9front kernel panic on bootstrap on rpi

9front kernel panic on bootstrap on rpi

[Daniel Morandini]

[-- Attachment #1: Type: text/plain, Size: 743 bytes --]

Hi folks,
After the mice problems I have on my rpi2 with 9front [1], I decided to check if I can have an instance working on another machine, i.e. an old rpi model A. I cannot get through the boostrap process, the kernel panics.

I can provide more context and information, but first I wanted to ask: am I picking a broken image? [2] Again, as for the mice problme, the legacy miller version works like a charm even on this rpi!

Bye,
jecoz

[1] https://9fans.topicbox.com/groups/9fans/Tf467e22d368cadb2-M65943bcc3de5a6a028956350/9front-amazon-basics-usb-mice-not-working-on-rpi-2
[2] http://bell-labs.co/9front/iso/9front-7408.1d345066125a.pi.img.gz.torrent <http://bell-labs.co/9front/iso/9front-7408.1d345066125a.pi.img.gz.torrent>

[-- Attachment #2: Type: text/html, Size: 1339 bytes --]

Re: [9fans] 9front kernel panic on bootstrap on rpi

[ori]

> Hi folks,
> After the mice problems I have on my rpi2 with 9front [1], I
> decided to check if I can have an instance working on another
> machine, i.e. an old rpi model A. I cannot get through the
> boostrap process, the kernel panics.

Thanks for the reports. Two things:

0) It's going to be very hard to debug this if you don't have
   a way to rebuild the software for testing. Getting net boot
   working (even from a VM) would make it much easier.

1) You're likely to get better results on the 9front mailing
   list. (9front@9front.org)

Unfortunately, I have neither version of the rasperry pi,
and the mouse model you have doens't seem to be available
on Amazon any more, so I can't even try it out on the PC
on my end.

> I can provide more context and information, but first I
> wanted to ask: am I picking a broken image?

cpu% sha1sum 9front.pi.img
ffb9f6569c7c239a58f0e6e508aac9d2418333af	9front.pi.img

> [2] Again,  as for the mice problme, the legacy miller
> version works like a charm even on this rpi!

From what I understand, the difference there is that 9front
will parse the HID descriptor, while 9legacy will hardcode
the boot protocol. There are a number of places that this can
be going wrong -- from bad HID descriptors to power brownouts
to USB bugs.

As far as booting: A picture of the panic message would help.

As far as the mouse: First question -- does it work if you
plug it into a PC? (You can test with a USB stick). If it
fails there, then it's likely a problem with HID parsing.

Second question: Does it work with a powered USB hub? If
yes, then it's likely to be the Pi's power limitations.

Third question: Does it work with a different mouse?

If not, then it'll take more work to investigate.

Re: [9fans] 9front kernel panic on bootstrap on rpi

[Daniel Morandini]

Dear ori,
thank you for the reply! First, the boot problem has been resolved with cinap’s answer, the mice problem persists though.

> 1) You're likely to get better results on the 9front mailing
>   list. (9front@9front.org)
I did actually, but I always get an error message back (containing only my message actually, and some protocol information) when I try to send. I’m subscribed to the list though, I can receive.

> cpu% sha1sum 9front.pi.img
> ffb9f6569c7c239a58f0e6e508aac9d2418333af        9front.pi.img
We got something on this ori!
% shasum -a 1 9front-7408.1d345066125a.pi.img 
d058861d64b02fd2d068f94b8f398e1a559c579c  9front-7408.1d345066125a.pi.img

> From what I understand, the difference there is that 9front
> will parse the HID descriptor, while 9legacy will hardcode
> the boot protocol. There are a number of places that this can
> be going wrong -- from bad HID descriptors to power brownouts
> to USB bugs.

> As far as the mouse: First question -- does it work if you
> plug it into a PC? (You can test with a USB stick). If it
> fails there, then it's likely a problem with HID parsing.
It does. With PC you mean microsoft? Anyway, I just tested out and it works on macOS 10.15, Windows 10, 9legacy.

> Second question: Does it work with a powered USB hub? If
> yes, then it's likely to be the Pi's power limitations.
> Third question: Does it work with a different mouse?
> If not, then it'll take more work to investigate.
I have just another mouse here, a nacon gaming mouse. It requires 5v though and it does not work on any rpi I have. I don’t have powered USB hub here for testing, but I can find one if needed!

Tell me what I can do to help you more.

jecoz

Re: [9fans] 9front kernel panic on bootstrap on rpi

[Marshall Conover]

Hi Dan! Just a note - you can only send plaintext emails to the 9front
mailing list, as a spyware prevention measure. It's likely what hung
you up.

On Fri, May 1, 2020 at 6:20 AM Daniel Morandini via 9fans
<9fans@9fans.net> wrote:
>
> Dear ori,
> thank you for the reply! First, the boot problem has been resolved with cinap’s answer, the mice problem persists though.
>
> > 1) You're likely to get better results on the 9front mailing
> >   list. (9front@9front.org)
> I did actually, but I always get an error message back (containing only my message actually, and some protocol information) when I try to send. I’m subscribed to the list though, I can receive.
>
> > cpu% sha1sum 9front.pi.img
> > ffb9f6569c7c239a58f0e6e508aac9d2418333af        9front.pi.img
> We got something on this ori!
> % shasum -a 1 9front-7408.1d345066125a.pi.img
> d058861d64b02fd2d068f94b8f398e1a559c579c  9front-7408.1d345066125a.pi.img
>
> > From what I understand, the difference there is that 9front
> > will parse the HID descriptor, while 9legacy will hardcode
> > the boot protocol. There are a number of places that this can
> > be going wrong -- from bad HID descriptors to power brownouts
> > to USB bugs.
> > As far as the mouse: First question -- does it work if you
> > plug it into a PC? (You can test with a USB stick). If it
> > fails there, then it's likely a problem with HID parsing.
> It does. With PC you mean microsoft? Anyway, I just tested out and it works on macOS 10.15, Windows 10, 9legacy.
>
> > Second question: Does it work with a powered USB hub? If
> > yes, then it's likely to be the Pi's power limitations.
> > Third question: Does it work with a different mouse?
> > If not, then it'll take more work to investigate.
> I have just another mouse here, a nacon gaming mouse. It requires 5v though and it does not work on any rpi I have. I don’t have powered USB hub here for testing, but I can find one if needed!
>
> Tell me what I can do to help you more.
>
> jecoz
> ------------------------------------------
> 9fans: 9fans
> Permalink: https://9fans.topicbox.com/groups/9fans/Td646dc8a14a94c8a-M7257c80b3dec1883120fb3e8
> Delivery options: https://9fans.topicbox.com/groups/9fans/subscription



-- 
Have a good day,

Marshall Conover

Re: [9fans] 9front kernel panic on bootstrap on rpi

[Daniel Morandini]

Hi Marshall!
Ahh you got it. In fact, I saw in the reply that my message was actually marshaled into html by Apple’s servers.

Thank you!
Dan

> Il giorno 1 mag 2020, alle ore 14:12, Marshall Conover <marzhall.o@gmail.com> ha scritto:
> 
> Hi Dan! Just a note - you can only send plaintext emails to the 9front
> mailing list, as a spyware prevention measure. It's likely what hung
> you up.
> 
>> On Fri, May 1, 2020 at 6:20 AM Daniel Morandini via 9fans
>> <9fans@9fans.net> wrote:
>> 
>> Dear ori,
>> thank you for the reply! First, the boot problem has been resolved with cinap’s answer, the mice problem persists though.
>> 
>>> 1) You're likely to get better results on the 9front mailing
>>>  list. (9front@9front.org)
>> I did actually, but I always get an error message back (containing only my message actually, and some protocol information) when I try to send. I’m subscribed to the list though, I can receive.
>> 
>>> cpu% sha1sum 9front.pi.img
>>> ffb9f6569c7c239a58f0e6e508aac9d2418333af        9front.pi.img
>> We got something on this ori!
>> % shasum -a 1 9front-7408.1d345066125a.pi.img
>> d058861d64b02fd2d068f94b8f398e1a559c579c  9front-7408.1d345066125a.pi.img
>> 
>>> From what I understand, the difference there is that 9front
>>> will parse the HID descriptor, while 9legacy will hardcode
>>> the boot protocol. There are a number of places that this can
>>> be going wrong -- from bad HID descriptors to power brownouts
>>> to USB bugs.
>>> As far as the mouse: First question -- does it work if you
>>> plug it into a PC? (You can test with a USB stick). If it
>>> fails there, then it's likely a problem with HID parsing.
>> It does. With PC you mean microsoft? Anyway, I just tested out and it works on macOS 10.15, Windows 10, 9legacy.
>> 
>>> Second question: Does it work with a powered USB hub? If
>>> yes, then it's likely to be the Pi's power limitations.
>>> Third question: Does it work with a different mouse?
>>> If not, then it'll take more work to investigate.
>> I have just another mouse here, a nacon gaming mouse. It requires 5v though and it does not work on any rpi I have. I don’t have powered USB hub here for testing, but I can find one if needed!
>> 
>> Tell me what I can do to help you more.
>> 
>> jecoz
> 
> 
> 
> -- 
> Have a good day,
> 
> Marshall Conover
> 
> ------------------------------------------
> 9fans: 9fans
> Permalink: https://9fans.topicbox.com/groups/9fans/Td646dc8a14a94c8a-Me4a2a9459b5230910087243b
> Delivery options: https://9fans.topicbox.com/groups/9fans/subscription

Re: [9fans] 9front kernel panic on bootstrap on rpi

[ori]

> It does. With PC you mean microsoft? Anyway, I just tested out and it works on macOS 10.15, Windows 10, 9legacy.

No. I mean if you boot 9front from a USB stick on x86
hardware, and then plug the mouse in to that, will the
mouse work?

The question I'm trying to answer with this is 'is it
the pi or the mouse that 9front is having trouble with'.
Taking the pi out of the equation helps answer that.

Re: [9fans] 9front kernel panic on bootstrap on rpi

[Daniel Morandini]

Ahh got it! I’ll test it tomorrow and let you know!

jecoz

> Il giorno 1 mag 2020, alle ore 17:06, ori@eigenstate.org ha scritto:
> 
> No. I mean if you boot 9front from a USB stick on x86
> hardware, and then plug the mouse in to that, will the
> mouse work?
> 
> The question I'm trying to answer with this is 'is it
> the pi or the mouse that 9front is having trouble with'.
> Taking the pi out of the equation helps answer that.

Re: [9fans] 9front kernel panic on bootstrap on rpi

[Daniel Morandini]

Hi ori,
I tried to run 9front from a live usb on my macbook, but I did not manage to make one working (I tried different approaches, with no success though. I see the EFI boot option, but the boot process does not start).

Nevertheless, I tested 9front using virtualbox. The mouse works there like a charm!

> On 1 May 2020, at 17:04, ori@eigenstate.org wrote:
> 
> No. I mean if you boot 9front from a USB stick on x86
> hardware, and then plug the mouse in to that, will the
> mouse work?
> 
> The question I'm trying to answer with this is 'is it
> the pi or the mouse that 9front is having trouble with'.
> Taking the pi out of the equation helps answer that.

Bye,
dan

Re: [9fans] 9front kernel panic on bootstrap on rpi

[cinap_lenrek]

yes, this is a known issue in this release and it has already
been fixed here:

https://code.9front.org/hg/plan9front/rev/cc5c37116d5e

i did not have the time to schedule a new binary release.

if you want you can try the "unofficial" kernels linked
below and copy them on the sdcard.

raw kenrel:

http://felloff.net/usr/cinap_lenrek/9pi

a.out with symbols:

http://felloff.net/usr/cinap_lenrek/s9pi

or you can build them yourself on any 9front installation
following the fqa instructions, which basically buils down
to: objtype=arm mk install in /sys/src and mk: 'CONF=pi'
in /sys/src/9/bcm

good luck.

--
cinap

Re: [9fans] 9front kernel panic on bootstrap on rpi

[Daniel Morandini]

[-- Attachment #1: Type: text/plain, Size: 565 bytes --]

Dear cinap,
thank you for the reply! Yes, now it works (the mice problem persists though).

> On 30 Apr 2020, at 20:24, cinap_lenrek@felloff.net wrote:
> 
> if you want you can try the "unofficial" kernels linked
> below and copy them on the sdcard.

As soon as I get a single workstation with 9front working properly I’ll start compiling the kernel myself. plan9 is exactly pointing towards the technological future I want to see. I want to start helping you guys too, as much as I can. I mostly do Go programming, but I’m an eager learner! :)

jecoz

[-- Attachment #2: Type: text/html, Size: 2179 bytes --]