Switch 2: window image sizes, queue semaphores, built-in fonts, memory budget

- gsv 1510 CalculateRequirements: block-linear size/alignment for window images.
- nvhost-gpu: Switch 2 channel event 7 is a real (unsignalled) event.
- Puller mirrors NVN2 queue semaphore releases into the page the CPU and GPU
  wait on (the first two-page nx2:gpu region); EDEN_NX2_METHOD_TRACE logs methods.
- pl:u 2000/2010 GetBuiltInFontSharedMemory/GetBuiltInFontData serve plain TTF
  copies of the shared fonts.
- Switch 2 processes count the alias region in their max memory, so the game's
  nninitStartup heap (total - 6.3 GiB) has room for its sound heap.
- gdbrsp.py: interrupt, threads, selthread, hits, continue_async, sleep.

1 files changed, 8 insertions, 0 deletions

diff --git a/src/video_core/nx2.h b/src/video_core/nx2.h
index b8258ca243..9fdf0a7c68 100644
--- a/src/video_core/nx2.h
+++ b/src/video_core/nx2.h
@@ -5,10 +5,18 @@
 
 #include <atomic>
 
+#include "common/common_types.h"
+
 namespace Tegra {
 
 // Set once a Switch 2 (Ampere, NVN2) driver is talking to the GPU. Its command streams use
 // Ampere classes and the Turing+ macro ISA, which Eden's Maxwell engines don't implement.
 inline std::atomic<bool> g_nx2_ampere{false};
 
+// NVN2 queues release their progress values into one GPU page and wait (GPU acquires and CPU
+// polls) on the same offsets in the page below it. Something on the console copies one into
+// the other; until that is known, the puller mirrors releases that land in the source page.
+inline std::atomic<u64> g_nx2_semaphore_source_page{0}; // GPU VA of the released-to page
+inline std::atomic<u64> g_nx2_semaphore_mirror_page{0}; // GPU VA of the waited-on page
+
 } // namespace Tegra