Hi Kim, On Sat, May 31, 2025 at 11:41:49PM +0200, Kim Alvefur wrote: > Hi Salvatore, > > On Sat, May 31, 2025 at 08:53:26PM +0200, Salvatore Bonaccorso wrote: > > One additional preliminary question, since I missed to ask it in the > > earlier reply: Is this a regression you are seeing when updating to > > 6.12.27-1 (and later)? If so can you pinpoint the Debian revision > > where you first saw this happening? > > Looking at earlier kernel logs in the journal, this happened for the > first time in March, which was after I upgraded to Debian 13. > > Using Debian snapshots I tried a few older kernels, the earliest being > 6.12.6-1 from late 2024 and it happened there too. > > > In case we have a regression, would you be able to bisect the > > respective upstream stable versions to pinpoint the breaking commit, > > which would be helpful for the upstream report? > > I can certainly try, at least using repository snapshots to narrow it > down. It has been many years since I last compiled a kernel, so I could > do with a refresher there :) > > Any suggestions for how far back to go?
If you updates from bookworm to trixie then start with the bookworm based kernel and bisect there the debian image versions first until we get a closer range where an upstream change happened. Note I have forwarded the report to usptream but so far no reaction. It might actually be that with the trixie system and the older kernel you will hit now the issue as well as it is just undovered by having iio-sensor-proxy running (you could double check in your apt/dpkg logs if it was already installed under bookworm though). Once we have a closer range, bisecting the upstream changes would be the next step. I still have some hope that we get on board upstream people :) Regards, Salvatore
