memory system tweaks

This reverts commit 20636ec951.

CMakeLists.txt

@@ -8,8 +8,8 @@ option(BUILD_SHARED_LIBS "Build shared library" ON)
 
 # Version
 SET(LIB_MAJOR_VERSION "1")
-SET(LIB_MINOR_VERSION "12")
-SET(LIB_REVISION "0")
+SET(LIB_MINOR_VERSION "15")
+SET(LIB_REVISION "4")
 SET(LIB_VERSION "${LIB_MAJOR_VERSION}.${LIB_MINOR_VERSION}.${LIB_REVISION}")
 
 # Build Type
@@ -60,6 +60,11 @@ if(NOT MSVC)
 	set_property(TARGET Refresh PROPERTY COMPILE_FLAGS "-std=gnu99 -Wall -Wno-strict-aliasing -pedantic")
 endif()
 
+# Windows is silly and we need to manually include the Vulkan SDK
+if(MSVC)
+	target_include_directories(Refresh PUBLIC $ENV{VULKAN_SDK}/include)
+endif()
+
 # Refresh folders as includes, for other targets to consume
 target_include_directories(Refresh PUBLIC
 	$<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/src>

include/Refresh.h

@@ -55,8 +55,8 @@ extern "C" {
 /* Version API */
 
 #define REFRESH_MAJOR_VERSION   1
-#define REFRESH_MINOR_VERSION	12
-#define REFRESH_PATCH_VERSION	0
+#define REFRESH_MINOR_VERSION	15
+#define REFRESH_PATCH_VERSION	4
 
 #define REFRESH_COMPILED_VERSION ( \
 	(REFRESH_MAJOR_VERSION * 100 * 100) + \
@@ -76,6 +76,7 @@ typedef struct Refresh_ShaderModule Refresh_ShaderModule;
 typedef struct Refresh_ComputePipeline Refresh_ComputePipeline;
 typedef struct Refresh_GraphicsPipeline Refresh_GraphicsPipeline;
 typedef struct Refresh_CommandBuffer Refresh_CommandBuffer;
+typedef struct Refresh_Fence Refresh_Fence;
 
 typedef enum Refresh_PresentMode
 {
@@ -778,15 +779,20 @@ REFRESHAPI void Refresh_SetTextureData(
 
 /* Uploads YUV image data to three R8 texture objects.
  *
- * y:		The texture storing the Y data.
- * u:		The texture storing the U (Cb) data.
- * v:		The texture storing the V (Cr) data.
- * yWidth:	The width of the Y plane.
- * yHeight:	The height of the Y plane.
- * uvWidth:	The width of the U/V planes.
- * uvHeight:	The height of the U/V planes.
- * data:	A pointer to the raw YUV image data.
- * dataLength:	The size of the image data in bytes.
+ * y:            The texture storing the Y data.
+ * u:            The texture storing the U (Cb) data.
+ * v:            The texture storing the V (Cr) data.
+ * yWidth:       The width of the Y plane.
+ * yHeight:      The height of the Y plane.
+ * uvWidth:      The width of the U/V planes.
+ * uvHeight:     The height of the U/V planes.
+ * yData:        A pointer to the raw Y image data.
+ * uData:        A pointer to the raw U image data.
+ * vData:        A pointer to the raw V image data.
+ * yDataLength:  The size of the Y image data in bytes.
+ * uvDataLength: The size of the UV image data in bytes.
+ * yStride:      The length of a Y image data row in bytes.
+ * uvStride:     The length of a UV image data row in bytes.
  */
 REFRESHAPI void Refresh_SetTextureDataYUV(
 	Refresh_Device *driverData,
@@ -798,8 +804,13 @@ REFRESHAPI void Refresh_SetTextureDataYUV(
 	uint32_t yHeight,
 	uint32_t uvWidth,
 	uint32_t uvHeight,
-	void* data,
-	uint32_t dataLength
+	void *yDataPtr,
+	void *uDataPtr,
+	void *vDataPtr,
+	uint32_t yDataLength,
+	uint32_t uvDataLength,
+	uint32_t yStride,
+	uint32_t uvStride
 );
 
 /* Performs an asynchronous texture-to-texture copy.
@@ -1196,15 +1207,50 @@ REFRESHAPI Refresh_Texture* Refresh_AcquireSwapchainTexture(
 /* Submits all of the enqueued commands. */
 REFRESHAPI void Refresh_Submit(
 	Refresh_Device* device,
-	uint32_t commandBufferCount,
-	Refresh_CommandBuffer **pCommandBuffers
+	Refresh_CommandBuffer *commandBuffer
 );
 
-/* Waits for all submissions to complete. */
+/* Submits a command buffer and acquires a fence.
+ * You can use the fence to check if or wait until the command buffer has finished processing.
+ * You are responsible for releasing this fence when you are done using it.
+ */
+REFRESHAPI Refresh_Fence* Refresh_SubmitAndAcquireFence(
+	Refresh_Device* device,
+	Refresh_CommandBuffer *commandBuffer
+);
+
+/* Waits for the device to become idle. */
 REFRESHAPI void Refresh_Wait(
 	Refresh_Device *device
 );
 
+/* Waits for given fences to be signaled.
+ *
+ * waitAll: If 0, waits for any fence to be signaled. If 1, waits for all fences to be signaled.
+ * fenceCount: The number of fences being submitted.
+ * pFences: An array of fences to be waited on.
+ */
+REFRESHAPI void Refresh_WaitForFences(
+	Refresh_Device *device,
+	uint8_t waitAll,
+	uint32_t fenceCount,
+	Refresh_Fence **pFences
+);
+
+/* Check the status of a fence. 1 means the fence is signaled. */
+REFRESHAPI int Refresh_QueryFence(
+	Refresh_Device *device,
+	Refresh_Fence *fence
+);
+
+/* Allows the fence to be reused by future command buffer submissions.
+ * If you do not release fences after acquiring them, you will cause unbounded resource growth.
+ */
+REFRESHAPI void Refresh_ReleaseFence(
+	Refresh_Device *device,
+	Refresh_Fence *fence
+);
+
 #ifdef __cplusplus
 }
 #endif /* __cplusplus */

include/Refresh_Image.h

@@ -44,115 +44,45 @@
 extern "C" {
 #endif /* __cplusplus */
 
-/* Decodes PNG data into raw RGBA8 texture data.
- *
- * w:		    Filled with the width of the image.
- * h:		    Filled with the height of the image.
- * numChannels: Filled with the number of channels in the image.
- *
- * Returns a block of memory suitable for use with Refresh_SetTextureData2D.
- * Be sure to free the memory with Refresh_Image_FreePNG after use!
- */
-REFRESHAPI uint8_t* Refresh_Image_LoadPNGFromFile(
-	char const *filename,
-	int32_t *w,
-	int32_t *h,
-	int32_t *numChannels
-);
+/* Image Read API */
 
-/* Decodes PNG data into raw RGBA8 texture data.
+/* Decodes image data into raw RGBA8 texture data.
  *
  * w:		    Filled with the width of the image.
  * h:		    Filled with the height of the image.
- * numChannels: Filled with the number of channels in the image.
+ * len:			Filled with the length of pixel data in bytes.
  *
  * Returns a block of memory suitable for use with Refresh_SetTextureData2D.
- * Be sure to free the memory with Refresh_Image_FreePNG after use!
+ * Be sure to free the memory with Refresh_Image_Free after use!
  */
-REFRESHAPI uint8_t* Refresh_Image_LoadPNGFromMemory(
-	uint8_t *buffer,
+REFRESHAPI uint8_t* Refresh_Image_Load(
+	uint8_t *bufferPtr,
 	int32_t bufferLength,
 	int32_t *w,
 	int32_t *h,
-	int32_t *numChannels
+	int32_t *len
 );
 
-/* Frees memory returned by Refresh_Image_LoadPNG functions. (Do NOT free the memory yourself!)
+/* Frees memory returned by Refresh_Image_Load. Do NOT free the memory yourself!
  *
  * mem: A pointer previously returned by Refresh_Image_LoadPNG.
  */
-REFRESHAPI void Refresh_Image_FreePNG(uint8_t *mem);
-
-/* Decodes QOI data into raw RGBA8 texture data.
- *
- * w:		    Filled with the width of the image.
- * h:		    Filled with the height of the image.
- * numChannels: Filled with the number of channels in the image.
- *
- * Returns a block of memory suitable for use with Refresh_SetTextureData2D.
- * Be sure to free the memory with Refresh_Image_FreeQOI after use!
- */
-REFRESHAPI uint8_t* Refresh_Image_LoadQOIFromFile(
-	char const *filename,
-	int32_t *w,
-	int32_t *h,
-	int32_t *numChannels
-);
-
-/* Decodes QOI data into raw RGBA8 texture data.
- *
- * bufferLength: The length of the input buffer to be decoded.
- * w:		     Filled with the width of the image.
- * h:		     Filled with the height of the image.
- * numChannels:  Filled with the number of channels in the image.
- *
- * Returns a block of memory suitable for use with Refresh_SetTextureData2D.
- * Be sure to free the memory with Refresh_Image_FreeQOI after use!
- */
-REFRESHAPI uint8_t* Refresh_Image_LoadQOIFromMemory(
-	uint8_t *buffer,
-	int32_t bufferLength,
-	int32_t *w,
-	int32_t *h,
-	int32_t *numChannels
-);
-
-/* Frees memory returned by Refresh_Image_LoadQOI functions. (Do NOT free the memory yourself!)
- *
- * mem: A pointer previously returned by Refresh_Image_LoadQOI.
- */
-REFRESHAPI void Refresh_Image_FreeQOI(uint8_t *mem);
+REFRESHAPI void Refresh_Image_Free(uint8_t *mem);
 
 /* Image Write API */
 
-/* Encodes 32-bit color data into PNG data.
+/* Returns a buffer of PNG encoded from RGBA8 color data.
  *
- * filename:    The filename that the image will be written to.
- * w:	        The width of the PNG data.
- * h:	        The height of the PNG data.
- * bgra:		Whether the data is in BGRA8 format. Otherwise will assume RBGA8.
- * data:	    The raw color data.
+ * data:	The raw color data.
+ * w:		The width of the color data.
+ * h:		The height of the color data.
+ * len:		Filled with the length of PNG data in bytes.
  */
 REFRESHAPI void Refresh_Image_SavePNG(
-	char const *filename,
+	const char* filename,
+	uint8_t* data,
 	int32_t w,
-	int32_t h,
-	uint8_t bgra,
-	uint8_t *data
-);
-
-/* Encodes 32-bit color data into PNG data.
- *
- * filename:    The filename that the image will be written to.
- * w:	        The width of the PNG data.
- * h:	        The height of the PNG data.
- * data:	    The raw color data.
- */
-REFRESHAPI void Refresh_Image_SaveQOI(
-	char const *filename,
-	int32_t w,
-	int32_t h,
-	uint8_t *data
+	int32_t h
 );
 
 #ifdef __cplusplus

shadercompiler/Program.cs

@@ -238,7 +238,7 @@ partial class Program
 	{
 		Process glslc = Process.Start(
 			"glslc",
-			$"{glslPath} -o {outputPath}"
+			$"\"{glslPath}\" -o \"{outputPath}\""
 		);
 		glslc.WaitForExit();
 		if (glslc.ExitCode != 0)
@@ -254,7 +254,7 @@ partial class Program
 	{
 		Process spirvcross = Process.Start(
 			"spirv-cross",
-			$"{spirvPath} --hlsl --shader-model 50 --output {outputPath}"
+			$"\"{spirvPath}\" --hlsl --shader-model 50 --output \"{outputPath}\""
 		);
 		spirvcross.WaitForExit();
 		if (spirvcross.ExitCode != 0)

src/Refresh.c

@@ -438,8 +438,13 @@ void Refresh_SetTextureDataYUV(
 	uint32_t yHeight,
 	uint32_t uvWidth,
 	uint32_t uvHeight,
-	void* data,
-	uint32_t dataLength
+	void *yDataPtr,
+	void *uDataPtr,
+	void *vDataPtr,
+	uint32_t yDataLength,
+	uint32_t uvDataLength,
+	uint32_t yStride,
+	uint32_t uvStride
 ) {
 	NULL_RETURN(device);
 	device->SetTextureDataYUV(
@@ -452,8 +457,13 @@ void Refresh_SetTextureDataYUV(
 		yHeight,
 		uvWidth,
 		uvHeight,
-		data,
-		dataLength
+		yDataPtr,
+		uDataPtr,
+		vDataPtr,
+		yDataLength,
+		uvDataLength,
+		yStride,
+		uvStride
 	);
 }
 
@@ -882,14 +892,23 @@ void Refresh_SetSwapchainPresentMode(
 
 void Refresh_Submit(
 	Refresh_Device *device,
-	uint32_t commandBufferCount,
-	Refresh_CommandBuffer **pCommandBuffers
+	Refresh_CommandBuffer *commandBuffer
 ) {
 	NULL_RETURN(device);
 	device->Submit(
 		device->driverData,
-		commandBufferCount,
-		pCommandBuffers
+		commandBuffer
+	);
+}
+
+Refresh_Fence* Refresh_SubmitAndAcquireFence(
+	Refresh_Device *device,
+	Refresh_CommandBuffer *commandBuffer
+) {
+	NULL_RETURN_NULL(device);
+	return device->SubmitAndAcquireFence(
+		device->driverData,
+		commandBuffer
 	);
 }
 
@@ -902,4 +921,44 @@ void Refresh_Wait(
 	);
 }
 
+void Refresh_WaitForFences(
+	Refresh_Device *device,
+	uint8_t waitAll,
+	uint32_t fenceCount,
+	Refresh_Fence **pFences
+) {
+	NULL_RETURN(device);
+	device->WaitForFences(
+		device->driverData,
+		waitAll,
+		fenceCount,
+		pFences
+	);
+}
+
+int Refresh_QueryFence(
+	Refresh_Device *device,
+	Refresh_Fence *fence
+) {
+	if (device == NULL) {
+		return 0;
+	}
+
+	return device->QueryFence(
+		device->driverData,
+		fence
+	);
+}
+
+void Refresh_ReleaseFence(
+	Refresh_Device *device,
+	Refresh_Fence *fence
+) {
+	NULL_RETURN(device);
+	device->ReleaseFence(
+		device->driverData,
+		fence
+	);
+}
+
 /* vim: set noexpandtab shiftwidth=8 tabstop=8: */

src/Refresh_Driver.h

@@ -268,7 +268,7 @@ struct Refresh_Device
 
 	/* Setters */
 
-	void(*SetTextureData)(
+	void (*SetTextureData)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_TextureSlice *textureSlice,
@@ -276,7 +276,7 @@ struct Refresh_Device
 		uint32_t dataLengthInBytes
 	);
 
-	void(*SetTextureDataYUV)(
+	void (*SetTextureDataYUV)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer* commandBuffer,
 		Refresh_Texture *y,
@@ -286,11 +286,16 @@ struct Refresh_Device
 		uint32_t yHeight,
 		uint32_t uvWidth,
 		uint32_t uvHeight,
-		void* data,
-		uint32_t dataLength
+		void *yDataPtr,
+		void *uDataPtr,
+		void *vDataPtr,
+		uint32_t yDataLength,
+		uint32_t uvDataLength,
+		uint32_t yStride,
+		uint32_t uvStride
 	);
 
-	void(*CopyTextureToTexture)(
+	void (*CopyTextureToTexture)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_TextureSlice *sourceTextureSlice,
@@ -298,14 +303,14 @@ struct Refresh_Device
 		Refresh_Filter filter
 	);
 
-	void(*CopyTextureToBuffer)(
+	void (*CopyTextureToBuffer)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_TextureSlice *textureSlice,
 		Refresh_Buffer *buffer
 	);
 
-	void(*SetBufferData)(
+	void (*SetBufferData)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_Buffer *buffer,
@@ -314,14 +319,14 @@ struct Refresh_Device
 		uint32_t dataLength
 	);
 
-	uint32_t(*PushVertexShaderUniforms)(
+	uint32_t (*PushVertexShaderUniforms)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		void *data,
 		uint32_t dataLengthInBytes
 	);
 
-	uint32_t(*PushFragmentShaderUniforms)(
+	uint32_t (*PushFragmentShaderUniforms)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		void *data,
@@ -335,14 +340,14 @@ struct Refresh_Device
 		uint32_t dataLengthInBytes
 	);
 
-	void(*BindVertexSamplers)(
+	void (*BindVertexSamplers)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_Texture **pTextures,
 		Refresh_Sampler **pSamplers
 	);
 
-	void(*BindFragmentSamplers)(
+	void (*BindFragmentSamplers)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_Texture **pTextures,
@@ -351,7 +356,7 @@ struct Refresh_Device
 
 	/* Getters */
 
-	void(*GetBufferData)(
+	void (*GetBufferData)(
 		Refresh_Renderer *driverData,
 		Refresh_Buffer *buffer,
 		void *data,
@@ -360,39 +365,39 @@ struct Refresh_Device
 
 	/* Disposal */
 
-	void(*QueueDestroyTexture)(
+	void (*QueueDestroyTexture)(
 		Refresh_Renderer *driverData,
 		Refresh_Texture *texture
 	);
 
-	void(*QueueDestroySampler)(
+	void (*QueueDestroySampler)(
 		Refresh_Renderer *driverData,
 		Refresh_Sampler *sampler
 	);
 
-	void(*QueueDestroyBuffer)(
+	void (*QueueDestroyBuffer)(
 		Refresh_Renderer *driverData,
 		Refresh_Buffer *buffer
 	);
 
-	void(*QueueDestroyShaderModule)(
+	void (*QueueDestroyShaderModule)(
 		Refresh_Renderer *driverData,
 		Refresh_ShaderModule *shaderModule
 	);
 
-	void(*QueueDestroyComputePipeline)(
+	void (*QueueDestroyComputePipeline)(
 		Refresh_Renderer *driverData,
 		Refresh_ComputePipeline *computePipeline
 	);
 
-	void(*QueueDestroyGraphicsPipeline)(
+	void (*QueueDestroyGraphicsPipeline)(
 		Refresh_Renderer *driverData,
 		Refresh_GraphicsPipeline *graphicsPipeline
 	);
 
 	/* Graphics State */
 
-	void(*BeginRenderPass)(
+	void (*BeginRenderPass)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_ColorAttachmentInfo *colorAttachmentInfos,
@@ -400,30 +405,30 @@ struct Refresh_Device
 		Refresh_DepthStencilAttachmentInfo *depthStencilAttachmentInfo
 	);
 
-	void(*EndRenderPass)(
+	void (*EndRenderPass)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer
 	);
 
-	void(*SetViewport)(
+	void (*SetViewport)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_Viewport *viewport
 	);
 
-	void(*SetScissor)(
+	void (*SetScissor)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_Rect *scissor
 	);
 
-	void(*BindGraphicsPipeline)(
+	void (*BindGraphicsPipeline)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_GraphicsPipeline *graphicsPipeline
 	);
 
-	void(*BindVertexBuffers)(
+	void (*BindVertexBuffers)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		uint32_t firstBinding,
@@ -432,7 +437,7 @@ struct Refresh_Device
 		uint64_t *pOffsets
 	);
 
-	void(*BindIndexBuffer)(
+	void (*BindIndexBuffer)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_Buffer *buffer,
@@ -440,19 +445,19 @@ struct Refresh_Device
 		Refresh_IndexElementSize indexElementSize
 	);
 
-	void(*BindComputePipeline)(
+	void (*BindComputePipeline)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_ComputePipeline *computePipeline
 	);
 
-	void(*BindComputeBuffers)(
+	void (*BindComputeBuffers)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_Buffer **pBuffers
 	);
 
-	void(*BindComputeTextures)(
+	void (*BindComputeTextures)(
 		Refresh_Renderer *driverData,
 		Refresh_CommandBuffer *commandBuffer,
 		Refresh_Texture **pTextures
@@ -464,7 +469,7 @@ struct Refresh_Device
 		Refresh_PresentMode presentMode
 	);
 
-	void(*UnclaimWindow)(
+	void (*UnclaimWindow)(
 		Refresh_Renderer *driverData,
 		void *windowHandle
 	);
@@ -492,16 +497,37 @@ struct Refresh_Device
 		Refresh_PresentMode presentMode
 	);
 
-	void(*Submit)(
+	void (*Submit)(
 		Refresh_Renderer *driverData,
-		uint32_t commandBufferCount,
-		Refresh_CommandBuffer **pCommandBuffers
+		Refresh_CommandBuffer *commandBuffer
 	);
 
-	void(*Wait)(
+	Refresh_Fence* (*SubmitAndAcquireFence)(
+		Refresh_Renderer *driverData,
+		Refresh_CommandBuffer *commandBuffer
+	);
+
+	void (*Wait)(
 		Refresh_Renderer *driverData
 	);
 
+	void (*WaitForFences)(
+		Refresh_Renderer *driverData,
+		uint8_t waitAll,
+		uint32_t fenceCount,
+		Refresh_Fence **pFences
+	);
+
+	int (*QueryFence)(
+		Refresh_Renderer *driverData,
+		Refresh_Fence *fence
+	);
+
+	void (*ReleaseFence)(
+		Refresh_Renderer *driverData,
+		Refresh_Fence *fence
+	);
+
 	/* Opaque pointer for the Driver */
 	Refresh_Renderer *driverData;
 };
@@ -555,7 +581,11 @@ struct Refresh_Device
 	ASSIGN_DRIVER_FUNC(GetSwapchainFormat, name) \
 	ASSIGN_DRIVER_FUNC(SetSwapchainPresentMode, name) \
 	ASSIGN_DRIVER_FUNC(Submit, name) \
-	ASSIGN_DRIVER_FUNC(Wait, name)
+	ASSIGN_DRIVER_FUNC(SubmitAndAcquireFence, name) \
+	ASSIGN_DRIVER_FUNC(Wait, name) \
+	ASSIGN_DRIVER_FUNC(WaitForFences, name) \
+	ASSIGN_DRIVER_FUNC(QueryFence, name) \
+	ASSIGN_DRIVER_FUNC(ReleaseFence, name)
 
 typedef struct Refresh_Driver
 {

src/Refresh_Driver_Template.txt

@@ -268,6 +268,19 @@ static void TEMPLATE_DrawPrimitives(
 	NOT_IMPLEMENTED
 }
 
+static void TEMPLATE_DrawPrimitivesIndirect(
+	Refresh_Renderer *driverData,
+	Refresh_CommandBuffer *commandBuffer,
+	Refresh_Buffer *buffer,
+	uint32_t offsetInBytes,
+	uint32_t drawCount,
+	uint32_t stride,
+	uint32_t vertexParamOffset,
+	uint32_t fragmentParamOffset
+) {
+	NOT_IMPLEMENTED
+}
+
 static void TEMPLATE_DispatchCompute(
 	Refresh_Renderer *device,
 	Refresh_CommandBuffer *commandBuffer,
@@ -347,8 +360,13 @@ static void TEMPLATE_SetTextureDataYUV(
 	uint32_t yHeight,
 	uint32_t uvWidth,
 	uint32_t uvHeight,
-	void* data,
-	uint32_t dataLength
+	void *yDataPtr,
+	void *uDataPtr,
+	void *vDataPtr,
+	uint32_t yDataLength,
+	uint32_t uvDataLength,
+	uint32_t yStride,
+	uint32_t uvStride
 ) {
 	NOT_IMPLEMENTED
 }
@@ -485,10 +503,15 @@ static void TEMPLATE_QueueDestroyGraphicsPipeline(
 
 /* Graphics State */
 
+static Refresh_CommandBuffer* TEMPLATE_AcquireCommandBuffer(
+	Refresh_Renderer *driverData
+) {
+	NOT_IMPLEMENTED
+}
+
 static void TEMPLATE_BeginRenderPass(
 	Refresh_Renderer *driverData,
 	Refresh_CommandBuffer *commandBuffer,
-	Refresh_Rect *renderArea,
 	Refresh_ColorAttachmentInfo *colorAttachmentInfos,
 	uint32_t colorAttachmentCount,
 	Refresh_DepthStencilAttachmentInfo *depthStencilAttachmentInfo
@@ -548,6 +571,8 @@ static void TEMPLATE_BindIndexBuffer(
 	NOT_IMPLEMENTED
 }
 
+/* Compute State */
+
 static void TEMPLATE_BindComputePipeline(
 	Refresh_Renderer *driverData,
 	Refresh_CommandBuffer *commandBuffer,
@@ -572,14 +597,24 @@ static void TEMPLATE_BindComputeTextures(
 	NOT_IMPLEMENTED
 }
 
-static Refresh_CommandBuffer* TEMPLATE_AcquireCommandBuffer(
+/* Window and Swapchain Management */
+
+static uint8_t TEMPLATE_ClaimWindow(
 	Refresh_Renderer *driverData,
-	uint8_t fixed
+	void *windowHandle,
+	Refresh_PresentMode presentMode
 ) {
 	NOT_IMPLEMENTED
 }
 
-Refresh_Texture* TEMPLATE_AcquireSwapchainTexture(
+static void TEMPLATE_UnclaimWindow(
+	Refresh_Renderer *driverData,
+	void *windowHandle
+) {
+	NOT_IMPLEMENTED
+}
+
+static Refresh_Texture* TEMPLATE_AcquireSwapchainTexture(
 	Refresh_Renderer *driverData,
 	Refresh_CommandBuffer *commandBuffer,
 	void *windowHandle,
@@ -589,17 +624,33 @@ Refresh_Texture* TEMPLATE_AcquireSwapchainTexture(
 	NOT_IMPLEMENTED
 }
 
-Refresh_TextureFormat TEMPLATE_GetSwapchainFormat(
+static Refresh_TextureFormat TEMPLATE_GetSwapchainFormat(
 	Refresh_Renderer *driverData,
 	void *windowHandle
 ) {
 	NOT_IMPLEMENTED
 }
 
+static void TEMPLATE_SetSwapchainPresentMode(
+	Refresh_Renderer *driverData,
+	void *windowHandle,
+	Refresh_PresentMode presentMode
+) {
+	NOT_IMPLEMENTED
+}
+
+/* Submission and Fences */
+
 static void TEMPLATE_Submit(
 	Refresh_Renderer *driverData,
-	uint32_t commandBufferCount,
-	Refresh_CommandBuffer **pCommandBuffers
+	Refresh_CommandBuffer *commandBuffer
+) {
+	NOT_IMPLEMENTED
+}
+
+static Refresh_Fence* TEMPLATE_SubmitAndAcquireFence(
+	Refresh_Renderer *driverData,
+	Refresh_CommandBuffer *commandBuffer
 ) {
 	NOT_IMPLEMENTED
 }
@@ -610,8 +661,38 @@ static void TEMPLATE_Wait(
 	NOT_IMPLEMENTED
 }
 
+static void TEMPLATE_WaitForFences(
+	Refresh_Renderer *driverData,
+	uint8_t waitAll,
+	uint32_t fenceCount,
+	Refresh_Fence **pFences
+) {
+	NOT_IMPLEMENTED
+}
+
+static int TEMPLATE_QueryFence(
+	Refresh_Renderer *driverData,
+	Refresh_Fence *fence
+) {
+	NOT_IMPLEMENTED
+}
+
+static void TEMPLATE_ReleaseFence(
+	Refresh_Renderer *driverData,
+	Refresh_Fence *fence
+) {
+	NOT_IMPLEMENTED
+}
+
+/* Device Creation */
+
+static uint8_t TEMPLATE_PrepareDriver(
+	uint32_t *flags
+) {
+	NOT_IMPLEMENTED
+}
+
 static Refresh_Device* TEMPLATE_CreateDevice(
-	Refresh_PresentationParameters *presentationParameters,
 	uint8_t debugMode
 ) {
 	NOT_IMPLEMENTED
@@ -619,6 +700,7 @@ static Refresh_Device* TEMPLATE_CreateDevice(
 
 Refresh_Driver TEMPLATEDriver = {
 	"TEMPLATE",
+	TEMPLATE_PrepareDriver,
 	TEMPLATE_CreateDevice
 };
 

src/Refresh_Driver_Vulkan_vkfuncs.h

@@ -89,6 +89,7 @@ VULKAN_DEVICE_FUNCTION(BaseVK, void, vkCmdClearDepthStencilImage, (VkCommandBuff
 VULKAN_DEVICE_FUNCTION(BaseVK, void, vkCmdCopyBuffer, (VkCommandBuffer commandBuffer, VkBuffer srcBuffer, VkBuffer dstBuffer, uint32_t regionCount, const VkBufferCopy* pRegions))
 VULKAN_DEVICE_FUNCTION(BaseVK, void, vkCmdCopyBufferToImage, (VkCommandBuffer commandBuffer, VkBuffer srcBuffer, VkImage dstImage, VkImageLayout dstImageLayout, uint32_t regionCount, const VkBufferImageCopy *pRegions))
 VULKAN_DEVICE_FUNCTION(BaseVK, void, vkCmdCopyImageToBuffer, (VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout, VkBuffer dstBuffer, uint32_t regionCount, const VkBufferImageCopy *pRegions))
+VULKAN_DEVICE_FUNCTION(BaseVK, void, vkCmdCopyImage, (VkCommandBuffer commandBuffer, VkImage srcImage, VkImageLayout srcImageLayout, VkImage dstImage, VkImageLayout dstImageLayout, uint32_t regionCount, const VkImageCopy* pRegions))
 VULKAN_DEVICE_FUNCTION(BaseVK, void, vkCmdDispatch, (VkCommandBuffer commandBuffer, uint32_t groupCountX, uint32_t groupCountY, uint32_t groupCountZ))
 VULKAN_DEVICE_FUNCTION(BaseVK, void, vkCmdDraw, (VkCommandBuffer commandBuffer, uint32_t vertexCount, uint32_t instanceCount, uint32_t firstVertex, uint32_t firstInstance))
 VULKAN_DEVICE_FUNCTION(BaseVK, void, vkCmdDrawIndexed, (VkCommandBuffer commandBuffer, uint32_t indexCount, uint32_t instanceCount, uint32_t firstIndex, int32_t vertexOffset, uint32_t firstInstance))

src/Refresh_Image.c

@@ -76,6 +76,7 @@
 
 /* These are per the Texture2D.FromStream spec */
 #define STBI_ONLY_PNG
+#define STBI_ONLY_QOI
 
 /* These are per the Texture2D.SaveAs* spec */
 #define STBIW_ONLY_PNG
@@ -146,12 +147,6 @@ SDL_SIMDRealloc(void *mem, const size_t len)
 #endif
 #include "stb_image.h"
 
-#define QOI_IMPLEMENTATION
-#define QOI_MALLOC SDL_SIMDAlloc
-#define QOI_FREE SDL_SIMDFree
-#define QOI_ZEROARR SDL_zero
-#include "qoi.h"
-
 #define MINIZ_NO_STDIO
 #define MINIZ_NO_TIME
 #define MINIZ_SDL_MALLOC
@@ -191,110 +186,73 @@ static unsigned char* dgibson_stbi_zlib_compress(
 
 /* Image Read API */
 
-uint8_t* Refresh_Image_LoadPNGFromFile(
-	char const *filename,
-	int32_t *w,
-	int32_t *h,
-	int32_t *numChannels
-) {
-	return stbi_load(filename, w, h, numChannels, STBI_rgb_alpha);
-}
-
-uint8_t* Refresh_Image_LoadPNGFromMemory(
-	uint8_t *buffer,
+uint8_t* Refresh_Image_Load(
+	uint8_t *bufferPtr,
 	int32_t bufferLength,
 	int32_t *w,
 	int32_t *h,
-	int32_t *numChannels
+	int32_t *len
 ) {
-	return stbi_load_from_memory(buffer, bufferLength, w, h, numChannels, STBI_rgb_alpha);
+	uint8_t* result;
+	uint8_t* pixels;
+	int32_t format;
+	int32_t i;
+
+	result = stbi_load_from_memory(
+		bufferPtr,
+		bufferLength,
+		w,
+		h,
+		&format,
+		STBI_rgb_alpha
+	);
+
+	if (result == NULL)
+	{
+		SDL_LogWarn(SDL_LOG_CATEGORY_ERROR, "Image loading failed: %s", stbi_failure_reason());
+	}
+
+	/* Ensure that the alpha pixels are... well, actual alpha.
+	 * You think this looks stupid, but be assured: Your paint program is
+	 * almost certainly even stupider.
+	 * -flibit
+	 */
+	pixels = result;
+	*len = (*w) * (*h) *4;
+	for (i = 0; i < *len; i += 4, pixels += 4)
+	{
+		if (pixels[3] == 0)
+		{
+			pixels[0] = 0;
+			pixels[1] = 1;
+			pixels[2] = 2;
+		}
+	}
+
+	return result;
 }
 
-void Refresh_Image_FreePNG(uint8_t *mem)
+void Refresh_Image_Free(uint8_t *mem)
 {
-	stbi_image_free(mem);
-}
-
-uint8_t *Refresh_Image_LoadQOIFromFile(
-	char const *filename,
-	int32_t *w,
-	int32_t *h,
-	int32_t *numChannels
-) {
-	qoi_desc desc;
-	uint8_t *pixels = qoi_read(filename, &desc, 0);
-	*w = desc.width;
-	*h = desc.height;
-	*numChannels = desc.channels;
-	return pixels;
-}
-
-uint8_t* Refresh_Image_LoadQOIFromMemory(
-	uint8_t *buffer,
-	int32_t bufferLength,
-	int32_t *w,
-	int32_t *h,
-	int32_t *numChannels
-) {
-	qoi_desc desc;
-	uint8_t *pixels = qoi_decode(buffer, bufferLength, &desc, 0);
-	*w = desc.width;
-	*h = desc.height;
-	*numChannels = desc.channels;
-	return pixels;
-}
-
-void Refresh_Image_FreeQOI(uint8_t *mem)
-{
-	QOI_FREE(mem);
+	SDL_SIMDFree(mem);
 }
 
 /* Image Write API */
 
 void Refresh_Image_SavePNG(
-	const char *filename,
+	const char* filename,
+	uint8_t* data,
 	int32_t w,
-	int32_t h,
-	uint8_t bgra,
-	uint8_t *data
+	int32_t h
 ) {
-	uint32_t i;
-	uint8_t *bgraData;
-
-	if (bgra)
-	{
-		bgraData = SDL_malloc(w * h * 4);
-
-		for (i = 0; i < w * h * 4; i += 4)
-		{
-			bgraData[i]     = data[i + 2];
-			bgraData[i + 1] = data[i + 1];
-			bgraData[i + 2] = data[i];
-			bgraData[i + 3] = data[i + 3];
-		}
-
-		stbi_write_png(filename, w, h, 4, bgraData, w * 4);
-
-		SDL_free(bgraData);
-	}
-	else
-	{
-		stbi_write_png(filename, w, h, 4, data, w * 4);
-	}
-}
-
-void Refresh_Image_SaveQOI(
-	char const *filename,
-	int32_t w,
-	int32_t h,
-	uint8_t *data
-) {
-	qoi_desc desc;
-	desc.width = w;
-	desc.height = h;
-	desc.channels = 4;
-	desc.colorspace = QOI_LINEAR;
-	qoi_write(filename, data, &desc);
+	stbi_write_png(
+		filename,
+		w,
+		h,
+		4,
+		data,
+		w * 4
+	);
 }
 
 /* vim: set noexpandtab shiftwidth=8 tabstop=8: */

src/qoi.h

@@ -1,649 +0,0 @@
-/*
-
-Copyright (c) 2021, Dominic Szablewski - https://phoboslab.org
-SPDX-License-Identifier: MIT
-
-
-QOI - The "Quite OK Image" format for fast, lossless image compression
-
--- About
-
-QOI encodes and decodes images in a lossless format. Compared to stb_image and
-stb_image_write QOI offers 20x-50x faster encoding, 3x-4x faster decoding and
-20% better compression.
-
-
--- Synopsis
-
-// Define `QOI_IMPLEMENTATION` in *one* C/C++ file before including this
-// library to create the implementation.
-
-#define QOI_IMPLEMENTATION
-#include "qoi.h"
-
-// Encode and store an RGBA buffer to the file system. The qoi_desc describes
-// the input pixel data.
-qoi_write("image_new.qoi", rgba_pixels, &(qoi_desc){
-	.width = 1920,
-	.height = 1080,
-	.channels = 4,
-	.colorspace = QOI_SRGB
-});
-
-// Load and decode a QOI image from the file system into a 32bbp RGBA buffer.
-// The qoi_desc struct will be filled with the width, height, number of channels
-// and colorspace read from the file header.
-qoi_desc desc;
-void *rgba_pixels = qoi_read("image.qoi", &desc, 4);
-
-
-
--- Documentation
-
-This library provides the following functions;
-- qoi_read    -- read and decode a QOI file
-- qoi_decode  -- decode the raw bytes of a QOI image from memory
-- qoi_write   -- encode and write a QOI file
-- qoi_encode  -- encode an rgba buffer into a QOI image in memory
-
-See the function declaration below for the signature and more information.
-
-If you don't want/need the qoi_read and qoi_write functions, you can define
-QOI_NO_STDIO before including this library.
-
-This library uses malloc() and free(). To supply your own malloc implementation
-you can define QOI_MALLOC and QOI_FREE before including this library.
-
-This library uses memset() to zero-initialize the index. To supply your own
-implementation you can define QOI_ZEROARR before including this library.
-
-
--- Data Format
-
-A QOI file has a 14 byte header, followed by any number of data "chunks" and an
-8-byte end marker.
-
-struct qoi_header_t {
-	char     magic[4];   // magic bytes "qoif"
-	uint32_t width;      // image width in pixels (BE)
-	uint32_t height;     // image height in pixels (BE)
-	uint8_t  channels;   // 3 = RGB, 4 = RGBA
-	uint8_t  colorspace; // 0 = sRGB with linear alpha, 1 = all channels linear
-};
-
-Images are encoded row by row, left to right, top to bottom. The decoder and
-encoder start with {r: 0, g: 0, b: 0, a: 255} as the previous pixel value. An
-image is complete when all pixels specified by width * height have been covered.
-
-Pixels are encoded as
- - a run of the previous pixel
- - an index into an array of previously seen pixels
- - a difference to the previous pixel value in r,g,b
- - full r,g,b or r,g,b,a values
-
-The color channels are assumed to not be premultiplied with the alpha channel
-("un-premultiplied alpha").
-
-A running array[64] (zero-initialized) of previously seen pixel values is
-maintained by the encoder and decoder. Each pixel that is seen by the encoder
-and decoder is put into this array at the position formed by a hash function of
-the color value. In the encoder, if the pixel value at the index matches the
-current pixel, this index position is written to the stream as QOI_OP_INDEX.
-The hash function for the index is:
-
-	index_position = (r * 3 + g * 5 + b * 7 + a * 11) % 64
-
-Each chunk starts with a 2- or 8-bit tag, followed by a number of data bits. The
-bit length of chunks is divisible by 8 - i.e. all chunks are byte aligned. All
-values encoded in these data bits have the most significant bit on the left.
-
-The 8-bit tags have precedence over the 2-bit tags. A decoder must check for the
-presence of an 8-bit tag first.
-
-The byte stream's end is marked with 7 0x00 bytes followed a single 0x01 byte.
-
-
-The possible chunks are:
-
-
-.- QOI_OP_INDEX ----------.
-|         Byte[0]         |
-|  7  6  5  4  3  2  1  0 |
-|-------+-----------------|
-|  0  0 |     index       |
-`-------------------------`
-2-bit tag b00
-6-bit index into the color index array: 0..63
-
-A valid encoder must not issue 2 or more consecutive QOI_OP_INDEX chunks to the
-same index. QOI_OP_RUN should be used instead.
-
-
-.- QOI_OP_DIFF -----------.
-|         Byte[0]         |
-|  7  6  5  4  3  2  1  0 |
-|-------+-----+-----+-----|
-|  0  1 |  dr |  dg |  db |
-`-------------------------`
-2-bit tag b01
-2-bit   red channel difference from the previous pixel between -2..1
-2-bit green channel difference from the previous pixel between -2..1
-2-bit  blue channel difference from the previous pixel between -2..1
-
-The difference to the current channel values are using a wraparound operation,
-so "1 - 2" will result in 255, while "255 + 1" will result in 0.
-
-Values are stored as unsigned integers with a bias of 2. E.g. -2 is stored as
-0 (b00). 1 is stored as 3 (b11).
-
-The alpha value remains unchanged from the previous pixel.
-
-
-.- QOI_OP_LUMA -------------------------------------.
-|         Byte[0]         |         Byte[1]         |
-|  7  6  5  4  3  2  1  0 |  7  6  5  4  3  2  1  0 |
-|-------+-----------------+-------------+-----------|
-|  1  0 |  green diff     |   dr - dg   |  db - dg  |
-`---------------------------------------------------`
-2-bit tag b10
-6-bit green channel difference from the previous pixel -32..31
-4-bit   red channel difference minus green channel difference -8..7
-4-bit  blue channel difference minus green channel difference -8..7
-
-The green channel is used to indicate the general direction of change and is
-encoded in 6 bits. The red and blue channels (dr and db) base their diffs off
-of the green channel difference and are encoded in 4 bits. I.e.:
-	dr_dg = (cur_px.r - prev_px.r) - (cur_px.g - prev_px.g)
-	db_dg = (cur_px.b - prev_px.b) - (cur_px.g - prev_px.g)
-
-The difference to the current channel values are using a wraparound operation,
-so "10 - 13" will result in 253, while "250 + 7" will result in 1.
-
-Values are stored as unsigned integers with a bias of 32 for the green channel
-and a bias of 8 for the red and blue channel.
-
-The alpha value remains unchanged from the previous pixel.
-
-
-.- QOI_OP_RUN ------------.
-|         Byte[0]         |
-|  7  6  5  4  3  2  1  0 |
-|-------+-----------------|
-|  1  1 |       run       |
-`-------------------------`
-2-bit tag b11
-6-bit run-length repeating the previous pixel: 1..62
-
-The run-length is stored with a bias of -1. Note that the run-lengths 63 and 64
-(b111110 and b111111) are illegal as they are occupied by the QOI_OP_RGB and
-QOI_OP_RGBA tags.
-
-
-.- QOI_OP_RGB ------------------------------------------.
-|         Byte[0]         | Byte[1] | Byte[2] | Byte[3] |
-|  7  6  5  4  3  2  1  0 | 7 .. 0  | 7 .. 0  | 7 .. 0  |
-|-------------------------+---------+---------+---------|
-|  1  1  1  1  1  1  1  0 |   red   |  green  |  blue   |
-`-------------------------------------------------------`
-8-bit tag b11111110
-8-bit   red channel value
-8-bit green channel value
-8-bit  blue channel value
-
-The alpha value remains unchanged from the previous pixel.
-
-
-.- QOI_OP_RGBA ---------------------------------------------------.
-|         Byte[0]         | Byte[1] | Byte[2] | Byte[3] | Byte[4] |
-|  7  6  5  4  3  2  1  0 | 7 .. 0  | 7 .. 0  | 7 .. 0  | 7 .. 0  |
-|-------------------------+---------+---------+---------+---------|
-|  1  1  1  1  1  1  1  1 |   red   |  green  |  blue   |  alpha  |
-`-----------------------------------------------------------------`
-8-bit tag b11111111
-8-bit   red channel value
-8-bit green channel value
-8-bit  blue channel value
-8-bit alpha channel value
-
-*/
-
-
-/* -----------------------------------------------------------------------------
-Header - Public functions */
-
-#ifndef QOI_H
-#define QOI_H
-
-#ifdef __cplusplus
-extern "C" {
-#endif
-
-/* A pointer to a qoi_desc struct has to be supplied to all of qoi's functions.
-It describes either the input format (for qoi_write and qoi_encode), or is
-filled with the description read from the file header (for qoi_read and
-qoi_decode).
-
-The colorspace in this qoi_desc is an enum where
-	0 = sRGB, i.e. gamma scaled RGB channels and a linear alpha channel
-	1 = all channels are linear
-You may use the constants QOI_SRGB or QOI_LINEAR. The colorspace is purely
-informative. It will be saved to the file header, but does not affect
-how chunks are en-/decoded. */
-
-#define QOI_SRGB   0
-#define QOI_LINEAR 1
-
-typedef struct {
-	unsigned int width;
-	unsigned int height;
-	unsigned char channels;
-	unsigned char colorspace;
-} qoi_desc;
-
-#ifndef QOI_NO_STDIO
-
-/* Encode raw RGB or RGBA pixels into a QOI image and write it to the file
-system. The qoi_desc struct must be filled with the image width, height,
-number of channels (3 = RGB, 4 = RGBA) and the colorspace.
-
-The function returns 0 on failure (invalid parameters, or fopen or malloc
-failed) or the number of bytes written on success. */
-
-int qoi_write(const char *filename, const void *data, const qoi_desc *desc);
-
-
-/* Read and decode a QOI image from the file system. If channels is 0, the
-number of channels from the file header is used. If channels is 3 or 4 the
-output format will be forced into this number of channels.
-
-The function either returns NULL on failure (invalid data, or malloc or fopen
-failed) or a pointer to the decoded pixels. On success, the qoi_desc struct
-will be filled with the description from the file header.
-
-The returned pixel data should be free()d after use. */
-
-void *qoi_read(const char *filename, qoi_desc *desc, int channels);
-
-#endif /* QOI_NO_STDIO */
-
-
-/* Encode raw RGB or RGBA pixels into a QOI image in memory.
-
-The function either returns NULL on failure (invalid parameters or malloc
-failed) or a pointer to the encoded data on success. On success the out_len
-is set to the size in bytes of the encoded data.
-
-The returned qoi data should be free()d after use. */
-
-void *qoi_encode(const void *data, const qoi_desc *desc, int *out_len);
-
-
-/* Decode a QOI image from memory.
-
-The function either returns NULL on failure (invalid parameters or malloc
-failed) or a pointer to the decoded pixels. On success, the qoi_desc struct
-is filled with the description from the file header.
-
-The returned pixel data should be free()d after use. */
-
-void *qoi_decode(const void *data, int size, qoi_desc *desc, int channels);
-
-
-#ifdef __cplusplus
-}
-#endif
-#endif /* QOI_H */
-
-
-/* -----------------------------------------------------------------------------
-Implementation */
-
-#ifdef QOI_IMPLEMENTATION
-#include <stdlib.h>
-#include <string.h>
-
-#ifndef QOI_MALLOC
-	#define QOI_MALLOC(sz) malloc(sz)
-	#define QOI_FREE(p)    free(p)
-#endif
-#ifndef QOI_ZEROARR
-	#define QOI_ZEROARR(a) memset((a),0,sizeof(a))
-#endif
-
-#define QOI_OP_INDEX  0x00 /* 00xxxxxx */
-#define QOI_OP_DIFF   0x40 /* 01xxxxxx */
-#define QOI_OP_LUMA   0x80 /* 10xxxxxx */
-#define QOI_OP_RUN    0xc0 /* 11xxxxxx */
-#define QOI_OP_RGB    0xfe /* 11111110 */
-#define QOI_OP_RGBA   0xff /* 11111111 */
-
-#define QOI_MASK_2    0xc0 /* 11000000 */
-
-#define QOI_COLOR_HASH(C) (C.rgba.r*3 + C.rgba.g*5 + C.rgba.b*7 + C.rgba.a*11)
-#define QOI_MAGIC \
-	(((unsigned int)'q') << 24 | ((unsigned int)'o') << 16 | \
-	 ((unsigned int)'i') <<  8 | ((unsigned int)'f'))
-#define QOI_HEADER_SIZE 14
-
-/* 2GB is the max file size that this implementation can safely handle. We guard
-against anything larger than that, assuming the worst case with 5 bytes per
-pixel, rounded down to a nice clean value. 400 million pixels ought to be
-enough for anybody. */
-#define QOI_PIXELS_MAX ((unsigned int)400000000)
-
-typedef union {
-	struct { unsigned char r, g, b, a; } rgba;
-	unsigned int v;
-} qoi_rgba_t;
-
-static const unsigned char qoi_padding[8] = {0,0,0,0,0,0,0,1};
-
-static void qoi_write_32(unsigned char *bytes, int *p, unsigned int v) {
-	bytes[(*p)++] = (0xff000000 & v) >> 24;
-	bytes[(*p)++] = (0x00ff0000 & v) >> 16;
-	bytes[(*p)++] = (0x0000ff00 & v) >> 8;
-	bytes[(*p)++] = (0x000000ff & v);
-}
-
-static unsigned int qoi_read_32(const unsigned char *bytes, int *p) {
-	unsigned int a = bytes[(*p)++];
-	unsigned int b = bytes[(*p)++];
-	unsigned int c = bytes[(*p)++];
-	unsigned int d = bytes[(*p)++];
-	return a << 24 | b << 16 | c << 8 | d;
-}
-
-void *qoi_encode(const void *data, const qoi_desc *desc, int *out_len) {
-	int i, max_size, p, run;
-	int px_len, px_end, px_pos, channels;
-	unsigned char *bytes;
-	const unsigned char *pixels;
-	qoi_rgba_t index[64];
-	qoi_rgba_t px, px_prev;
-
-	if (
-		data == NULL || out_len == NULL || desc == NULL ||
-		desc->width == 0 || desc->height == 0 ||
-		desc->channels < 3 || desc->channels > 4 ||
-		desc->colorspace > 1 ||
-		desc->height >= QOI_PIXELS_MAX / desc->width
-	) {
-		return NULL;
-	}
-
-	max_size =
-		desc->width * desc->height * (desc->channels + 1) +
-		QOI_HEADER_SIZE + sizeof(qoi_padding);
-
-	p = 0;
-	bytes = (unsigned char *) QOI_MALLOC(max_size);
-	if (!bytes) {
-		return NULL;
-	}
-
-	qoi_write_32(bytes, &p, QOI_MAGIC);
-	qoi_write_32(bytes, &p, desc->width);
-	qoi_write_32(bytes, &p, desc->height);
-	bytes[p++] = desc->channels;
-	bytes[p++] = desc->colorspace;
-
-
-	pixels = (const unsigned char *)data;
-
-	QOI_ZEROARR(index);
-
-	run = 0;
-	px_prev.rgba.r = 0;
-	px_prev.rgba.g = 0;
-	px_prev.rgba.b = 0;
-	px_prev.rgba.a = 255;
-	px = px_prev;
-
-	px_len = desc->width * desc->height * desc->channels;
-	px_end = px_len - desc->channels;
-	channels = desc->channels;
-
-	for (px_pos = 0; px_pos < px_len; px_pos += channels) {
-		px.rgba.r = pixels[px_pos + 0];
-		px.rgba.g = pixels[px_pos + 1];
-		px.rgba.b = pixels[px_pos + 2];
-
-		if (channels == 4) {
-			px.rgba.a = pixels[px_pos + 3];
-		}
-
-		if (px.v == px_prev.v) {
-			run++;
-			if (run == 62 || px_pos == px_end) {
-				bytes[p++] = QOI_OP_RUN | (run - 1);
-				run = 0;
-			}
-		}
-		else {
-			int index_pos;
-
-			if (run > 0) {
-				bytes[p++] = QOI_OP_RUN | (run - 1);
-				run = 0;
-			}
-
-			index_pos = QOI_COLOR_HASH(px) % 64;
-
-			if (index[index_pos].v == px.v) {
-				bytes[p++] = QOI_OP_INDEX | index_pos;
-			}
-			else {
-				index[index_pos] = px;
-
-				if (px.rgba.a == px_prev.rgba.a) {
-					signed char vr = px.rgba.r - px_prev.rgba.r;
-					signed char vg = px.rgba.g - px_prev.rgba.g;
-					signed char vb = px.rgba.b - px_prev.rgba.b;
-
-					signed char vg_r = vr - vg;
-					signed char vg_b = vb - vg;
-
-					if (
-						vr > -3 && vr < 2 &&
-						vg > -3 && vg < 2 &&
-						vb > -3 && vb < 2
-					) {
-						bytes[p++] = QOI_OP_DIFF | (vr + 2) << 4 | (vg + 2) << 2 | (vb + 2);
-					}
-					else if (
-						vg_r >  -9 && vg_r <  8 &&
-						vg   > -33 && vg   < 32 &&
-						vg_b >  -9 && vg_b <  8
-					) {
-						bytes[p++] = QOI_OP_LUMA     | (vg   + 32);
-						bytes[p++] = (vg_r + 8) << 4 | (vg_b +  8);
-					}
-					else {
-						bytes[p++] = QOI_OP_RGB;
-						bytes[p++] = px.rgba.r;
-						bytes[p++] = px.rgba.g;
-						bytes[p++] = px.rgba.b;
-					}
-				}
-				else {
-					bytes[p++] = QOI_OP_RGBA;
-					bytes[p++] = px.rgba.r;
-					bytes[p++] = px.rgba.g;
-					bytes[p++] = px.rgba.b;
-					bytes[p++] = px.rgba.a;
-				}
-			}
-		}
-		px_prev = px;
-	}
-
-	for (i = 0; i < (int)sizeof(qoi_padding); i++) {
-		bytes[p++] = qoi_padding[i];
-	}
-
-	*out_len = p;
-	return bytes;
-}
-
-void *qoi_decode(const void *data, int size, qoi_desc *desc, int channels) {
-	const unsigned char *bytes;
-	unsigned int header_magic;
-	unsigned char *pixels;
-	qoi_rgba_t index[64];
-	qoi_rgba_t px;
-	int px_len, chunks_len, px_pos;
-	int p = 0, run = 0;
-
-	if (
-		data == NULL || desc == NULL ||
-		(channels != 0 && channels != 3 && channels != 4) ||
-		size < QOI_HEADER_SIZE + (int)sizeof(qoi_padding)
-	) {
-		return NULL;
-	}
-
-	bytes = (const unsigned char *)data;
-
-	header_magic = qoi_read_32(bytes, &p);
-	desc->width = qoi_read_32(bytes, &p);
-	desc->height = qoi_read_32(bytes, &p);
-	desc->channels = bytes[p++];
-	desc->colorspace = bytes[p++];
-
-	if (
-		desc->width == 0 || desc->height == 0 ||
-		desc->channels < 3 || desc->channels > 4 ||
-		desc->colorspace > 1 ||
-		header_magic != QOI_MAGIC ||
-		desc->height >= QOI_PIXELS_MAX / desc->width
-	) {
-		return NULL;
-	}
-
-	if (channels == 0) {
-		channels = desc->channels;
-	}
-
-	px_len = desc->width * desc->height * channels;
-	pixels = (unsigned char *) QOI_MALLOC(px_len);
-	if (!pixels) {
-		return NULL;
-	}
-
-	QOI_ZEROARR(index);
-	px.rgba.r = 0;
-	px.rgba.g = 0;
-	px.rgba.b = 0;
-	px.rgba.a = 255;
-
-	chunks_len = size - (int)sizeof(qoi_padding);
-	for (px_pos = 0; px_pos < px_len; px_pos += channels) {
-		if (run > 0) {
-			run--;
-		}
-		else if (p < chunks_len) {
-			int b1 = bytes[p++];
-
-			if (b1 == QOI_OP_RGB) {
-				px.rgba.r = bytes[p++];
-				px.rgba.g = bytes[p++];
-				px.rgba.b = bytes[p++];
-			}
-			else if (b1 == QOI_OP_RGBA) {
-				px.rgba.r = bytes[p++];
-				px.rgba.g = bytes[p++];
-				px.rgba.b = bytes[p++];
-				px.rgba.a = bytes[p++];
-			}
-			else if ((b1 & QOI_MASK_2) == QOI_OP_INDEX) {
-				px = index[b1];
-			}
-			else if ((b1 & QOI_MASK_2) == QOI_OP_DIFF) {
-				px.rgba.r += ((b1 >> 4) & 0x03) - 2;
-				px.rgba.g += ((b1 >> 2) & 0x03) - 2;
-				px.rgba.b += ( b1       & 0x03) - 2;
-			}
-			else if ((b1 & QOI_MASK_2) == QOI_OP_LUMA) {
-				int b2 = bytes[p++];
-				int vg = (b1 & 0x3f) - 32;
-				px.rgba.r += vg - 8 + ((b2 >> 4) & 0x0f);
-				px.rgba.g += vg;
-				px.rgba.b += vg - 8 +  (b2       & 0x0f);
-			}
-			else if ((b1 & QOI_MASK_2) == QOI_OP_RUN) {
-				run = (b1 & 0x3f);
-			}
-
-			index[QOI_COLOR_HASH(px) % 64] = px;
-		}
-
-		pixels[px_pos + 0] = px.rgba.r;
-		pixels[px_pos + 1] = px.rgba.g;
-		pixels[px_pos + 2] = px.rgba.b;
-
-		if (channels == 4) {
-			pixels[px_pos + 3] = px.rgba.a;
-		}
-	}
-
-	return pixels;
-}
-
-#ifndef QOI_NO_STDIO
-#include <stdio.h>
-
-int qoi_write(const char *filename, const void *data, const qoi_desc *desc) {
-	FILE *f = fopen(filename, "wb");
-	int size;
-	void *encoded;
-
-	if (!f) {
-		return 0;
-	}
-
-	encoded = qoi_encode(data, desc, &size);
-	if (!encoded) {
-		fclose(f);
-		return 0;
-	}
-
-	fwrite(encoded, 1, size, f);
-	fclose(f);
-
-	QOI_FREE(encoded);
-	return size;
-}
-
-void *qoi_read(const char *filename, qoi_desc *desc, int channels) {
-	FILE *f = fopen(filename, "rb");
-	int size, bytes_read;
-	void *pixels, *data;
-
-	if (!f) {
-		return NULL;
-	}
-
-	fseek(f, 0, SEEK_END);
-	size = ftell(f);
-	if (size <= 0) {
-		fclose(f);
-		return NULL;
-	}
-	fseek(f, 0, SEEK_SET);
-
-	data = QOI_MALLOC(size);
-	if (!data) {
-		fclose(f);
-		return NULL;
-	}
-
-	bytes_read = fread(data, 1, size, f);
-	fclose(f);
-
-	pixels = qoi_decode(data, bytes_read, desc, channels);
-	QOI_FREE(data);
-	return pixels;
-}
-
-#endif /* QOI_NO_STDIO */
-#endif /* QOI_IMPLEMENTATION */