takoroka@users.noreply.github.com> Date: Fri, 30 Oct 2026 05:16:34 -0400 Subject: texture_cache: give images over unwritten sparse blocks no device memory Mario Kart World creates nine terrain layer pools in sparse regions (four R8 16384x16384, four RG16 8192x8192, one R16 8192x8192: 2.9 GB of device memory). In 180 s of title and attract demo not one of their blocks is written and nothing renders into them, so they hold zeros; on the console only the few tiles the game maps have memory. nx2:gpu now publishes which 4 MiB blocks of each sparse region have been written (nx2.h, a leaf lock the texture cache may take). An image of 32 MiB or more lying in one sparse region with no written block is created as a sparse resident image with nothing bound: reads return zero, as an unmapped tile does. The first write into its range, or the first time the GPU renders into it (that range then counts as written), marks it remapped, so a backed image replaces it at its next use. Mario Kart World, 180 s: the collector deletes 0-80 images per 600 frames instead of thousands, frames over 100 ms per 600 presents drop from 1-9 to 0-1, peak working set 14.7 GB. One render target in a small sparse region gets replaced once (logged). --- src/video_core/texture_cache/image_base.h | 3 +++ src/video_core/texture_cache/texture_cache.h | 14 ++++++++++++++ 2 files changed, 17 insertions(+) (limited to 'src/video_core/texture_cache') diff --git a/src/video_core/texture_cache/image_base.h b/src/video_core/texture_cache/image_base.h index 20d3883130..cfb839d17e 100644 --- a/src/video_core/texture_cache/image_base.h +++ b/src/video_core/texture_cache/image_base.h @@ -50,6 +50,9 @@ enum class ImageFlagBits : u32 { /// in VRAM); uploads skip the rest, downloads are never done. TrimmedLevels = 1 << 19, ZeroFilled = 1 << 20, ///< Cleared to zero blocks before its first asynchronous decode. + /// Over Switch 2 sparse blocks nobody wrote: all zeros, so the host image has no memory + /// (Unbacked) until the first write, which makes a backed image replace it. + UnwrittenSparse = 1 << 21, }; DECLARE_ENUM_FLAG_OPERATORS(ImageFlagBits) diff --git a/src/video_core/texture_cache/texture_cache.h b/src/video_core/texture_cache/texture_cache.h index d582f520f7..168df50c8b 100644 --- a/src/video_core/texture_cache/texture_cache.h +++ b/src/video_core/texture_cache/texture_cache.h @@ -28,6 +28,7 @@ #include "video_core/texture_cache/samples_helper.h" #include "video_core/texture_cache/texture_cache_base.h" #include "video_core/texture_cache/util.h" +#include "video_core/nx2.h" #include "video_core/nx2_profile.h" #include "video_core/textures/decoders.h" @@ -717,6 +718,11 @@ FramebufferId TextureCache

::GetFramebufferId(const RenderTargets& key) { template void TextureCache

::WriteMemory(DAddr cpu_addr, size_t size) { ForEachImageInRegion(cpu_addr, size, [this, cpu_addr, size](ImageId image_id, Image& image) { + if (True(image.flags & ImageFlagBits::UnwrittenSparse)) { + // Written at last: a backed image replaces it where it is used next. + image.flags |= ImageFlagBits::CpuModified | ImageFlagBits::Remapped; + return; + } if (UploadsByLayer(image)) { MarkLayersWritten(image, cpu_addr, size); return; @@ -2918,6 +2924,14 @@ void TextureCache

::PrepareImage(ImageId image_id, bool is_modification, bool } } if (is_modification) { + if (True(image.flags & ImageFlagBits::UnwrittenSparse) && + False(image.flags & ImageFlagBits::Remapped)) { + // The GPU renders into it: these writes are lost, later uses get a backed image. + LOG_WARNING(HW_GPU, "GPU write into unwritten sparse image {:#x}; replacing it", + image.gpu_addr); + Tegra::Nx2SparseMarkWritten(image.gpu_addr, image.guest_size_bytes); + image.flags |= ImageFlagBits::CpuModified | ImageFlagBits::Remapped; + } MarkModification(image); } lru_cache.Touch(image.lru_index, frame_tick); -- cgit v1.2.3