[plug] Intel igb module issues

Onno Benschop onno at itmaze.com.au
Thu Dec 14 10:27:08 AWST 2023


You said that it works with a version of Kubuntu, but not with a version of
Debian.

Whilst there are undoubtedly many source code differences, my first "best
guess" would be that there is a specific kernel option (perhaps even more
than one) that's causing this. Comparing those is a much smaller job than
diffing the two kernels and as you already pointed out, using a back-ported
kernel did not solve the problem.

The way I'd approach this is to use the source package (dpkg-src) for both
kernels, running in their respective OS, and compile each kernel package
from source using the standard tools to do so.

If you then discover that the same happens, ie. your source compiled
Kubuntu kernel works and the Debian compiled kernel does not, I'd start
comparing the two config files for those two kernels.

You could then potentially compile the same Debian kernel using the Kubuntu
config and determine if that solves the problem. If it does not. the issue
is likely in the kernel source, if it does, the issue is likely in the
kernel config.

If what you find is that neither works after compiling from source, explore
what the difference is between a compiled from source package is when
compared with the supplied binary package.

Ultimately you're looking for a needle in a haystack. Starting with git and
logs and source code is in my experience not the way to a swift resolution.

Finally, git blame is a useful way to discover who changed a relevant file
and how it was changed. While it's unlikely that the change was caused by a
patch of the actual kernel driver - you know, that would be a "doh" moment
- you can use tools to create a chart of dependencies and see what links
into the driver file.

For example: make -Bnd | make2graph | dot -Tsvg -o out.svg
Source: https://unix.stackexchange.com/a/283501

Disclaimer: There are MANY moving parts here. This could use as easily be a
tool difference, for example a different version of the compiler tool
chain, so tread wearily.


On Thu, 14 Dec 2023 at 09:51, Nick Bannon <nick at ucc.gu.uwa.edu.au> wrote:

> On Thu, Dec 14, 2023 at 09:18:02AM +0800, Dean Bergin wrote:
> > Seems as though the backports kernel 6.5.10-1~bpo12+1 unfortunately did
> not
> > solve the problem.
>
> As the driver loads, it may have kernel log messages about having to
> load some firmware blob(s). Are those different? try copying the "good"
> firmware files to the "bad" system? also:
>
> dmesg -T |grep firmware
> dpkg -l '*firmware*'
> ls /lib/firmware/intel
>
> Nick.
>
> --
>    Nick Bannon   | "I made this letter longer than usual because
> nick-sig at rcpt.to | I lack the time to make it shorter." - Pascal
> _______________________________________________
> PLUG discussion list: plug at plug.org.au
> http://lists.plug.org.au/mailman/listinfo/plug
> Committee e-mail: committee at plug.org.au
> PLUG Membership: http://www.plug.org.au/membership
>


-- 
Onno Benschop

()/)/)()        ..ASCII for Onno..
|>>?            ..EBCDIC for Onno..
--- -. -. ---   ..Morse for Onno..

If you need to know: "What computer should I buy?" http://goo.gl/spsb66

ITmaze   -   ABN: 56 178 057 063   -  ph: 04 1219 8888   -
onno at itmaze.com.au
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.plug.org.au/pipermail/plug/attachments/20231214/a9f571df/attachment.html>


More information about the plug mailing list