Resource sharing using process delay
View Patent ↗Methods and systems that reduce the number of instance of a shared resource needed for a processor to perform an operation and/or execute a process without impacting function are provided. a method of processing in a processor is provided. Aspects include determining that an operation to be performed by the processor will require the use of a shared resource. A command can be issued to cause a second operation to not use the shared resources N cycles later. The shared resource can then be used for a first aspect of the operation at cycle X and then used for a second aspect of the operation at cycle X+N. The second operation may be rescheduled according to embodiments.
1. A method of processing in a computer system, comprising:
determining that a first operation and a second operation will require the use of a shared resource;
issuing a command configured to cause restriction of access to the shared resource for the second operation at an Nth cycle after a cycle X;
using the shared resource for a first aspect of the first operation at cycle X;
scheduling the second aspect of the first operation to use the shared resource at cycle X+N, wherein cycle X+N is a time when the second operation would have otherwise been scheduled to use the shared resource; and
re-scheduling the second operation to use the shared resource at a next opportunity without rescheduling other waiting operations,
wherein X and N are positive integers.
2. The method of claim 1 , wherein re-scheduling the second operation comprises allowing the second operation to use the shared resource immediately following the use of the shared resource by the second aspect of the first operation.
3. The method of claim 1 , wherein the first operation comprises a store operation requiring that both an address and data be read from an internal memory.
4. The method of claim 3 , wherein the first aspect of the first operation comprises reading an address from the internal memory.
5. The method of claim 3 , wherein the second aspect of the first operation comprises reading data from the internal memory.
6. The method of claim 3 , further comprising writing the address to a load/store queue.
7. The method of claim 3 , further comprising writing the data to a load/store queue.
8. The method of claim 1 , wherein the shared resource is a read port of internal memory.
9. The method of claim 1 , wherein the command is communicated in fewer than N cycles.
10. The method of claim 1 , wherein N is greater than or equal to 2 cycles.
11. A processor, comprising:
a shared resource configured to be used by a plurality of operations during different processor cycles;
a scheduler configured to:
determine that a first operation will require the use of a shared resource and to cause restriction of access to the shared resource for the second operation at an Nth cycle after a cycle X, and
reschedule the second operation to use the shared resource at a next opportunity without rescheduling other waiting operations; and
a processing component configured to:
perform a first aspect of the first operation at cycle X, and
perform a second aspect of the first operation at cycle X+N, wherein cycle X+N is a time when the second operation would have otherwise been scheduled to use the shared resource,
wherein X and N are positive integers.
12. The processor of claim 11 , wherein the scheduler is configured to reschedule the second operation by allowing the second operation to use the shared resource immediately following the use of the shared resource by the second aspect of the first operation.
13. The processor of claim 11 , wherein the first operation comprises a store operation requiring that both an address and data be read from an internal memory.
14. The processor of claim 13 , wherein the processing component is configured to read an address from the internal memory as the first aspect of the first operation.
15. The processor of claim 13 , wherein the processing component is configured to read data from the internal memory as the second aspect of the first operation.
16. The processor of claim 13 , wherein the processing component is configured to write the address to a load/store queue.
17. The processor of claim 13 , wherein the processing component is configured to write the data to a load/store queue.
18. The processor of claim 11 , wherein the shared resource is a read port of an internal memory.
19. The processor of claim 18 , wherein the internal memory comprises a register file and/or a cache.
20. The processor of claim 11 , wherein the N is greater than or equal to 2 cycles.