Re: [PATCH/RFC] spi: bitbang: add PM QoS support

[Guennadi Liakhovetski <g.liakhovetski-Mmb7MZpHnFY@public.gmane.org>]

On Thu, 15 Mar 2012, Grant Likely wrote:

> On Wed, 14 Mar 2012 22:04:25 +0100 (CET), Guennadi Liakhovetski <g.liakhovetski-Mmb7MZpHnFY@public.gmane.org> wrote:
> > This patch adds a PM QoS requirement to the spi-bitbang driver, preventing
> > the underlying SPI hardware driver to suspend for too long a time, as long
> > as there are transfers on the queue.
> > 
> > Signed-off-by: Guennadi Liakhovetski <g.liakhovetski-Mmb7MZpHnFY@public.gmane.org>
> 
> Shouldn't this be part of the core spi infrastructure?  Particularly since queuing
> is moving into the core.

Maybe, can you point me out to commit IDs / subject lines / archive 
threads?

Anyway, this patch has to be reworked and extended. OTOH, if we 
universally agree, that no runtime-suspending of SPI controllers should 
take place with active CSs, it might happen, that we don't need this PM 
QoS at all - client drivers tend to hold CS lines for essentially longer 
periods of time, than what the SPI subsystem can recognise as a 
low-latency transfer sequence. The only example, that I can think of, 
where PM QoS could be useful is, if we have to keep latency low while 
communicating with different SPI devices on different chip-selects (like 
multiple audio-codecs, where their adjustment has to happen as 
synchronously as possible... No, I have no idea what I'm talking 
about:-)).

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/

------------------------------------------------------------------------------
This SF email is sponsosred by:
Try Windows Azure free for 90 days Click Here 
http://p.sf.net/sfu/sfd2d-msazure