[RFC,PATCH] RCUify single-thread case of clock_gettime()

From: Paul E. McKenney (paulmck_at_us.ibm.com)
Date: 10/31/05

  • Next message: Allen Martin: "RE: PCI-DMA: high address but no IOMMU - nForce4" 
    Date:	Mon, 31 Oct 2005 09:44:16 -0800
To: mingo@elte.hu

    Hello!

    The attached patch uses RCU to avoid the need to acquire tasklist_lock
    in the single-thread case of clock_gettime(). Still acquires tasklist_lock
    when asking for the time of a (potentially multithreaded) process.

    Experimental, has been touch-tested on x86 and POWER. Requires RCU on
    task_struct. Further more focused testing in progress.

    Thoughts? (Why? Some off-list users want to be able to monitor CPU
    consumption of specific threads. They need to do so quite frequently,
    so acquiring tasklist_lock is inappropriate.)

                                                            Thanx, Paul

    diff -urpNa -X dontdiff linux-2.6.14-rc5-rt2/kernel/posix-cpu-timers.c linux-2.6.14-rc5-rt2-RCUgettime/kernel/posix-cpu-timers.c
    --- linux-2.6.14-rc5-rt2/kernel/posix-cpu-timers.c 2005-10-22 14:45:56.000000000 -0700
    +++ linux-2.6.14-rc5-rt2-RCUgettime/kernel/posix-cpu-timers.c 2005-10-22 15:44:27.000000000 -0700
    @@ -302,7 +302,7 @@ int posix_cpu_clock_get(clockid_t which_
                      * should be able to see it.
                      */
                     struct task_struct *p;
    - read_lock(&tasklist_lock);
    + rcu_read_lock();
                     p = find_task_by_pid(pid);
                     if (p) {
                             if (CPUCLOCK_PERTHREAD(which_clock)) {
    @@ -311,11 +311,13 @@ int posix_cpu_clock_get(clockid_t which_
                                                                      p, &rtn);
                                     }
                             } else if (p->tgid == pid && p->signal) {
    + read_lock(&tasklist_lock);
                                     error = cpu_clock_sample_group(which_clock,
                                                                    p, &rtn);
    + read_unlock(&tasklist_lock);
                             }
                     }
    - read_unlock(&tasklist_lock);
    + rcu_read_unlock();
             }
     
             if (error)
    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

  • Next message: Allen Martin: "RE: PCI-DMA: high address but no IOMMU - nForce4"