lwn.git/include/linux, branch v2.6.33-rt7

Merge branch 'linus' into rt/head

2010-02-25T09:59:56+00:00

Conflicts: Makefile Signed-off-by: Thomas Gleixner

perf_events: defer poll() wakeups to softirq

2010-02-24T16:35:13+00:00

Use timer softirq for wakeups on preempt_rt Normally pending work is work that cannot be done from NMI context, such as wakeups and disabling the counter. The pending work is a single linked list using atomic ops so that it functions from NMI context. Normally this is called from IRQ context through use of an self-IPI (x86) or upon enabling hard interrupts (powerpc). Architectures that do not implement perf_event_set_pending() nor call perf_event_do_pending() upon leaving NMI context will get a polling fallback from the timer softirq. However, in -rt we cannot do the wakeup from IRQ context because its a wait_queue wakup, which can be O(n), so defer all wakeups to the softirq fallback by creating a second pending list that's only processed from there. Signed-off-by: Peter Zijlstra Signed-off-by: Thomas Gleixner

sched: Fix taskstates in sched_switch and proc

2010-02-24T11:10:52+00:00

The sched_switch trace event displays erroneous character codes of task states, after a new task state was added in the scheduler code but omitted to add in the trace event code. Define character codes of task states individually. In addition, define task state descriptions needed in /proc at the same place. This will help to keep the task state bits, characters and descriptions in sync should they ever need to be changed again. Signed-off-by: Carsten Emde Signed-off-by: Thomas Gleixner

tracing: Fix the merge fallout from 31-rt to 33

2010-02-24T11:10:52+00:00

Apply upgrade changes from 2.6.31 to the histograms. Signed-off-by: Carsten Emde Signed-off-by: Thomas Gleixner

rwsem: Make inner lock raw

2010-02-24T11:10:52+00:00

There is no reason to convert the lock protecting rwsems (the ownerless variant) to a sleeping spinlock on -rt. Convert it to raw. Signed-off-by: Thomas Gleixner

semaphore: Make inner lock raw

2010-02-24T11:10:52+00:00

There is no reason to have the spin_lock protecting the semaphore being converted to a sleeping spinlock in -rt. Convert it to a raw_spinlock. That also solves lockdep complaining about the rt_mutex.wait_lock being not initialized. Signed-off-by: Thomas Gleixner

Merge branch 'master' of ssh://master.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6 into rt/head

2010-02-23T16:17:43+00:00

core: Revert the in_irq() in_softirq() modifications

2010-02-23T16:16:29+00:00

RT added a check check for running in an interrupt handler thread or a softirq thread to in_irq() and in_softirq(). Most of the users of these checks do not care, but RCU and perf_events end up making wrong decisions. Especially the check in rcu_read_unlock_special() /* Hardware IRQ handlers cannot block. */ if (in_irq()) { local_irq_restore(flags); return; } falls flat on his nose on -RT due to this. Revert the changes and simply check for hardirq_count() resp. softirq_count() as mainline does. I checked all the users and the only dubious one is dev_kfree_skb_any() but some hysteric research and talking to acme makes me sure that it has no weird side effects. Signed-off-by: Thomas Gleixner

Revert "block: improve queue_should_plug() by looking at IO depths"

2010-02-23T07:40:43+00:00

This reverts commit fb1e75389bd06fd5987e9cda1b4e0305c782f854. "Benjamin S." reports that the patch in question causes a big drop in sequential throughput for him, dropping from 200MB/sec down to only 70MB/sec. Needs to be investigated more fully, for now lets just revert the offending commit. Conflicts: include/linux/blkdev.h Signed-off-by: Jens Axboe

sched: Extend enqueue_task to allow head queueing

2010-02-21T19:20:10+00:00

The ability of enqueueing a task to the head of a SCHED_FIFO priority list is required to fix some violations of POSIX scheduling policy. Extend the related functions with a "head" argument. Signed-off-by: Thomas Gleixner