On Wed, 25 Oct 2006 19:12:01 -0700,
[email protected] wrote:
> Subject: drivers: wait for threaded probes between initcall levels
> From: Andrew Morton <[email protected]>
>
> The multithreaded-probing code has a problem: after one initcall level (eg,
> core_initcall) has been processed, we will then start processing the next
> level (postcore_initcall) while the kernel threads which are handling
> core_initcall are still executing. This breaks the guarantees which the
> layered initcalls previously gave us.
>
> IOW, we want to be multithreaded _within_ an initcall level, but not between
> different levels.
>
> Fix that up by causing the probing code to wait for all outstanding probes at
> one level to complete before we start processing the next level.
>
> Cc: Greg KH <[email protected]>
> Signed-off-by: Andrew Morton <[email protected]>
Makes a lot of sense. I guess we could also get rid of
driver_probe_done() in prepare_namespace() with this patch...
> +#ifdef CONFIG_PCI_MULTITHREAD_PROBE
> +static int __init wait_for_probes(void)
> +{
> + DEFINE_WAIT(wait);
> +
> + if (!atomic_read(&probe_count))
> + return 0;
> + printk(KERN_INFO "%s: waiting for %d threads\n", __FUNCTION__,
> + atomic_read(&probe_count));
> + while (atomic_read(&probe_count)) {
> + prepare_to_wait(&probe_waitqueue, &wait, TASK_UNINTERRUPTIBLE);
> + if (atomic_read(&probe_count))
> + schedule();
> + }
> + finish_wait(&probe_waitqueue, &wait);
> + return 0;
> +}
> +
> +core_initcall_sync(wait_for_probes);
> +postcore_initcall_sync(wait_for_probes);
> +arch_initcall_sync(wait_for_probes);
> +subsys_initcall_sync(wait_for_probes);
> +fs_initcall_sync(wait_for_probes);
> +device_initcall_sync(wait_for_probes);
> +late_initcall_sync(wait_for_probes);
> +#endif
...if we get rid of this #ifdef.
--
Cornelia Huck
Linux for zSeries Developer
Tel.: +49-7031-16-4837, Mail: [email protected]
On Thu, Oct 26, 2006 at 10:09:39AM +0200, Cornelia Huck wrote:
> On Wed, 25 Oct 2006 19:12:01 -0700,
> [email protected] wrote:
>
> > Subject: drivers: wait for threaded probes between initcall levels
> > From: Andrew Morton <[email protected]>
> >
> > The multithreaded-probing code has a problem: after one initcall level (eg,
> > core_initcall) has been processed, we will then start processing the next
> > level (postcore_initcall) while the kernel threads which are handling
> > core_initcall are still executing. This breaks the guarantees which the
> > layered initcalls previously gave us.
> >
> > IOW, we want to be multithreaded _within_ an initcall level, but not between
> > different levels.
> >
> > Fix that up by causing the probing code to wait for all outstanding probes at
> > one level to complete before we start processing the next level.
> >
> > Cc: Greg KH <[email protected]>
> > Signed-off-by: Andrew Morton <[email protected]>
>
> Makes a lot of sense. I guess we could also get rid of
> driver_probe_done() in prepare_namespace() with this patch...
>
>
> > +#ifdef CONFIG_PCI_MULTITHREAD_PROBE
> > +static int __init wait_for_probes(void)
> > +{
> > + DEFINE_WAIT(wait);
> > +
> > + if (!atomic_read(&probe_count))
> > + return 0;
> > + printk(KERN_INFO "%s: waiting for %d threads\n", __FUNCTION__,
> > + atomic_read(&probe_count));
> > + while (atomic_read(&probe_count)) {
> > + prepare_to_wait(&probe_waitqueue, &wait, TASK_UNINTERRUPTIBLE);
> > + if (atomic_read(&probe_count))
> > + schedule();
> > + }
> > + finish_wait(&probe_waitqueue, &wait);
> > + return 0;
> > +}
> > +
> > +core_initcall_sync(wait_for_probes);
> > +postcore_initcall_sync(wait_for_probes);
> > +arch_initcall_sync(wait_for_probes);
> > +subsys_initcall_sync(wait_for_probes);
> > +fs_initcall_sync(wait_for_probes);
> > +device_initcall_sync(wait_for_probes);
> > +late_initcall_sync(wait_for_probes);
> > +#endif
>
> ...if we get rid of this #ifdef.
Yeah, let me play with this a bit, along with your proposed change, I
think it can be cleaned up to be a little more cleaner.
thanks,
greg k-h