On Sun, 2018-10-28 at 12:04 +0100, Hans de Goede wrote: > Hi, > > On 26-10-18 17:29, Andy Shevchenko wrote: > > On Fri, Oct 26, 2018 at 03:02:10PM +0200, Hans de Goede wrote: > > > Hi, > > > > > > On 26-10-18 14:54, Andy Shevchenko wrote: > > > > On Thu, Jul 26, 2018 at 09:15:00PM +0200, Lukas Wunner wrote: > > > > > > > > > Andy, why did > > > > > you dig this out now? Cleaning stale bugzilla/patchwork > > > > > entries? :-) > > > > > > > > Lukas, thanks a lot for your testing! > > > > > > > > The patch popped up in regard to my search for some other stuff > > > > in Fedora > > > > kernels. Fedora kernels (used to?) have this patch for years. > > > > > > I just checked and we (Fedora) still have it, for some reason it > > > is grouped > > > together with a bunch of secure boot patches which are very hard > > > to > > > get upstream (we are still working on this, but it is a very slow > > > and painful process). > > > > > > I guess we never re-tried pushing this upstream because it was > > > grouped > > > together with the secure-boot stuff. > > > > > > But it was originally submitted upstream and still is valid > > > AFAICT. > > > > The question is how to reproduce the issue? > > I don't know I guess it speeds up the boot and maybe suppresses a > bunch > of errors from trying to probe the non existent i8042 controller. > > Bastien, do you remember what the issue was this patch fixes? It's the same bug that Lukas mentioned: https://www.spinics.net/lists/linux-input/msg57621.html with the firmware that was available at the time claiming that the laptop had a i8042 controller when it didn't have anything there. The problem was a close to 10 seconds pause on boot, not just 1 second as Lukas tested, which is why I actually went to the trouble of disabling it. It might be that there were changes in the kernel that reduced that pause in any case. If Lukas says that the newer firmwares fix this bug, I'm fine with dropping this patch. I would expect folks running Linux on Macs to keep a MacOS partition around for the purpose of updating their firmwares, as that's currently the only way to fetch firmware security updates. And if users haven't updated their firmware, it looks like the pause will also be reduced. If that's not the case, I would expect us to get bug reports soon enough. Cheers
