Processor system with predicate register, computer system, method for managing predicates and computer program product
View Patent ↗A processor system is adapted to carry out a predicate swap instruction of an instruction set to swap, via a data pathway, predicate data in a first predicate data location of a predicate register with data in a corresponding additional predicate data location of a first additional predicate data container and to swap, via a data pathway, predicate data in a second predicate storage location of the predicate register with data in a corresponding additional predicate data location in a second additional predicate data container.
1. A processor system comprising:
a processing unit with a predetermined instruction set;
a predicate register comprising a first plurality of predicate data locations,
including PL 1 , PL 2 , and PL 3 , each predicate data location being adapted to store predicate data, the first plurality of predicate data locations is an integer number N;
a general purpose register comprising a second plurality of general purpose data locations, the second plurality of general purpose data locations are configured to be partitioned into sets of general purpose register locations, each set of the M sets having N general purpose data locations;
a data pathway for transmitting data between the predicate register and the general purpose register; and
the processor system configured to carry out a predicate swap instruction of the predetermined instruction set to:
swap, via the data pathway, predicate data at PL 1 of the predicate register with data at a corresponding general purpose data location of a first set of the M sets based on a first portion of the predicate swap instruction identifying PL 1 and a second portion of the predicate swap instruction identifying the first set of the M sets; and to
swap, via the data pathway, predicate data at PL 2 of the predicate register with data at a corresponding general purpose data location of a second set of the M sets based on a third portion of the predicate swap instruction identifying PL 2 and a fourth portion of the predicate swap instruction identifying the second set of the M sets.
2. The processor system according to claim 1 , wherein the predicate swap instruction is defined to swap predicate data at more than two predicate data locations of the first plurality of predicate data locations of the predicate register with data at more than two general purpose data locations of the second plurality of general purpose data locations of the general purpose register.
3. The processor system according to claim 1 , wherein each predicate data location of the first plurality of predicate data locations of the predicate register stores one predicate bit, and each general purpose data location of the second plurality of general purpose data locations stores one predicate bit.
4. The processor system according to claim 1 , wherein the predicate data locations of the first plurality of predicate data locations of the predicate register are equally sized.
5. The processor system according to claim 1 , wherein the processor system carries out the predicate swap instruction within one clock cycle of the processing unit.
6. The processor system according to claim 1 , wherein the processor system comprises more than one general purpose register comprising the second plurality of general purpose data locations.
7. The processor system according to claim 6 , wherein more than one general purpose register is addressable by the predicate swap instruction.
8. The processor system according to claim 1 , comprising more than one predicate register.
9. The processor system according to claim 8 , wherein more than one predicate register is addressable by the predicate swap instruction.
10. A method, comprising:
executing, by a processing unit of a processor system,
a predicate swap instruction of a predetermined instruction set to perform:
swapping, by the processing unit, via a data pathway, predicate data at a first predicate data location of a first plurality of predicate data locations of a predicate register with data at a corresponding general purpose data location of a first set of M sets of general purpose register locations of a second plurality of general purpose data locations based on a first portion of the predicate swap instruction identifying the first predicate data location and a second portion of the predicate swap instruction identifying the first set of the M sets; and
swapping, by the processing unit, via the data pathway, predicate data at a second predicate data location of the first plurality of predicate data locations with data at a corresponding general purpose data location of a second set of the M sets based on a third portion of the predicate swap instruction identifying the second predicate data location and a fourth portion of the predicate swap instruction identifying the second set of the M sets,
wherein the first plurality of predicate data locations is an integer N, and the second plurality of general purpose data locations are configured to be partitioned into the M sets of general purpose register locations, each of the M sets having N general purpose data locations.
11. The method according to claim 10 , wherein the predicate swap instruction is defined to swap predicate data at more than two predicate data locations of the first plurality of predicate data locations with data at two or more corresponding general purpose data locations of the second set of the M sets.
12. The method of claim 10 , wherein each predicate data location of the first plurality of predicate data locations of the predicate register stores one predicate bit, and each general purpose data location of the second plurality of general purpose data locations stores one predicate bit.
13. The method of claim 10 , wherein the predicate data locations of the first plurality of predicate data locations of the predicate register are equally sized.
14. The method of claim 10 , wherein the processor system carries out the predicate swap instruction within one clock cycle of the processing unit.
15. The method of claim 10 , wherein the processor system comprises more than one general purpose register comprising the second plurality of general purpose data locations.
16. The method of claim 10 , wherein more than one general purpose register is addressable by the predicate swap instruction.
17. The method of claim 10 , wherein the processor system comprising more than one predicate register.
18. A method, comprising:
executing, by a processing unit of a processor system, a predicate swap instruction of a predetermined instruction set to perform:
swapping, by the processing unit, predicate data at a first predicate data location of a first plurality of predicate data locations of a predicate register with data at a corresponding general purpose data location of a first set of M sets of general purpose register locations of a second plurality of general purpose data locations based on a first portion of the predicate swap instruction identifying the first predicate data location and a second portion of the predicate swap instruction identifying the first set of the M sets; and
swapping, by the processing unit, predicate data at a second predicate data location of the first plurality of predicate data locations with data at a corresponding general purpose data location of a second set of the M sets based on a third portion of the predicate swap instruction identifying the second predicate data location and a fourth portion of the predicate swap instruction identifying the second set of the M sets.