diff mbox

Linus HEAD build break

Message ID 87y4ebyznc.fsf@saruman.tx.rr.com (mailing list archive)
State New, archived
Headers show

Commit Message

Felipe Balbi Nov. 6, 2015, 2:24 p.m. UTC
Hi again,

Felipe Balbi <balbi@ti.com> writes:
> Hi Marek,
>
> your commit af19161aaed7 ("ARM: dts: twl4030: Add iio properties for bci
> subnode") breaks build on current linus/master (which current sits in

this commit cannot be found in next. How come it's in linus/master ?

If it had been in next, we'd find out about it a lot sooner and avoid
the build regression. It seems author didn't build his own patch,
considering the error is clear. Here's a patch, please merge it or
provide a better version. Below has been build tested
(omap2plus_defconfig) and boot tested (am437x SK, am437x IDK, beaglebone
black) (yeah, I don't have boards which sport MADC around).

8<---------------------------------------------------------------

From 7af26f7123bad3e9c9dcfe9dfe9f4ad58e56f7cb Mon Sep 17 00:00:00 2001
From: Felipe Balbi <balbi@ti.com>
Date: Fri, 6 Nov 2015 08:14:45 -0600
Subject: [PATCH] ARM: dts: twl4030: fix Monitoring ADC label

commit af19161aaed7 ("ARM: dts: twl4030: Add iio
properties for bci subnode") breaks build of several
DTS files by referring to a label that doesn't
exist. Fix it.

Signed-off-by: Felipe Balbi <balbi@ti.com>
---
 arch/arm/boot/dts/twl4030.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Belisko Marek Nov. 6, 2015, 2:31 p.m. UTC | #1
Hi,

On Fri, Nov 6, 2015 at 3:24 PM, Felipe Balbi <balbi@ti.com> wrote:
>
> Hi again,
>
> Felipe Balbi <balbi@ti.com> writes:
>> Hi Marek,
>>
>> your commit af19161aaed7 ("ARM: dts: twl4030: Add iio properties for bci
>> subnode") breaks build on current linus/master (which current sits in
>
> this commit cannot be found in next. How come it's in linus/master ?
I did post fix but Tony seems doesn't merge it:
https://lkml.org/lkml/2015/10/13/816
>
> If it had been in next, we'd find out about it a lot sooner and avoid
> the build regression. It seems author didn't build his own patch,
> considering the error is clear. Here's a patch, please merge it or
> provide a better version. Below has been build tested
> (omap2plus_defconfig) and boot tested (am437x SK, am437x IDK, beaglebone
> black) (yeah, I don't have boards which sport MADC around).
>
> 8<---------------------------------------------------------------
>
> From 7af26f7123bad3e9c9dcfe9dfe9f4ad58e56f7cb Mon Sep 17 00:00:00 2001
> From: Felipe Balbi <balbi@ti.com>
> Date: Fri, 6 Nov 2015 08:14:45 -0600
> Subject: [PATCH] ARM: dts: twl4030: fix Monitoring ADC label
>
> commit af19161aaed7 ("ARM: dts: twl4030: Add iio
> properties for bci subnode") breaks build of several
> DTS files by referring to a label that doesn't
> exist. Fix it.
>
> Signed-off-by: Felipe Balbi <balbi@ti.com>
> ---
>  arch/arm/boot/dts/twl4030.dtsi | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/arch/arm/boot/dts/twl4030.dtsi b/arch/arm/boot/dts/twl4030.dtsi
> index 482b7aa37808..c8197f209efc 100644
> --- a/arch/arm/boot/dts/twl4030.dtsi
> +++ b/arch/arm/boot/dts/twl4030.dtsi
> @@ -22,7 +22,7 @@
>         charger: bci {
>                 compatible = "ti,twl4030-bci";
>                 interrupts = <9>, <2>;
> -               io-channels = <&twl4030_madc 11>;
> +               io-channels = <&twl_madc 11>;
>                 io-channel-name = "vac";
>                 bci3v1-supply = <&vusb3v1>;
>         };
> --
> 2.6.2
>
>
>
> --
> balbi

BR,

marek
Felipe Balbi Nov. 6, 2015, 2:40 p.m. UTC | #2
Hi,

Belisko Marek <marek.belisko@gmail.com> writes:
> Hi,
>
> On Fri, Nov 6, 2015 at 3:24 PM, Felipe Balbi <balbi@ti.com> wrote:
>>
>> Hi again,
>>
>> Felipe Balbi <balbi@ti.com> writes:
>>> Hi Marek,
>>>
>>> your commit af19161aaed7 ("ARM: dts: twl4030: Add iio properties for bci
>>> subnode") breaks build on current linus/master (which current sits in
>>
>> this commit cannot be found in next. How come it's in linus/master ?
> I did post fix but Tony seems doesn't merge it:
> https://lkml.org/lkml/2015/10/13/816

looking at that thread, I have no idea how come the old version was
merged in the first place. Tony was clear that it broke build and yet
this patch has made its way to mainline and it didn't even go through
linux-next, which makes the problem worse.

Now we have a bisection point where the tree (well, some DTS files)
won't even build. This is quite messy.
Tony Lindgren Nov. 6, 2015, 2:53 p.m. UTC | #3
* Felipe Balbi <balbi@ti.com> [151106 06:41]:
> 
> Hi,
> 
> Belisko Marek <marek.belisko@gmail.com> writes:
> > Hi,
> >
> > On Fri, Nov 6, 2015 at 3:24 PM, Felipe Balbi <balbi@ti.com> wrote:
> >>
> >> Hi again,
> >>
> >> Felipe Balbi <balbi@ti.com> writes:
> >>> Hi Marek,
> >>>
> >>> your commit af19161aaed7 ("ARM: dts: twl4030: Add iio properties for bci
> >>> subnode") breaks build on current linus/master (which current sits in
> >>
> >> this commit cannot be found in next. How come it's in linus/master ?
> > I did post fix but Tony seems doesn't merge it:
> > https://lkml.org/lkml/2015/10/13/816
> 
> looking at that thread, I have no idea how come the old version was
> merged in the first place. Tony was clear that it broke build and yet
> this patch has made its way to mainline and it didn't even go through
> linux-next, which makes the problem worse.
> 
> Now we have a bisection point where the tree (well, some DTS files)
> won't even build. This is quite messy.

Yes didn't I drop the patch and pointed out it breaks the build?

Guys, please stop doing this. Do not merge driver code that has
not been sitting in linux next at least a week. And for the dts
changes, please make sure you have proper acks.

And now we have yet another merge window where things unexpectedly
break during the merge window because of untested driver changes.

Please repost the fix and I'll ack it and you guys send a new pull
request to fix it ASAP explaining how it happened.

Regards,

Tony
Sebastian Reichel Nov. 6, 2015, 3:34 p.m. UTC | #4
Hi,

On Fri, Nov 06, 2015 at 06:53:52AM -0800, Tony Lindgren wrote:
> * Felipe Balbi <balbi@ti.com> [151106 06:41]:
> > Belisko Marek <marek.belisko@gmail.com> writes:
> > > On Fri, Nov 6, 2015 at 3:24 PM, Felipe Balbi <balbi@ti.com> wrote:
> > >> Felipe Balbi <balbi@ti.com> writes:
> > >>> your commit af19161aaed7 ("ARM: dts: twl4030: Add iio properties for bci
> > >>> subnode") breaks build on current linus/master (which current sits in
> > >>
> > >> this commit cannot be found in next. How come it's in linus/master ?
> > > I did post fix but Tony seems doesn't merge it:
> > > https://lkml.org/lkml/2015/10/13/816
> > 
> > looking at that thread, I have no idea how come the old version was
> > merged in the first place. Tony was clear that it broke build and yet
> > this patch has made its way to mainline and it didn't even go through
> > linux-next, which makes the problem worse.
> > 
> > Now we have a bisection point where the tree (well, some DTS files)
> > won't even build. This is quite messy.
> 
> Yes didn't I drop the patch and pointed out it breaks the build?
> 
> Guys, please stop doing this. Do not merge driver code that has
> not been sitting in linux next at least a week. And for the dts
> changes, please make sure you have proper acks.
> 
> And now we have yet another merge window where things unexpectedly
> break during the merge window because of untested driver changes.
> 
> Please repost the fix and I'll ack it and you guys send a new pull
> request to fix it ASAP explaining how it happened.

mh for some reason I accidently added it to my next branch. I'm not
sure why, since I did not intend to queue it at all. Sorry for the
mess.

I wonder why it has not been in next, though. Checking my next tree,
it should have been there since end of September.

-- Sebastian
Sebastian Reichel Nov. 6, 2015, 4:09 p.m. UTC | #5
Hi,

On Fri, Nov 06, 2015 at 04:34:12PM +0100, Sebastian Reichel wrote:
> On Fri, Nov 06, 2015 at 06:53:52AM -0800, Tony Lindgren wrote:
> > * Felipe Balbi <balbi@ti.com> [151106 06:41]:
> > > Belisko Marek <marek.belisko@gmail.com> writes:
> > > > On Fri, Nov 6, 2015 at 3:24 PM, Felipe Balbi <balbi@ti.com> wrote:
> > > >> Felipe Balbi <balbi@ti.com> writes:
> > > >>> your commit af19161aaed7 ("ARM: dts: twl4030: Add iio properties for bci
> > > >>> subnode") breaks build on current linus/master (which current sits in
> > > >>
> > > >> this commit cannot be found in next. How come it's in linus/master ?
> > > > I did post fix but Tony seems doesn't merge it:
> > > > https://lkml.org/lkml/2015/10/13/816
> > > 
> > > looking at that thread, I have no idea how come the old version was
> > > merged in the first place. Tony was clear that it broke build and yet
> > > this patch has made its way to mainline and it didn't even go through
> > > linux-next, which makes the problem worse.
> > > 
> > > Now we have a bisection point where the tree (well, some DTS files)
> > > won't even build. This is quite messy.
> > 
> > Yes didn't I drop the patch and pointed out it breaks the build?
> > 
> > Guys, please stop doing this. Do not merge driver code that has
> > not been sitting in linux next at least a week. And for the dts
> > changes, please make sure you have proper acks.
> > 
> > And now we have yet another merge window where things unexpectedly
> > break during the merge window because of untested driver changes.
> > 
> > Please repost the fix and I'll ack it and you guys send a new pull
> > request to fix it ASAP explaining how it happened.
> 
> mh for some reason I accidently added it to my next branch. I'm not
> sure why, since I did not intend to queue it at all. Sorry for the
> mess.
> 
> I wonder why it has not been in next, though. Checking my next tree,
> it should have been there since end of September.

I missed Stephen's messages about the build failure [0] and he
switched to using a checkout from 20150925, which is directly before
the patch. As a result this patch and all following patches have not
been tested in linux-next :(

I will send a revert to you guys in a few minutes, so that I can get
your ACK for sending it as pull request to Torvalds.

[0] https://lkml.org/lkml/2015/9/28/844

-- Sebastian
Tony Lindgren Nov. 6, 2015, 7:03 p.m. UTC | #6
* Sebastian Reichel <sre@kernel.org> [151106 08:09]:
> Hi,
> 
> On Fri, Nov 06, 2015 at 04:34:12PM +0100, Sebastian Reichel wrote:
> > On Fri, Nov 06, 2015 at 06:53:52AM -0800, Tony Lindgren wrote:
> > > * Felipe Balbi <balbi@ti.com> [151106 06:41]:
> > > > Belisko Marek <marek.belisko@gmail.com> writes:
> > > > > On Fri, Nov 6, 2015 at 3:24 PM, Felipe Balbi <balbi@ti.com> wrote:
> > > > >> Felipe Balbi <balbi@ti.com> writes:
> > > > >>> your commit af19161aaed7 ("ARM: dts: twl4030: Add iio properties for bci
> > > > >>> subnode") breaks build on current linus/master (which current sits in
> > > > >>
> > > > >> this commit cannot be found in next. How come it's in linus/master ?
> > > > > I did post fix but Tony seems doesn't merge it:
> > > > > https://lkml.org/lkml/2015/10/13/816
> > > > 
> > > > looking at that thread, I have no idea how come the old version was
> > > > merged in the first place. Tony was clear that it broke build and yet
> > > > this patch has made its way to mainline and it didn't even go through
> > > > linux-next, which makes the problem worse.
> > > > 
> > > > Now we have a bisection point where the tree (well, some DTS files)
> > > > won't even build. This is quite messy.
> > > 
> > > Yes didn't I drop the patch and pointed out it breaks the build?
> > > 
> > > Guys, please stop doing this. Do not merge driver code that has
> > > not been sitting in linux next at least a week. And for the dts
> > > changes, please make sure you have proper acks.
> > > 
> > > And now we have yet another merge window where things unexpectedly
> > > break during the merge window because of untested driver changes.
> > > 
> > > Please repost the fix and I'll ack it and you guys send a new pull
> > > request to fix it ASAP explaining how it happened.
> > 
> > mh for some reason I accidently added it to my next branch. I'm not
> > sure why, since I did not intend to queue it at all. Sorry for the
> > mess.
> > 
> > I wonder why it has not been in next, though. Checking my next tree,
> > it should have been there since end of September.
> 
> I missed Stephen's messages about the build failure [0] and he
> switched to using a checkout from 20150925, which is directly before
> the patch. As a result this patch and all following patches have not
> been tested in linux-next :(
> 
> I will send a revert to you guys in a few minutes, so that I can get
> your ACK for sending it as pull request to Torvalds.

Sounds good to me.

Thanks

Tony

> [0] https://lkml.org/lkml/2015/9/28/844
diff mbox

Patch

diff --git a/arch/arm/boot/dts/twl4030.dtsi b/arch/arm/boot/dts/twl4030.dtsi
index 482b7aa37808..c8197f209efc 100644
--- a/arch/arm/boot/dts/twl4030.dtsi
+++ b/arch/arm/boot/dts/twl4030.dtsi
@@ -22,7 +22,7 @@ 
 	charger: bci {
 		compatible = "ti,twl4030-bci";
 		interrupts = <9>, <2>;
-		io-channels = <&twl4030_madc 11>;
+		io-channels = <&twl_madc 11>;
 		io-channel-name = "vac";
 		bci3v1-supply = <&vusb3v1>;
 	};