IP Library › Granted Patent US 7,124,224
Granted Patent B2
US 7,124,224 · App. 09/746,487 · Granted Oct 17, 2006

Method and apparatus for shared resource management in a multiprocessing system

Assignee: Intel Corporation
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 7,124,224
App. No.
09/746,487
Granted
Oct 17, 2006
Kind
B2
Abstract

In a multiprocessor, access to shared resources is provided by a semaphore control mechanism, herein disclosed. The semaphore control mechanism provides for a high degree of programmable firmware reuse requiring relatively few modifications from a uniprocessor. A machine check abort (MCA) handling mechanism is disclosed, which works with the semaphore control mechanism in the multiprocessor to provide improved system availability and reliability. The MCA handling mechanism provides for synchronization of multiple processors and shared resources and for timely execution resumption within the processors that remain on-line.

Claims (47)

1. A multiprocessor system including a semaphore checker comprising:

means for receiving one or more semaphore modification requests from one or more requesting devices;

means for identifying an ownership state of a semaphore corresponding to the one or more semaphore modification requests;

means for arbitrating to identify a first modification request of the one or more semaphore modification requests, the first modification request from a first requesting device;

means for granting the first modification request if the identified ownership state corresponds to the first requesting device; and

means for granting the first modification request if the identified ownership state corresponds to no owner.

2. The multiprocessor system recited in claim 1 further comprising:

means for receiving a semaphore read requests from one of the one or more requesting devices;

means for transmitting the identified ownership state in response to the semaphore read request; and

means for allowing the first requesting device to access a shared resource.

3. The multiprocessor system recited in claim 1 wherein the one or more requesting devices are fabricated on a single die.

4. A multiprocessor comprising:

a logical plurality of processors;

a resource scheduling device coupled to one or more of the logical plurality of processors to provide access to a set of resources;

a shared resource of the set of resources having a semaphore; and

a semaphore checker coupled to the resource scheduling device and to the semaphore to:

receive one or more semaphore modification requests from the one or more of the logical plurality of processors,

identify an ownership state of the semaphore,

arbitrate the one or more semaphore modification requests and identify a first modification request from a first requesting processor of the one or more of the logical plurality of processors,

allow the first modification request to succeed if the identified ownership state corresponds to the first requesting processor and

allow the first modification request to succeed if the identified ownership state corresponds to no ownership.

5. The multiprocessor recited in claim 4 wherein the semaphore checker is further to:

decline a second modification request of the one or more semaphore modification requests.

6. The multiprocessor recited in claim 4 wherein the semaphore checker is further to:

receive a semaphore read requests from one of the one or more of the logical plurality of processors;

transmit the identified ownership state in response to the semaphore read request; and

allow the first requesting processor to access a shared resource.

7. The multiprocessor recited in claim 4 wherein each of the one or more semaphore modification requests received identify a corresponding requesting processor of the one or more of the logical plurality of processors.

8. The multiprocessor recited in claim 4 wherein the multiprocessor is fabricated on a single die.

9. The multiprocessor recited in claim 4 wherein arbitration is resolved on a round-robin basis.

10. The multiprocessor recited in claim 4 wherein arbitration is resolved on a priority basis.

11. An apparatus comprising:

a register to access a shared resource of a set of resources;

a semaphore corresponding to the shared resource; and

a semaphore checker coupled to the semaphore to allow access to the shared resource through the register, the semaphore checker further to:

receive one or more semaphore modification requests from one or more of a logical plurality of processing devices, wherein each of the one or more semaphore modification requests received identify a corresponding requesting device of the one or more of the logical plurality of processing devices,

identify an ownership state of the semaphore,

arbitrate the one or more semaphore modification requests and identify a first modification request from a first requesting device of the one or more of the logical plurality of processing devices to succeed if the identified ownership state corresponds to the first requesting device; and

allow the first modification request to succeed if the identified ownership state corresponds to no ownership.

12. The apparatus of claim 11 wherein the semaphore checker is further to:

decline a second modification request of the one or more semaphore modification requests.

13. The apparatus of claim 11 wherein the semaphore checker is further to:

receive a semaphore read requests from one of the one or more of the logical plurality of processors;

transmit the identified ownership state in response to the semaphore read request; and

allow the first requesting processor to access a shared resource.

14. The apparatus of claim 11 wherein the logical plurality of processing devices are integrated on a single die.

15. The apparatus of claim 11 wherein arbitration is resolved on a round-robin basis.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Aug 6, 2001
From: TU, STEVEN; NGUYEN, HANG
To: INTEL CORPORATION
Reel/Frame 012047/0191 →
Continuity (1)
Related Publication 20020116438A1 · Aug 22, 2002