Le 12.05.2019 à 22.14, Theo de Raadt a écrit :
Stéphane Guedon <[email protected]> wrote:
Le 08.05.2019 à 22.28, Stefan Sperling a écrit :
On Wed, May 08, 2019 at 10:19:48PM +0200, Stéphane Guedon wrote:
Actually, just to check, you ask me to download kernels from the past and
try to boot them. Just like that.
Right ?
Yes.
I did that with two kernels : first the one from the 1st of February. Then
one from the 6th of January and they both react this way (device timeout)
when I activate the device with ifconfig. Either I do it wrong, or I should
find kernels from November-ish times.
(I wish to help and not being too annoying. :) )
Does a 6.4 kernel actually still work?
If you can't find binaries please try to build old kernels and test them.
You can also try back-dating the files I have changed in my athn commits.
How exactly you go about this isn't important; we need to find out somehow
which commit broke things.
I have some answers !
Today, I downloaded via cvs kernel sources of various dates and compiled.
A kernel build from sources of 15 weeks ago boots without trouble,
whereas a kernel from sources 14 weeks ago produces the bug !
Please provide exact dates. If you want to really have a moment of joy,
you are 90% of the way so 10% more effort should find the specific commit!
So I actually did it again with exact dates this time. And narrowed the
thing just a bit more (sic).
I have a kernel built from the sources from the 1st of february running
now, no trouble (cvs status for the Makefile in amd64/compile/GENERIC.MP
indicate Sticky Date: 2019.01.31.23.00.00 ).
The kernel running from the 3rd of february sources is buggy (athn
device timeout).
I hope it can help.
