[LEDE-DEV] LEDE new image on gw5520

[Outback Dingo]

On Thu, Nov 9, 2017 at 7:33 PM, Outback Dingo <outbackdingo at gmail.com> wrote:
> sorry that was a local merge from my tree so previous known working
> lede commit would be c88770c766fdc5599efc4672bca230017f52e8e4
>
> On Thu, Nov 9, 2017 at 7:21 PM, Outback Dingo <outbackdingo at gmail.com> wrote:
>> 7af22af0e9ebc0f69091af19a3104692c18fbe73 is the last known good commit
>> running on my board now..... time to walk the tree to narrow it down
>> it guess
>>
>> On Thu, Nov 9, 2017 at 5:57 PM, Koen Vandeputte
>> <koen.vandeputte at ncentric.com> wrote:
>>>
>>>
>>> On 2017-11-09 17:46, Outback Dingo wrote:
>>>>
>>>> On Thu, Nov 9, 2017 at 5:38 PM, Koen Vandeputte
>>>> <koen.vandeputte at ncentric.com> wrote:
>>>>>
>>>>>
>>>>> On 2017-11-09 16:40, Outback Dingo wrote:
>>>>>>
>>>>>> On Thu, Nov 9, 2017 at 4:16 PM, Koen Vandeputte
>>>>>> <koen.vandeputte at ncentric.com> wrote:
>>>>>>>
>>>>>>>
>>>>>>> On 2017-11-09 15:55, Outback Dingo wrote:
>>>>>>>>
>>>>>>>> heres more data, we still have a no interfaces issue with recent LEDE
>>>>>>>> trunk
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Thu, Nov 2, 2017 at 6:42 PM, Outback Dingo <outbackdingo at gmail.com>
>>>>>>>> wrote:
>>>>>>>>>
>>>>>>>>> seems that ive lost my eth0 / eth1 interfeaces on a RECENT ... today
>>>>>>>>> image built for the gw5520
>>>>>>>>> i do however see the ifb0 and ifb1 interfaces... any ideas?
>>>>>>>
>>>>>>>
>>>>>>> Hi,
>>>>>>>
>>>>>>> As I have similar hardware, maybe I can help out. (GW 5510 and others)
>>>>>>>
>>>>>>> I'm not seeing the FEC driver being loaded in your log: (Freescale
>>>>>>> Ethernet)
>>>>>>>
>>>>>>>
>>>>>>> Snip from my 5510:
>>>>>>>
>>>>>>>
>>>>>>> [    0.692354] libphy: Fixed MDIO Bus: probed
>>>>>>> [    0.693357] libphy: mdio_driver_register: mv88e6085
>>>>>>> [    0.695036] 2188000.ethernet supply phy not found, using dummy
>>>>>>> regulator
>>>>>>> [    0.697541] pps pps0: new PPS source ptp0
>>>>>>> [    0.701333] libphy: fec_enet_mii_bus: probed
>>>>>>>
>>>>>>> --> [    0.702005] fec 2188000.ethernet eth0: registered PHC device 0
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> Please run  "make kernel_menuconfig"  and check if CONFIG_FEC is
>>>>>>> selected.
>>>>>>>
>>>>>>>
>>>>>>
>>>>>> Hi, the following were selected under drivers / ethernet
>>>>>>
>>>>>>     [*]   Freescale devices
>>>>>>                                    │ │
>>>>>>    │ │                                <*>     FEC ethernet controller
>>>>>> (of ColdFire and some i.MX CPUs)
>>>>>> │ │
>>>>>>    │ │                                < >     Freescale PQ MDIO
>>>>>>                                                                      │ │
>>>>>>    │ │                                < >     Freescale XGMAC MDIO
>>>>>>                                                                      │ │
>>>>>>    │ │                                < >     Gianfar Ethernet
>>>>>>                                                                      │ │
>>>>>>    │ │                                [*]   Hisilicon devices
>>>>>>
>>>>>>> Kind Regards,
>>>>>>>
>>>>>>> Koen
>>>>>


well this is working now....... seems someone forgot to add the Intel
gigabit drivers to the kernel config when they updated the kernel

   2.045840] igb: Intel(R) Gigabit Ethernet Network Driver - version 5.4.0-k
[    2.052835] igb: Copyright (c) 2007-2014 Intel Corporation.
[    2.058544] igb 0000:03:00.0: enabling device (0140 -> 0142)
[    2.094991] pps pps0: new PPS source ptp0
[    2.099075] igb 0000:03:00.0: added PHC on eth0
[    2.103620] igb 0000:03:00.0: Intel(R) Gigabit Ethernet Network Connection
[    2.110528] igb 0000:03:00.0: eth0: (PCIe:2.5Gb/s:Width x1) 00:d0:12:16:f3:5f
[    2.117675] igb 0000:03:00.0: eth0: PBA No: FFFFFF-0FF
[    2.122842] igb 0000:03:00.0: Using legacy interrupts. 1 rx
queue(s), 1 tx queue(s)
[    2.130634] igb 0000:07:00.0: enabling device (0140 -> 0142)
[    2.167038] pps pps1: new PPS source ptp1
[    2.171120] igb 0000:07:00.0: added PHC on eth1
[    2.175663] igb 0000:07:00.0: Intel(R) Gigabit Ethernet Network Connection
[    2.182569] igb 0000:07:00.0: eth1: (PCIe:2.5Gb/s:Width x1) 00:d0:12:16:f3:60
[    2.189729] igb 0000:07:00.0: eth1: PBA No: FFFFFF-0FF
[    2.194880] igb 0000:07:00.0: Using legacy interrupts. 1 rx
queue(s), 1 tx queue(s)


>>>>>
>>>>> Hi,
>>>>>
>>>>> Quickly checked a little more in-depth.
>>>>>
>>>>> (My board is a GW52xx.)
>>>>>
>>>>> Looking at dtsi, it seems that the fec part is not specified for the 552x
>>>>>
>>>>> As a test, could you add attached patch to your build?
>>>>> Just put it in  target/linux/imx6/patches-4.9 and do a full rebuild (make
>>>>> clean; make)
>>>>
>>>> This board was working fine with LEDE in September? so im curious how
>>>> a dtsi file would regress in a few weeks
>>>
>>> Thats the reason for the quick test.  Forcing it (yes, I checked dtsi
>>> history upstream. No, it didn't change)
>>> Otherwise .. please identify the commit causing the regression ;)
>>>>>
>>>>>
>>>>> Thanks,
>>>>>
>>>>> Koen
>>>
>>>