diff options
author | Truxton Fulton <trux@truxton.com> | 2005-09-09 13:02:18 -0700 |
---|---|---|
committer | Linus Torvalds <torvalds@g5.osdl.org> | 2005-09-09 13:57:35 -0700 |
commit | 59f4e7d572980a521b7bdba74ab71b21f5995538 (patch) | |
tree | 65004b9875c8fa830de2879229ecf871de5525f4 /drivers | |
parent | 6e8dcee3e63f5a2cba4affff4bbb6e228f4b258a (diff) | |
download | op-kernel-dev-59f4e7d572980a521b7bdba74ab71b21f5995538.zip op-kernel-dev-59f4e7d572980a521b7bdba74ab71b21f5995538.tar.gz |
[PATCH] fix reboot via keyboard controller reset
I have a system (Biostar IDEQ210M mini-pc with a VIA chipset) which will
not reboot unless a keyboard is plugged in to it. I have tried all
combinations of the kernel "reboot=x,y" flags to no avail. Rebooting by
any method will leave the system in a wedged state (at the "Restarting
system" message).
I finally tracked the problem down to the machine's refusal to fully reboot
unless the keyboard controller status register had bit 2 set. This is the
"System flag" which when set, indicates successful completion of the
keyboard controller self-test (Basic Assurance Test, BAT).
I suppose that something is trying to protect against sporadic reboots
unless the keyboard controller is in a good state (a keyboard is present),
but I need this machine to be headless.
I found that setting the system flag (via the command byte) before giving
the "pulse reset line" command will allow the reboot to proceed. The patch
is simple, and I think it should be fine for everybody whether they have
this type of machine or not. This affects the "hard" reboot (as done when
the kernel boot flags "reboot=c,h" are used).
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Diffstat (limited to 'drivers')
0 files changed, 0 insertions, 0 deletions