Caching data in a memory system having memory nodes at different hierarchical levels
US-2015378913-A1 · Dec 31, 2015 · US
US10095623B2 · US · B2
| Field | Value |
|---|---|
| Publication number | US-10095623-B2 |
| Application number | US-201615297084-A |
| Country | US |
| Kind code | B2 |
| Filing date | Oct 18, 2016 |
| Priority date | Sep 26, 2014 |
| Publication date | Oct 9, 2018 |
| Grant date | Oct 9, 2018 |
A practical reading order for non-experts. Skip the full description unless you need deep technical detail.
What the patent document calls the invention.
A short plain-language summary of the technical disclosure.
Who owns or filed the patent and who is credited as inventor.
Filing, priority, publication, and grant dates set the timeline.
The legal scope of protection — read this for what is actually claimed.
Technology tags used to group this patent with similar filings.
Prior art links and similar publications in this corpus.
Official abstract text for this publication.
Methods and apparatuses to control access to a multiple bank data cache are described. In one embodiment, a processor includes conflict resolution logic to detect multiple instructions scheduled to access a same bank of a multiple bank data cache in a same clock cycle and to grant access priority to an instruction of the multiple instructions scheduled to access a highest total of banks of the multiple bank data cache. In another embodiment, a method includes detecting multiple instructions scheduled to access a same bank of a multiple bank data cache in a same clock cycle, and granting access priority to an instruction of the multiple instructions scheduled to access a highest total of banks of the multiple bank data cache.
Opening claim text (preview).
What is claimed is: 1. A hardware apparatus comprising: a conflict resolution circuit to detect multiple instructions scheduled to access a same bank of a multiple bank data cache in a same time period and to grant access priority to an instruction of the multiple instructions with a largest access width of the multiple bank data cache. 2. The hardware apparatus of claim 1 , wherein the conflict resolution circuit is to grant the access priority to the instruction scheduled to access a highest total of banks of the multiple bank data cache. 3. The hardware apparatus of claim 1 , wherein a queue entry for each of the multiple instructions include an access width value. 4. The hardware apparatus of claim 1 , wherein the conflict resolution circuit is to grant access priority to an earliest instruction when the multiple instructions are scheduled to access a same access width of the multiple bank data cache. 5. The hardware apparatus of claim 4 , wherein a queue entry for each of the multiple instructions include a relative program order. 6. The hardware apparatus of claim 1 , wherein an instruction of the multiple instructions that is not to be granted access priority is to be redispatched to the conflict resolution circuit. 7. The hardware apparatus of claim 1 , wherein the conflict resolution circuit is to detect the multiple instructions scheduled to access the same bank of the multiple bank data cache in the same time period by comparing a bank access mask of each of a plurality of instructions. 8. The hardware apparatus of claim 1 , wherein the conflict resolution circuit is to grant access to the multiple bank data cache to multiple instructions that are not scheduled to access the same bank of the multiple bank data cache in the same time period. 9. A method to control access to a multiple bank data cache comprising: detecting multiple instructions scheduled to access a same bank of the multiple bank data cache in a same time period; and granting access priority to an instruction of the multiple instructions with a largest access width of the multiple bank data cache. 10. The method of claim 9 , wherein the granting comprises granting the access priority to the instruction scheduled to access a highest total of banks of the multiple bank data cache. 11. The method of claim 9 , further comprising providing a queue entry for each of the multiple instructions that includes an access width value. 12. The method of claim 9 , wherein the granting comprises granting access priority to an earliest instruction when the multiple instructions are scheduled to access a same access width of the multiple bank data cache. 13. The method of claim 12 , further comprising providing a queue entry for each of the multiple instructions that include a relative program order. 14. The method of claim 9 , further comprising redispatching an instruction of the multiple instructions that is not to be granted access priority. 15. The method of claim 9 , wherein the detecting comprises comparing a bank access mask of each of a plurality of instructions. 16. The method of claim 9 , further comprising granting access to the multiple bank data cache to multiple instructions that are not scheduled to access the same bank of the multiple bank data cache in the same time period. 17. A hardware system comprising: a core to execute a plurality of instructions in a same time period; and a conflict resolution circuit to detect multiple instructions scheduled to access a same bank of a multiple bank data cache in the same time period and to grant access priority to an instruction of the multiple instructions with a largest access width of the multiple bank data cache. 18. The hardware system of claim 17 , wherein the conflict resolution circuit is to grant the access priority to the instruction scheduled to access a highest total of banks of the multiple bank data cache. 19. The hardware system of claim 17 , wherein a queue entry for each of the multiple instructions include an access width value. 20. The hardware system of claim 17 , wherein the conflict resolution circuit is to grant access priority to an earliest instruction when the multiple instructions are scheduled to access a same access width of the multiple bank data cache. 21. The hardware system of claim 20 , wherein a queue entry for each of the multiple instructions include a relative program order. 22. The hardware system of claim 17 , wherein an instruction of the multiple instructions that is not to be granted access priority is to be redispatched to the conflict resolution circuit. 23. The hardware system of claim 17 , wherein the conflict resolution circuit is to detect the multiple instructions scheduled to access the same bank of the multiple bank data cache in the same time period by comparing a bank access mask of each of the plurality of instructions. 24. The hardware system of claim 17 , wherein the conflict resolution circuit is to grant access to the multiple bank data cache to multiple instructions that are not scheduled to access the same bank of the multiple bank data cache in the same time period.
with a shared cache · CPC title
Permissions · CPC title
Multiple simultaneous or quasi-simultaneous cache accessing · CPC title
with dedicated cache, e.g. instruction or stack · CPC title
with multilevel cache hierarchies · CPC title
Related publications grouped by family.
Answers are generated from the same data shown on this page.