[U-Boot] [PATCH v3 00/35] dm: tegra: Move nyan-big, jetson-tk1, beaver to livetree

[Simon Glass]

Hi Stephen,

On 12 June 2017 at 17:15, Simon Glass <sjg at chromium.org> wrote:
> Hi Stephen,
>
> On 12 June 2017 at 15:39, Stephen Warren <swarren at wwwdotorg.org> wrote:
>> On 06/12/2017 03:15 PM, Simon Glass wrote:
>>>
>>> [trimming the cc list a bit]
>>>
>>> Hi Stephen,
>>>
>>> On 12 June 2017 at 15:02, Stephen Warren <swarren at wwwdotorg.org> wrote:
>>>>
>>>> On 06/12/2017 02:09 PM, Simon Glass wrote:
>>>>>
>>>>>
>>>>> Hi Stephen,
>>>>>
>>>>> On 12 June 2017 at 13:52, Stephen Warren <swarren at wwwdotorg.org> wrote:
>>>>>>
>>>>>>
>>>>>> On 06/12/2017 06:21 AM, Simon Glass wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> This moves three entire boards to use a live device tree as an example
>>>>>>> of
>>>>>>> the impact.
>>>>>>
>>>>>>
>>>>>> This series appears to cause problems on a couple of boards:
>>>>>>
>>>>>> On both Jetson TK1 and Jetson TX1 (p2372-180), "pci enum" fails with:
>>>>>>
>>>>>> ERROR: tegra-pcie: failed to power on PHY: -110
>>>>>>
>>>>>> at .../pci_tegra.c:774/tegra_pcie_enable_controller()
>>>>>> ERROR: tegra-pcie: failed to enable controller
>>>>>> at .../pci_tegra.c:1154/pci_tegra_probe()
>>>>>
>>>>>
>>>>> I seem to get this error always (before these patches) on TK1. I
>>>>> thought it was because I corrupted my PMIC. Do it happen on vanilla
>>>>> mainline for you?
>>>>
>>>>
>>>> No, this error doesn't occur in mainline AFAIK; git hash 8cb3ce64f936
>>>> from
>>>> mainline ran through testing a couple of days ago without any issue.
>>>
>>>
>>> That does make me think my board is broken as mentioned. Here's what I
>>> get with that commit which looks very similar to you:
>>>
>>
>>> U-Boot 2017.07-rc1-00163-g8cb3ce6 (Jun 12 2017 - 15:12:24 -0600)
>>
>>
>> OK, that's the same commit that ran OK on my test system.
>>
>>> Tegra124 (Jetson TK1) # boot
>>> MMC: no card present
>>> mmc_init: -123, time 0
>>> tegra_mmc_send_cmd_bounced: MMC Timeout
>>>      Interrupt status        0x00000001
>>>      Interrupt status enable 0xffff003b
>>>      Interrupt signal enable 0xffff0002
>>>      Present status          0x01fb02f6
>>> mmc_init: -1, time 8042
>>
>>
>> I haven't seen that error occur in any commit/branch; perhaps something is
>> wrong with your system.
>
> This is fixed by this patch:
>
> http://patchwork.ozlabs.org/patch/772771/
>
> I'm unsure how your board could work without that patch.
>
>>
>> ...
>>>
>>> USB device 0: unknown device
>>> ERROR: tegra-pcie: failed to power on PHY: -110
>>>
>>> at
>>> /scratch/sglass/cosarm/src/third_party/u-boot/files/drivers/pci/pci_tegra.c:779/tegra_pcie_enable_controller()
>>> ERROR: tegra-pcie: failed to enable controller
>>> at
>>> /scratch/sglass/cosarm/src/third_party/u-boot/files/drivers/pci/pci_tegra.c:1159/pci_tegra_probe()
>>
>>
>> That's the issue I saw in your branch though, even though your build is for
>> an upstream commit.
>>
>> ...
>>>
>>> ERROR: tegra-pcie: failed to power on PHY: -110
>>>
>>> at
>>> /scratch/sglass/cosarm/src/third_party/u-boot/files/drivers/pci/pci_tegra.c:779/tegra_pcie_enable_controller()
>>> ERROR: tegra-pcie: failed to enable controller
>>> at
>>> /scratch/sglass/cosarm/src/third_party/u-boot/files/drivers/pci/pci_tegra.c:1159/pci_tegra_probe()
>>> No ethernet found.
>>> No ethernet found.
>>
>>
>> Same there.
>>
>> It seems coincidental that your broken board would give the exact same error
>> message as my test system sees with your branch. I wonder if you have an
>> incremental build issue; have you tried deleting all the build results and
>> trying again?
>
> Yes, but same result. It works OK on beaver though. Can you try it on beaver?
>
> Regards,
> Simon

At this point I'm thinking of applying the patches except for the one
that actually enables livetree. Once we figure out what is happening
with jetson-tk1, or you can confirm that it works on beaver, we can
look at the final patch.

Regards,
Simon