6.1.y Regression found on AM3517

6.1.y Regression found on AM3517

[Adam Ford]

I have an AM3517-EVM board that I am trying to use the latest 6.1.y
stable, currently 6.1.51.

With git bisect, I narrowed the regression between 6.1.15 and 6.1.16 to commit
eaf9b5612a47 ("driver core: fw_devlink: Don't purge child fwnode's
consumer links")

In the regression, 48002000.scm appears as not ready.  This affects a
variety of dependent peripherals making them unavailable:

wl12xx_buf platform: supplier 48002000.scm not ready
wl12xx_vmmc2 platform: supplier wl12xx_buf not ready
48050000.dss platform: supplier display@0 not ready
48064800.ehci platform: supplier hsusb1_phy not ready
backlight platform: supplier 48002000.scm not ready
display@0 platform: supplier backlight not ready
dmtimer-pwm@11 platform: supplier 48002000.scm not ready
hsusb1_phy platform: supplier 48002000.scm not ready
gpio-leds platform: supplier 48002000.scm not ready
480b4000.mmc platform: supplier wl12xx_vmmc2 not ready

If I build 6.1.51 but I checkout drivers/base/core.c from commit
2455b81afe68 ("driver core: fw_devlink: Add DL_FLAG_CYCLE support to
device links"),
the regression is gone.

I checked the 6.5 kernel, and it appears fine, so I think there is a
possible backport commit missing, and I was hoping Saravana or Tony
might have a suggestion as to which one(s) I should try.  I don't know
if this is found on other OMAP3 boards, but I wouldn't be surprised.

adam

Re: 6.1.y Regression found on AM3517

[Bagas Sanjaya]

[-- Attachment #1: Type: text/plain, Size: 1809 bytes --]

On Mon, Sep 04, 2023 at 11:37:46AM -0500, Adam Ford wrote:
> I have an AM3517-EVM board that I am trying to use the latest 6.1.y
> stable, currently 6.1.51.
> 
> With git bisect, I narrowed the regression between 6.1.15 and 6.1.16 to commit
> eaf9b5612a47 ("driver core: fw_devlink: Don't purge child fwnode's
> consumer links")
> 
> In the regression, 48002000.scm appears as not ready.  This affects a
> variety of dependent peripherals making them unavailable:
> 
> wl12xx_buf platform: supplier 48002000.scm not ready
> wl12xx_vmmc2 platform: supplier wl12xx_buf not ready
> 48050000.dss platform: supplier display@0 not ready
> 48064800.ehci platform: supplier hsusb1_phy not ready
> backlight platform: supplier 48002000.scm not ready
> display@0 platform: supplier backlight not ready
> dmtimer-pwm@11 platform: supplier 48002000.scm not ready
> hsusb1_phy platform: supplier 48002000.scm not ready
> gpio-leds platform: supplier 48002000.scm not ready
> 480b4000.mmc platform: supplier wl12xx_vmmc2 not ready
> 
> If I build 6.1.51 but I checkout drivers/base/core.c from commit
> 2455b81afe68 ("driver core: fw_devlink: Add DL_FLAG_CYCLE support to
> device links"),
> the regression is gone.
> 
> I checked the 6.5 kernel, and it appears fine, so I think there is a
> possible backport commit missing, and I was hoping Saravana or Tony
> might have a suggestion as to which one(s) I should try.  I don't know
> if this is found on other OMAP3 boards, but I wouldn't be surprised.
> 

Thanks for the regression report. I'm adding it to regzbot as stable-specific
entry:

#regzbot ^introduced: eaf9b5612a47f0
#regzbot title: keeping consumer links of child fwnode doesn't prepare AM3517-EVM suppliers

-- 
An old man doll... just what I always wanted! - Clara

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 228 bytes --]

Re: 6.1.y Regression found on AM3517

[Tony Lindgren]

* Bagas Sanjaya <bagasdotme@gmail.com> [230905 00:31]:
> On Mon, Sep 04, 2023 at 11:37:46AM -0500, Adam Ford wrote:
> > I have an AM3517-EVM board that I am trying to use the latest 6.1.y
> > stable, currently 6.1.51.
> > 
> > With git bisect, I narrowed the regression between 6.1.15 and 6.1.16 to commit
> > eaf9b5612a47 ("driver core: fw_devlink: Don't purge child fwnode's
> > consumer links")
> > 
> > In the regression, 48002000.scm appears as not ready.  This affects a
> > variety of dependent peripherals making them unavailable:
> > 
> > wl12xx_buf platform: supplier 48002000.scm not ready
> > wl12xx_vmmc2 platform: supplier wl12xx_buf not ready
> > 48050000.dss platform: supplier display@0 not ready
> > 48064800.ehci platform: supplier hsusb1_phy not ready
> > backlight platform: supplier 48002000.scm not ready
> > display@0 platform: supplier backlight not ready
> > dmtimer-pwm@11 platform: supplier 48002000.scm not ready
> > hsusb1_phy platform: supplier 48002000.scm not ready
> > gpio-leds platform: supplier 48002000.scm not ready
> > 480b4000.mmc platform: supplier wl12xx_vmmc2 not ready
> > 
> > If I build 6.1.51 but I checkout drivers/base/core.c from commit
> > 2455b81afe68 ("driver core: fw_devlink: Add DL_FLAG_CYCLE support to
> > device links"),
> > the regression is gone.

Adam, maybe check if 6.1 stable series is missing something to be
backported from this mailing list thread:

https://lore.kernel.org/lkml/20230207014207.1678715-2-saravanak@google.com/

That email thread version seemed to work for me based on my reply in the
thread. I recall issues with the earlier revisions but don't remember
what fixed them.

> > I checked the 6.5 kernel, and it appears fine, so I think there is a
> > possible backport commit missing, and I was hoping Saravana or Tony
> > might have a suggestion as to which one(s) I should try.  I don't know
> > if this is found on other OMAP3 boards, but I wouldn't be surprised.
> > 
> 
> Thanks for the regression report. I'm adding it to regzbot as stable-specific
> entry:
> 
> #regzbot ^introduced: eaf9b5612a47f0
> #regzbot title: keeping consumer links of child fwnode doesn't prepare AM3517-EVM suppliers

Regards,

Tony

Re: 6.1.y Regression found on AM3517

[Adam Ford]

On Tue, Sep 5, 2023 at 11:25 AM Tony Lindgren <tony@atomide.com> wrote:
>
> * Bagas Sanjaya <bagasdotme@gmail.com> [230905 00:31]:
> > On Mon, Sep 04, 2023 at 11:37:46AM -0500, Adam Ford wrote:
> > > I have an AM3517-EVM board that I am trying to use the latest 6.1.y
> > > stable, currently 6.1.51.
> > >
> > > With git bisect, I narrowed the regression between 6.1.15 and 6.1.16 to commit
> > > eaf9b5612a47 ("driver core: fw_devlink: Don't purge child fwnode's
> > > consumer links")
> > >
> > > In the regression, 48002000.scm appears as not ready.  This affects a
> > > variety of dependent peripherals making them unavailable:
> > >
> > > wl12xx_buf platform: supplier 48002000.scm not ready
> > > wl12xx_vmmc2 platform: supplier wl12xx_buf not ready
> > > 48050000.dss platform: supplier display@0 not ready
> > > 48064800.ehci platform: supplier hsusb1_phy not ready
> > > backlight platform: supplier 48002000.scm not ready
> > > display@0 platform: supplier backlight not ready
> > > dmtimer-pwm@11 platform: supplier 48002000.scm not ready
> > > hsusb1_phy platform: supplier 48002000.scm not ready
> > > gpio-leds platform: supplier 48002000.scm not ready
> > > 480b4000.mmc platform: supplier wl12xx_vmmc2 not ready
> > >
> > > If I build 6.1.51 but I checkout drivers/base/core.c from commit
> > > 2455b81afe68 ("driver core: fw_devlink: Add DL_FLAG_CYCLE support to
> > > device links"),
> > > the regression is gone.
>
> Adam, maybe check if 6.1 stable series is missing something to be
> backported from this mailing list thread:
>
> https://lore.kernel.org/lkml/20230207014207.1678715-2-saravanak@google.com/
>
> That email thread version seemed to work for me based on my reply in the
> thread. I recall issues with the earlier revisions but don't remember
> what fixed them.

Tony,

Indeed, it does appear that a patch was missing from the back-ported series.

Applying commit 4a032827daa8 ("of: property: Simplify
of_link_to_phandle()") to the 6.1.y branch appears
to fix the regression.

Thanks for the pointer to the series.

Stable group,

Do I need to send a separate e-mail requesting the backporting of that
commit to the 6.1.y branch?

adam
>
> > > I checked the 6.5 kernel, and it appears fine, so I think there is a
> > > possible backport commit missing, and I was hoping Saravana or Tony
> > > might have a suggestion as to which one(s) I should try.  I don't know
> > > if this is found on other OMAP3 boards, but I wouldn't be surprised.
> > >
> >
> > Thanks for the regression report. I'm adding it to regzbot as stable-specific
> > entry:
> >
> > #regzbot ^introduced: eaf9b5612a47f0
> > #regzbot title: keeping consumer links of child fwnode doesn't prepare AM3517-EVM suppliers
>
> Regards,
>
> Tony

Re: 6.1.y Regression found on AM3517

[Linux regression tracking #update (Thorsten Leemhuis)]

[TLDR: This mail in primarily relevant for Linux kernel regression
tracking. See link in footer if these mails annoy you.]

On 06.09.23 12:59, Adam Ford wrote:
> On Tue, Sep 5, 2023 at 11:25 AM Tony Lindgren <tony@atomide.com> wrote:
>> * Bagas Sanjaya <bagasdotme@gmail.com> [230905 00:31]:
>>> On Mon, Sep 04, 2023 at 11:37:46AM -0500, Adam Ford wrote:
>>>> I have an AM3517-EVM board that I am trying to use the latest 6.1.y
>>>> stable, currently 6.1.51.
>>>>
>>>> With git bisect, I narrowed the regression between 6.1.15 and 6.1.16 to commit
>>>> eaf9b5612a47 ("driver core: fw_devlink: Don't purge child fwnode's
>>>> consumer links")
>>>>
>>>> In the regression, 48002000.scm appears as not ready.  This affects a
>>>> variety of dependent peripherals making them unavailable:
> [...]
> Indeed, it does appear that a patch was missing from the back-ported series.
> 
> Applying commit 4a032827daa8 ("of: property: Simplify
> of_link_to_phandle()") to the 6.1.y branch appears
> to fix the regression.
> 
> Thanks for the pointer to the series.
> 
> Stable group,
> 
> Do I need to send a separate e-mail requesting the backporting of that
> commit to the 6.1.y branch?

Stable team picked this up, hence:

#regzbot fix: of: property: Simplify of_link_to_phandle()
#regzbot ignore-activity

Ciao, Thorsten (wearing his 'the Linux kernel's regression tracker' hat)
--
Everything you wanna know about Linux kernel regression tracking:
https://linux-regtracking.leemhuis.info/about/#tldr
That page also explains what to do if mails like this annoy you.