[U-Boot] Olimex-A20-EVB ethernet stops working with latest U-boot, works with tag v2016.01

[U-Boot] Olimex-A20-EVB ethernet stops working with latest U-boot, works with tag v2016.01

[Ing. Damiano Bolla]

Good morning all

I updated to latest u-boot yesterday (git pull) and ethernet was not working

reverted to tag v2016.01 with the command

$ git checkout tags/v2016.01

rebuilt, uploaded and ethernet works

As far as I can tell the only difference on the two tests (just a simple ping 192.168.1.1) is the u-boot version
Ethernet cable was not moved and I manage to load u-boot using FEL

I can do further tests is any of you is interested.

Thanks, Damiano

[U-Boot] Olimex-A20-EVB ethernet stops working with latest U-boot, works with tag v2016.01

[Peng Fan]

Hi Damiano,

Can you try revert the following patch and retest on your board?
"
commit a058052c358c3ecf5f394ff37def6a45eb26768c
Author: Stefan Agner <stefan@agner.ch>
Date:   Wed Dec 9 11:21:25 2015 -0800

    net: phy: do not read configuration register on reset
        
    When doing a software reset, the reset flag should be written without
    other bits set. Writing the current state will lead to restoring the
    state of the PHY (e.g. Powerdown), which is not what is expected from
    a software reset.
			    
    Signed-off-by: Stefan Agner <stefan@agner.ch>
"

This patch breaks ethernet function of i.MX6 Quad Sabresd/Quad Sabreauto.
Not sure whether reverting this patch can help in your case or not. Please
try.

Thanks,
Peng.

On Wed, Feb 24, 2016 at 09:12:02AM +0100, Ing. Damiano Bolla wrote:
>Good morning all
>
>I updated to latest u-boot yesterday (git pull) and ethernet was not working
>
>reverted to tag v2016.01 with the command
>
>$ git checkout tags/v2016.01
>
>rebuilt, uploaded and ethernet works
>
>As far as I can tell the only difference on the two tests (just a simple ping 192.168.1.1) is the u-boot version
>Ethernet cable was not moved and I manage to load u-boot using FEL
>
>I can do further tests is any of you is interested.
>
>Thanks, Damiano
>_______________________________________________
>U-Boot mailing list
>U-Boot at lists.denx.de
>http://lists.denx.de/mailman/listinfo/u-boot

[U-Boot] Olimex-A20-EVB ethernet stops working with latest U-boot, works with tag v2016.01

[Ing. Damiano Bolla]

Good morning Peng

I did revert the commit, see attached image showing git repository status

Unfortunately the result is still negative, see the following log

U-Boot 2016.03-rc2-00117-gde889c0 (Feb 25 2016 - 08:12:50 +0100) Allwinner Technology
CPU:   Allwinner A20 (SUN7I)
I2C:   ready
DRAM:  1 GiB
MMC:   SUNXI SD/MMC: 0, SUNXI SD/MMC: 1
In:    serial
Out:   serial
Err:   serial
SCSI:  SUNXI SCSI INIT
SATA link 0 timeout.
AHCI 0001.0100 32 slots 1 ports 3 Gbps 0x1 impl SATA mode
flags: ncq stag pm led clo only pmp pio slum part ccc apst
Net:   eth0: ethernet at 01c50000
starting USB...
USB0:   USB EHCI 1.00
USB1:   USB OHCI 1.0
USB2:   USB EHCI 0.00
USB3:   USB OHCI 0.0
scanning bus 0 for devices... 1 USB Device(s) found
scanning bus 2 for devices... 1 USB Device(s) found
Hit any key to stop autoboot:  0
ethernet at 01c50000 Waiting for PHY auto negotiation to complete......... TIMEOUT !
ethernet at 01c50000: No link.
ethernet at 01c50000 Waiting for PHY auto negotiation to complete......... TIMEOUT !
ethernet at 01c50000: No link.
Bad Linux ARM zImage magic!

Can I assist further ?

Thanks, Damiano


In data 25 febbraio 2016 alle ore 02:11:00, Peng Fan <van.freenix@gmail.com> ha scritto:

> Hi Damiano,
>
> Can you try revert the following patch and retest on your board?
> "
> commit a058052c358c3ecf5f394ff37def6a45eb26768c
> Author: Stefan Agner <stefan@agner.ch>
> Date:   Wed Dec 9 11:21:25 2015 -0800
>
>     net: phy: do not read configuration register on reset
>    When doing a software reset, the reset flag should be written without
>     other bits set. Writing the current state will lead to restoring the
>     state of the PHY (e.g. Powerdown), which is not what is expected from
>     a software reset.
> 			
>     Signed-off-by: Stefan Agner <stefan@agner.ch>
> "
>
> This patch breaks ethernet function of i.MX6 Quad Sabresd/Quad Sabreauto.
> Not sure whether reverting this patch can help in your case or not. Please
> try.
>
> Thanks,
> Peng.
>
> On Wed, Feb 24, 2016 at 09:12:02AM +0100, Ing. Damiano Bolla wrote:
>> Good morning all
>>
>> I updated to latest u-boot yesterday (git pull) and ethernet was not working
>>
>> reverted to tag v2016.01 with the command
>>
>> $ git checkout tags/v2016.01
>>
>> rebuilt, uploaded and ethernet works
>>
>> As far as I can tell the only difference on the two tests (just a simple ping 192.168.1.1) is the u-boot version
>> Ethernet cable was not moved and I manage to load u-boot using FEL
>>
>> I can do further tests is any of you is interested.
>>
>> Thanks, Damiano
>> _______________________________________________
>> U-Boot mailing list
>> U-Boot at lists.denx.de
>> http://lists.denx.de/mailman/listinfo/u-boot
>


-- 
Ing. Damiano Bolla, 3486947264
-------------- next part --------------
A non-text attachment was scrubbed...
Name: git-status-after-revert.png
Type: image/png
Size: 143879 bytes
Desc: not available
URL: <http://lists.denx.de/pipermail/u-boot/attachments/20160225/fc9d12b8/attachment.png>

[U-Boot] Olimex-A20-EVB ethernet stops working with latest U-boot, works with tag v2016.01

[Hans de Goede]

Hi,

On 25-02-16 08:31, Ing. Damiano Bolla wrote:
> Good morning Peng
>
> I did revert the commit, see attached image showing git repository status
>
> Unfortunately the result is still negative, see the following log
>
> U-Boot 2016.03-rc2-00117-gde889c0 (Feb 25 2016 - 08:12:50 +0100) Allwinner Technology
> CPU:   Allwinner A20 (SUN7I)
> I2C:   ready
> DRAM:  1 GiB
> MMC:   SUNXI SD/MMC: 0, SUNXI SD/MMC: 1
> In:    serial
> Out:   serial
> Err:   serial
> SCSI:  SUNXI SCSI INIT
> SATA link 0 timeout.
> AHCI 0001.0100 32 slots 1 ports 3 Gbps 0x1 impl SATA mode
> flags: ncq stag pm led clo only pmp pio slum part ccc apst
> Net:   eth0: ethernet at 01c50000
> starting USB...
> USB0:   USB EHCI 1.00
> USB1:   USB OHCI 1.0
> USB2:   USB EHCI 0.00
> USB3:   USB OHCI 0.0
> scanning bus 0 for devices... 1 USB Device(s) found
> scanning bus 2 for devices... 1 USB Device(s) found
> Hit any key to stop autoboot:  0
> ethernet at 01c50000 Waiting for PHY auto negotiation to complete......... TIMEOUT !
> ethernet at 01c50000: No link.
> ethernet at 01c50000 Waiting for PHY auto negotiation to complete......... TIMEOUT !
> ethernet at 01c50000: No link.
> Bad Linux ARM zImage magic!
>
> Can I assist further ?

Can you please git bisect this, and find out which commit breaks things ?

Thanks & Regards,

Hans

>
> Thanks, Damiano
>
>
> In data 25 febbraio 2016 alle ore 02:11:00, Peng Fan <van.freenix@gmail.com> ha scritto:
>
>> Hi Damiano,
>>
>> Can you try revert the following patch and retest on your board?
>> "
>> commit a058052c358c3ecf5f394ff37def6a45eb26768c
>> Author: Stefan Agner <stefan@agner.ch>
>> Date:   Wed Dec 9 11:21:25 2015 -0800
>>
>>     net: phy: do not read configuration register on reset
>>    When doing a software reset, the reset flag should be written without
>>     other bits set. Writing the current state will lead to restoring the
>>     state of the PHY (e.g. Powerdown), which is not what is expected from
>>     a software reset.
>>
>>     Signed-off-by: Stefan Agner <stefan@agner.ch>
>> "
>>
>> This patch breaks ethernet function of i.MX6 Quad Sabresd/Quad Sabreauto.
>> Not sure whether reverting this patch can help in your case or not. Please
>> try.
>>
>> Thanks,
>> Peng.
>>
>> On Wed, Feb 24, 2016 at 09:12:02AM +0100, Ing. Damiano Bolla wrote:
>>> Good morning all
>>>
>>> I updated to latest u-boot yesterday (git pull) and ethernet was not working
>>>
>>> reverted to tag v2016.01 with the command
>>>
>>> $ git checkout tags/v2016.01
>>>
>>> rebuilt, uploaded and ethernet works
>>>
>>> As far as I can tell the only difference on the two tests (just a simple ping 192.168.1.1) is the u-boot version
>>> Ethernet cable was not moved and I manage to load u-boot using FEL
>>>
>>> I can do further tests is any of you is interested.
>>>
>>> Thanks, Damiano
>>> _______________________________________________
>>> U-Boot mailing list
>>> U-Boot at lists.denx.de
>>> http://lists.denx.de/mailman/listinfo/u-boot
>>
>
>
>
>
> _______________________________________________
> U-Boot mailing list
> U-Boot at lists.denx.de
> http://lists.denx.de/mailman/listinfo/u-boot
>

[U-Boot] Olimex-A20-EVB ethernet stops working with latest U-boot, works with tag v2016.01

[Simon Glass]

Hi,

On 9 March 2016 at 14:04, Karsten Merker <merker@debian.org> wrote:
> [Olimex A20-SOM-EVB: ethernet non-functional with current u-boot
>                      git head, but no problems with v2016.01]
>
> On Tue, Mar 08, 2016 at 11:54:03AM +0100, Hans de Goede wrote:
>> On 25-02-16 08:31, Ing. Damiano Bolla wrote:
>> >Can I assist further ?
>>
>> Can you please git bisect this, and find out which commit breaks things ?
>
> Hello Hans, hello Simon,
>
> I am not the original poster, but I have bisected the issue down
> to the following commit:
>
> commit c32a6fd07b1839e4a45729587ebc8e1c55601a4d
> Author: Simon Glass <sjg@chromium.org>
> Date:   Sun Jan 17 14:51:56 2016 -0700
>
>     net: Don't call board/cpu_eth_init() with driver model
>
>     We should avoid weak functions with driver model. Existing boards that use
>     driver model don't need them, so let's kill them off.
>
>     Signed-off-by: Simon Glass <sjg@chromium.org>
>     Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
>     Acked-by: Joe Hershberger <joe.hershberger@ni.com>

Is this board-specific init, or sunxi-specific?

Can you just call the init on start-up, instead of from the Ethernet driver?

Regards,
Simon

[U-Boot] Olimex-A20-EVB ethernet stops working with latest U-boot, works with tag v2016.01

[Hans de Goede]

Hi,

On 13-03-16 12:14, Karsten Merker wrote:
> On Wed, Mar 09, 2016 at 02:39:11PM -0700, Simon Glass wrote:
>> On 9 March 2016 at 14:04, Karsten Merker <merker@debian.org> wrote:
>>> [Olimex A20-SOM-EVB: ethernet non-functional with current u-boot
>>>                       git head, but no problems with v2016.01]
> [...]
>>> I am not the original poster, but I have bisected the issue down
>>> to the following commit:
>>>
>>> commit c32a6fd07b1839e4a45729587ebc8e1c55601a4d
>>> Author: Simon Glass <sjg@chromium.org>
>>> Date:   Sun Jan 17 14:51:56 2016 -0700
>>>
>>>      net: Don't call board/cpu_eth_init() with driver model
>>>
>>>      We should avoid weak functions with driver model. Existing boards that use
>>>      driver model don't need them, so let's kill them off.
>>>
>>>      Signed-off-by: Simon Glass <sjg@chromium.org>
>>>      Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
>>>      Acked-by: Joe Hershberger <joe.hershberger@ni.com>
>>
>> Is this board-specific init, or sunxi-specific?
>
> Hello Simon,
>
> I am unsure whether I understand your question correctly. There
> is only one cpu_eth_init() function for sunxi-based devices, but
> its behaviour depends on several configuration options. Some of
> those depend on the SoC type (and are thereby indirectly
> board-specific) and some of those (e.g. the GPIO pin controlling
> the regulator that powers the PHY) are directly board-specific.
>
>> Can you just call the init on start-up, instead of from the
>> Ethernet driver?
>
>  From a look at the source in arch/arm/cpu/armv7/sunxi/board.c and
> board/sunxi/gmac.c I guess that should be possible, but I don't
> know enough about u-boot's internals to really assess that.
>
> Hans, could you take a look at the issue?

Yes I've had this thread on my radar / todo the whole time already,
not sure when I'll get around to it. I'll make sure to get this
fixed soon though, as we'll have a new u-boot release soonish.

Regards,

Hans

[U-Boot] Olimex-A20-EVB ethernet stops working with latest U-boot, works with tag v2016.01

[Ing. Damiano Bolla]

Just to say that I am back on this board now and can test a new U-Boot version, when available

Thanks you all

In data 14 marzo 2016 alle ore 08:46:18, Hans de Goede <hdegoede@redhat.com> ha scritto:

> Hi,
>
> On 13-03-16 12:14, Karsten Merker wrote:
>> On Wed, Mar 09, 2016 at 02:39:11PM -0700, Simon Glass wrote:
>>> On 9 March 2016 at 14:04, Karsten Merker <merker@debian.org> wrote:
>>>> [Olimex A20-SOM-EVB: ethernet non-functional with current u-boot
>>>>                       git head, but no problems with v2016.01]
>> [...]
>>>> I am not the original poster, but I have bisected the issue down
>>>> to the following commit:
>>>>
>>>> commit c32a6fd07b1839e4a45729587ebc8e1c55601a4d
>>>> Author: Simon Glass <sjg@chromium.org>
>>>> Date:   Sun Jan 17 14:51:56 2016 -0700
>>>>
>>>>      net: Don't call board/cpu_eth_init() with driver model
>>>>
>>>>      We should avoid weak functions with driver model. Existing boards that use
>>>>      driver model don't need them, so let's kill them off.
>>>>
>>>>      Signed-off-by: Simon Glass <sjg@chromium.org>
>>>>      Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
>>>>      Acked-by: Joe Hershberger <joe.hershberger@ni.com>
>>>
>>> Is this board-specific init, or sunxi-specific?
>>
>> Hello Simon,
>>
>> I am unsure whether I understand your question correctly. There
>> is only one cpu_eth_init() function for sunxi-based devices, but
>> its behaviour depends on several configuration options. Some of
>> those depend on the SoC type (and are thereby indirectly
>> board-specific) and some of those (e.g. the GPIO pin controlling
>> the regulator that powers the PHY) are directly board-specific.
>>
>>> Can you just call the init on start-up, instead of from the
>>> Ethernet driver?
>>
>>  From a look at the source in arch/arm/cpu/armv7/sunxi/board.c and
>> board/sunxi/gmac.c I guess that should be possible, but I don't
>> know enough about u-boot's internals to really assess that.
>>
>> Hans, could you take a look at the issue?
>
> Yes I've had this thread on my radar / todo the whole time already,
> not sure when I'll get around to it. I'll make sure to get this
> fixed soon though, as we'll have a new u-boot release soonish.
>
> Regards,
>
> Hans
>


-- 
Ing. Damiano Bolla, 3486947264

[U-Boot] Olimex-A20-EVB ethernet stops working with latest U-boot, works with tag v2016.01

[Ing. Damiano Bolla]

Good morning Peng, I cannot attache an image, moderators will stop it forever

I did revert the commit, see attached image showing git repository status

Unfortunately the result is still negative, see the following log

U-Boot 2016.03-rc2-00117-gde889c0 (Feb 25 2016 - 08:12:50 +0100) Allwinner Technology
CPU:   Allwinner A20 (SUN7I)
I2C:   ready
DRAM:  1 GiB
MMC:   SUNXI SD/MMC: 0, SUNXI SD/MMC: 1
In:    serial
Out:   serial
Err:   serial
SCSI:  SUNXI SCSI INIT
SATA link 0 timeout.
AHCI 0001.0100 32 slots 1 ports 3 Gbps 0x1 impl SATA mode
flags: ncq stag pm led clo only pmp pio slum part ccc apst
Net:   eth0: ethernet at 01c50000
starting USB...
USB0:   USB EHCI 1.00
USB1:   USB OHCI 1.0
USB2:   USB EHCI 0.00
USB3:   USB OHCI 0.0
scanning bus 0 for devices... 1 USB Device(s) found
scanning bus 2 for devices... 1 USB Device(s) found
Hit any key to stop autoboot:  0
ethernet at 01c50000 Waiting for PHY auto negotiation to complete......... TIMEOUT !
ethernet at 01c50000: No link.
ethernet at 01c50000 Waiting for PHY auto negotiation to complete......... TIMEOUT !
ethernet at 01c50000: No link.
Bad Linux ARM zImage magic!

Can I assist further ?

Thanks, Damiano


In data 25 febbraio 2016 alle ore 02:11:00, Peng Fan <van.freenix@gmail.com> ha scritto:

> Hi Damiano,
>
> Can you try revert the following patch and retest on your board?
> "
> commit a058052c358c3ecf5f394ff37def6a45eb26768c
> Author: Stefan Agner <stefan@agner.ch>
> Date:   Wed Dec 9 11:21:25 2015 -0800
>
>     net: phy: do not read configuration register on reset
>    When doing a software reset, the reset flag should be written without
>     other bits set. Writing the current state will lead to restoring the
>     state of the PHY (e.g. Powerdown), which is not what is expected from
>     a software reset.
> 			
>     Signed-off-by: Stefan Agner <stefan@agner.ch>
> "
>
> This patch breaks ethernet function of i.MX6 Quad Sabresd/Quad Sabreauto.
> Not sure whether reverting this patch can help in your case or not. Please
> try.
>
> Thanks,
> Peng.
>
> On Wed, Feb 24, 2016 at 09:12:02AM +0100, Ing. Damiano Bolla wrote:
>> Good morning all
>>
>> I updated to latest u-boot yesterday (git pull) and ethernet was not working
>>
>> reverted to tag v2016.01 with the command
>>
>> $ git checkout tags/v2016.01
>>
>> rebuilt, uploaded and ethernet works
>>
>> As far as I can tell the only difference on the two tests (just a simple ping 192.168.1.1) is the u-boot version
>> Ethernet cable was not moved and I manage to load u-boot using FEL
>>
>> I can do further tests is any of you is interested.
>>
>> Thanks, Damiano
>> _______________________________________________
>> U-Boot mailing list
>> U-Boot at lists.denx.de
>> http://lists.denx.de/mailman/listinfo/u-boot
>


-- 
Ing. Damiano Bolla, 3486947264