- sanitized dependencies of the softpoly render backend.

This included half the game state and lots of unneeded parts of the software renderer.
The two modules that are shared between softpoly and the classic software renderer have been moved to a neutral place.

src/common/platform/posix/i_system_posix.cpp

@@ -1,24 +1,33 @@
-//-----------------------------------------------------------------------------
-//
-// Copyright 1993-1996 id Software
-// Copyright 1999-2016 Randy Heit
-//
-// 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 3 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, see http://www.gnu.org/licenses/
-//
-//-----------------------------------------------------------------------------
-//
-
+/*
+**---------------------------------------------------------------------------
+** Copyright 2016 Randy Heit
+** All rights reserved.
+**
+** Redistribution and use in source and binary forms, with or without
+** modification, are permitted provided that the following conditions
+** are met:
+**
+** 1. Redistributions of source code must retain the above copyright
+**    notice, this list of conditions and the following disclaimer.
+** 2. Redistributions in binary form must reproduce the above copyright
+**    notice, this list of conditions and the following disclaimer in the
+**    documentation and/or other materials provided with the distribution.
+** 3. The name of the author may not be used to endorse or promote products
+**    derived from this software without specific prior written permission.
+**
+** THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
+** IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
+** OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
+** IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
+** INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
+** NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+** DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+** THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+** (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
+** THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+**---------------------------------------------------------------------------
+**
+*/
 #include <fnmatch.h>
 
 #ifdef __APPLE__

src/common/rendering/r_thread.h

@@ -0,0 +1,224 @@
+/*
+**  Renderer multithreading framework
+**  Copyright (c) 2016 Magnus Norddahl
+**
+**  This software is provided 'as-is', without any express or implied
+**  warranty.  In no event will the authors be held liable for any damages
+**  arising from the use of this software.
+**
+**  Permission is granted to anyone to use this software for any purpose,
+**  including commercial applications, and to alter it and redistribute it
+**  freely, subject to the following restrictions:
+**
+**  1. The origin of this software must not be misrepresented; you must not
+**     claim that you wrote the original software. If you use this software
+**     in a product, an acknowledgment in the product documentation would be
+**     appreciated but is not required.
+**  2. Altered source versions must be plainly marked as such, and must not be
+**     misrepresented as being the original software.
+**  3. This notice may not be removed or altered from any source distribution.
+**
+*/
+
+#pragma once
+
+#include <vector>
+#include <memory>
+#include <thread>
+#include <mutex>
+#include <condition_variable>
+#include "templates.h"
+#include "c_cvars.h"
+#include "basics.h"
+
+// Use multiple threads when drawing
+EXTERN_CVAR(Int, r_multithreaded)
+
+class PolyTriangleThreadData;
+
+namespace swrenderer { class WallColumnDrawerArgs; }
+
+// Worker data for each thread executing drawer commands
+class DrawerThread
+{
+public:
+	std::thread thread;
+	size_t current_queue = 0;
+
+	// Thread line index of this thread
+	int core = 0;
+
+	// Number of active threads
+	int num_cores = 1;
+
+	// NUMA node this thread belongs to
+	int numa_node = 0;
+
+	// Number of active NUMA nodes
+	int num_numa_nodes = 1;
+
+	// Active range for the numa block the cores are part of
+	int numa_start_y = 0;
+	int numa_end_y = MAXHEIGHT;
+
+	// Working buffer used by the tilted (sloped) span drawer
+	const uint8_t *tiltlighting[MAXWIDTH];
+
+	std::shared_ptr<PolyTriangleThreadData> poly;
+	std::shared_ptr<swrenderer::WallColumnDrawerArgs> columndrawer;
+
+	size_t debug_draw_pos = 0;
+
+	// Checks if a line is rendered by this thread
+	bool line_skipped_by_thread(int line)
+	{
+		return line < numa_start_y || line >= numa_end_y || line % num_cores != core;
+	}
+
+	// The number of lines to skip to reach the first line to be rendered by this thread
+	int skipped_by_thread(int first_line)
+	{
+		int clip_first_line = MAX(first_line, numa_start_y);
+		int core_skip = (num_cores - (clip_first_line - core) % num_cores) % num_cores;
+		return clip_first_line + core_skip - first_line;
+	}
+
+	// The number of lines to be rendered by this thread
+	int count_for_thread(int first_line, int count)
+	{
+		count = MIN(count, numa_end_y - first_line);
+		int c = (count - skipped_by_thread(first_line) + num_cores - 1) / num_cores;
+		return MAX(c, 0);
+	}
+
+	// Calculate the dest address for the first line to be rendered by this thread
+	template<typename T>
+	T *dest_for_thread(int first_line, int pitch, T *dest)
+	{
+		return dest + skipped_by_thread(first_line) * pitch;
+	}
+
+	// The first line in the dc_temp buffer used this thread
+	int temp_line_for_thread(int first_line)
+	{
+		return (first_line + skipped_by_thread(first_line)) / num_cores;
+	}
+};
+
+// Task to be executed by each worker thread
+class DrawerCommand
+{
+public:
+	virtual ~DrawerCommand() { }
+
+	virtual void Execute(DrawerThread *thread) = 0;
+};
+
+// Wait for all worker threads before executing next command
+class GroupMemoryBarrierCommand : public DrawerCommand
+{
+public:
+	void Execute(DrawerThread *thread);
+
+private:
+	std::mutex mutex;
+	std::condition_variable condition;
+	size_t count = 0;
+};
+
+// Copy finished rows to video memory
+class MemcpyCommand : public DrawerCommand
+{
+public:
+	MemcpyCommand(void *dest, int destpitch, const void *src, int width, int height, int srcpitch, int pixelsize);
+	void Execute(DrawerThread *thread);
+
+private:
+	void *dest;
+	const void *src;
+	int destpitch;
+	int width;
+	int height;
+	int srcpitch;
+	int pixelsize;
+};
+
+class DrawerCommandQueue;
+typedef std::shared_ptr<DrawerCommandQueue> DrawerCommandQueuePtr;
+
+class DrawerThreads
+{
+public:
+	// Runs the collected commands on worker threads
+	static void Execute(DrawerCommandQueuePtr queue);
+
+	// Waits for all commands to finish executing
+	static void WaitForWorkers();
+
+	static void ResetDebugDrawPos();
+	
+private:
+	DrawerThreads();
+	~DrawerThreads();
+	
+	void StartThreads();
+	void StopThreads();
+	void WorkerMain(DrawerThread *thread);
+
+	static DrawerThreads *Instance();
+	
+	std::mutex threads_mutex;
+	std::vector<DrawerThread> threads;
+
+	std::mutex start_mutex;
+	std::condition_variable start_condition;
+	std::vector<DrawerCommandQueuePtr> active_commands;
+	bool shutdown_flag = false;
+
+	std::mutex end_mutex;
+	std::condition_variable end_condition;
+	size_t tasks_left = 0;
+
+	size_t debug_draw_end = 0;
+
+	DrawerThread single_core_thread;
+	
+	friend class DrawerCommandQueue;
+};
+
+class RenderMemory;
+
+class DrawerCommandQueue
+{
+public:
+	DrawerCommandQueue(RenderMemory *memoryAllocator);
+	
+	void Clear() { commands.clear(); }
+	
+	// Queue command to be executed by drawer worker threads
+	template<typename T, typename... Types>
+	void Push(Types &&... args)
+	{
+		DrawerThreads *threads = DrawerThreads::Instance();
+		if (r_multithreaded != 0)
+		{
+			void *ptr = AllocMemory(sizeof(T));
+			T *command = new (ptr)T(std::forward<Types>(args)...);
+			commands.push_back(command);
+		}
+		else
+		{
+			T command(std::forward<Types>(args)...);
+			command.Execute(&threads->single_core_thread);
+		}
+	}
+	
+private:
+	// Allocate memory valid for the duration of a command execution
+	void *AllocMemory(size_t size);
+	
+	std::vector<DrawerCommand *> commands;
+	RenderMemory *FrameMemory;
+	
+	friend class DrawerThreads;
+};

src/common/utility/basics.h

@@ -3,6 +3,9 @@
 #include <stddef.h>
 #include <stdint.h>
 
+#define MAXWIDTH 12000
+#define MAXHEIGHT 5000
+
 //
 // fixed point, 32bit as 16.16.
 //

src/common/utility/m_bbox.cpp

@@ -1,47 +0,0 @@
-//-----------------------------------------------------------------------------
-//
-// Copyright 1993-1996 id Software
-// Copyright 1999-2016 Randy Heit
-// Copyright 2002-2016 Christoph Oelckers
-//
-// 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 3 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, see http://www.gnu.org/licenses/
-//
-//-----------------------------------------------------------------------------
-//
-// DESCRIPTION:
-//		bounding box class
-//
-//-----------------------------------------------------------------------------
-
-#include "m_bbox.h"
-
-//==========================================================================
-//
-//
-//
-//==========================================================================
-
-void FBoundingBox::AddToBox (const DVector2 &pos)
-{
-	if (pos.X < m_Box[BOXLEFT])
-		m_Box[BOXLEFT] = pos.X;
-	if (pos.X > m_Box[BOXRIGHT])
-		m_Box[BOXRIGHT] = pos.X;
-
-	if (pos.Y < m_Box[BOXBOTTOM])
-		m_Box[BOXBOTTOM] = pos.Y;
-	if (pos.Y > m_Box[BOXTOP])
-		m_Box[BOXTOP] = pos.Y;
-}
-

src/common/utility/m_bbox.h

@@ -1,28 +1,3 @@
-//-----------------------------------------------------------------------------
-//
-// Copyright 1993-1996 id Software
-// Copyright 1999-2016 Randy Heit
-// Copyright 2002-2016 Christoph Oelckers
-//
-// 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 3 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, see http://www.gnu.org/licenses/
-//
-//-----------------------------------------------------------------------------
-//
-// DESCRIPTION:
-//	  Nil.
-//	  
-//-----------------------------------------------------------------------------
 
 #ifndef __M_BBOX_H__
 #define __M_BBOX_H__
@@ -84,7 +59,18 @@ public:
 							m_Box[BOXTOP] > box2.m_Box[BOXTOP] ? m_Box[BOXTOP] : box2.m_Box[BOXTOP]);
 	}
 
-	void AddToBox(const DVector2 &pos);
+	void AddToBox(const DVector2 &pos)
+	{
+		if (pos.X < m_Box[BOXLEFT])
+			m_Box[BOXLEFT] = pos.X;
+		if (pos.X > m_Box[BOXRIGHT])
+			m_Box[BOXRIGHT] = pos.X;
+
+		if (pos.Y < m_Box[BOXBOTTOM])
+			m_Box[BOXBOTTOM] = pos.Y;
+		if (pos.Y > m_Box[BOXTOP])
+			m_Box[BOXTOP] = pos.Y;
+	}
 
 	inline double Top () const { return m_Box[BOXTOP]; }
 	inline double Bottom () const { return m_Box[BOXBOTTOM]; }

src/common/utility/r_memory.cpp

@@ -0,0 +1,101 @@
+/*
+**  Render memory allocation
+**  Copyright (c) 2016-2020 Magnus Norddahl
+**
+**  This software is provided 'as-is', without any express or implied
+**  warranty.  In no event will the authors be held liable for any damages
+**  arising from the use of this software.
+**
+**  Permission is granted to anyone to use this software for any purpose,
+**  including commercial applications, and to alter it and redistribute it
+**  freely, subject to the following restrictions:
+**
+**  1. The origin of this software must not be misrepresented; you must not
+**     claim that you wrote the original software. If you use this software
+**     in a product, an acknowledgment in the product documentation would be
+**     appreciated but is not required.
+**  2. Altered source versions must be plainly marked as such, and must not be
+**     misrepresented as being the original software.
+**  3. This notice may not be removed or altered from any source distribution.
+**
+*/
+
+#include <stdlib.h>
+#include "templates.h"
+#include "r_memory.h"
+#include <stdlib.h>
+
+void *RenderMemory::AllocBytes(int size)
+{
+	size = (size + 15) / 16 * 16; // 16-byte align
+		
+	if (UsedBlocks.empty() || UsedBlocks.back()->Position + size > BlockSize)
+	{
+		if (!FreeBlocks.empty())
+		{
+			auto block = std::move(FreeBlocks.back());
+			block->Position = 0;
+			FreeBlocks.pop_back();
+			UsedBlocks.push_back(std::move(block));
+		}
+		else
+		{
+			UsedBlocks.push_back(std::unique_ptr<MemoryBlock>(new MemoryBlock()));
+		}
+	}
+		
+	auto &block = UsedBlocks.back();
+	void *data = block->Data + block->Position;
+	block->Position += size;
+
+	return data;
+}
+	
+void RenderMemory::Clear()
+{
+	while (!UsedBlocks.empty())
+	{
+		auto block = std::move(UsedBlocks.back());
+		UsedBlocks.pop_back();
+		FreeBlocks.push_back(std::move(block));
+	}
+}
+
+static void* Aligned_Alloc(size_t alignment, size_t size)
+{
+	void* ptr;
+#if defined (_MSC_VER) || defined (__MINGW32__)
+	ptr = _aligned_malloc(size, alignment);
+	if (!ptr)
+		throw std::bad_alloc();
+#else
+	// posix_memalign required alignment to be a min of sizeof(void *)
+	if (alignment < sizeof(void*))
+		alignment = sizeof(void*);
+
+	if (posix_memalign((void**)&ptr, alignment, size))
+		throw std::bad_alloc();
+#endif
+	return ptr;
+}
+
+static void Aligned_Free(void* ptr)
+{
+	if (ptr)
+	{
+#if defined _MSC_VER
+		_aligned_free(ptr);
+#else
+		free(ptr);
+#endif
+	}
+}
+
+RenderMemory::MemoryBlock::MemoryBlock() : Data(static_cast<uint8_t*>(Aligned_Alloc(16, BlockSize))), Position(0)
+{
+}
+
+RenderMemory::MemoryBlock::~MemoryBlock()
+{
+	Aligned_Free(Data);
+}

src/common/utility/r_memory.h

@@ -0,0 +1,43 @@
+#pragma once
+
+#include <memory>
+#include <vector>
+
+// Memory needed for the duration of a frame rendering
+class RenderMemory
+{
+public:
+	void Clear();
+		
+	template<typename T>
+	T *AllocMemory(int size = 1)
+	{
+		return (T*)AllocBytes(sizeof(T) * size);
+	}
+		
+	template<typename T, typename... Types>
+	T *NewObject(Types &&... args)
+	{
+		void *ptr = AllocBytes(sizeof(T));
+		return new (ptr)T(std::forward<Types>(args)...);
+	}
+		
+private:
+	void *AllocBytes(int size);
+		
+	enum { BlockSize = 1024 * 1024 };
+		
+	struct MemoryBlock
+	{
+		MemoryBlock();
+		~MemoryBlock();
+			
+		MemoryBlock(const MemoryBlock &) = delete;
+		MemoryBlock &operator=(const MemoryBlock &) = delete;
+			
+		uint8_t *Data;
+		uint32_t Position;
+	};
+	std::vector<std::unique_ptr<MemoryBlock>> UsedBlocks;
+	std::vector<std::unique_ptr<MemoryBlock>> FreeBlocks;
+};