nx2: map GPU memory before later command lists are read

837133792d queues the GMMU maps on the GPU thread (GPU::RunInOrder) so the
service thread never waits for the texture cache. With the GPU split
(Low GPU accuracy) the GPU thread defers queued functions to its back, but
its front reads the command lists: it read the lists submitted after a map
through GPU addresses that weren't mapped yet. Pokopia stopped after its
first frame.

Add GPU::MapInOrder for the GMMU maps and the sparse tile remaps. It runs
in order like RunInOrder, and with the split it runs on the front after
draining the back, before the front reads further command lists. The
service thread still never takes the texture cache lock, and the back is
idle while the map changes the caches. Unmaps and the other queued work
still run on the back.

Assisted-by: Claude Opus 5.5 (bisect, fix, testing and commit message)

1 files changed, 8 insertions, 0 deletions

diff --git a/src/video_core/gpu.cpp b/src/video_core/gpu.cpp
index a6f31359a3..444dc6d015 100644
--- a/src/video_core/gpu.cpp
+++ b/src/video_core/gpu.cpp
@@ -210,6 +210,10 @@ struct GPU::Impl {
         gpu_thread.Run(std::move(func), is_async);
     }
 
+    void MapInOrder(std::function<void()>&& func) {
+        gpu_thread.Run(std::move(func), is_async, true);
+    }
+
     /// Notify rasterizer that any caches of the specified region should be flushed to Switch memory
     void FlushRegion(DAddr addr, u64 size) {
         gpu_thread.FlushRegion(addr, size, is_async);
@@ -553,6 +557,10 @@ void GPU::RunInOrder(std::function<void()>&& func) {
     impl->RunInOrder(std::move(func));
 }
 
+void GPU::MapInOrder(std::function<void()>&& func) {
+    impl->MapInOrder(std::move(func));
+}
+
 void GPU::InvalidateRegion(DAddr addr, u64 size, bool preserve_gpu_writes) {
     impl->InvalidateRegion(addr, size, preserve_gpu_writes);
 }