nvidia/cuda/context: Merge context-stack into context header

CMakeLists.txt

@@ -950,8 +950,6 @@ if(HAVE_NVIDIA_CUDA)
 		"source/nvidia/cuda/nvidia-cuda.cpp"
 		"source/nvidia/cuda/nvidia-cuda-context.hpp"
 		"source/nvidia/cuda/nvidia-cuda-context.cpp"
-		"source/nvidia/cuda/nvidia-cuda-context-stack.hpp"
-		"source/nvidia/cuda/nvidia-cuda-context-stack.cpp"
 		"source/nvidia/cuda/nvidia-cuda-gs-texture.hpp"
 		"source/nvidia/cuda/nvidia-cuda-gs-texture.cpp"
 		"source/nvidia/cuda/nvidia-cuda-memory.hpp"

source/nvidia/cuda/nvidia-cuda-context-stack.cpp

@@ -1,48 +0,0 @@
-/*
- * Modern effects for a modern Streamer
- * Copyright (C) 2020 Michael Fabian Dirks
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
- */
-
-#include "nvidia-cuda-context-stack.hpp"
-#include <stdexcept>
-
-nvidia::cuda::context_stack::context_stack(std::shared_ptr<::nvidia::cuda::cuda>    cuda,
-										   std::shared_ptr<::nvidia::cuda::context> context)
-	: _cuda(cuda), _ctx(context)
-{
-	using namespace ::nvidia::cuda;
-
-	if (!cuda)
-		throw std::invalid_argument("cuda");
-	if (!context)
-		throw std::invalid_argument("context");
-
-	if (result res = _cuda->cuCtxPushCurrent(_ctx->get()); res != result::SUCCESS) {
-		throw std::runtime_error("Failed to push context.");
-	}
-}
-
-nvidia::cuda::context_stack::~context_stack()
-{
-	using namespace ::nvidia::cuda;
-
-	context_t ctx;
-	_cuda->cuCtxGetCurrent(&ctx);
-	if (ctx == _ctx->get()) {
-		_cuda->cuCtxPopCurrent(&ctx);
-	}
-}

source/nvidia/cuda/nvidia-cuda-context-stack.hpp

@@ -1,34 +0,0 @@
-/*
- * Modern effects for a modern Streamer
- * Copyright (C) 2020 Michael Fabian Dirks
- *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA
- */
-
-#pragma once
-#include <memory>
-#include "nvidia-cuda-context.hpp"
-#include "nvidia-cuda.hpp"
-
-namespace nvidia::cuda {
-	class context_stack {
-		std::shared_ptr<::nvidia::cuda::cuda>    _cuda;
-		std::shared_ptr<::nvidia::cuda::context> _ctx;
-
-		public:
-		context_stack(std::shared_ptr<::nvidia::cuda::cuda> cuda, std::shared_ptr<::nvidia::cuda::context> context);
-		~context_stack();
-	};
-} // namespace nvidia::cuda

source/nvidia/cuda/nvidia-cuda-context.cpp

@@ -18,6 +18,7 @@
  */
 
 #include "nvidia-cuda-context.hpp"
+#include <cassert>
 #include <stdexcept>
 #include "util/util-logging.hpp"
 
@@ -46,6 +47,8 @@
 #endif
 #endif
 
+#define ENABLE_STACK_CHECKS
+
 nvidia::cuda::context::context() : _cuda(::nvidia::cuda::cuda::get()), _ctx(), _has_device(false), _device()
 {
 	D_LOG_DEBUG("Initializating... (Addr: 0x%" PRIuPTR ")", this);
@@ -93,3 +96,41 @@ nvidia::cuda::context::context(ID3D11Device* device) : context()
 {
 	return _ctx;
 }
+
+std::shared_ptr<::nvidia::cuda::context_stack> nvidia::cuda::context::enter()
+{
+	return std::make_shared<::nvidia::cuda::context_stack>(shared_from_this());
+}
+
+void nvidia::cuda::context::push()
+{
+	if (auto res = _cuda->cuCtxPushCurrent(_ctx); res != ::nvidia::cuda::result::SUCCESS) {
+		throw ::nvidia::cuda::cuda_error(res);
+	}
+}
+
+void nvidia::cuda::context::pop()
+{
+#ifdef ENABLE_STACK_CHECKS
+	::nvidia::cuda::context_t ctx;
+	if (_cuda->cuCtxGetCurrent(&ctx) == ::nvidia::cuda::result::SUCCESS)
+		assert(ctx == _ctx);
+#endif
+
+	assert(_cuda->cuCtxPopCurrent(&ctx) == ::nvidia::cuda::result::SUCCESS);
+}
+
+void nvidia::cuda::context::synchronize()
+{
+	D_LOG_DEBUG("Synchronizing... (Addr: 0x%" PRIuPTR ")", this);
+
+#ifdef ENABLE_STACK_CHECKS
+	::nvidia::cuda::context_t ctx;
+	if (_cuda->cuCtxGetCurrent(&ctx) == ::nvidia::cuda::result::SUCCESS)
+		assert(ctx == _ctx);
+#endif
+
+	if (auto res = _cuda->cuCtxSynchronize(); res != ::nvidia::cuda::result::SUCCESS) {
+		throw ::nvidia::cuda::cuda_error(res);
+	}
+}

source/nvidia/cuda/nvidia-cuda-context.hpp

@@ -22,7 +22,9 @@
 #include "nvidia-cuda.hpp"
 
 namespace nvidia::cuda {
-	class context {
+	class context_stack;
+
+	class context : public std::enable_shared_from_this<::nvidia::cuda::context> {
 		std::shared_ptr<::nvidia::cuda::cuda> _cuda;
 		::nvidia::cuda::context_t             _ctx;
 		bool                                  _has_device;
@@ -39,5 +41,27 @@ namespace nvidia::cuda {
 #endif
 
 		::nvidia::cuda::context_t get();
+
+		void push();
+		void pop();
+
+		void synchronize();
+
+		public:
+		std::shared_ptr<::nvidia::cuda::context_stack> enter();
+	};
+
+	class context_stack {
+		std::shared_ptr<::nvidia::cuda::context> _ctx;
+
+		public:
+		inline ~context_stack()
+		{
+			_ctx->pop();
+		}
+		inline context_stack(std::shared_ptr<::nvidia::cuda::context> ctx) : _ctx(ctx)
+		{
+			_ctx->push();
+		}
 	};
 } // namespace nvidia::cuda