Methods and apparatus for distributing system management signals

1. A computer system, comprising:

a plurality of local resources including a local cluster of processors in a local box, the local cluster of processors interconnected in a point-to-point architecture;

a plurality of remote resources including remote clusters of remote processors in remote boxes, the remote clusters of processors interconnected in point-to-point architectures;

a routing table having partition information used to configure the local and remote resources into one or more partitions, wherein partition information can configure multiple partitions within a single box and a single partition across multiple boxes;

wherein a first line between the local box and a remote box is configured to distribute system management signals based on partition information in the routing table and a second line between the local box and the remote box is configured to distribute memory data based on partition information in the routing table.

2. The computer system of claim 1 , wherein resources in the local box and the remote boxes are configurable as one or more partitions.

3. The computer system of claim 1 , wherein the first line is configured to carry system management signals to only resources in a particular partition, even if multiple partitions reside on a single box.

4. The computer system of claim 1 , wherein the first line is configured to carry system management signals to only resources in a particular partition, even if the same partition resides on a plurality of boxes.

5. The computer system of claim 1 , wherein the system management signals comprise power management signals.

6. The computer system of claim 1 , wherein the system management signals consist of power management signals.

7. The computer system of claim 1 , wherein the system management signals comprise data transport stop and data transport request signals.

8. The computer system of claim 5 , wherein the system management signals comprise LDTSTOP signals.

9. The computer system of claim 8 , wherein LDTSTOP signals are used to synchronize frequency and voltage changes in resources associated with a partition.

10. The computer system of claim 9 , wherein an LDTSTOP routing table is used to determine partition information associated with the LDTSTOP signals.

11. The computer system of claim 10 , wherein the LDTSTOP routing table is used to forward the LDTSTOP signal to a subset of remote resources in remote boxes.

12. A method for distributing system management signals, the method comprising:

receiving a system management signal at a configuration system module associated with a local cluster of processors configured in a point-to-point architecture in a local box;

referencing a routing table having partition information used to configure the local and remote resources into one or more partitions, wherein partition information can configure multiple partitions within a single box and a single partition across multiple boxes;

determining using the routing table which local resources and remote resources associated with remote boxes should receive the system management signal; and

distributing the system management signal to selected local and remote resources using a first line configured to carry system management information, wherein a second line is configured to carry memory data, wherein the system management information and the memory data is distributed using partition information.

13. The method of claim 12 , wherein the local and remote resources comprise processors, controllers, and I/O resources.

14. The method of claim 12 , wherein the routing table is an LDTSTOP routing table.

15. The method of claim 12 , wherein resources in the local box and the remote boxes are configurable as one or more partitions.

16. The method of claim 12 , wherein the lines can be configured to carry system management signals to only resources in a particular partition, even if multiple partitions reside on a single box.

17. The method of claim 12 , wherein the dedicated lines are configurable to carry system management signals to only resources in a particular partition, even if the same partition resides on a plurality of boxes.

18. The method of claim 12 , wherein the system management signals comprise power management signals.

19. The method of claim 12 , wherein the system management signals consist of power management signals.

20. The method of claim 12 , wherein the system management signals comprise data transport stop and data transport request signals.

21. The method of claim 18 , wherein the system management signals comprise LDTSTOP signals.

22. The method of claim 21 , wherein LDTSTOP signals are used to synchronize frequency and voltage changes in resources associated with a partition.

23. The method of claim 22 , wherein an LDTSTOP routing table is used to determine partition information associated with the LDTSTOP signals.

24. The method of claim 23 , wherein the LDTSTOP routing table is used to forward the LDTSTOP signal to a subset of remote resources in remote boxes.

25. A multiple processor box, comprising:

means for receiving a system management signal at a configuration system module associated with a local cluster of processors configured in a point-to-point architecture in a local box;

means for referencing a routing table having partition information used to configure the local and remote resources into one or more partitions, wherein partition information can configure multiple partitions within a single box and a single partition across multiple boxes;

means for determining using the routing table which local resources and remote resources associated with remote boxes should receive the system management signal; and

means for distributing the system management signal to selected local and remote resources using a first line configured to carry system management information, wherein a second line is configured to carry memory data, wherein the system management information and the memory data is distributed using partition information.

26. The multiple processor box of claim 25 , wherein the local and remote resources comprise processors, controllers, and I/O resources.

27. The multiple processor box of claim 25 , wherein the routing table is an LDTSTOP routing table.

28. A method for distributing system management information, the method comprising:

identifying a plurality of local components including a local cluster of processors in a local box, the local box coupled to a remote box having a remote cluster of processors;

receiving an instruction to generate a data transport signal; and

accessing a data transport routing table, the data transport routing table having partition information used to configure the local and remote resources into one or more partitions, wherein partition information can configure multiple partitions within a single box and a single partition across multiple boxes, the data transport routing table configured to allow distribution of system management information to resources associated with particular partitions;

forwarding the data transport signal to the remote box, wherein a first line between the local box and a remote box is configured to distribute system management signals based on partition information in the routing table and a second line between the local box and the remote box is configured to distribute memory data based on partition information in the routing table.

29. The method of claim 28 , wherein the data transport signal is a request to change processor voltage and frequency.

30. The method of claim 28 , wherein the data transport signal is a request to change link width and frequency.

31. The method of claim 28 , wherein the data transport signal is a request to enter a particular sleep state.

32. The method of claim 28 , wherein the local cluster of processors are interconnected using point-to-point links.

33. A computer system, comprising:

means for identifying a plurality of local components including a local cluster of processors in a local box, the local box coupled to a remote box having a remote cluster of processors;

means for receiving an instruction to generate a data transport signal; and

means for accessing a data transport routing table, the data transport routing table having partition information used to configure the local and remote resources into one or more partitions, wherein partition information can configure multiple partitions within a single box and a single partition across multiple boxes, the data transport routing table configured to allow distribution of system management information to resources associated with particular partitions;

means for forwarding the data transport signal to the remote box, wherein a first line between the local box and a remote box is configured to distribute system management signals based on partition information in the routing table and a second line between the local box and the remote box is configured to distribute memory data based on partition information in the routing table.

34. The computer system of claim 33 , wherein the data transport signal is a request to change processor voltage and frequency.

35. The computer system of claim 33 , wherein the data transport signal is a request to change link width and frequency.

36. The computer system of claim 33 , wherein the data transport signal is a request to enter a particular sleep state.

37. The computer system of claim 33 , wherein the local cluster of processors are interconnected using point-to-point links.