[PATCH 00/02] sh: Panning fixes

[PATCH 00/02] sh: Panning fixes

[Phil Edworthy]

sh: Panning fixes

[PATCH 01/02] sh: Add wait for vsync
[PATCH 02/02] sh: Fix panning

These patches add wait for vsync and fix the panning
so that apps can just do ioctl(FBIOPAN_DISPLAY) followed
by ioctl(FBIO_WAITFORVSYNC) in order to ensure that the
pan has happened.

Signed-off-by: Phil Edworthy <phil.edworthy@renesas.com>
---

Re: [PATCH 00/02] sh: Panning fixes

[Magnus Damm]

Hi Phil,

On Thu, Feb 11, 2010 at 7:23 PM, Phil Edworthy
<Phil.Edworthy@renesas.com> wrote:
> sh: Panning fixes
>
> [PATCH 01/02] sh: Add wait for vsync
> [PATCH 02/02] sh: Fix panning
>
> These patches add wait for vsync and fix the panning
> so that apps can just do ioctl(FBIOPAN_DISPLAY) followed
> by ioctl(FBIO_WAITFORVSYNC) in order to ensure that the
> pan has happened.
>
> Signed-off-by: Phil Edworthy <phil.edworthy@renesas.com>
> ---

Thanks for your patches. Please make sure that the SYS panel case with
deferred io is working. Migo-R is a good board to test this on.

Cheers,

/ magnus

RE: [PATCH 00/02] sh: Panning fixes

[Phil Edworthy]

Hi Magnus,

> Thanks for your patches. Please make sure that the SYS panel case with
> deferred io is working. Migo-R is a good board to test this on.

I thought the Migo-R doesn't have vsync connected to the SH, so maybe this
patch won't show any difference. Is that your understanding?

Thanks
Phil

Re: [PATCH 00/02] sh: Panning fixes

[Magnus Damm]

Hi Phil,

On Fri, Feb 12, 2010 at 5:45 AM, Phil Edworthy
<Phil.Edworthy@renesas.com> wrote:
> Hi Magnus,
>
>> Thanks for your patches. Please make sure that the SYS panel case with
>> deferred io is working. Migo-R is a good board to test this on.
>
> I thought the Migo-R doesn't have vsync connected to the SH, so maybe this
> patch won't show any difference. Is that your understanding?

Something like that, I think the QVGA LCD panel on Migo-R does not
expose the VSYNC pin of the on-panel LCD controller so there is
nothing to connect to the sh7722.

It may be difficult to test your patch because of this, I'm not sure.
My main concern was just to double check that you've thought about
dealing with both the deferred io case and the "regular" non-deferred
io case. The sh7724 KFR2R09 board is using deferred io and I think
some upcoming boards may make use of that as well, so I'd prefer to
keep that case working. =)

Cheers,

/ magnus

RE: [PATCH 00/02] sh: Panning fixes

[Phil Edworthy]

Hi Magnus,

> >> Thanks for your patches. Please make sure that the SYS panel case with
> >> deferred io is working. Migo-R is a good board to test this on.
> >
> > I thought the Migo-R doesn't have vsync connected to the SH, so maybe
> this
> > patch won't show any difference. Is that your understanding?
> 
> Something like that, I think the QVGA LCD panel on Migo-R does not
> expose the VSYNC pin of the on-panel LCD controller so there is
> nothing to connect to the sh7722.
That's what I thought.


> It may be difficult to test your patch because of this, I'm not sure.
> My main concern was just to double check that you've thought about
> dealing with both the deferred io case and the "regular" non-deferred
> io case. The sh7724 KFR2R09 board is using deferred io and I think
> some upcoming boards may make use of that as well, so I'd prefer to
> keep that case working. =)

How do you want to proceed then? I can't say that I am familiar with
deferred io and I don't have a board I can test it on... Is there anyone you
know who can try this?

Thanks
Phil

Re: [PATCH 00/02] sh: Panning fixes

[Magnus Damm]

Hi Phil,

On Mon, Feb 15, 2010 at 5:41 PM, Phil Edworthy
<Phil.Edworthy@renesas.com> wrote:
>> >> Thanks for your patches. Please make sure that the SYS panel case with
>> >> deferred io is working. Migo-R is a good board to test this on.
>> >
>> > I thought the Migo-R doesn't have vsync connected to the SH, so maybe
>> this
>> > patch won't show any difference. Is that your understanding?
>>
>> Something like that, I think the QVGA LCD panel on Migo-R does not
>> expose the VSYNC pin of the on-panel LCD controller so there is
>> nothing to connect to the sh7722.
> That's what I thought.

So there will be tearing on Migo-R regardless how smart our software is...

>> It may be difficult to test your patch because of this, I'm not sure.
>> My main concern was just to double check that you've thought about
>> dealing with both the deferred io case and the "regular" non-deferred
>> io case. The sh7724 KFR2R09 board is using deferred io and I think
>> some upcoming boards may make use of that as well, so I'd prefer to
>> keep that case working. =)
>
> How do you want to proceed then? I can't say that I am familiar with
> deferred io and I don't have a board I can test it on... Is there anyone you
> know who can try this?

... but it should be possible to test on Migo-R even though the vsync
pin is missing.

Usually people who submit patches do the testing to make sure other
platforms won't break. I realize that it may be difficult for you if
you don't have any platform to test on.

I may be able to find a Migo-R board somewhere, and I'd be happy to
add this to my TODO list. Beware that my list is pretty long, so if
you're in a rush then I suggest you to test yourself or find someone
else to do it for you.

Cheers,

/ magnus

RE: [PATCH 00/02] sh: Panning fixes

[Phil Edworthy]

Hi Magnus,

> On Mon, Feb 15, 2010 at 5:41 PM, Phil Edworthy
> <Phil.Edworthy@renesas.com> wrote:
> >> >> Thanks for your patches. Please make sure that the SYS panel case
> with
> >> >> deferred io is working. Migo-R is a good board to test this on.
> >> >
> >> > I thought the Migo-R doesn't have vsync connected to the SH, so maybe
> >> this
> >> > patch won't show any difference. Is that your understanding?
> >>
> >> Something like that, I think the QVGA LCD panel on Migo-R does not
> >> expose the VSYNC pin of the on-panel LCD controller so there is
> >> nothing to connect to the sh7722.
> > That's what I thought.
> 
> So there will be tearing on Migo-R regardless how smart our software is...
Yes, there is nothing we can do about that.

 
> >> It may be difficult to test your patch because of this, I'm not sure.
> >> My main concern was just to double check that you've thought about
> >> dealing with both the deferred io case and the "regular" non-deferred
> >> io case. The sh7724 KFR2R09 board is using deferred io and I think
> >> some upcoming boards may make use of that as well, so I'd prefer to
> >> keep that case working. =)
> >
> > How do you want to proceed then? I can't say that I am familiar with
> > deferred io and I don't have a board I can test it on... Is there anyone
> you
> > know who can try this?
> 
> ... but it should be possible to test on Migo-R even though the vsync
> pin is missing.
> 
> Usually people who submit patches do the testing to make sure other
> platforms won't break. I realize that it may be difficult for you if
> you don't have any platform to test on.
Ah, I misunderstood... I can test on MigoR but the output will tear
regardless. Is there anything I should look for on MigoR to see if deferred
io is working or not?

Thanks
Phil

Re: [PATCH 00/02] sh: Panning fixes

[Magnus Damm]

Hey Phil,

>> Usually people who submit patches do the testing to make sure other
>> platforms won't break. I realize that it may be difficult for you if
>> you don't have any platform to test on.
> Ah, I misunderstood... I can test on MigoR but the output will tear
> regardless. Is there anything I should look for on MigoR to see if deferred
> io is working or not?

Nothing special, just make sure that the console comes up as usual and
that fsync is still working.

Cheers,

/ magnus

RE: [PATCH 00/02] sh: Panning fixes

[Phil Edworthy]

Hi Magnus,

> >> Usually people who submit patches do the testing to make sure other
> >> platforms won't break. I realize that it may be difficult for you if
> >> you don't have any platform to test on.
> > Ah, I misunderstood... I can test on MigoR but the output will tear
> > regardless. Is there anything I should look for on MigoR to see if
> deferred
> > io is working or not?
> 
> Nothing special, just make sure that the console comes up as usual and
> that fsync is still working.

Done, all appears to work as it did before the patch.

I just need to address Paul's comments now.

Thanks
Phil

RE: [PATCH 00/02] sh: Panning fixes

[Guennadi Liakhovetski]

Hi Phil

On Mon, 15 Feb 2010, Phil Edworthy wrote:

> Hi Magnus,
> 
> > >> Usually people who submit patches do the testing to make sure other
> > >> platforms won't break. I realize that it may be difficult for you if
> > >> you don't have any platform to test on.
> > > Ah, I misunderstood... I can test on MigoR but the output will tear
> > > regardless. Is there anything I should look for on MigoR to see if
> > deferred
> > > io is working or not?
> > 
> > Nothing special, just make sure that the console comes up as usual and
> > that fsync is still working.
> 
> Done, all appears to work as it did before the patch.

Have you also tested panning on Migo-R? I presume, its behaviour shouldn't 
change?

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

RE: [PATCH 00/02] sh: Panning fixes

[Phil Edworthy]

Hi Guennadi,

> > > >> Usually people who submit patches do the testing to make sure other
> > > >> platforms won't break. I realize that it may be difficult for you
> if
> > > >> you don't have any platform to test on.
> > > > Ah, I misunderstood... I can test on MigoR but the output will tear
> > > > regardless. Is there anything I should look for on MigoR to see if
> > > deferred
> > > > io is working or not?
> > >
> > > Nothing special, just make sure that the console comes up as usual and
> > > that fsync is still working.
> >
> > Done, all appears to work as it did before the patch.
> 
> Have you also tested panning on Migo-R? I presume, its behaviour shouldn't
> change?

Yes, I have checked that panning behaves correctly on MigoR.

Thanks
Phil