Re: [PATCH 1/3] spi/qspi: Add memory mapped read support.

[Mark Brown <broonie@kernel.org>]

[-- Attachment #1.1: Type: text/plain, Size: 1532 bytes --]

On Thu, Oct 10, 2013 at 04:53:46PM -0700, Trent Piepho wrote:
> On Thu, Oct 10, 2013 at 3:10 AM, Mark Brown <broonie@kernel.org> wrote:

> > Yeah, though of course at the minute the implementation of that thread
> > is pretty much up to the individual drivers which isn't triumph - and
> > the quality of implementation does vary rather a lot. I'm currently
> > working on trying to factor more of this out, hopefully then it'll be
> > easier to push out improvements. It may be nice to be able to kick off
> > the first DMA transfer from within the caller for example.

> I did testing with the mxs driver, which uses transfer_one
> _message and the spi core queue pumping code.  For small messages the
> overhead of queuing work to the pump_messages queue and waiting for
> completion is rather more than the time the actual transfer takes.  Which
> makes using a kthread rather pointless.  Part of the problem could be the
> high context switch cost for ARMv5.

Indeed - that's what I'd expect to happen.  Sufficiently small PIO
transfers and DMA initiation should both be faster if done from the
calling context.  The thread starts to get more useful once you start
building up work (at which point it should reduce context switches,
especially if multiple threads are using the bus) and if the system is
very heavily loaded when you can raise the thread priority so it can
drive data in more quickly.  

That said it's also helpful in terms of making the code simpler to just
have the one data path.

[-- Attachment #1.2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]

[-- Attachment #2: Type: text/plain, Size: 144 bytes --]

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/