Routing interrupts in a multi-node system

• US 7,552,236 B2
• Filed: 07/14/2005
• Issued: 06/23/2009
• Est. Priority Date: 07/14/2005
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Kriegel, Jon K., Greenfield, Todd Alan
• Primary Examiner(s): DENNISON, JERRY B
• Application Number:
  US11/181,710
• Publication Number:
  US 20070038818A1
• Time in Patent Office: 1,440 Days
• Field of Search: 709/238, 709/239, 709/240, 710/48, 710/260, 710/264, 710/268, 719/311, 719/313, 718/103
• US Class Current: 709/238
• CPC Class Codes: G06F 12/0831 using a bus scheme, e.g. with bus monitoring or watching means
• CPC Class Codes: G06F 13/24 using interrupt G06F13/32 takes precedence

Abstract

A method, apparatus, system, and signal-bearing medium that, in an embodiment, detect a new task priority for a processor, where the processor is connected to a first node, find a home node for the processor via a cluster to which the processor belongs, and send the new task priority to the home node if the home node is different from the first node. In another embodiment, an interrupt directed to a first processor is detected, the interrupt is determined to be redirectable, a home node for the first process is found via a cluster to which the first processor belongs, and an interrupt vector is sent to the home node if the home node is different from the first node.

Claims

1. A method comprising:

detecting a new task priority for a processor, wherein the processor is connected to a first node, and wherein the detecting the new task priority for the processor comprises detecting that a priority of a task tat is currently executing on the processor is changed to the new task priority, wherein the priority of the task is relative to a plurality of priorities of other tasks that execute on the processor;

finding a home node for the processor via a cluster to which the processor belongs; and

sending the new task priority to the home node if the home node is different from the first node.

2. The method of claim 1, further comprising:

receiving the new task priority at the home node; and

storing the new task priority at the home node.

3. The method of claim 2, further comprising:

storing the new task priority at the home node if the home node is identical to the first node.

4. The method of claim 2, wherein the sending further comprises:

sending an identification of the processor to the home node.

5. The method of claim 4, wherein the storing further comprises:

associating the new task priority with the identification of the processor at the home node.

6. The method of claim 1, wherein the home node and the first node are connected via a scalability fabric.

7. The method of claim 1, wherein the home node is responsible for the cluster.

8. A method for configuring a computer, comprising:

configuring the commuter to detect a new task priority for a third processor, wherein the third processor is connected to a third node, and wherein the configuring the computer to detect the new task priority for the third processor comprises configuring the computer to detect that a priority of a task that is currently executing on the third processor is changed to the new task priority, wherein the priority of the task is relative to a plurality of priorities of other tasks that execute on the third processor;

configuring the computer to find a home node for the third processor via a cluster to which the third processor belongs;

configuring the computer to send the new task priority to the home node for the third processor if the home node for the third processor is different from the third node;

configuring the computer to detect an interrupt directed to a first processor, wherein the first processor is connected to a first node;

configuring the computer to determine that the interrupt is redirectable;

configuring the computer to find a home node for the first processor via a cluster to which the first processor belongs; and

configuring the computer to send an interrupt vector to the home node for the first processor if the home node for the first processor is different from the first node.

9. The method of claim 8, further comprising:

configuring the computer to receive the interrupt vector at the home node; and

configuring the computer to select a second processor with a lowest task priority in the cluster.

10. The method of claim 9, further comprising:

configuring the computer to determine a second node to which the second processor is connected.

11. The method of claim 10, further comprising:

configuring the computer to send the interrupt vector to the second node.

12. The method of claim 11, further comprising:

configuring the computer to process the interrupt vector at the second node.

13. The method of claim 8, wherein the computer comprises the home node, the first node, and the second node, and wherein the home node, the first node, and the second node are connected via a scalability fabric.