DIRECTING DATA UNITS TO A CORE SUPPORTING TASKS

• US 20100083259A1
• Filed: 09/29/2008
• Published: 04/01/2010
• Est. Priority Date: 09/29/2008
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Veal, Bryan, Foong, Annie
• Granted Patent: US 8,626,955 B2
• Time in Patent Office: Days
• US Class Current: 718/102
• CPC Class Codes: G06F 9/5088 involving task migration
• CPC Class Codes: H04L 12/56 Packet switching systems
• CPC Class Codes: H04L 47/10 Flow control; Congestion control

Abstract

A computer system may comprise a plurality of cores that may process the tasks determined by the operating system. A network device may direct a first set of packets to a first core using a flow-spreading technique such as receive side scaling (RSS). However, the operating system may re-provision a task from the first core to a second core to balance the load, for example, on the computer system. The operating system may determine an identifier of the second core using a new data field in the socket calls to track the identifier of the second core. The operating system may provide the identifier of the second core to a network device. The network device may then direct a second set of packets to the second core using the identifier of the second core.

Claims

1. A method comprising:

directing a first set of packets to a first core using a flow-spreading technique,

re-provisioning a task from the first core to a second core,

determining an identifier of the second core,

providing the identifier of the second core to a network device

directing a second set of packets to the second core using the identifier of the second core.

2. The method of claim 1, wherein the identifier of the second core is determined using a socket system call initiated from the task to a socket, wherein the task and the socket is supported by the second core.

3. The method of claim 2, wherein the socket system call comprises a data field to track the identifier of the second core.

4. The method of claim 2, wherein the identifier of the second core is used to store the second set of packets in a buffer associated with the second core.

5. The method of claim 4 further comprises generating a flow table using flow identifiers extracted from the first set of packets and the second set of packets and associating the flow identifiers with identifiers of the core generated by the operating system.

6. The method of claim 5, wherein the flow table is combined with an indirection table generated using the flow-spreading technique,

wherein a flow identifier of an incoming packet is compared with the entries of the flow table to determine the identifier of the core, and

wherein the flow identifier of the incoming packet is used to determine the identifier of the core included in the indirection table if the flow identifier does not match an entry in the flow table.

8. The method of claim 6, wherein the entries of the flow table and the indirection table is updated using least recently used policy.

9. The method of claim 6, wherein the entries of the flow table and the indirection table are stored using set-associativity.

10. The method of claim 9, wherein the entries of the flow table is tagged using the flow identifiers and sets are indexed using the flow identifiers.

11. An apparatus comprising:

an interface to receive a first set of packets and a second set of packets, and

a logic, wherein the logic is to,

direct the first set of packets to a first core using a flow-spreading technique,

receive an identifier of the second core, wherein the identifier of the second core is determined after a task is re-provisioned from the first core to a second core, and

direct the second set of packets to the second core using the identifier of the second core.

12. The apparatus of claim 11, wherein the identifier of the second core is determined using a socket system call initiated from the task to a socket, wherein the task and the socket is supported by the second core.

13. The apparatus of claim 12, wherein the socket system call comprises a data field to track the identifier of the second core.

14. The apparatus of claim 12 further comprises a first buffer, wherein the logic is to store the first set of packets in a first buffer, wherein the first buffer is associated with the first core.

15. The apparatus of claim 12 further comprises a second buffer, wherein the logic is to store the second set of packets in the second buffer in response to receiving the identifier of the second core, wherein the second buffer is associated with the second core.

16. The apparatus of claim 14 further comprises a flow table, wherein the flow table stores a plurality of entries, wherein the plurality of entries is generated using flow identifiers extracted from the first set of packets and the second set of packets and associating the flow identifiers with identifiers of the plurality of cores.

17. The apparatus of claim 16, wherein the flow table is combined with an indirection table generated using the flow-spreading technique,

wherein a flow identifier of an incoming packet is compared with the plurality of entries of the flow table to determine the identifier of the core to which the incoming packets is to be sent, and

wherein the flow identifier of the incoming packet is used to determine the identifier of the core included in the indirection table if the flow identifier does not match an entry in the flow table.

18. The apparatus of claim 17, wherein the entries of the flow table is updated using least recently used policy.

19. The apparatus of claim 18, wherein the entries of the flow table is stored using set-associativity.

20. The apparatus of claim 19, wherein the entries of the flow table is tagged using the flow identifiers and sets are indexed using the flow identifiers.

21. A system comprising:

a plurality of processing cores, wherein the plurality of processing cores comprise a first core and a second core

an operating system coupled to the plurality of processing cores, wherein the operating system is to,

re-provision a task from the first core to the second core,

determine an identifier of the second core, and

provide the identifier of the second core,

a network device coupled to the operating system, wherein the network device is to

direct a first set of packets to a first core using a flow-spreading technique before receiving the identifier of the second core, and

direct a second set of packets to the second core in response to receiving the identifier of the second core.

22. The system of claim 21, wherein the identifier of the second core is determined using a socket system call initiated from the task to a socket, wherein the socket system call comprises a data field to track the identifier of the second core, wherein the task and the socket is supported by the second core.

23. The system of claim 22 further comprises a first buffer and a second buffer,

wherein the network device is to store the first set of packets in a first buffer, wherein the first buffer is associated with the first core.

wherein the network device is to store the second set of packets in the second buffer in response to receiving the identifier of the second core, wherein the second buffer is associated with the second core.

24. The system of claim 22 further comprises a flow table, wherein the flow table stores a plurality of entries, wherein the plurality of entries is generated using flow identifiers extracted from the first set of packets and the second set of packets and associating the flow identifiers with identifiers of the plurality of cores.

25. The system of claim 24, wherein the flow table is combined with an indirection table generated using the flow-spreading technique,

wherein a flow identifier of an incoming packet is compared with the plurality of entries of the flow table to determine the identifier of the core to which the incoming packets is to be sent, and

wherein the flow identifier of the incoming packet is used to determine the identifier of the core included in the indirection table if the flow identifier does not match an entry in the flow table.