Le 24/02/2024 à 14:08, Thorsten Leemhuis a écrit :
Raphaël: would nevertheless still be good if you could identify which of the three causes the problem, as then the developers might consider simply reverting it.
Hi,
It can't be the third one (43527a0) since I clearly remember that I experienced the regression before it was applied to the Debian kernel.
So I'll try applying only the first one (46a0a2c), and report.
(in the meantime I crafted a quick and dirty Debian package to build the module with DKMS, so it will be easy)
Regards,