Raphael Hertzog wrote:
clone 481312 -1
reassign -1 acpid 1.0.6-5.1
retitle -1 invoke-rc.d acpid restart fails
severity -1 important
thanks
On Thu, 15 May 2008, Bart Samwel wrote:
Hi Leo,
Thanks for the information. I will look into this. I actually think that
a combination of the first and third solution would be appropriate: we
should tolerate the absence, but we should put a very strong
recommendation in the manual to keep it. I actually think that the real
problem is in acpid -- if ACPI events are being delivered differently,
it should be able to handle that, and it should not fail if the old
mechanism is no longer present! Would you be so kind to report this as a
bug for the acpid package?
Bart, the proper way is to clone the bug yourself and reassign it to acpid
(like I just did). The postinst is doing nothing more than:
invoke-rc.d acpid restart >/dev/null
ACK, thanks for the reminder.
So the acpi init script shouldn't fail on restart really. This might be
somewhat related to #447339, not sure though. There might be a bigger
problem.
Definitely. I'm thinking that it simply doesn't work without
/proc/acpi/events, and that perhaps the newer mechanisms aren't
supported. This is not a simple bug. Depending on how badly the kernel
interfaces have been changed/broken, this threatens acpi-support as well.
On the acpi-support-base side, I believe it's safe to ignore failures from
that script so we should add "|| true" to our line.
Will do.
Cheers,
Bart
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
