>Hi Tom,
> Thanks for your help to review these changes.
>
>> Just one comment from me. The rest looks good.
>>
>> usr/src/cmd/acpihpd/svc-acpihpd:
>> line 22 I'm not sure that I understand the purpose of this
>> sleep.
>It's used to work around a race condition between SMF restarter
>and the acpihpd service. Without the sleep at line 22, the SMF
>restarter may keep on restarting the acpihpd service and finally
>put it into maintain status.
I think this points to a bug somewhere else.
In my experience a sleep is always too short to make sure you always
miss the window.
But perhaps this is a known bug; there are many strange uses of sleep
svc-hal has this:
if smf_is_nonglobalzone; then
/usr/sbin/svcadm disable $SMF_FMRI
echo "$SMF_FMRI is not supported in a local zone"
sleep 5 &
exit $SMF_EXIT_OK
fi
Casper
