[lkml]   [2011]   [Jul]   [10]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
Subject[PATCH 0/6 v2] PM / Domains: Generic PM domains improvements

The majority of patches have been updated since the last submission and there's
one new patch, so it's the time for a refresh. :-)

All patches in this series have been posted already (the last one without
a changelog, though).

On Wednesday, July 06, 2011, Rafael J. Wysocki wrote:
> For the last few days I've been working on modifications of the generic
> PM domains code allowing .start_device()/.stop_device() and drivers'
> .runtime_suspend()/.runtime_resume() callbacks to execute things like
> pm_runtime_resume() for other devices in the same PM domain safely
> without deadlocking (the original motivation was a console driver on
> shmobile that pm_runtime_get_*()/pm_runtime_put_*() are called for from
> the other devices' .runtime_suspend()/.runtime_resume() callbacks).
> I think I solved that problem, but in the process I found a few places
> where fixes and/on improvements were necessary. The patches in this
> series address those issues.

[1/6] - Set device state to "active" during system resume.

[2/6] - Make failing pm_genpd_prepare() clean up properly.

[3/6] - Rework generic PM domains locking so that device callbacks are not
executed under the genpd lock.

[4/6] - Allow device callbacks executed by the generic PM domains code to
run runtime PM helpers safely (ie. without deadlocking).

[5/6] - Avoid restoring the states of all devices on failing domain power off.

[6/6] - Imorove the handling of system wakeup devices during system suspend.

The series is on top of the branch at:

git:// pm-domains

All patches in this series are regarded as 3.1 material.


 \ /
  Last update: 2011-07-10 11:13    [W:0.235 / U:12.780 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site