Date:   Fri, 10 Mar 2023 18:20:53 +0100
From:   Fabrice Gasnier <fabrice.gasnier@...s.st.com>
To:     Saravana Kannan <saravanak@...gle.com>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        "Rafael J. Wysocki" <rafael@...nel.org>,
        Heikki Krogerus <heikki.krogerus@...ux.intel.com>,
        Guenter Roeck <linux@...ck-us.net>,
        Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
        Daniel Scally <djrscally@...il.com>,
        Sakari Ailus <sakari.ailus@...ux.intel.com>,
        Len Brown <lenb@...nel.org>
CC:     Yongqin Liu <yongqin.liu@...aro.org>,
        Sumit Semwal <sumit.semwal@...aro.org>,
        Martin Kepplinger <martin.kepplinger@...i.sm>,
        Amelie Delaunay <amelie.delaunay@...s.st.com>,
        <kernel-team@...roid.com>, <linux-kernel@...r.kernel.org>,
        <linux-usb@...r.kernel.org>, <linux-acpi@...r.kernel.org>,
        Alexandre Torgue <alexandre.torgue@...s.st.com>,
        <linux-stm32@...md-mailman.stormreply.com>
Subject: Re: [PATCH v1 0/4] Remove use of fw_devlink_purge_absent_suppliers()

On 3/1/23 22:49, Saravana Kannan wrote:
> Yongqin, Martin, Amelie,
> 
> We recent refactor of fw_devlink that ends with commit fb42378dcc7f
> ("mtd: mtdpart: Don't create platform device that'll never probe"),
> fw_devlink is smarter and doesn't depend on compatible property. So, I
> don't think these calls are needed anymore. But I don't have these
> devices to test on and be sure and the hardware I use to test changes
> doesn't have this issue either.
> 
> Can you please test these changes on the hardware where you hit the
> issue to make sure things work as expected?


Hi Saravana,

Sorry for the late reply,
On behalf of Amelie, I did some testing on STM32MP15 DK2 board, on top
of commit fb42378dcc7f, and also with your series applied.
For reference, it's based on: arch/arm/boot/dts/stm32mp15xx-dkx.dtsi

I noticed some error messages on this board, since the 12 patch series,
around the I2C PMIC device links:

[    3.585514] i2c 1-0033: Failed to create device link with 1-0033
[    3.590115] i2c 1-0033: Failed to create device link with 1-0033
[    3.596278] i2c 1-0033: Failed to create device link with 1-0033
[    3.602188] i2c 1-0033: Failed to create device link with 1-0033
[    3.608165] i2c 1-0033: Failed to create device link with 1-0033
[    3.614278] i2c 1-0033: Failed to create device link with 1-0033
[    3.620256] i2c 1-0033: Failed to create device link with 1-0033
[    3.626253] i2c 1-0033: Failed to create device link with 1-0033
[    3.632252] i2c 1-0033: Failed to create device link with 1-0033
[    3.639001] stpmic1 1-0033: PMIC Chip Version: 0x10
[    3.645398] platform 5c002000.i2c:stpmic@33:regulators: Fixed
dependency cycle(s) with /soc/i2c@...0200
0/stpmic@...regulators/boost
[    3.655937] platform 5c002000.i2c:stpmic@33:regulators: Fixed
dependency cycle(s) with /soc/i2c@...0200
0/stpmic@...regulators/buck2
[    3.667824] platform 5c002000.i2c:stpmic@33:regulators: Fixed
dependency cycle(s) with /soc/i2c@...0200
0/stpmic@...regulators/buck4
[    3.719751] stpmic1-regulator 5c002000.i2c:stpmic@33:regulators:
Failed to create device link with 1-0033
[    3.728099] stpmic1-regulator 5c002000.i2c:stpmic@33:regulators:
Failed to create device link with 1-0033
[    3.737576] stpmic1-regulator 5c002000.i2c:stpmic@33:regulators:
Failed to create device link with 1-0033
[    3.747216] stpmic1-regulator 5c002000.i2c:stpmic@33:regulators:
Failed to create device link with 1-0033
[    3.756750] stpmic1-regulator 5c002000.i2c:stpmic@33:regulators:
Failed to create device link with 1-0033
[    3.766382] stpmic1-regulator 5c002000.i2c:stpmic@33:regulators:
Failed to create device link with 1-0033
[    3.775914] stpmic1-regulator 5c002000.i2c:stpmic@33:regulators:
Failed to create device link with 1-0033
[    3.785545] stpmic1-regulator 5c002000.i2c:stpmic@33:regulators:
Failed to create device link with 1-0033

Strangely some of the regulators seems to have "Fixed dependency", but
not all.

Regarding the typec stusb160x I noticed the message below. It seems
correct, right ?

[   15.962771] typec port0: Fixed dependency cycle(s) with
/soc/usb-otg@...00000/port/endpoint

But sometimes (lets say 1/5 times) during boot, when I have a cable
already plugged in, it looks like there's some race condition. The dwc2
driver reports some error logs in a loop, indefinitely, up to the
watchdog resets the platform :-(.

[   16.288458] dwc2 49000000.usb-otg: Mode Mismatch Interrupt: currently
in Host mode
[   16.288490] dwc2 49000000.usb-otg: Mode Mismatch Interrupt: currently
in Host mode
[   16.310429] dwc2 49000000.usb-otg: Mode Mismatch Interrupt: currently
in Host mode

It probably just points some already existing race condition here. Maybe
it isn't even linked to this patch. But I have no evidence at this
stage. I hope I can investigate further on this one, hopefully I can
free up some time for that.

Best Regards,
Fabrice

> 
> Yongqin, If you didn't have the context, this affected hikey960.
> 
> Greg,
> 
> Let's wait for some tests before we land these.
> 
> Thanks,
> Saravana
> 
> Cc: Yongqin Liu <yongqin.liu@...aro.org>
> Cc: Sumit Semwal <sumit.semwal@...aro.org>
> Cc: Martin Kepplinger <martin.kepplinger@...i.sm>
> Cc: Amelie Delaunay <amelie.delaunay@...s.st.com>
> 
> Saravana Kannan (4):
>   usb: typec: stusb160x: Remove use of
>     fw_devlink_purge_absent_suppliers()
>   usb: typec: tipd: Remove use of fw_devlink_purge_absent_suppliers()
>   usb: typec: tcpm: Remove use of fw_devlink_purge_absent_suppliers()
>   driver core: Delete fw_devlink_purge_absent_suppliers()
> 
>  drivers/base/core.c           | 16 ----------------
>  drivers/usb/typec/stusb160x.c |  9 ---------
>  drivers/usb/typec/tcpm/tcpm.c |  9 ---------
>  drivers/usb/typec/tipd/core.c |  9 ---------
>  include/linux/fwnode.h        |  1 -
>  5 files changed, 44 deletions(-)
> 

Powered by blists - more mailing lists