Arbitration for an embedded processor block core in an integrated circuit

• US 7,673,087 B1
• Filed: 03/27/2008
• Issued: 03/02/2010
• Est. Priority Date: 03/27/2008
• Status: Active Grant

• Current Assignee: Xilinx, Inc. (Advanced Micro Devices, Inc.)
• Original Assignee: Xilinx, Inc. (Advanced Micro Devices, Inc.)
• Inventors: Shenoy, Kunal R., Applebaum, Jeffery H., Ansari, Ahmad R.
• Primary Examiner(s): Rinehart; Mark
• Assistant Examiner(s): Spittle; Matthew D
• Application Number:
  US12/057,322
• Time in Patent Office: 705 Days
• Field of Search: 710/109, 710/200, 710/220, 710/240, 710/241, 710/305, 710/317, 710/309, 370/351, 370/447, 370/449, 370/911, 711/151
• US Class Current: 710/241
• CPC Class Codes: G06F 13/366 using a centralised polling arbiter
• CPC Class Codes: Y10S 370/911 Bridge, e.g. brouter, bus extender

Abstract

Arbitration for a processor block core is described. Master devices are associated with a processor block core embedded in a host integrated circuit (“IC”). The master devices are coupled to core logic of the host IC via a crossbar switch and a bridge, which are part of the processor block core. The crossbar switch includes an arbiter. An arbitration protocol is selected from among a plurality of arbitration protocols for use by the arbiter. Pending transactions having are polled for access to the bridge for arbitration using the arbitration protocol selected.

Claims

1. A method for arbitration, comprising:

associating master devices with a processor block core embedded in a host integrated circuit;

the master devices coupled to core logic of the host integrated circuit via a crossbar switch and a bridge;

the bridge being for access to and from the core logic and for access from and to the crossbar switch;

the crossbar switch and the bridge being part of the processor block core;

the crossbar switch including an arbiter;

selecting an arbitration protocol from among a plurality of arbitration protocols for use by the arbiter;

polling the master devices responsive to the arbitration protocol selected to identify a first pending transaction with a ready status of a first master device of the master devices;

sending the first pending transaction from the first master device to the arbiter; and

arbitrating by the arbiter the first pending transaction for access to the bridge using the arbitration protocol selected.

2. The method according to claim 1, wherein the polling further includes:

identifying a second pending transaction with the ready status of a second master device of the master devices; and

placing the first pending transaction and the second pending transaction in a command queue of the bridge.

3. The method according to claim 2, further comprising:

assigning request priorities respectively to the master devices;

identifying a first request priority of the first pending transaction and a second request priority of the second pending transaction; and

the placing of the first pending transaction and the second pending transaction in the command queue being ordered responsive to the first request priority and the second request priority.

4. The method according to claim 3, wherein the arbitrating includes selecting the first pending transaction for access to the bridge prior to the second pending transaction responsive to the first request priority being higher than the second request priority.

5. The method according to claim 4, wherein the arbitration protocol is selected from a group consisting of round robin and least recently used.

6. The method according to claim 5, wherein the first request priority overrides the arbitration protocol.

7. The method according to claim 2, further comprising:

assigning request priorities respectively to the master devices;

setting master priorities respectively to the master devices;

identifying a first request priority of the first pending transaction and a second request priority of the second pending transaction as being equal;

identifying a first master priority of the master device and a second master priority of the second master device; and

the placing of the first pending transaction and the second pending transaction in the command queue being ordered responsive to the first master priority and the second master priority.

8. The method according to claim 7, wherein:

the master priorities are set for the arbitration protocol; and

the arbitration protocol is selected from a group consisting of round robin, and least recently used.

9. A method for arbitration, comprising:

associating master devices with a processor block core embedded in a host integrated circuit;

the master devices coupled to core logic of the host integrated circuit via a crossbar switch, a first bridge, and a second bridge;

each of the first bridge and the second bridge being for access to and from the core logic and for access from and to the crossbar switch;

the crossbar switch, the first bridge, and the second bridge being part of the processor block core;

the crossbar switch including a first arbiter and a second arbiter respectively for access to the first bridge and the second bridge;

the first arbiter and the second arbiter being independent of one another;

selecting a first arbitration protocol and a second arbitration protocol from among a plurality of arbitration protocols respectively for use by the first arbiter and the second arbiter;

the first arbitration protocol and the second arbitration protocol capable of being same or different from one another for respective use by the first arbiter and the second arbiter; and

polling the master devices to identify a master device of the master devices having a pending transaction for either the first bridge or the second bridge.

10. The method according to claim 9, wherein the arbitration protocol is selected from a group consisting of fixed, round robin, and least recently used.

11. The method according to claim 9, further comprising:

granting a locked status of the first bridge to the master device of the master devices for passing commands from the master device to the first bridge without arbitrating for access thereto with at least one other master device of the master devices having another pending transaction.

12. The method according to claim 11, wherein:

the polling includes identification of the at least one other master device having the other pending transaction; and

the polling is only of pending transactions having a ready status.

13. The method according to claim 11, further comprising:

the arbitrating access to the first bridge being suspended until the locked status of the first bridge is removed;

removing the locked status of the first bridge; and

arbitrating first pending transactions of the master devices for access to the first bridge.

14. The method according to claim 13, wherein the commands are for atomic operation.

15. The method according to claim 13, further comprising:

arbitrating second pending transactions of the master devices for access to the second bridge while the first bridge has the locked status.

16. The method according to claim 15, wherein the polling is only of the second pending transactions while the first bridge has the locked status.

17. The method according to claim 16, wherein the polling is only of the second pending transactions having a ready status.

18. The method according to claim 17, wherein the arbitration protocol is selected from a group consisting of round robin and least recently used.