Optimizing preemptible read-copy update for low-power usage by avoiding ...

(75) Inventors: Paul E. McKenney, Beaverton, OR (US); Joshua A. Triplett, Hillsboro, OR (US) (73) Assignee: International Business Machines Corporation, Armonk, NY (US) ( * ) Notice: Subject to any disclaimer, the term of this patent is extended or adjusted under 35 U.S.C. 154(b) by 888 days. This patent is subject to a terminal disclaimer. (21) Appl.No.: 12/061,819 (22) Filed: Apr. 3, 2008 (65) Prior Publication Data US 2009/0254764 A1 Oct. 8, 2009 (51) Int. Cl. G06F 1/26 (2006.01) (52) U.S. C1. ...................................... .. 713/300; 713/323 (58) Field of Classification Search ................ .. 713/300,

2006/ 0100996 A1 5/2006 McKenney 2006/0112121 A1 5/2006 McKenney et al. 2006/0117072 A1 6/2006 McKenney 2006/ 0123 100 A1 6/2006 McKenney 2006/0130061 A1 6/2006 McKenney 2006/0265373 A1 11/2006 McKenney et al. 2007/0083565 A1 4/2007 McKenney 2007/0101071 A1 5/2007 McKenney 2007/0226431 A1 9/2007 McKenney et al. 2007/ 0226440 A1 9/2007 McKenney et al. 2007/0266209 A1 11/2007 McKenney et al. 2008/0033952 A1 2/2008 McKenney et al. 2008/0040720 A1 2/2008 McKenney et al. OTHER PUBLICATIONS

J. Seigh, “RCU + SMR for preemptive kernel/user threads,” Linux Kernel Mailing List, May 9, 2005, 2 pages.

M. Michael, “Hazard Pointers: Safe Memory Reclamation for LockFree Objects,” IEEE Transactions on Parallel and Distributed Systems, Jun. 2004, vol. 15, No. 6, pp. 491-504.

D. Sarma et al., “Making RCU Safe for Deep Sub-Millisecond Response Realtime Applications,” 2004 USENIX (UseLinux track) Jun. 2004, 9 pages.

P. McKenney, “RCU vs. Locking Performance on Different CPUs,” 2004 Linux.conf.au, 2004, 18 pages.

(Continued)

Primary Examiner * Nitin Patel (74) Attorney, Agent, or Firm * Walter W. Duft

(57) ABSTRACT

A technique for 1oW-poWer detection of a grace period for deferring the destruction of a shared data element until preexisting references to the data element have been removed. A grace period processing action is implemented that requires a response from a processor that may be running a preemptible reader of said shared data element before further grace period processing can proceed. A poWer and reader status of the processor is also determined. Grace period processing may proceed despite the absence of a response from the processor if the poWer and reader status indicates that an actual response from the processor is unnecessary.

20 Claims, 7 Drawing Sheets

42

OTHER PUBLICATIONS

P. McKenney et al., “Scaling dcache With RCU,” Linux Journal, Jan. 1,2004, 12 pages.

P. McKenney et al., “Using RCU in the Linux 2.5 Kernel,” Linux Journal, Oct. 1, 2003, 11 pages.

P. McKenney et al.,“Read-Copy Update,” 2002 Ottawa Linux Symposium, Jul. 8, 2002, 28 pages.

H. Linder et al., “Scalability of the Directory Entry Cache,” 2002 Ottawa Linux Symposium, Jun. 26, 2002, pp. 289-300.

P. Mckenney et al., “Read-Copy Update,” 2001 Ottawa Linux symposium, Jul. 2001, 22 pages.

P. McKenney et al., “Read-Copy Update: Using Execution History to Solve Concurrency Problems,” PDCS, Oct. 1998, 11 pages.

S. Dietrich et al., “Evolution of Real-Time Linux,” 7th RTL Workshop, Nov. 17, 2005, 18 pages.

B. Gamsa, “Tornado: Maximizing Locality and Concurrency in a Shared Memory Multiprocessor Operating System,” 1999, 14 pages. Molnar et al., “Realtime and Linux,” 2005 Linux Kernel Summit, 8 pages.

H. Boehm, “The Space Cost of Lazy Reference Counting,” ACM SIGPLAN Notices, Proceedings of the 31st ACM SIGPLANSIGACT Symposium on Principles of Programming Languages, POPL ’04, vol. 39, Issue 1, Jan. 2004, p. 210-219.

M. Michael, “Scalable Lock-Free Dynamic Memory Allocation,” ACM SIGPLAN Notices, Proceedings of the ACM SIGPLAN 2004 Conference on Pro gramming Language Design and Implementation; PLDI ’04, vol. 39, Issue 6, Jun. 2004, p. 35-46.

D. Dice eta1., “Mostly Lock-Free Malloc,” ACM SIGPLAN Notices, Proceedings of the 3rd International Symposium on Memory Management, ISMM ’02, vol. 38, Issue 2 Supplement, Jun. 2002, p. 163-174.

J . Corbet, “Read-copy-update for realtime,” LWN.net, Sep. 26, 2006, 3 pages.

“Priority Inversion,” Wikipedia, Feb. 4, 2007, 3 pages.

McKenney, “Seven real-time Linux approaches (Part C)”, LinuxDevices.com, Jun. 7, 2005, 13 pages.

P. McKenney, “ RCU and CONFIGiPREEMPTiRT progress,” Linux Kernel Mailing List, May 5, 2005, 2 pages.

O. Nesterov, QRCU: ‘Quick’ SRCU Implementation, Linux Kernel Mailing List, Dec. 1, 2005, 3 pages.

T. Gleixner, “High Resolution Timers/Dynamic Ticks-V2”, LWN. net, Oct. 1, 2006, 6 pages.

P. Mckenney, “Sleepable RCU”, LWN.net, Oct. 9, 2006, 10 pages. O. Nesterov, “cpufreqitsc() as coreiinitcallisync”, LKML.org, Nov. 19, 2006, 2 pages.

P. McKenney, “Using RCU in the Linux 2.5 Kernel”, Kernel Korner, Oct. 1, 2003,11 pages.

A. Kleen, “How to do nothing efficiently or better laziness: No Idle tick on x86-64”, 2005, 16 pages.

P. McKenney, “Read-Copy Update Implementations”, 2001, 3 pages. Non-Final Office Action dated Apr. 1, 2011 from copending U.S. Appl. No. 12/178,752.

Amendment and Response to Office Action dated Jul. 1, 2011 from copending U.S. Appl. No. 12/178,572.