pxa2xx-flash again, please?

pxa2xx-flash again, please?

[Guennadi Liakhovetski]

David, everybody

Are there plans to integrate the pxa2xx-flash unification patch from 
Nicolas Pitre?

Thanks
Guennadi
---------------------------------
Guennadi Liakhovetski, Ph.D.
DSA Daten- und Systemtechnik GmbH
Pascalstr. 28
D-52076 Aachen
Germany

Re: pxa2xx-flash again, please?

[Jörn Engel]

On Thu, 8 June 2006 09:25:07 +0200, Guennadi Liakhovetski wrote:
> 
> Are there plans to integrate the pxa2xx-flash unification patch from 
> Nicolas Pitre?

Not by me, as I don't know anything about this patch.

[ Hint: a link to the patch or even inlining it might help. ]

Jörn

-- 
Good warriors cause others to come to them and do not go to others.
-- Sun Tzu

Re: pxa2xx-flash again, please?

[Guennadi Liakhovetski]

[-- Attachment #1: Type: TEXT/PLAIN, Size: 880 bytes --]

On Thu, 8 Jun 2006, Jörn Engel wrote:

> On Thu, 8 June 2006 09:25:07 +0200, Guennadi Liakhovetski wrote:
>>
>> Are there plans to integrate the pxa2xx-flash unification patch from
>> Nicolas Pitre?
>
> Not by me, as I don't know anything about this patch.
>
> [ Hint: a link to the patch or even inlining it might help. ]

Sorry. This patch has been committed to mtd CVS (I think so, as it 
appeared on mtd-cvs ML: 
http://lists.infradead.org/pipermail/linux-mtd-cvs/2005-December/005252.html) 
and I asked this question already a few weeks ago, so, I thought it might 
ring the bell... The problem is - the patch has been submitted last 
December still it is not in stock kernel nor in 2 mtd git-trees on 
kernel.org/git, so...

Thanks
Guennadi
---------------------------------
Guennadi Liakhovetski, Ph.D.
DSA Daten- und Systemtechnik GmbH
Pascalstr. 28
D-52076 Aachen
Germany

Re: pxa2xx-flash again, please?

[David Woodhouse]

On Thu, 2006-06-08 at 09:25 +0200, Guennadi Liakhovetski wrote:
> Are there plans to integrate the pxa2xx-flash unification patch from 
> Nicolas Pitre?

I have a distinct recollection of looking at this when I went through
the CVS tree, and leaving it. I don't actually remember _why_ I did
that. Perhaps I just wanted to leave it to Nico.

Nico?

-- 
dwmw2

Re: pxa2xx-flash again, please?

[Jörn Engel]

On Thu, 8 June 2006 09:59:10 +0200, Guennadi Liakhovetski wrote:
> 
> Sorry. This patch has been committed to mtd CVS (I think so, as it 
> appeared on mtd-cvs ML: 
> http://lists.infradead.org/pipermail/linux-mtd-cvs/2005-December/005252.html) 
> and I asked this question already a few weeks ago, so, I thought it might 
> ring the bell... The problem is - the patch has been submitted last 
> December still it is not in stock kernel nor in 2 mtd git-trees on 
> kernel.org/git, so...

Well, the cvs has basically been abandoned.  There was some effort to
pull the important patches from cvs into git, but as you noticed,
things were lost.

Not sure whether the patch makes sense or not.  Maybe Todd can comment
that.

Jörn

-- 
It does not matter how slowly you go, so long as you do not stop.
-- Confucius

Re: pxa2xx-flash again, please?

[Guennadi Liakhovetski]

[-- Attachment #1: Type: TEXT/PLAIN, Size: 987 bytes --]

On Thu, 8 Jun 2006, Jörn Engel wrote:

> On Thu, 8 June 2006 09:59:10 +0200, Guennadi Liakhovetski wrote:
>>
>> Sorry. This patch has been committed to mtd CVS (I think so, as it
>> appeared on mtd-cvs ML:
>> http://lists.infradead.org/pipermail/linux-mtd-cvs/2005-December/005252.html)
>> and I asked this question already a few weeks ago, so, I thought it might
>> ring the bell... The problem is - the patch has been submitted last
>> December still it is not in stock kernel nor in 2 mtd git-trees on
>> kernel.org/git, so...
>
> Well, the cvs has basically been abandoned.  There was some effort to
> pull the important patches from cvs into git, but as you noticed,
> things were lost.
>
> Not sure whether the patch makes sense or not.  Maybe Todd can comment
> that.

I think it does. It makes one file per (pxa) platform fewer.

Thanks
Guennadi
---------------------------------
Guennadi Liakhovetski, Ph.D.
DSA Daten- und Systemtechnik GmbH
Pascalstr. 28
D-52076 Aachen
Germany

Re: pxa2xx-flash again, please?

[Todd Poynor]

Guennadi Liakhovetski wrote:
> On Thu, 8 Jun 2006, Jörn Engel wrote:
> 
>> On Thu, 8 June 2006 09:59:10 +0200, Guennadi Liakhovetski wrote:
>>>
>>> Sorry. This patch has been committed to mtd CVS (I think so, as it
>>> appeared on mtd-cvs ML:
>>> http://lists.infradead.org/pipermail/linux-mtd-cvs/2005-December/005252.html) 
>>>
>>> and I asked this question already a few weeks ago, so, I thought it 
>>> might
>>> ring the bell... The problem is - the patch has been submitted last
>>> December still it is not in stock kernel nor in 2 mtd git-trees on
>>> kernel.org/git, so...
>>
>> Well, the cvs has basically been abandoned.  There was some effort to
>> pull the important patches from cvs into git, but as you noticed,
>> things were lost.
>>
>> Not sure whether the patch makes sense or not.  Maybe Todd can comment
>> that.
> 
> I think it does. It makes one file per (pxa) platform fewer.

When I last checked the patch was a good thing, collapsing two 
practically identical XScale PXA2xx mapping drivers into one and 
converting to platform devices.  It had previously been acked by PXA2xx 
and MTD guy Nicolas Pitre, and Russell King accepted the ARM-side 
changes, which are in kernel.org waiting to become useful.  I think the 
patch happened to hit a window in which it wasn't merged upstream and 
has sat in limbo since then.

(And then the driver is almost totally generic other than relying on the 
ARM-only (last I knew) asm/mach/flash.h stuff and needs pxa-specific 
cache invalidate functions, in case anybody's tackling those problems.)


-- 
Todd

Re: pxa2xx-flash again, please?

[Lennert Buytenhek]

On Thu, Jun 08, 2006 at 12:42:04PM -0700, Todd Poynor wrote:

> >>> Sorry. This patch has been committed to mtd CVS (I think so, as it
> >>> appeared on mtd-cvs ML:
> >>> http://lists.infradead.org/pipermail/linux-mtd-cvs/2005-December/005252.html) 
> >>>
> >>> and I asked this question already a few weeks ago, so, I thought it 
> >>> might
> >>> ring the bell... The problem is - the patch has been submitted last
> >>> December still it is not in stock kernel nor in 2 mtd git-trees on
> >>> kernel.org/git, so...
> >>
> >> Well, the cvs has basically been abandoned.  There was some effort to
> >> pull the important patches from cvs into git, but as you noticed,
> >> things were lost.
> >>
> >> Not sure whether the patch makes sense or not.  Maybe Todd can comment
> >> that.
> > 
> > I think it does. It makes one file per (pxa) platform fewer.
> 
> When I last checked the patch was a good thing, collapsing two 
> practically identical XScale PXA2xx mapping drivers into one and 
> converting to platform devices.

I'd also like to see the unified pxa2xx flash mapping driver go in, so
that it can be used by the LogicPD PXA270 board support code (merged in
2.6.17) too, instead of making Yet Another copy of mainstone-flash.c.


cheers,
Lennert

Re: pxa2xx-flash again, please?

[David Woodhouse]

On Fri, 2006-06-09 at 02:16 +0200, Lennert Buytenhek wrote:
> I'd also like to see the unified pxa2xx flash mapping driver go in, so
> that it can be used by the LogicPD PXA270 board support code (merged
> in 2.6.17) too, instead of making Yet Another copy of
> mainstone-flash.c.

Can't we use platform_devices and physmap for all of it?

-- 
dwmw2

Re: pxa2xx-flash again, please?

[Nicolas Pitre]

On Fri, 9 Jun 2006, David Woodhouse wrote:

> On Fri, 2006-06-09 at 02:16 +0200, Lennert Buytenhek wrote:
> > I'd also like to see the unified pxa2xx flash mapping driver go in, so
> > that it can be used by the LogicPD PXA270 board support code (merged
> > in 2.6.17) too, instead of making Yet Another copy of
> > mainstone-flash.c.
> 
> Can't we use platform_devices and physmap for all of it?

We probably could if it wasn't for the cached mapping.


Nicolas

Re: pxa2xx-flash again, please?

[Guennadi Liakhovetski]

On Fri, 9 Jun 2006, Lennert Buytenhek wrote:

> On Thu, Jun 08, 2006 at 12:42:04PM -0700, Todd Poynor wrote:
>
>>>>> Sorry. This patch has been committed to mtd CVS (I think so, as it
>>>>> appeared on mtd-cvs ML:
>>>>> http://lists.infradead.org/pipermail/linux-mtd-cvs/2005-December/005252.html)
>>>>>
>>>>> and I asked this question already a few weeks ago, so, I thought it
>>>>> might
>>>>> ring the bell... The problem is - the patch has been submitted last
>>>>> December still it is not in stock kernel nor in 2 mtd git-trees on
>>>>> kernel.org/git, so...
>>>>
>>>> Well, the cvs has basically been abandoned.  There was some effort to
>>>> pull the important patches from cvs into git, but as you noticed,
>>>> things were lost.
>>>>
>>>> Not sure whether the patch makes sense or not.  Maybe Todd can comment
>>>> that.
>>>
>>> I think it does. It makes one file per (pxa) platform fewer.
>>
>> When I last checked the patch was a good thing, collapsing two
>> practically identical XScale PXA2xx mapping drivers into one and
>> converting to platform devices.
>
> I'd also like to see the unified pxa2xx flash mapping driver go in, so
> that it can be used by the LogicPD PXA270 board support code (merged in
> 2.6.17) too, instead of making Yet Another copy of mainstone-flash.c.

This patch is still not upstream, as far as I can see. Can we please push 
it up?

Thanks
Guennadi
---------------------------------
Guennadi Liakhovetski, Ph.D.
DSA Daten- und Systemtechnik GmbH
Pascalstr. 28
D-52076 Aachen
Germany

Re: pxa2xx-flash again, please?

[Lennert Buytenhek]

On Tue, Nov 21, 2006 at 08:52:46AM +0100, Guennadi Liakhovetski wrote:

> >I'd also like to see the unified pxa2xx flash mapping driver go in, so
> >that it can be used by the LogicPD PXA270 board support code (merged in
> >2.6.17) too, instead of making Yet Another copy of mainstone-flash.c.
> 
> This patch is still not upstream, as far as I can see. Can we please push 
> it up?

We should probably just use physmap instead.  The only thing that
physmap lacks, compared to mainstone-flash (as far as I can see), is
->cached support, but that should be easy enough to add.