On Fri, 2010-07-09 at 15:57 -0700, Darren Hart wrote: > This version pulls in the bits mistakenly left in 3/4. > > > >From 9f8b4faac79518f98131464c2d21a1c64fb841d2 Mon Sep 17 00:00:00 2001 > From: Darren Hart <dvhltc@xxxxxxxxxx> > Date: Fri, 9 Jul 2010 16:44:47 -0400 > Subject: [PATCH 4/4 V2] futex: convert hash_bucket locks to raw_spinlock_t > > The requeue_pi mechanism introduced proxy locking of the rtmutex. This creates > a scenario where a task can wake-up, not knowing it has been enqueued on an > rtmutex. In order to detect this, the task would have to be able to take either > task->pi_blocked_on->lock->wait_lock and/or the hb->lock. Unfortunately, > without already holding one of these, the pi_blocked_on variable can change > from NULL to valid or from valid to NULL. Therefor, the task cannot be allowed > to take a sleeping lock after wakeup or it could end up trying to block on two > locks, the second overwriting a valid pi_blocked_on value. This obviously > breaks the pi mechanism. > > This patch increases latency, while running the ltp pthread_cond_many test > which Michal reported the bug with, I see double digit hrtimer latencies > (typically only on the first run after boo): > > kernel: hrtimer: interrupt took 75911 ns > > This might be addressed by changing the various loops in the futex code to be > incremental, probably at an additional throughput hit. The private hash_bucket > lists discussed in the past could reduce hb->lock contention in some scenarios. > It should be noted that pthread_cond_many is a rather pathological case. > > This also introduces problems for plists which want a spinlock_t rather > than a raw_spinlock_t. Any thoughts on how to address this? > > Signed-off-by: Darren Hart <dvhltc@xxxxxxxxxx> > Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx> > Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx> > Cc: Ingo Molnar <mingo@xxxxxxx> > Cc: Eric Dumazet <eric.dumazet@xxxxxxxxx> > Cc: John Kacur <jkacur@xxxxxxxxxx> Acked-by: Steven Rostedt <rostedt@xxxxxxxxxxx> -- Steve > Cc: Mike Galbraith <efault@xxxxxx> > --- > kernel/futex.c | 73 ++++++++++++++++++++++++++++++-------------------------- > 1 files changed, 39 insertions(+), 34 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
