nx2: semaphore releases latch the acquires they satisfy

Switch 2 channels that wait for an exact semaphore value poll on the console
and see every release. Eden resumes a waiting channel only between command
lists and writes releases when fences signal, often several back to back, so
an acquire-equal could miss its value and wait forever. Title 4's 3D queue
ticks a counter that two other queues wait on value by value; under the
loading after the post-customization save one of them missed a tick and the
game froze (channel 4 waiting for 0x5293 with the counter at 0x5310).

Pending acquires now register with Nx2SemaphoreWaiters before fences run,
and every query-cache payload write and QMD add/max release latches the
waiters it meets at the moment it is written.

1 files changed, 4 insertions, 1 deletions

diff --git a/src/video_core/query_cache/query_cache.h b/src/video_core/query_cache/query_cache.h
index f94407b026..245ab1996c 100644
--- a/src/video_core/query_cache/query_cache.h
+++ b/src/video_core/query_cache/query_cache.h
@@ -23,6 +23,7 @@
 #include "video_core/gpu.h"
 #include "video_core/host1x/gpu_device_memory_manager.h"
 #include "video_core/memory_manager.h"
+#include "video_core/nx2.h"
 #include "video_core/query_cache/bank_base.h"
 #include "video_core/query_cache/query_base.h"
 #include "video_core/query_cache/query_cache_base.h"
@@ -273,7 +274,7 @@ void QueryCacheBase<Traits>::CounterReport(GPUVAddr addr, QueryType counter_type
     u8* pointer_timestamp = impl->device_memory.template GetPointer<u8>(cpu_addr + 8);
     bool is_synced = (Settings::IsGPUFenceBehaviorDefault() ? !Settings::IsGPULevelHigh() : !Settings::IsGPUFenceBehaviorBalanced() && !Settings::IsGPUFenceBehaviorAccurate()) && is_fence;
     std::function<void()> operation([this, is_synced, streamer, query_base = query, query_location,
-                                     pointer, pointer_timestamp] {
+                                     pointer, pointer_timestamp, cpu_addr] {
         if (True(query_base->flags & QueryFlagBits::IsInvalidated)) {
             if (!is_synced) [[likely]] {
                 impl->pending_unregister.push_back(query_location);
@@ -298,6 +299,7 @@ void QueryCacheBase<Traits>::CounterReport(GPUVAddr addr, QueryType counter_type
             u32 value = static_cast<u32>(query_base->value);
             std::memcpy(pointer, &value, sizeof(value));
         }
+        Tegra::Nx2SemaphoreWaiters::Get().OnWrite(cpu_addr, static_cast<u32>(query_base->value));
         if (!is_synced) [[likely]] {
             impl->pending_unregister.push_back(query_location);
         }
@@ -314,6 +316,7 @@ void QueryCacheBase<Traits>::CounterReport(GPUVAddr addr, QueryType counter_type
             } else {
                 std::memcpy(pointer, &payload, sizeof(payload));
             }
+            Tegra::Nx2SemaphoreWaiters::Get().OnWrite(cpu_addr, payload);
             streamer->Free(new_query_id);
             return;
         }