IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0267778
(2005-11-04)
|
등록번호 |
US-8424012
(2013-04-16)
|
발명자
/ 주소 |
- Karandikar, Ashish
- Gadre, Shirish
- Gruner, Frederick R.
- Sijstermans, Franciscus W.
|
출원인 / 주소 |
|
인용정보 |
피인용 횟수 :
6 인용 특허 :
149 |
초록
▼
A method for context switching on a video processor having a scalar execution unit and a vector execution unit. The method includes executing a first task and a second task on a vector execution unit. The first task in the second task can be from different respective contexts. The first task and the
A method for context switching on a video processor having a scalar execution unit and a vector execution unit. The method includes executing a first task and a second task on a vector execution unit. The first task in the second task can be from different respective contexts. The first task and the second task are each allocated to the vector execution unit from a scalar execution unit. The first task and the second task each comprise a plurality of work packages. In response to a switch notification, a work package boundary of the first task is designated. A context switch from the first task to the second task is then executed on the work package boundary.
대표청구항
▼
1. A method for implementing context switching on a video processor, comprising: executing a first task on a vector execution unit, the first task allocated to the vector execution unit from a scalar execution unit;executing a second task on the vector execution unit, the second task allocated to th
1. A method for implementing context switching on a video processor, comprising: executing a first task on a vector execution unit, the first task allocated to the vector execution unit from a scalar execution unit;executing a second task on the vector execution unit, the second task allocated to the vector execution unit from the scalar execution unit;in response to a switch notification, designating a work package boundary of the first task to create a work package, wherein the work package boundary is operable to reduce an amount of state data associated with the first task to be saved during a context switch from the first task to the second task, wherein the vector execution unit is operable to continue executing the work package upon receiving the switch notification until the work package boundary is encountered; andexecuting the context switch from the first task to the second task on the work package boundary. 2. The method of claim 1, wherein the work package is completely executed by the vector execution unit prior to the context switch from the first task to the second task. 3. The method of claim 1, wherein the work package comprises a plurality of video processing commands for implementing at least one task of a video processing application. 4. The method of claim 1, further comprising: storing a plurality of commands comprising the first task in a command FIFO;for each of the commands comprising read instructions, fetching the read data corresponding to the read instructions from a frame buffer memory; andstoring the read data in an instruction cache of the vector execution unit or a data cache of the vector execution unit. 5. The method of claim 4, further comprising: for each of the commands comprising read instructions, ensuring the read data corresponding to the read instructions are successfully fetched from the frame buffer memory and stored in the instruction cache or the data cache prior to commencing execution of the work package by the vector execution unit. 6. The method of claim 4, wherein a DMA engine is configured to fetch the read data from the frame buffer memory. 7. The method of claim 6, wherein a vector control unit is coupled to the command FIFO and is coupled to the DMA engine and is configured to control the DMA engine to fetch the read data from the frame buffer memory. 8. The method of claim 7, further comprising: in response to a switch notification, saving vector control unit state data; andsaving command FIFO state data. 9. The method of claim 8, wherein state data for a vector datapath of the vector execution unit is not saved. 10. A video processor for implementing context switching and having a scalar execution unit and a vector execution unit, comprising: a memory interface for establishing communication between the video processor and a frame buffer memory;a vector execution unit coupled to the memory interface and configured to execute vector video processing operations; anda scalar execution unit coupled to the memory interface and configured to control the vector execution unit, wherein the vector execution unit is configured to execute a first task allocated from the scalar execution unit, and execute a second task allocated from the scalar execution unit, and wherein, in response to a switch notification, designate a work package boundary of the first task to create a work package, wherein the work package boundary is operable to reduce an amount of state data associated with the first task to be saved during a context switch from the first task to the second task, wherein the vector execution unit is operable to continue executing the work package upon receiving the switch notification until the work package boundary is encountered, and execute the context switch from the first task to the second task on the work package boundary. 11. The video processor of claim 10, wherein the work package is completely executed by the vector execution unit prior to the context switch from the first task to the second task. 12. The video processor of claim 10, further comprising: a command FIFO configured to store a plurality of commands comprising the first task,an instruction cache; anda data cache, wherein for each of the commands comprising read instructions, the read data corresponding to the read instructions is fetched from a frame buffer memory and stored in the instruction cache or the data cache. 13. The video processor of claim 12, wherein the video processor is configured to ensure, for each of the commands comprising read instructions, the read data corresponding to the read instructions are successfully fetched from the frame buffer memory and stored in the instruction cache or the data cache prior to commencing execution of the work package by the vector execution unit. 14. The video processor of claim 13, further comprising: a vector data path;a DMA engine;a command FIFO; anda vector control unit coupled to the command FIFO and coupled to the DMA engine, wherein, in response to a switch notification, vector control unit state data and command FIFO state data are saved. 15. The video processor of claim 14, wherein, in response to the switch notification, vector data path state data is not saved. 16. A system for executing video processing operations, comprising: a CPU;a video processor coupled to the CPU, comprising:a memory interface for establishing communication between the video processor and a frame buffer memory;a vector execution unit coupled to the memory interface and configured to execute vector video processing operations; anda scalar execution unit coupled to the memory interface and configured to control the vector execution unit, wherein the vector execution unit is configured to execute a first task allocated from the scalar execution unit, and execute a second task allocated from the scalar execution unit, and wherein in response to a switch notification, designate a work package boundary of the first task to create a work package, wherein the work package boundary is operable to reduce the amount of state data associated with the first task to be saved during a context switch from the first task to the second task, wherein the vector execution unit is operable to continue executing the work package upon receiving the switch notification until the work package boundary is encountered, and execute the context switch from the first task to the second task on the work package boundary. 17. The system of claim 16, wherein the vector execution unit further comprises: a command FIFO configured to store a plurality of commands comprising the first task;an instruction cache; and a data cache, wherein for each of the commands comprising read instructions, the read data corresponding to the read instructions is fetched from a frame buffer memory and stored in the instruction cache or the data cache. 18. The system of claim 17, wherein the video processor is configured to ensure, for each of the commands comprising read instructions, the read data corresponding to the read instructions are successfully fetched from the frame buffer memory and stored in the instruction cache or the data cache prior to commencing execution of the work package by the vector execution unit, and wherein, in response to a switch notification, state data of a vector data path of the vector execution unit state data is not saved.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.