2006/1/28 patches to pxa2xx-spi in 2.6.25 tree

2006/1/28 patches to pxa2xx-spi in 2.6.25 tree

[Ned Forrester]

David:

I wonder if you have seen the two patches submitted by Eric Miao that 
appeared in the 2.6.25 git tree on 1/26/08 for drivers/spi/pxa2xx-spi.c 
and related files.  They do not appear to have been passed through 
spi-devel-general, which is why I did not see them before. They were 
signed off by Russell King and not by you.  These patches make 
significant changes in way pxa2xx_spi plugs into the kernel, and appear 
to require changes in the platform setup that will likely affect all 
users.  They also make drivers/spi/pxa2xx-spi.c dependent on 
arch/arm/mach-pxa/ssp.c.

Are these changes compatible with the grand scheme of things?  I don't 
have much visibility of the direction in which the kernel or the spi 
core is heading, so I can't tell.

Eric:

It seems there should have been a related patch to 
Documentation/spi/pxa2xx to change the examples for platform setup, 
along the lines of what your patch calls for in 
arch/arm/mach-pxa/lubbock.c, and to add information about struct ssp_device.

-- 
Ned Forrester                                       nforrester-/d+BM93fTQY@public.gmane.org
Oceanographic Systems Lab                                  508-289-2226
Applied Ocean Physics and Engineering Dept.
Woods Hole Oceanographic Institution          Woods Hole, MA 02543, USA
http://www.whoi.edu/sbl/liteSite.do?litesiteid=7212
http://www.whoi.edu/hpb/Site.do?id=1532
http://www.whoi.edu/page.do?pid=10079


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/

Re: 2006/1/28 patches to pxa2xx-spi in 2.6.25 tree

[Eric Miao]

Let me clarify:

Yes, I apologize I break the potential rules here for two reasons:

1. I was not aware of the existence of spi-devel-general mailinglist
at that time (I know this now :)

2. This patch is one of the series of patches refactoring PXA/ARM SSP
code, and the other patches are tightly PXA/ARM related, and David
reviewed part of the patch if not all.

And indeed yes, I will fire another patch correcting the doc, maybe
after another round of SSP code clean-up.

Any other concerns/issues about the driver, please make me CC'ed, so
I'll try fix them if Stephen isn't available.

Thanks
- eric

> -----Original Message-----
> From: Ned Forrester [mailto:nforrester-/d+BM93fTQY@public.gmane.org]
> Sent: Friday, February 15, 2008 5:46 AM
> To: David Brownell; stephen-nl6u4wocdmy51APUEpUfAkEOCMrvLtNR@public.gmane.org
> Cc: spi-devel-general-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org; Eric Miao
> Subject: 2006/1/28 patches to pxa2xx-spi in 2.6.25 tree
> 
> David:
> 
> I wonder if you have seen the two patches submitted by Eric Miao that
> appeared in the 2.6.25 git tree on 1/26/08 for drivers/spi/pxa2xx-spi.c
> and related files.  They do not appear to have been passed through
> spi-devel-general, which is why I did not see them before. They were
> signed off by Russell King and not by you.  These patches make
> significant changes in way pxa2xx_spi plugs into the kernel, and appear
> to require changes in the platform setup that will likely affect all
> users.  They also make drivers/spi/pxa2xx-spi.c dependent on
> arch/arm/mach-pxa/ssp.c.
> 
> Are these changes compatible with the grand scheme of things?  I don't
> have much visibility of the direction in which the kernel or the spi
> core is heading, so I can't tell.
> 
> Eric:
> 
> It seems there should have been a related patch to
> Documentation/spi/pxa2xx to change the examples for platform setup,
> along the lines of what your patch calls for in
> arch/arm/mach-pxa/lubbock.c, and to add information about struct ssp_device.
> 
> --
> Ned Forrester                                       nforrester-/d+BM93fTQY@public.gmane.org
> Oceanographic Systems Lab                                  508-289-2226
> Applied Ocean Physics and Engineering Dept.
> Woods Hole Oceanographic Institution          Woods Hole, MA 02543, USA
> http://www.whoi.edu/sbl/liteSite.do?litesiteid=7212
> http://www.whoi.edu/hpb/Site.do?id=1532
> http://www.whoi.edu/page.do?pid=10079

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/

Re: 2006/1/28 patches to pxa2xx-spi in 2.6.25 tree

[Ned Forrester]

Eric Miao wrote:
> Let me clarify:
> 
> Yes, I apologize I break the potential rules here for two reasons:
> 
> 1. I was not aware of the existence of spi-devel-general mailinglist
> at that time (I know this now :)
> 
> 2. This patch is one of the series of patches refactoring PXA/ARM SSP
> code, and the other patches are tightly PXA/ARM related, and David
> reviewed part of the patch if not all.

I guess change is the price of progress.

> And indeed yes, I will fire another patch correcting the doc, maybe
> after another round of SSP code clean-up.

Thanks.

> Any other concerns/issues about the driver, please make me CC'ed, so
> I'll try fix them if Stephen isn't available.

The only outstanding issue that I am aware of is a minor problem with 
the sequence of setting chip select that was discussed on 
spi-devel-general this week.  Stephen has not chimed in on that issue, 
so I will submit a patch for that myself, either tomorrow or early next 
week.  I will do this patch so that it won't break a heavily modified 
version of pxa2xx_spi.c that I am working on, which deals with chained 
transfers using DMA descriptors (don't expect to ever see it submitted 
for inclusion in the kernel, but I am trying to stay in sync, anyway).

I have examined the patches you submitted, so far, and I think they do 
not interfere with my version.  Someday I will be able to test, but I 
work from the distribution provided with the Gumstix computers, and they 
are currently at 2.6.21 (and I am still at 2.6.20), so it may be a year 
or more before I see 2.6.25.

-- 
Ned Forrester                                       nforrester-/d+BM93fTQY@public.gmane.org
Oceanographic Systems Lab                                  508-289-2226
Applied Ocean Physics and Engineering Dept.
Woods Hole Oceanographic Institution          Woods Hole, MA 02543, USA
http://www.whoi.edu/sbl/liteSite.do?litesiteid=7212
http://www.whoi.edu/hpb/Site.do?id=1532
http://www.whoi.edu/page.do?pid=10079


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/

Re: 2006/1/28 patches to pxa2xx-spi in 2.6.25 tree

[David Brownell]

On Thursday 14 February 2008, Eric Miao topquoted (sigh):

> 2. This patch is one of the series of patches refactoring PXA/ARM SSP
> code, and the other patches are tightly PXA/ARM related, and David
> reviewed part of the patch if not all.

I skimmed that bit.  Time to track all the details of the PXA overhaul
is lacking.


> And indeed yes, I will fire another patch correcting the doc, maybe
> after another round of SSP code clean-up.

That doc update should happen before 2.6.25-final ...


> > From: Ned Forrester [mailto:nforrester-/d+BM93fTQY@public.gmane.org]
> > 
> > I wonder if you have seen the two patches submitted by Eric Miao that
> > appeared in the 2.6.25 git tree on 1/26/08 for drivers/spi/pxa2xx-spi.c
> > and related files.  ...
> >
> > Are these changes compatible with the grand scheme of things?  I don't
> > have much visibility of the direction in which the kernel or the spi
> > core is heading, so I can't tell.

It's more "PXA overhaul" stuff.  The *SSP modules should become more
generally usable, not unlike what drivers/misc/atmel-ssc.c does, so
that they can easily be used for SPI, I2S, or whatever.  It'd probably
be good to have a dma-capable core, for example, that gets re-used.

- Dave


> > Eric:
> > 
> > It seems there should have been a related patch to
> > Documentation/spi/pxa2xx to change the examples for platform setup,
> > along the lines of what your patch calls for in
> > arch/arm/mach-pxa/lubbock.c, and to add information about struct ssp_device.
> > 

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/