Re: [PATCH 0/8] Remove deprecated workqueue interface users

[Sebastian Reichel]

Hi,

On Mon, Aug 15, 2016 at 06:59:48PM -0400, Tejun Heo wrote:
> On Sat, Aug 13, 2016 at 09:41:21PM +0530, Bhaktipriya Shridhar wrote:
> > This patch set removes the instances of deprecated
> > create_singlethread_workqueues in drivers/power by making the appropriate
> > conversions.
> > 
> > Bhaktipriya Shridhar (8):
> >   power: abx500_chargalg: Remove deprecated create_singlethread_workqueue
> >   power: ab8500_btemp: Remove deprecated create_singlethread_workqueue
> >   power: pm2301_charger: Remove deprecated create_singlethread_workqueue
> >   power: intel_mid_battery: Remove deprecated create_singlethread_workqueue
> >   power: ab8500_charger: Remove deprecated create_singlethread_workqueue
> >   power: ipaq_micro_battery: Remove deprecated create_singlethread_workqueue
> >   power: ab8500_fg: Remove deprecated create_singlethread_workqueue
> >   power: ds2760_battery: Remove deprecated create_singlethread_workqueue
>
> Patches look good to me.  I'm a bit curious about WQ_MEM_RECLAIM
> part tho. I suppose the reasoning is that as the hardware in
> question is involved in battery management which may be time
> critical, WQ_MEM_RECLAIM is added to ensure (timely) forward
> progress under memory pressure, right? It'd be great if someone
> who's more familiar with these hardware can confirm whether this
> is actually necessary.

At least the charger monitoring is timing critical: Many charger
chips must be monitored (e.g. for over-temp. events). I think
all of them have a self-protection, or a watchdog, so they would
actually stop charging instead of overheating, but that may result
in system failure due to power loss. Battery monitors also include
some temperture safety checks. So without careful case-by-case
checking the WQ_MEM_RECLAIM should not be removed IMHO.

Anyways, I queued this patchset, since create_singlethread_workqueue()
did use WQ_MEM_RECLAIM, so removing that bit should be in another patch.

> Oh, it'd also be nice to put the target subsystem in the subject of
> the patchset - e.g. "[PATCH 0/8] power: Remove..."

Yes please. Also it seems like PATCH2-8 have wrong reference to
PATCH0/8 mail.

-- Sebastian

Attachment: signature.asc
Description: PGP signature