Protecting shared resources using shared memory and sockets

• US 8,341,643 B2
• Filed: 03/29/2010
• Issued: 12/25/2012
• Est. Priority Date: 03/29/2010
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Fulton, Michael, Lin, Angela, Low, Andrew R., Nageshappa, Prashanth K.
• Primary Examiner(s): Truong, Lechi
• Application Number:
  US12/749,392
• Publication Number:
  US 20110239219A1
• Time in Patent Office: 1,002 Days
• Field of Search: 719/312
• US Class Current: 719/312
• CPC Class Codes: G06F 9/526 Mutual exclusion algorithms

Abstract

Shared memory and sockets are used to protect shared resources where multiple operating systems execute concurrently on the same hardware. Rather than using spinlocks for serializing access, when a thread is unable to acquire a shared resource because that resource is already held by another thread, the thread creates a socket with which it will wait to be notified that the shared resource has been released. The sockets may be network sockets or in-memory sockets that are accessible across the multiple operating systems; if sockets are not available, communication technology that provides analogous services between operating systems may be used instead. Optionally, fault tolerance is provided to address socket failures, in which case one or more threads may fall back (at least temporarily) to using spinlocks. A locking service may execute on each operating system to provide a programming interface through which threads can invoke lock operations.

Claims

1. A method of protecting a shared resource in a computing system, comprising:

storing, in shared memory accessible to at least a first thread and a second thread which execute concurrently in the computing system, a lock data structure for serializing access to the shared resource, the first thread executing under control of a first operating system hosted by the computing system and the second thread executing under control of a second operating system hosted by the computing system; and

upon detecting, by the first thread, that a lock provided by the lock data structure is already held by the second thread, performing, by the first thread;

using a socket to block while waiting for the lock to be released;

adding a waiter entry to a queue, the waiter entry specifying an address of the socket on which the first thread is blocking; and

receiving a notification from the second thread, on the socket, when the second thread releases the lock.

2. The method according to claim 1, wherein the socket is a network socket.

3. The method according to claim 1, wherein the socket is a cross-guest shared-memory socket.

4. The method according to claim 1, further comprising:

obtaining, by the second thread, the waiter entry from the queue when the second thread is releasing the lock;

using the address of the socket specified in the obtained waiter entry, by the second thread, to connect to the socket on which the first thread blocked; and

sending a message on the socket on which the first thread blocked, by the second thread, to notify the first thread when the second thread releases the lock.

5. The method according to claim 1, wherein the first thread uses a spinlock to suspend execution while waiting for the lock to be released, rather than using the socket to block, if the first thread encounters a socket error.

6. The method according to claim 1, wherein the first thread uses a spinlock to suspend execution while waiting for the lock to be released, rather than using the socket to block, if the first thread observes a variable setting that indicates that the second thread encountered a socket failure.

7. The method according to claim 1, wherein the first operating system executes a first cross-guest locking service and the second operating system executes a second cross-guest locking service, the first cross-guest locking service and the second cross-guest locking service providing a programming interface through which threads executing on the first and second operating systems invoke operations for holding and releasing the lock on the shared resource.

8. A system for protecting shared resources in a computing system, comprising:

a computer comprising a processor; and

instructions which are executable, using the processor, to implement functions comprising;

storing, in shared memory accessible to at least a first thread and a second thread which execute concurrently in the computing system, a lock data structure for serializing access to one of the shared resources, the first thread executing under control of a first operating system hosted by the computing system and the second thread executing under control of a second operating system hosted by the computing system; and

upon detecting, by the first thread, that a lock provided by the lock data structure is already held by the second thread, performing, by the first thread;

using a socket to block while waiting for the lock to be released;

adding a waiter entry to a queue, the waiter entry specifying an address of the socket on which the first thread is blocking; and

receiving a notification from the second thread, on the socket, when the second thread releases the lock.

9. The system according to claim 8, wherein the socket comprises one of a network socket, a cross-guest shared-memory socket, or communication technology that provides socket-like service between the first operating system and the second operating system.

10. The system according to Claim 8,

wherein the functions further comprise:

obtaining, by the second thread, the waiter entry from the queue when the second thread is releasing the lock;

using the address of the socket specified in the obtained waiter entry, by the second thread, to connect to the socket on which the first thread blocked; and

sending a message on the socket on which the first thread blocked, by the second thread, to notify the first thread when the second thread releases the lock.

11. The system according to claim 8, wherein the first thread uses a spinlock to suspend execution while waiting for the lock to be released, rather than using the socket to block, if the first thread encounters a socket error or observes a variable setting that indicates that the second thread encountered a socket failure.

12. A computer program product for protecting shared resources in a computing system, the computer program product comprising:

a computer readable storage medium having computer readable program code embodied therein, the computer readable program code configured for;

storing, in shared memory accessible to at least a first thread and a second thread which execute concurrently in the computing system, a lock data structure for serializing access to one of the shared resources, the first thread executing under control of a first operating system hosted by the computing system and the second thread executing under control of a second operating system hosted by the computing system; and

upon detecting, by the first thread, that a lock provided by the lock data structure is already held by the second thread, performing, by the first thread;

using a socket to block while waiting for the lock to be released;

adding a waiter entry to a queue, the waiter entry specifying an address of the socket on which the first thread is blocking; and

receiving a notification from the second thread, on the socket, when the second thread releases the lock.

13. The computer program product according to claim 12, wherein the socket comprises one of a network socket or a cross-guest shared-memory socket.

14. The computer program product according to Claim 12, wherein the computer readable program code is further configured for:

obtaining, by the second thread, the waiter entry from the queue when the second thread is releasing the lock;

using the address of the socket specified in the obtained waiter entry, by the second thread, to connect to the socket on which the first thread blocked; and

sending a message on the socket on which the first thread blocked, by the second thread, to notify the first thread when the second thread releases the lock.

15. The computer program product according to claim 12, wherein the first thread uses a spinlock to suspend execution while waiting for the lock to be released, rather than using the socket to block, if the first thread encounters a socket error or observes a variable setting that indicates that the second thread encountered a socket failure.

16. The computer program product according to claim 12, wherein the first operating system executes a first cross-guest locking service and the second operating system executes a second cross-guest locking service, the first cross-guest locking service and the second cross-guest locking service providing a programming interface through which threads executing on the first and second operating systems invoke operations for holding and releasing the lock on the shared resource.