If you dig deep into pm-hibernate code, you'll see that the no-chvt quirk is
used in this case. According to my experience this is bad, I think this
switches keyboard into xlate mode (same as you would doing alt-sysrq-R, but
can't be really switched back into normal mode). The only way I discovered to
prevent this is disabling the no-chvt quirk. My ugly hack consists of
commenting a line like this (comment already included):
# add_parameters --quirk-no-chvt
in the file /usr/lib/pm-utils/sleep.d/98video-quirk-db-handler: 360 (on my
Debian sid). I have pm-utils 1.3.0
This code is part of the suspend|hibernate case when using_kms."
For the record, on the tuxonice mailing list, the following possible solution was given, see http:// lists.tuxonice. net/pipermail/ tuxonice- users/2010- August/ 000511. html:
" If you are on intel graphic chipset(KMS) and you use pm-utils, I think I
have some clues about the issue.
If you dig deep into pm-hibernate code, you'll see that the no-chvt quirk is pm-utils/ sleep.d/ 98video- quirk-db- handler: 360 (on my
used in this case. According to my experience this is bad, I think this
switches keyboard into xlate mode (same as you would doing alt-sysrq-R, but
can't be really switched back into normal mode). The only way I discovered to
prevent this is disabling the no-chvt quirk. My ugly hack consists of
commenting a line like this (comment already included):
# add_parameters --quirk-no-chvt
in the file /usr/lib/
Debian sid). I have pm-utils 1.3.0
This code is part of the suspend|hibernate case when using_kms."