KR102254119B1 - Method and apparatus for processing graphics command - Google Patents

Abstract

A method of processing a graphics command may receive the graphics command from a device executing an application. At least one shader included in the graphics instruction to be processed by the graphics processor may be selected, and a shader program may be generated using the selected at least one shader. Among at least one shader program compiled previously, a shader program identical to the generated shader program can be searched and the searched shader program can be output.

Description

METHOD AND APPARATUS FOR PROCESSING GRAPHICS COMMAND}

It relates to a method and apparatus for processing graphics commands.

The Graphics Processing Unit (GPU) processes graphics APIs such as OpenGL ES and OpenCL (CUDA) to realize gorgeous graphics effects on mobile devices and TVs.

In order for a program to be processed by the GPU, the program must be written according to the rules set by the standard API. The program processed on the GPU consists of a vertex shader and a pixel shader, which are essential for image generation. Vertex shaders and pixel shaders are compiled and linked to be used in programs.

As colorful graphics effects are implemented in applications used in portable devices and TVs, the program size of applications processed by GPUs is increasing. As the size of the program processed in the GPU increases, it takes a lot of time to compile and link the vertex shaders and pixel shaders that make up the program. In particular, in a program used in a game, multiple shaders are repeatedly used to achieve the same effect. .

It is to provide a method and apparatus for processing graphics commands. Further, it is to provide a computer-readable recording medium in which a program for executing the method on a computer is recorded. The technical problem to be achieved by the present embodiment is not limited to the technical problems as described above, and other technical problems may be inferred from the following embodiments.

According to an aspect, a method of processing a graphics command, the method comprising: receiving the graphics command from a device executing an application; Selecting at least one shader included in the graphics instruction to be processed by a graphics processor; Generating a shader program using the selected at least one shader; Searching for a shader program identical to the generated shader program from among at least one shader program compiled previously; And outputting the searched shader program.

According to another aspect, in a graphics command processing apparatus, a graphics command is received from a device executing an application, and at least one shader to be processed by a graphics processor included in the graphics command is selected, and the selected at least one shader A processor for generating a shader program by using, searching for a shader program identical to the generated shader program among at least one shader program compiled previously, and outputting the found shader program; And a cache for storing previously compiled shader programs.

According to another aspect, in a GPU, a graphics command is received from a device executing an application, a graphics processor included in the graphics command selects at least one shader to be processed, and a shader program using the selected shader A graphics driver for generating and searching for a shader program identical to the generated shader program among at least one shader program compiled previously, and outputting the found shader program; And a cache for storing previously compiled shader programs.

According to another aspect, a computer-readable non-transitory recording medium storing a program for execution on a computer records the method of claim 1 as a program for execution on a computer.

As described above, the operation can be omitted by delaying the graphics command. In addition, when regenerating the shader program, the amount of computation can be reduced by reusing the shader program binaries previously stored and omitting the process of compiling the shader program.

1 is a flowchart of a method of processing graphics commands according to an exemplary embodiment.
FIG. 2 is a diagram illustrating an example in which graphics commands are processed in a graphics command processing apparatus located outside a GPU according to an exemplary embodiment.
3 is a diagram illustrating an example in which graphics commands are processed in a graphics command processing apparatus located inside a GPU according to an exemplary embodiment.
4 is a block diagram of a shader program manager according to an embodiment.
5 is a diagram illustrating an example of searching for a shader program in a work processor according to an embodiment.
6 is a diagram illustrating an example in which a graphics command processing apparatus according to an exemplary embodiment searches for and outputs a shader program.
7 is a diagram illustrating a detailed example of a process of processing data in a graphics command processing apparatus according to an exemplary embodiment.
8 is a detailed flowchart of a method of processing a graphics command according to an exemplary embodiment.
9 is a flowchart illustrating a method of processing an API related to a shader program by a graphics driver according to an embodiment.
10 is a flowchart of a method for a shader program manager to process an API related to a shader program according to an exemplary embodiment.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement the present invention. However, the present invention may be implemented in various different forms and is not limited to the embodiments described herein. In the drawings, parts not related to the description are omitted in order to clearly describe the present invention, and redundant descriptions are omitted by denoting similar reference numerals for similar parts throughout the specification.

Throughout the specification, when a part is said to be "connected" with another part, this includes not only "directly connected" but also "electrically connected" with another element interposed therebetween. . In addition, when a part "includes" a certain component, it means that other components may be further included rather than excluding other components unless specifically stated to the contrary.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a flowchart of a method of processing graphics commands according to an exemplary embodiment. In step 101, when an application requests to process a graphics command, the graphics command processing device receives the graphics command from the device executing the application. Graphics instructions may include instructions that can be processed by the GPU. For example, APIs related to shader programs may be included.

In step 102, the graphics command processing apparatus may select at least one shader included in the graphics command received in step 101 to be processed by the graphics processor. Shaders to be processed in the graphics processor may include vertex shaders and pixel shaders. However, it is not limited thereto.

In step 103, the graphics command processing apparatus may generate a shader program using the at least one shader selected in step 102. Shader programs are units or types of programs that can be processed by the GPU. The shader program may include the shader selected in step 102 and uniform data received by the graphics command processing apparatus. However, it is not limited thereto.

In step 104, the graphics command processing apparatus may search for a shader program identical to the shader program generated in step 103 from among at least one shader program compiled previously. The graphics instruction processing unit uses the vertex shader and pixel shader source code (or compilation code), uniform data, compiler version, and GPU version information included in the shader program to be the same shader as the shader program generated in step 103. You can search for a program. The graphics command processing apparatus may search for a shader program identical to the shader program generated in step 103 by using the text and the length of the text of the shader source code and the text and the length of the text of the uniform data. The graphics command processing apparatus may generate a key using text and text length of a shader source code, text and text length of uniform data, a compiler version, and GPU version information. The graphics command processing device may search for the same key as the generated key. The key is unique data to identify the shader program. In other words, different keys are generated for different shader programs.

In step 105, the graphics command processing apparatus may output the shader program retrieved in step 104. The graphics command processing apparatus may output the searched shader program to a graphics driver or a graphics processor. The shader program manager may output the shader program searched in step 104 as binary.

The steps of FIG. 1 are stored in a memory in the form of a program or instruction, and may be implemented by a main processor, a GPU, a shader program manager, a graphics driver, or the like. However, it is not limited thereto.

FIG. 2 is a diagram illustrating an example in which graphics commands are processed in a graphics command processing apparatus located outside a GPU according to an exemplary embodiment.

Referring to FIG. 2, the graphics command processing apparatus may be implemented by the main processor 10 or the shader program manager 100. The shader program manager 100 may be located outside the GPU 20.

The GPU 20 may include a graphics driver 200, a graphics processor 300, and a compiler 500.

The main processor 10 may include a shader program manager 100 and a cache 400.

At least one previously compiled shader program may be stored in the cache 400. The shader program may be stored in the cache 400 as a binary. A key generated by using information included in the shader program 1000a may be stored in the cache 400. The key may be generated using text and text length of the shader source code, text and text length of uniform data, compiler version, and GPU version information. However, it is not limited thereto.

The shader program manager 100 may receive graphics commands. The shader program manager 100 may select at least one shader included in the received graphics command to be processed by the graphics processor. The shader program manager 100 may generate a shader program using at least one selected shader. The shader program manager 100 may search for a shader program identical to the generated shader program from among at least one shader program compiled previously, and may output the searched shader program.

The shader program manager 100 may use the key in the step of searching for the same shader program as the generated shader program 1000a. In an embodiment, the shader program manager 100 may generate a key by using information included in the generated shader program 1000a. Also, the key generated by the shader program manager 100 may be stored in the cache 400. When the key is stored in the cache, the shader program manager 100 compares the key of the generated shader program 1000a with the key of the shader program stored in the cache 400, thereby creating the same shader program as the generated shader program 1000a. It can be searched in the cache 400.

When the shader program manager 100 searches for the same shader program as the generated shader program 1000a, the shader program manager 100 may output the searched shader program. When the same shader program is searched, the shader program manager 100 may output the searched shader program to the graphics driver 200. The shader program manager 100 may output the searched shader program as a binary. The binary is data in a form that can be recognized by the graphics processor 300.

When the shader program manager 100 fails to search for the same shader program as the generated shader program 1000a, the shader program manager 100 may output the generated shader program 1000a to the graphics driver 200. When the generated shader program 1000a is output to the graphics driver 200, the graphics driver 200 may generate a binary by compiling the shader program 1000a received using the compiler 500. The graphics driver 200 may output the generated binary to the graphics processor 300. The graphics driver 200 may store the generated binary in the cache 400. The shader program manager 100 may store the key of the shader program 1000a used for searching in the cache 400.

3 is a diagram illustrating an example in which graphics commands are processed in a graphics command processing apparatus located inside a GPU according to an exemplary embodiment.

Referring to FIG. 3, the graphics command processing apparatus may be implemented by the graphics driver 200 or the shader program manager 100. However, it is not limited thereto. The shader program manager 100 may be located inside the GPU 20.

The GPU 20 may include a graphics driver 200, a graphics processor 300, a cache 400, and a compiler 500.

At least one previously compiled shader program may be stored in the cache 400. The shader program may be stored in the cache 400 as a binary. A key generated by using information included in the shader program 1000a may be stored in the cache 400. The key may be generated using text and text length of the shader source code, text and text length of uniform data, compiler version, and GPU version information. However, it is not limited thereto.

The shader program manager 100 may receive graphics commands. The shader program manager 100 may select at least one shader included in the received graphics command to be processed by the graphics processor. The shader program manager 100 may generate a shader program using at least one selected shader. The shader program manager 100 may search for a shader program identical to the generated shader program from among at least one shader program compiled previously, and may output the searched shader program.

The shader program manager 100 may use the key in the step of searching for the same shader program as the generated shader program 1000a. In an embodiment, the shader program manager 100 may generate a key by using information included in the generated shader program 1000a. Also, the key generated by the shader program manager 100 may be stored in the cache 400. When the key is stored in the cache, the shader program manager 100 compares the key of the generated shader program 1000a with the key of the shader program stored in the cache 400, thereby creating the same shader program as the generated shader program 1000a. It can be searched in the cache 400.

When the shader program manager 100 searches for the same shader program as the generated shader program 1000a, the shader program manager 100 may output the searched shader program. The shader program manager 100 may output the searched shader program to the graphics processor 300. The shader program manager 100 may output the searched shader program as a binary.

When the shader program manager 100 fails to search for the same shader program as the generated shader program 1000a, the shader program manager 100 may output the generated shader program 1000a to the compiler 500. The compiler 500 may generate a binary by compiling the input shader program 1000a. The graphics driver 200 may output the generated binary to the graphics processor 300. The generated binary may be stored in the cache 400. When the shader program manager 100 fails to search for the same shader program as the generated shader program 1000a, the key of the generated shader program 1000a may be stored in the cache 400. However, it is not limited thereto.

4 is a block diagram of a shader program manager according to an embodiment.

Referring to FIG. 4, the shader program manager 100 according to an embodiment may include a shader memory 110 and a work manager 120.

Graphics commands received by the shader program manager 100 may be stored in the shader memory 110. At least one shader included in the graphics instruction may be stored in the shader memory 110. Shaders stored in the shader memory 110 may include vertex shaders and pixel shaders.

In one embodiment, the shader memory 110 may store a vertex shader and a pixel shader including a source API. The source API may include source code. The shader memory 110 may store source codes of a vertex shader including a source API and a pixel shader. The source code of the vertex shader and the pixel shader may be stored in a text format. Source codes of each of the vertex shader and the pixel shader may be independently stored in the TEXT_source space of the shader memory 110. However, it is not limited thereto.

In addition, the shader memory 110 may store a vertex shader and a pixel shader including a source API and a compilation API. The compilation API may include compiled code. Compilation codes of vertex shaders and pixel shaders may be stored in the shader memory 110. Compilation codes of vertex shaders and pixel shaders can be stored in text format. Compilation codes for each of the vertex shader and the pixel shader may be independently stored in the TEXT_compile space of the shader memory 110.

Or, when one shader is received, the compilation API can be processed later than the source API. The vertex shader and the pixel shader source code independently stored in the TEXT_source space of the shader memory 110 are copied, and the copied source code may be independently stored in the TEXT_compile space of the shader memory 110. . However, it is not limited thereto.

When the shader program manager 100 sequentially receives at least one vertex shader and pixel shader including a source API, the shader program manager 100 stores the received vertex shader and pixel shader in the shader memory 110. Can be saved.

In one embodiment, regardless of whether the source code is stored in the TEXT_source space in the shader memory 110, the shader program manager 100 stores the source codes of each of the vertex shaders and pixel shaders received later in the shader memory ( 110) can be stored independently in the TEXT_source space. Finally, in the TEXT_source space in the shader memory 110, the vertex shader and the source code of the pixel shader that the shader program manager 100 received last among the vertex shaders and pixel shaders including the source API may be independently stored. .

When the shader program manager 100 sequentially receives at least one vertex shader and pixel shader including a source API and a compile API, the shader program manager 100 stores the received vertex shader and pixel shader in a shader memory ( 110).

In one embodiment, regardless of whether the source code (or compiled code) is stored in the TEXT_compile space in the shader memory 110, the shader program manager 100 is Code (or compiled code) may be independently stored in the TEXT_compile space in the shader memory 110. When the shader program manager 100 independently stores the source code in the TEXT_compile space in the shader memory 110, the vertex shader and pixel shader sources independently stored in the TEXT_source space of the shader memory 110 Since the code is copied, the copied source code may be independently stored in the TEXT_compile space of the shader memory 110.

In one embodiment, the operation may be omitted by delaying the graphics command. In another embodiment, a shader generation operation that is not to be used in a shader program may be omitted by delaying a received shader for selection of a shader program. In another embodiment, shaders to be delayed may be vertex shaders and pixel shaders.

The work manager 120 may search for the same shader program as the generated shader program 1000a.

The work manager 120 may include a work memory 121 and a work processor 122.

The work memory 121 may store information used to search for a shader program.

The work memory 121 may include vertex shader, pixel shader, and uniform data information. In one embodiment, the work memory 121 may include text and text length of vertex shader and pixel shader source code (or compilation code), text and text length of uniform data, compiler version and GPU version information.

The work processor 122 may search for a shader program. The work processor 122 will be described with reference to FIG. 5 below.

5 is a diagram illustrating an example of searching for a shader program in a work processor according to an embodiment.

The work processor 122 may use information stored in the work memory 121 to search for a shader program identical to the generated shader program 1000a from among at least one compiled shader program.

The work processor 122 may perform a search using a key. When searching using a key, the work processor 122 may include a key generator 123 and a key comparison unit 124.

The key generator 123 may generate a key using information stored in the work memory 121 and used for search. For example, the text and text length of the vertex shader and pixel shader source code (or compilation code) stored in the work memory 121, the text and the length of the text of the uniform data, the compiler version, and the GPU version information are used to determine the key. Can be generated.

The key comparison unit 124 may compare a key generated by the key generation unit 123 with a key of at least one shader program compiled previously. The key of at least one shader program compiled previously may be stored in the cache 400. When the key is stored in the cache 400, the key comparison unit 124 may compare the key generated by the key generation unit 123 with the key of at least one shader program compiled previously stored in the cache 400. have.

6 is a diagram illustrating an example in which a graphics command processing apparatus according to an exemplary embodiment searches for and outputs a shader program.

Referring to FIG. 6, the graphics command processing apparatus according to an embodiment may be implemented by the shader program manager 100. However, it is not limited thereto.

When a compiled shader program identical to the shader program 1000a generated by the work processor 122 is searched, the shader program manager 100 may output the searched shader program to the graphics driver 200. Alternatively, when the graphics command processing apparatus is implemented by the graphics driver 200, the graphics driver 200 may output the searched shader program to the graphics processor 300. The searched shader program can be output in binary.

When the work processor 122 fails to search for the same compiled shader program as the generated shader program 1000a, the shader program manager 100 may output the generated shader program 1000a to the compiler 500, The generated shader program 1000a may be compiled by the compiler 500. Alternatively, when the graphics command processing device is implemented by the graphics driver 200, the graphics driver 200 may output the generated shader program 1000a to the compiler 500, and the generated shader program 1000a It can be compiled in the compiler 500. The compiled shader program 1000b may be used when the work processor 122 searches for the same shader program as the newly generated shader program 1000a. When the work processor 122 fails to search for the same compiled shader program as the generated shader program 1000a, the compiled shader program 1000b may be stored in the cache 400. The compiled shader program 1000b may be stored in the cache 400 as a binary.

When the work processor 122 searches for the same compiled shader program as the generated shader program 1000a, the shader program may be searched using a key.

When the work processor 122 fails to search for a compiled shader program having the same key as the key of the generated shader program 1000a, the key of the generated shader program 1000a may be stored in the cache 400. After the generated shader program 1000a is compiled, the compiled shader program 1000b may be stored in the cache 400 as a binary. After the binary is stored in the cache 400, the key of the shader program 1000a stored in the cache 400 may be linked with the binary.

7 is a diagram illustrating a detailed example of a process of processing data in a graphics command processing apparatus according to an exemplary embodiment. But now it is not limited.

8 is a detailed flowchart of a method of processing a graphics command according to an exemplary embodiment.

The steps below in FIG. 8 may be implemented by a main processor, a GPU, a shader program manager, a graphics driver, or the like. However, it is not limited thereto.

Hereinafter, a case where each step is executed by the shader program manager 100 will be described.

The shader program manager 100 may determine whether a graphics command has been received. When it is determined that the graphics command has been received, the shader program manager 100 may start the following operation. In operation 810, the shader program manager 100 may determine whether the received graphics command is an API related to a shader program.

As a result of the determination in step 810, if it is determined that an API other than a shader program related API has been received, an API other than the shader program related API may not be processed by the shader program manager 100 and may be directly output to the graphics driver 200. The existing graphics driver 200 operation may be performed on an API other than an API related to a shader program output to the graphics driver 200.

Alternatively, as a result of the determination in step 810, if it is determined that the shader program related API has been received, step 820 may be performed.

As a result of the determination in step 820, if it is determined that the shader program manager 100 has received the shader program creation or deletion API, in step 821, the shader program manager 100 may allocate or delete space in memory related to the shader and the shader program. . In operation 821, the shader program manager 100 may allocate or delete spaces in the shader memory 110 and the work memory 121.

Alternatively, if it is determined in step 820 that the shader program generation or shader program deletion API is not received, step 830 may be performed.

As a result of the determination in step 830, if it is determined that the shader program manager 100 has received the shader source API, the shader program manager 100 may store the shader including the shader source API in step 831. Shaders may be stored in the shader memory 110. The shader source code may be stored in the shader memory 110. The shader source code may be stored in the TEXT_source space of the shader memory 110. Source codes of each of the vertex shader and the pixel shader may be independently stored in the TEXT_source space of the shader memory 110. However, it is not limited thereto.

Alternatively, as a result of the determination in step 830, if it is determined that the shader program manager 100 has not received the shader source API, step 840 may be performed.

As a result of the determination in step 840, if it is determined that the shader program manager 100 has received the shader compilation API, the shader program manager 100 may store the shader in step 841. Shaders may be stored in the shader memory 110. The shader source code (or compiled code) may be stored in the shader memory 110. The shader source code (or compilation code) may be stored in the TEXT_compile space of the shader memory 110. When the shader source code is stored, the source code stored in the TEXT_source space is copied, and the copied source code may be stored in the TEXT_compile space of the shader memory 110. Source codes of each of the vertex shader and the pixel shader may be independently stored in the TEXT_compile space of the shader memory 110.

In addition, in step 841, if the source code is stored in the TEXT_source space in the shader memory 110 or the source code (or compiled code) is stored in the TEXT_compile space, the TEXT_source space or the TEXT_compile space Source code (or compilation code) of the shader received later by the shader program manager 100 may be stored therein.

Alternatively, as a result of the determination in step 840, if it is determined that the shader program manager 100 has not received the shader compilation API, step 850 may be performed.

As a result of the determination in step 850, if it is determined that the shader program manager 100 has received the shader attach/detach API, in step 851, the shader program manager 100 transfers the shader stored in the shader memory 110 to the work manager 120. You can connect (or save). When the shader is connected (or stored) to the work manager 120, the source code (or compiled code) stored in the TEXT_compile space may be connected (or stored) to the work manager 120. When the shader is stored in the work manager 120, the shader program manager 100 may store the source code (or compiled code) in the work memory 121. The text of the source code (or the compilation code) and the length of the text may be stored in the work memory 121. Alternatively, if it is determined that the shader program manager 100 has not received the shader attach/detach API as a result of the determination in step 850, step 860 may be performed.

As a result of the determination in step 860, if it is determined that the shader program manager 100 has received the uniform data storage API, the shader program manager 100 may connect (or store) the uniform data to the work manager 120 in step 861. have. When the shader is stored in the work manager 120, the shader program manager 100 may store uniform data in the work memory 121. The text of the uniform data and the length of the text may be stored in the work memory 121. However, it is not limited thereto.

Alternatively, as a result of the determination in step 860, if it is determined that the shader program manager 100 has not received the uniform data storage API, step 870 may be performed.

As a result of the determination in step 870, if it is determined that the shader program manager 100 has received the shader program link API, in step 871, the shader program manager 100 uses information connected (or stored) to the work manager 120 You can generate keys. When a key is generated, the shader program manager 100 includes text and text length of at least one shader source code connected (or stored) to the work manager 120, text and text length of uniform data, compiler version, And it is possible to generate a key using the GPU version information.

Alternatively, as a result of the determination in step 870, if it is determined that the shader program manager 100 has not received the shader program link API, the existing graphics driver 200 may perform an operation on the graphics command for the received API.

As a result of the determination in step 872, if it is determined that a shader program having the same key as the key generated in step 871 is found, the shader program manager 100 may output the found shader program in step 873. An existing graphics driver 200 operation may be performed on the output shader program. The shader program's key can be stored in a cache. The searched shader program can be output in binary.

Alternatively, as a result of the determination in step 872, if it is determined that a shader program having the same key as the key generated in step 871 is not found, the shader program manager 100 is connected (or stored) to the work manager 120 in step 874. Using the information, information necessary for compiling a shader program may be generated and a compilation request may be made to the graphics driver 200.

9 is a flowchart illustrating a method of processing an API related to a shader program by a graphics driver according to an exemplary embodiment. 10 is a flowchart of a method for a shader program manager to process an API related to a shader program according to an exemplary embodiment.

Hereinafter, compared with FIG. 9, FIG. 10 will be described.

Referring to FIG. 10, when the shader program manager 100 receives at least one vertex shader and at least one pixel shader, the vertex shader and the pixel shader are processed by the shader program manager 100 in the order in which the shaders are received. Can be.

Unlike in FIG. 9, the graphics driver 200 processes all APIs included in the shader each time a shader is received, in FIG. 10, the shader program manager 100 may process only the shader generation API and the source code storage API. . Processing can be suspended for other APIs included in the shader. The shader program manager 100 may process only the shader generation API and the source code storage API until the shader program link API is received, and may suspend processing for other APIs. However, it is not limited thereto.

In FIG. 10, when the shader program manager 100 receives the shader program link API, it may process an API included in each of the vertex shader and pixel shader that is received last. Vertex shaders and pixel shaders can include source APIs and compilation APIs. When a source API and a compile API are included in the vertex shader and the pixel shader, the shader program manager 100 may process APIs other than the compile API. After APIs other than the compile API are processed, the compile API may be processed by the graphics driver 200.

Referring to FIG. 10, when the graphics driver 200 processes a compile API command, the vertex shader and the pixel shader may be compiled by the compiler 500. Each of the vertex shader and pixel shader compiled by the compiler 500 may be a vertex shader and a pixel shader that the shader program manager 100 receives last. When the vertex shader and the pixel shader are compiled, the graphics driver 200 may generate a shader program 1000b including the compiled vertex shader and the pixel shader. An existing graphics driver 200 operation may be performed on the generated shader program 1000b.

Some embodiments may also be implemented in the form of a recording medium including instructions executable by a computer, such as a program module executed by a computer. Computer-readable media can be any available media that can be accessed by a computer, and includes both volatile and nonvolatile media, removable and non-removable media. Further, the computer-readable medium may include both computer storage media and communication media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Communication media typically includes computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave, or other transmission mechanism, and includes any information delivery media.

In addition, in this specification, the "unit" may be a hardware component such as a processor or a circuit, and/or a software component executed by a hardware configuration such as a processor.

The above description of the present invention is for illustrative purposes only, and those of ordinary skill in the art to which the present invention pertains will be able to understand that other specific forms can be easily modified without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative and non-limiting in all respects. For example, each component described as a single type may be implemented in a distributed manner, and similarly, components described as being distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the claims to be described later rather than the detailed description, and all changes or modified forms derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention do.

Claims (15)

In the method of processing graphics commands,
Receiving the graphics command from a device executing an application;
Selecting a vertex shader and a pixel shader to be processed by a graphics processor from among a plurality of vertex shaders and a plurality of pixel shaders included in the graphics command;
Processing a source command and a compile command included in the selected vertex shader and pixel shader;
Generating a shader program using a vertex shader and a pixel shader on which the source command and the compile command are processed;
Searching for a shader program identical to the generated shader program among at least one shader program compiled before the generated shader program is generated; And
Outputting the searched shader program;
Including,
The step of selecting the vertex shader and the pixel shader,
For each of the plurality of vertex shaders and the plurality of pixel shaders, processing only a shader generation instruction and a shader source code storage instruction from among a plurality of instructions included in each of the vertex shaders and pixel shaders;
Receiving a shader program link instruction for compiling and linking at least some of a plurality of shaders included in the graphics instruction; And
Selecting a vertex shader and a pixel shader received last from among a plurality of vertex shaders and a plurality of pixel shaders included in the graphics command when the shader program link command is received; Including,
Wherein the plurality of vertex shaders and the plurality of pixel shaders include the source instruction and the compile instruction. delete delete The method of claim 1,
The generating of the shader program comprises generating a shader program using the vertex shader, the pixel shader, and uniform data. The method of claim 1,
The searching may include generating a key including text and text length of the at least one shader, text and text length of uniform data, a compiler version, and GPU version information; And
Searching using the generated key;
Containing, the method. The method of claim 1,
The outputting step is to output a binary (binary) of the searched shader program. In the graphics command processing device,
Receives a graphics command from a device executing an application, selects a vertex shader and a pixel shader to be processed in a graphics processor from among a plurality of vertex shaders and a plurality of pixel shaders included in the graphics command, and the selected vertex shader and pixel shader At least one compiled before the generation time of the generated shader program by processing the source command and the compile command included in and generating a shader program using the vertex shader and pixel shader on which the source command and the compile command are processed. A processor that searches for a shader program identical to the generated shader program among shader programs and outputs the found shader program; And
A cache for storing previously compiled shader programs;
Including,
The processor, for each of the plurality of vertex shaders and the plurality of pixel shaders, processes only a shader generation instruction and a shader source code storage instruction among a plurality of instructions included in the respective vertex shaders and pixel shaders, and the graphics instruction When a shader program link instruction for compiling and linking at least some of the plurality of shaders included in is received, and when the shader program link instruction is received, the last of a plurality of vertex shaders and a plurality of pixel shaders included in the graphics instruction is received. Is to select the received vertex shader and pixel shader,
Wherein the plurality of vertex shaders and the plurality of pixel shaders include the source instruction and the compile instruction. delete delete The method of claim 7,
Wherein the processor generates a shader program using the vertex shader, the pixel shader, and uniform data. The method of claim 7,
The processor generates a key using data including text and text length of the at least one shader, text and text length of uniform data, compiler version, and GPU version information, and searches using the generated key The device. The method of claim 7,
Wherein the cache stores information including keys and binaries corresponding to compiled shader programs. The method of claim 7,
The apparatus, wherein the processor does not perform compilation even when receiving a shader including a compiled code. Receives a graphics command from a device executing an application, selects a vertex shader and a pixel shader to be processed in a graphics processor from among a plurality of vertex shaders and a plurality of pixel shaders included in the graphics command, and the selected vertex shader and pixel shader At least one compiled before the generation time of the generated shader program by processing the source command and the compile command included in and generating a shader program using the vertex shader and pixel shader on which the source command and the compile command are processed. A graphics driver for searching for a shader program identical to the generated shader program among shader programs and outputting the found shader program; And
A cache for storing previously compiled shader programs;
Including
The graphics driver, for each of the plurality of vertex shaders and the plurality of pixel shaders, processes only a shader generation command and a shader source code storage command among a plurality of commands included in the respective vertex shaders and pixel shaders, and the graphics A shader program link instruction for compiling and linking at least some of the plurality of shaders included in the instruction is received, and when the shader program link instruction is received, the most of a plurality of vertex shaders and a plurality of pixel shaders included in the graphics instruction is received. Is to select the vertex shader and pixel shader received later,
Wherein the plurality of vertex shaders and the plurality of pixel shaders include the source instruction and the compile instruction. A computer-readable non-transitory recording medium storing a program for executing the method of claim 1 on a computer.

Images