IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0120161
(2002-04-11)
|
발명자
/ 주소 |
|
출원인 / 주소 |
|
대리인 / 주소 |
Sterne Kessler Goldstein &
|
인용정보 |
피인용 횟수 :
6 인용 특허 :
22 |
초록
▼
A method for conditionally performing a SIMD operation causing a predetermined number of result objects to be held in a combination of different ones of a plurality of destination stores, the method comprising receiving and decoding instruction fields to determine at least one source store, a plural
A method for conditionally performing a SIMD operation causing a predetermined number of result objects to be held in a combination of different ones of a plurality of destination stores, the method comprising receiving and decoding instruction fields to determine at least one source store, a plurality of destination stores and at least one control store, said source and destination stores being capable of holding one or a plurality of objects, each object defining a SIMD lane. Conditional execution of the operation on a per SIMD lane basis is controlled using a plurality of pre-set indicators of the at least one control store designated in the instruction, wherein each said pre-set indicator i controls a predetermined number of result lanes p, where p takes a value greater than or equal to two. A predetermined number of result objects are sent to the destination stores such that the predetermined number of result objects are held by a combination of different ones of the plurality of destination stores.
대표청구항
▼
What is claimed is: 1. A method for conditionally performing a SIMD operation using operands from at least one source store and causing a predetermined number of result objects to be held in a combination of different ones of a plurality of destination stores, the method comprising: receiving and d
What is claimed is: 1. A method for conditionally performing a SIMD operation using operands from at least one source store and causing a predetermined number of result objects to be held in a combination of different ones of a plurality of destination stores, the method comprising: receiving and decoding an instruction to determine at least one source store, a plurality of destination stores, and at least one control store, wherein said source and destination stores are capable of holding one or more source objects and result objects, respectively, each source object associated with a corresponding SIMD lane, and wherein said at least one control store comprises a plurality of pre-set indicators for controlling per SIMD lane conditional execution; controlling conditional execution of the operation on a per SIMD lane basis using said plurality of pre-set indicators of the at least one control store designated in the instruction, wherein each said pre-set indicator is operable to control a plurality of result lanes; and sending a predetermined number of result objects to said destination stores such that the predetermined number of said result objects are held by a combination of different ones of the plurality of destination stores. 2. A method as in claim 1, wherein said controlling step comprises controlling conditional execution of the operation on a per SIMD lane basis using said plurality of pre-set indicators of the at least one control store designated in the instruction, wherein each said pre-set indicator is operable to control two result lanes. 3. A method as in claim 1, wherein the at least one control store comprises a predicate register. 4. A method as in claim 3, wherein each pre-set indicator for controlling conditional execution of a SIMD lane comprises a single bit of the predicate register. 5. A method as in claim 1, wherein the source and result objects are selected from one or more of the following: a byte; a half word comprising 16 bits; a word comprising 32 bits; a long comprising 64 bits and a double long comprising 128 bits. 6. A method as in claim 1, wherein each source and result object comprises a bit sequence, and wherein a first result object comprises a longer bit sequence than a corresponding source object. 7. A method as in claim 1, wherein a further instruction indicates first and second sets of source stores, each capable of holding one or a plurality of objects each defining a SIMD lane. 8. A method as in claim 7, wherein the per SIMD lane conditional execution of said further instruction is controlled by respective pre-set indicators of first and second control stores in combination, each said pre-set indicator controlling a single result lane. 9. A method as in claim 8, wherein the further instruction comprises two or more portions defining different operations. 10. A method as in claim 9, wherein a first portion of the further instruction includes a field indicating data processing operations are to be performed on source objects in said first and second sets of source stores contemporaneously. 11. A method as in claim 9, wherein a second portion of the further instruction defines a load/store operation to effect loading or storing accesses to a plurality of stores contemporaneously. 12. A method as in claim 10 or 11, wherein a field of the further instruction causes corresponding even and odd store address values to be issued. 13. A method as in claim 10 or 11, wherein a field of the further instruction causes a first source store address to be issued and a second source store address to be generated in connection therewith. 14. A computer system for conditionally performing a SIMD operation using operands from at least one source store and causing a predetermined number of result objects to be held in a combination of different ones of a plurality of a destination stores, the computer system comprising: a plurality of stores; decode circuitry for receiving and decoding an instruction and operable to determine at least one source store, a plurality of destination stores, and at least one control store, wherein each said source and destination store is capable of holding one or a plurality of source objects and result objects, respectively, each source object defining a SIMD lane, and wherein said at least one control store comprises a plurality of pre-set indicators for controlling per SIMD lane conditional execution; a plurality of operators operable to perform the operation defined in the instruction on each said SIMD lane; control circuitry for controlling per SIMD lane conditional execution of the operation, wherein the control circuitry comprises said at least one control store and said plurality of pre-set indicators are operable to indicate for each of a plurality of result lanes whether or not a result should be generated on that result lane, each said pre-set indicator being operable to control a plurality of result lanes; and a plurality of destination stores connected to receive a predetermined number of result objects, wherein the predetermined number of result objects are held by a combination of different ones of the plurality of destination stores. 15. A method for conditionally performing a SIMD operation using operands from at least one source store and causing result objects to be sent to different ones of a plurality of destination stores, the method comprising: receiving and decoding an instruction defining an operation, the instruction indicating at least one source store, a plurality of destination stores, and at least one control store, each said source and destination store being capable of holding one or a plurality of source and result objects, respectively, each source object defining a SIMD lane; controlling per SIMD lane conditional execution of the operation, wherein respective pre-set indicators of said at least one control store are operable to control a plurality of corresponding result lanes; and sending objects resulting from the operation on respective predetermined lanes to said plurality of destination stores. 16. A method for conditionally performing multiple data processing operations and multiple load/store operations contemporaneously, the method comprising: providing first and second SIMD execution channels, each being capable of processing a data processing operation and a load/store operation contemporaneously; receiving and decoding a first portion of an instruction responsive to parameters defining first and second data processing operations; receiving and decoding a second portion of an instruction responsive to parameters defining first and second load/store operations; and conditionally performing said first and second data processing operations and said first and second load/store operations contemporaneously on a per SIMD lane basis, wherein said first data processing operation and said first load/store operation are contemporaneously performed by said first SIMD execution channel and said second data processing operation and said second load/store operation are contemporaneously performed by said second SIMD execution channel. 17. A method for conditionally performing a SIMD operation using operands from source stores and causing a predetermined number of result objects to be held in a combination of different ones of a plurality of destination stores; receiving and decoding instruction fields to determine first and second sets of source stores, a plurality of destination stores, each of said source and destination stores being capable of holding one or a plurality of source objects and result objects, respectively, each source object defining a SIMD lane, and first and second control stores each comprising a plurality of pre-set indicators; controlling per SIMD lane conditional execution using said pluralities of pre-set indicators of said first and second control stores, wherein respective pre-set indicators in said first and second control stores are operable to control respective corresponding result lanes contemporaneously; sending a predetermined number of result objects to said destination stores such that said predetermined number of said result objects are held by a combination of different ones of the plurality of destination stores. 18. A computer system for conditionally performing a SIMD operation using operands from source stores and causing a predetermined number of result objects to be held in a combination of different ones of a plurality of destination stores, the computer system comprising: a plurality of stores including source stores, destination stores and control stores; decode circuitry for receiving and decoding instruction fields to determine first and second sets of source stores, a plurality of destination stores, and first and second control stores each comprising a plurality of pre-set indicators, wherein each said source and destination store is capable of holding a source object and a result object, respectively, each source object defining a SIMD lane; a plurality of operators operable to perform the operation defined in the instruction on each SIMD lane; wherein said pluralities of pre-set indicators of said first and second control stores are operable to control per SIMD lane conditional execution, respective pre-set indicators in said first and second control stores being operable to control conditional execution of corresponding respective result lanes contemporaneously. 19. A computer program product stored in a computer readable medium comprising program code means adapted to perform the method of claim 1, 15, 16, or 17.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.