Re: What's inside the pxa tree for this merge window

Re: What's inside the pxa tree for this merge window

[Guennadi Liakhovetski]

On Fri, 18 Sep 2009, Russell King - ARM Linux wrote:

> On Fri, Sep 18, 2009 at 08:43:47AM +0200, Matej Kenda wrote:
> > On Thu, 2009-09-10 at 19:46 +0800, Eric Miao wrote:
> > > Here's the preview of request-pull result, please let me know if there
> > > is anything missed, thanks.
> > > 
> > > URL: git://git.kernel.org/pub/scm/linux/kernel/git/ycmiao/pxa-linux-2.6.git
> > > 
> > > 
> > > Matej Kenda (2):
> > >       [ARM] pxa: add support for the IskraTel XCEP board
> > >       [ARM] pxa: add defconfig for IskraTel XCEP board
> > 
> > Eric, what is the current status of this request? Do I need to repost
> > the patches to included into 2.6.32?
> > 
> > Russell discarded them from his patch system, because they are on your
> > list.
> 
> I discarded them _because_ Eric handled them, which is what I said in the
> comments when I discarded them.

Ok, I did do my best to get patches in the right order in the mainline, 
but it all failed. AFAICS, v4l and sh are already in the mainline with a 
_wrongly_ resolved mefge conflict, which, most likely, breaks the 
sh_mobile_ceu_camera.c driver, and the three PXA platforms, patches for 
which should have been applied before both those trees and still haven't 
been applied are broken until the patches do get in and the later those 
patches get applied the longer the interval with the broken for them 
bisection is going to be.

Thanks
Guennadi
---
Guennadi Liakhovetski

Re: What's inside the pxa tree for this merge window

[Russell King - ARM Linux]

On Mon, Sep 21, 2009 at 09:19:46PM +0200, Guennadi Liakhovetski wrote:
> On Fri, 18 Sep 2009, Russell King - ARM Linux wrote:
> 
> > On Fri, Sep 18, 2009 at 08:43:47AM +0200, Matej Kenda wrote:
> > > On Thu, 2009-09-10 at 19:46 +0800, Eric Miao wrote:
> > > > Here's the preview of request-pull result, please let me know if there
> > > > is anything missed, thanks.
> > > > 
> > > > URL: git://git.kernel.org/pub/scm/linux/kernel/git/ycmiao/pxa-linux-2.6.git
> > > > 
> > > > 
> > > > Matej Kenda (2):
> > > >       [ARM] pxa: add support for the IskraTel XCEP board
> > > >       [ARM] pxa: add defconfig for IskraTel XCEP board
> > > 
> > > Eric, what is the current status of this request? Do I need to repost
> > > the patches to included into 2.6.32?
> > > 
> > > Russell discarded them from his patch system, because they are on your
> > > list.
> > 
> > I discarded them _because_ Eric handled them, which is what I said in the
> > comments when I discarded them.
> 
> Ok, I did do my best to get patches in the right order in the mainline, 
> but it all failed. AFAICS, v4l and sh are already in the mainline with a 
> _wrongly_ resolved mefge conflict, which, most likely, breaks the 
> sh_mobile_ceu_camera.c driver, and the three PXA platforms, patches for 
> which should have been applied before both those trees and still haven't 
> been applied are broken until the patches do get in and the later those 
> patches get applied the longer the interval with the broken for them 
> bisection is going to be.

Meanwhile I have to consider that we have several bug fixes outstanding,
and since I can't send Linus a pull request every day (max once a week)
I have to be very careful about when I send stuff.

So I only get _two_ opportunities during a merge window to send a pull
request.

I'm going to wait until tomorrow before sending my final pull for this
window, which is the penultimate day before the window closes.

Don't blame me for these delays - it's not my choice to impose such
delays.  I'd really like to fix those broken platforms right now.  I
just can't do so without causing additional delays for other issues.
Blame Linus for imposing the "max one pull a week" rule on me.

Re: What's inside the pxa tree for this merge window

[Jamie Lokier]

Russell King - ARM Linux wrote:
> Don't blame me for these delays - it's not my choice to impose such
> delays.  I'd really like to fix those broken platforms right now.  I
> just can't do so without causing additional delays for other issues.
> Blame Linus for imposing the "max one pull a week" rule on me.

Fwiw,I'm thinking the clrex-on-exception-return fix, at least, should
go to stable@kernel.org too if it's proving to be reliable, as it
affects userspace correctness on ARMv6+.

-- Jamie

arm tree in broken state (was Re: What's inside the pxa tree for this merge window)

[Pavel Machek]

> > > I discarded them _because_ Eric handled them, which is what I said in the
> > > comments when I discarded them.
> > 
> > Ok, I did do my best to get patches in the right order in the mainline, 
> > but it all failed. AFAICS, v4l and sh are already in the mainline with a 
> > _wrongly_ resolved mefge conflict, which, most likely, breaks the 
> > sh_mobile_ceu_camera.c driver, and the three PXA platforms, patches for 
> > which should have been applied before both those trees and still haven't 
> > been applied are broken until the patches do get in and the later those 
> > patches get applied the longer the interval with the broken for them 
> > bisection is going to be.
> 
> Meanwhile I have to consider that we have several bug fixes outstanding,
> and since I can't send Linus a pull request every day (max once a week)
> I have to be very careful about when I send stuff.
> 
> So I only get _two_ opportunities during a merge window to send a pull
> request.

Well, you should certainly try to keep your tree  unbroken, but when
it breaks, fixing it asap should be a priority. I don't know where you
got the 'once a week' rule, but it seems stupid.

> I'm going to wait until tomorrow before sending my final pull for this
> window, which is the penultimate day before the window closes.
> 
> Don't blame me for these delays - it's not my choice to impose such
> delays.  I'd really like to fix those broken platforms right now.  I
> just can't do so without causing additional delays for other issues.
> Blame Linus for imposing the "max one pull a week" rule on me.

Do you have maillist reference? Not even Linus should slow down
development like that.

If Linus really insists on that, perhaps possible solution would be to
make subarch maintainers send pull requests for simple fixes directly
to Linus?

								Pavel


-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

Re: arm tree in broken state (was Re: What's inside the pxa tree for this merge window)

[Guennadi Liakhovetski]

Hi Pavel

On Wed, 30 Sep 2009, Pavel Machek wrote:

> > > > I discarded them _because_ Eric handled them, which is what I said in the
> > > > comments when I discarded them.
> > > 
> > > Ok, I did do my best to get patches in the right order in the mainline, 
> > > but it all failed. AFAICS, v4l and sh are already in the mainline with a 
> > > _wrongly_ resolved mefge conflict, which, most likely, breaks the 
> > > sh_mobile_ceu_camera.c driver, and the three PXA platforms, patches for 
> > > which should have been applied before both those trees and still haven't 
> > > been applied are broken until the patches do get in and the later those 
> > > patches get applied the longer the interval with the broken for them 
> > > bisection is going to be.
> > 
> > Meanwhile I have to consider that we have several bug fixes outstanding,
> > and since I can't send Linus a pull request every day (max once a week)
> > I have to be very careful about when I send stuff.
> > 
> > So I only get _two_ opportunities during a merge window to send a pull
> > request.
> 
> Well, you should certainly try to keep your tree  unbroken, but when
> it breaks, fixing it asap should be a priority. I don't know where you
> got the 'once a week' rule, but it seems stupid.
> 
> > I'm going to wait until tomorrow before sending my final pull for this
> > window, which is the penultimate day before the window closes.
> > 
> > Don't blame me for these delays - it's not my choice to impose such
> > delays.  I'd really like to fix those broken platforms right now.  I
> > just can't do so without causing additional delays for other issues.
> > Blame Linus for imposing the "max one pull a week" rule on me.
> 
> Do you have maillist reference? Not even Linus should slow down
> development like that.
> 
> If Linus really insists on that, perhaps possible solution would be to
> make subarch maintainers send pull requests for simple fixes directly
> to Linus?

Thanks for your concern, but the patches are long in mainline, no reason 
to worry any more.

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