Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 2 Oct 2022 23:36:42 +0200
From:      =?utf-8?Q?Klaus_K=C3=BCchemann?= <maciphone2@googlemail.com>
To:        Mark Millard <marklmi@yahoo.com>, freebsd-arm@freebsd.org, bob prohaska <fbsd@www.zefox.net>
Subject:   Re: u-boot debug, was: Re: U-boot on RPI3, sees disk but won't boot it
Message-ID:  <CA1A0345-C1CF-47F9-B684-DE1CD3DFC584@googlemail.com>
In-Reply-To: <64C1085D-D3F8-45A3-80FB-4B88F81E480E@googlemail.com>
References:  <20220929002131.GA77106@www.zefox.net> <197D3C46-063B-4C67-AB1A-A3A072521D7F@yahoo.com> <A8C2BA4E-4520-4B34-9614-DDC4D8BEB097@yahoo.com> <6AA65AE6-41F1-405F-A592-7D641EA4C9CF@yahoo.com> <E3A1C678-8C47-4283-9F9F-4C9011DB8A2B@yahoo.com> <20221001174724.GA98055@www.zefox.net> <ABFDD634-5CB6-4DAE-B4DE-629CE7E4FE06@yahoo.com> <20221001193033.GA98348@www.zefox.net> <46226720-D867-4AD3-9559-A4365FAC28C4@yahoo.com> <6DB88FC9-629C-43E6-9673-32640FC547F7@yahoo.com> <20221002182049.GA2255@www.zefox.net> <5FFDAA6A-AD8C-4E40-A2EB-4082E5086679@googlemail.com> <38DFEB91-AC60-4FD1-8088-95B0A06C5E5D@yahoo.com> <EEC43DA1-6B68-4FDD-A68A-A3055E86E407@googlemail.com> <64C1085D-D3F8-45A3-80FB-4B88F81E480E@googlemail.com>

next in thread | previous in thread | raw e-mail | index | archive | help


> Am 02.10.2022 um 22:46 schrieb Klaus K=C3=BCchemann =
<maciphone2@googlemail.com>:
>=20
>=20
>=20
>> Am 02.10.2022 um 22:18 schrieb Klaus K=C3=BCchemann =
<maciphone2@googlemail.com>:
>>=20
>>=20
>>=20
>>> Am 02.10.2022 um 21:58 schrieb Mark Millard <marklmi@yahoo.com>:
>>>=20
>>> On 2022-Oct-2, at 12:35, Klaus K=C3=BCchemann =
<maciphone2@googlemail.com> wrote:
>>>=20
>>>> Am 02.10.2022 um 20:20 schrieb bob prohaska <fbsd@www.zefox.net>:
>>>>>=20
>>>>> On Sat, Oct 01, 2022 at 02:21:42PM -0700, Mark Millard wrote:
>>>>>>=20
>>>>>> http://nemesis.zefox.com/~fbsd/pelorus_console.txt7_orig_fragment
>>>>>>=20
>>>>>> still shows all the debug output. It did not
>>>>>> avoid the timing changes.
>>>>>>=20
>>>>>> You might need to not use either of:
>>>>>>=20
>>>>>> patch-common_usb__hub.c
>>>>>> patch-common_usb__storage.c
>>>>>>=20
>>>>>> and to disable the LOG_DEBUG and DEBUG lines in:
>>>>>>=20
>>>>>> patch-common_usb.c
>>>>>>=20
>>>>>> via turning them into comments by adding // as
>>>>>> indicated below:
>>>>>>=20
>>>>>> +//#define LOG_DEBUG
>>>>>> +//#define DEBUG
>>>>>>=20
>>>>>=20
>>>>> I think the changes were successful, u-boot compiles and
>>>>> runs. There's no extra output, and unfortunately only one=20
>>>>> successful reboot so far. Bus scanning seems quite slow.
>>>>> Storage devices are rarely found on reset, but usb reset
>>>>> does sometimes work. Run bootcmd_usb0 paused for minutes
>>>>> at Device 0: and paused again after reporting ..current device.
>>>>> No echo from the console, ctrl-C did nothing.=20
>>>>>=20
>>>>> The attempt sequence was
>>>>> SRBSPRMRPRRPUPPRRUPUCUUC
>>>>> where=20
>>>>> S is shutdown -r
>>>>> R is reset of u-boot
>>>>> U is usb reset
>>>>> P is powercycle
>>>>> M is stop at mountroot
>>>>> C is run bootcmd_usb0
>>>>>=20
>>>>> The console log is at
>>>>> http://nemesis.zefox.com/~fbsd/pelorus_console.txt8_no_debug
>>>>>=20
>>>>> It now appears that the run bootcmd_usb0 rather reliably gets
>>>>> stuck, with the disk LED on steadily (no activity). Maybe in
>>>>> one of the loops seen earlier?=20
>>>>>=20
>>>>> Thanks again for all your help!
>>>>>=20
>>>>> bob prohaska
>>>>>=20
>>>>=20
>>>>=20
>>>> So if you now reapply the  #define DEBUG  patches(while keeping the =
mdelay-patch) and the reboot issues definitely went away
>>>> we have a typical so called Heisenbug, hopefully more or less  now =
a fixed issue.
>>>=20
>>> No. Bob has more than one problem: more problems observed
>>> after "1 Storage Device(s) found". The DEBUG/mdelay
>>> combination only seemed to cause the "1 Storage Device(s)
>>> found" to be at least more reliable, not later stages.
>>>=20
>>> It is not obvious if earlier activity contributes or not
>>> to the problems observed after "1 Storage Device(s) found".
>>>=20
>>> So far nothing has gotten near having things just work for
>>> booting without manual intervention, multiple retries
>>> being involved sometimes.
>>>=20
>>>> Well, USB-boot problems on earlier Pi models( afaik all except the =
4) are commonly known, from defective HW to power cycle issues we will =
find a lot of discussions on the WWW and we will see that even the =
debug-message =E2=80=9Eis your USB cable bad?=E2=80=9C  did fix issues =
in some cases. Others applied RNG devices or external clock or even =
plugging a mouse fixed it( to change usb enumeration).
>>>>=20
>>>> I think with the working u-boot.bin after 1500 successful reboots =
you can be sure it=E2=80=99s working =E2=80=A6.
>>>> just kidding=E2=80=A6 :-)
>>>>=20
>>>=20
>>>=20
>>> =3D=3D=3D
>>> Mark Millard
>>> marklmi at yahoo.com
>>=20
>> hard to read and remember every log but I thought Bob wrote about =
aprox. 30 successful reboots after the mdelay patch,
>> while of course that could be coincidence, who really knows what =
happens in this untrackable inconsistent behavior of the usb-boot?!
>>=20
>>> Am 02.10.2022 um 21:48 schrieb Mark Millard <marklmi@yahoo.com>:
>>>=20
>>> (RaspiOS and Ubuntu do not use U-Boot last I knew. So
>>> they do not make for good comparisons for the purpose
>>> as far as I know.)
>>=20
>> RaspiOS doesn=E2=80=99t , Ubuntu(and others) use u-boot since years =
=E2=80=A6
>> while possible Ubuntu(or others) have own u-boot patches ,
>> from guessing it seems more probable that they also will sometimes =
hang after (re)boot.
>>=20
>> If I would want to keep such a device as an online server, like Bob =
does, for whatever reason I would=20
>> Implement something like an =E2=80=9EIPMI=E2=80=9C or simpler said:
>> An immediate console remote access after being warned by a script =
that the machine is offline.
>> But I would remove it from cluster if there are known Hardware =
problems.=20
>>=20
>>=20
>> Regards
>>=20
>> Klaus
>>=20
>>=20
>=20
> =E2=80=A6 but of course, Mark, that is correct :
> overwriting parts of the msdos-partition by linux ones could be the =
last resort to save something=E2=80=A6
> but if linux had patched inside u-boot, as you did it for Bob, I would =
see other problems coming=E2=80=A6
>=20
> Regards
>=20
> Klaus
>=20

Not debugging related, so off-topic:
Bob, if you care about remotely rebooting without losing access to the =
console:
You can plug a known functioning online machine(e.g. your Pi4) v ia GPIO =
directly to Pi 3b.
ssh into the Pi4 in this example  , then cu / minicom /xy into the 3b =
console and manually reboot the 3b
over the u-boot console if it hangs.

For unexpected 3b-crashes while uptime, as said, you could use a warning =
script or whatever.

Seems to be easier this way than continue hunting an untraceable bug   .

But (at least my) main intention was to come closer to the u-boot source =
code,

So many thanks to you,Mark ,giving so much POWER(and SPEED) to these =
things, great work!
/Bob knows what I mean with POWER & SPEED terminology :-)


Regards=20
Thanks=20

Klaus=20


=20






Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CA1A0345-C1CF-47F9-B684-DE1CD3DFC584>