TL;DR: it seems to be a plymouth issue. On Sun, Nov 22, 2015 at 10:19:12PM +0000, Ben Hutchings wrote: > Can you try this: > > - Boot with the added parameter: break=top > - In the shell, run: > echo 1 >/proc/sys/kernel/sysrq > openvt > exit > - Once the system has hung, press Alt-SysRq, Alt-W, Alt-SysRq, Alt-L > - Switch to VT 2 then show the kernel log with: dmesg | less > - Send photos of all the call traces I haven't done all of the above, because I don't know what generates a SysRq on ThinkPad's keyboards. Do you know? I've a T440s. (I guess the web would probably tell me, but I'm writing this on a plane w/o net connection..., sorry about that.) Nonetheless I've used the shell obtained with openvt to compare the running processes in the linux 4.1 and 4.3 cases. In 4.1 (where the boot works) the processes asking for the luks password is "/lib/cryptsetup/askpass Please unlock disk sda3_crypt:" ; in 4.3 (where the boot doesn't work) is "plymouth ask-for-password --prompt Please unlock disk sda3_crypt". In both cases, the process with the next PID, after askpass/plymouth, is "/sbin/cryptsetup -T 1 --allow-discards open --type luks <DEVICE_NAME>". So this bug might deserve a reassign to plymouth, I guess. Still, why are kernels 4.2 and 4.3 triggering plymouth usage, whereas 4.1 is not? (FWIW, all my grub entries have "splash" on the kernel cmdline.) Cheers. -- Stefano Zacchiroli . . . . . . . zack@upsilon.cc . . . . o . . . o . o Maître de conférences . . . . . http://upsilon.cc/zack . . . o . . . o o Former Debian Project Leader . . . . . @zacchiro . . . . o o o . . . o . « the first rule of tautology club is the first rule of tautology club »
Attachment:
signature.asc
Description: PGP signature