Switch 2: suspend channels on semaphore acquires; fix pushbuffer desyncs

NVN2 gates command buffers with semaphores the CPU releases after writing
them. Eden read past unmet SEMAPHORED acquires, so it executed stale commands
and desynchronised. Channels now suspend on an unmet acquire and resume from
the same word once it is met (scheduler keeps the waiting channels, the GPU
thread polls them), as in changes.patch.

Also: methods split across GPFIFO entries around a called command buffer are
set aside and resumed; subchannel 2 defaults to inline-to-memory; unbound
subchannels and out-of-range compute/DMA methods are dropped instead of
crashing; 4-byte semaphore releases no longer write 16 bytes; Ampere
SEM_PAYLOAD_HI/SEM_EXECUTE no longer spin as Kepler acquires; unmapped GPU
reads log instead of asserting. Debug aids: EDEN_NX2_PB_CHECK (pushbuffer
changed after submit), EDEN_NX2_QUERY_TRACE, a one-time pushbuffer dump.

1 files changed, 32 insertions, 0 deletions

diff --git a/src/video_core/control/scheduler.cpp b/src/video_core/control/scheduler.cpp
index 8af13e415c..24f1db1f7c 100644
--- a/src/video_core/control/scheduler.cpp
+++ b/src/video_core/control/scheduler.cpp
@@ -4,7 +4,9 @@
 // SPDX-FileCopyrightText: 2021 yuzu Emulator Project
 // SPDX-License-Identifier: GPL-3.0-or-later
 
+#include <algorithm>
 #include <memory>
+#include <vector>
 
 #include "common/assert.h"
 #include "video_core/control/channel_state.h"
@@ -26,6 +28,36 @@ void Scheduler::Push(GPU& gpu, s32 channel, CommandList&& entries) {
     // Multiple channels can prepare their commands in parallel.
     channel_state->payload->dma_pusher.Push(std::move(entries));
     channel_state->payload->dma_pusher.DispatchCalls();
+    if (channel_state->payload->dma_pusher.HasPendingAcquire() &&
+        std::find(pending_channels.begin(), pending_channels.end(), channel_state) ==
+            pending_channels.end()) {
+        pending_channels.push_back(channel_state);
+    }
+    if (!pending_channels.empty()) {
+        ResumePending(gpu);
+    }
+}
+
+void Scheduler::ResumePending(GPU& gpu) {
+    // Running one channel can release the semaphore another waits for: retry until none moves.
+    for (bool progress = true; progress;) {
+        progress = false;
+        for (const auto& state : pending_channels) {
+            auto& dma_pusher = state->payload->dma_pusher;
+            if (!dma_pusher.HasPendingAcquire() || !dma_pusher.CanResumeAcquire()) {
+                continue;
+            }
+            {
+                std::scoped_lock lk{scheduling_guard};
+                gpu.BindChannel(state->bind_id);
+            }
+            dma_pusher.DispatchCalls();
+            progress = true;
+        }
+    }
+    std::erase_if(pending_channels, [](const auto& state) {
+        return !state->payload->dma_pusher.HasPendingAcquire();
+    });
 }
 
 void Scheduler::DeclareChannel(std::shared_ptr<ChannelState> new_channel) {