Orderability of operations

Patent metadata

Field	Value
Publication number	US-12306772-B2
Application number	US-202318133022-A
Country	US
Kind code	B2
Filing date	Apr 11, 2023
Priority date	Dec 2, 2022
Publication date	May 20, 2025
Grant date	May 20, 2025

Abstract

A method for managing orders of operations between one or more clients and one or more servers is disclosed. The method includes partitioning addressable regions of logical servers on or within an interconnect link into multiple regions including a first orderable region, and providing logical client an ability to push ordering responsibility within the first orderable region to a server. Over the first orderable region, two request messages for access to memory-mapped sources including two respective operations are transmitted, and the two request messages originate from a same logical client. The ordering responsibility can include a first rule for order of operations between the two request messages.

First claim

What is claimed is:

1. A method for managing orders of operations between one or more clients and one or more servers through an interconnect link, comprising:
partitioning addressable regions of one or more logical servers on or within an interconnect link into multiple regions including at least a first orderable region, wherein:
two request messages for access to memory-mapped sources including two respective operations are transmitted over the first orderable region; and
the two request messages originate from a logical client;

providing the logical client an ability to push ordering responsibility within the first orderable region to a server, wherein the ordering responsibility includes a first rule for order of operations between the two request messages; and
attaching, when the two operations are non-overlapping operations, an acquire flag to a first request message of the two request messages or a release flag to a second request message of the two request messages, wherein the acquire flag and the release flag provide a second rule for the order of operations between the two request messages.

2. The method of claim 1 , wherein the first rule includes ordering the order of operations included in the two request messages by ordering a read operation before a write operation when the server accepts the read operation and the write operation on a same cycle.

3. The method of claim 2 , wherein the second rule includes, when the acquire flag or the release flag is attached, ordering the order of operations between the two request messages by ordering an operation of the read operation and the write operation based on an accepted sequence by the server.

4. The method of claim 1 , further comprising, ordering the order of operations between the two request messages according to the first rule and the second rule.

5. The method of claim 1 , wherein the two operations are two first operations, and the multiple regions further comprise a second orderable region, wherein:
a third request message for access to memory-mapped sources and a fourth request message for access to memory-mapped sources including two respective second operations are transmitted over the second orderable region;
the third request message and the fourth request messages originate from a second logical client; and
the two respective second operations of the third request message and the fourth request message are overlapping operations.

6. The method of claim 5 , further comprising:
providing the second logical client an ability to push the ordering responsibility within the second orderable region to the server, wherein the ordering responsibility includes the first rule for order of operations between the third request message and the fourth request message.

7. The method of claim 6 , wherein the multiple regions further comprise a third unordered region, wherein a third logical client enforces an order between operations included in request messages for access to memory-mapped sources within the third unordered region.

8. The method of claim 7 , further comprising:
providing an assertion error when the server receives a request message with the acquire flag or the release flag from the second orderable region or the third unordered region.

9. An integrated circuit for executing instructions comprising:
an interconnect configured to:
partition addressable regions of one or more logical servers on or within the interconnect link into multiple regions including at least a first orderable region, wherein:
two request messages for access to memory-mapped sources including two respective operations are transmitted over the first orderable region; and
the two request messages originate from a logical client;

provide the logical client an ability to push ordering responsibility within the first orderable region to a server, wherein the ordering responsibility includes a first rule for order of operations between the two request messages; and
attach, when the two operations are non-overlapping operations, an acquire flag to a first request message of the two request messages or a release flag to a second request message of the two request messages, wherein the acquire flag and the release flag provide a second rule for the order of operations between the two request messages.

10. The integrated circuit of claim 9 , wherein the first rule includes ordering the order of operations included in the two request messages by ordering a read operation before a write operation when the server accepts the read operation and the write operation on a same cycle.

11. The integrated circuit of claim 10 , wherein the second rule includes, when the acquire flag or the release flag is attached, ordering the order of operations between the two request messages by ordering an operation of the read operation and the write operation based on an accepted sequence by the server.

12. The integrated circuit of claim 9 , wherein the interconnect is further configured to:
order the order of operations between the two request messages according to the first rule and the second rule.

13. The integrated circuit of claim 9 , wherein the two operations are two first operations, the multiple regions further comprise:
a second orderable region, wherein:
a third request message for access to memory-mapped sources and a fourth request message for access to memory-mapped sources including two respective second operations are transmitted over the second orderable region;
the third request message and the fourth request messages originate from a second logical client; and
the two respective second operations of the third request message and the fourth request message are overlapping operations; and

a third unordered region, wherein a third logical client enforces an order between operations included in request messages for access to memory-mapped sources within the third unordered region.

14. A non-transitory computer readable medium comprising a circuit representation that, when processed by a computer, is used to program or manufacture an integrated circuit comprising:
an interconnect configured to:
partition addressable regions of one or more logical servers on or within the interconnect link into multiple regions including at least a first orderable region, wherein:
two request messages for access to memory-mapped sources including two respective operations are transmitted over the first orderable region; and
the two request messages originate from a logical client;

provide the logical client an ability to push ordering responsibility within the first orderable region to a server, wherein the ordering responsibility includes a first rule for order of operations between the two request messages; and
attach, when the two operations are non-overlapping operations, an acquire flag to a first request message of the two request messages or a release flag to a second request message of the two request messages, wherein the acquire flag and the release flag provide a second rule for the order of operations between the two request messages.

15. The non-transitory computer readable medium of claim 14 , comprising the circuit representation that, when processed by the computer, is used to program or manufacture the integrated circuit, wherein the first rule includes ordering the order of operations included in the two request messages by ordering a read operation before a write operation when the server accepts the read operation and the write operation on a same cycle.

16. The non-transitory computer readable medium of claim 15 , comprising

Assignees

• Sifive Inc

Inventors

• Gouldey Eric Andrew
• Klinglesmith Michael
• Cook Henry
• Terpstra Wesley Waylon

Classifications

• G06F13/1626 Primary

  by reordering requests · CPC title