> I've been trying to get pmud working again for the
> last couple of weeks with no luck after it
> mysteriously died.
>
> Basically, I've got a G3 14" running kernel 2.6.6
> and...
G3 Pismo, or what sort of hardware?
> 1) when i shut the lid, my hard disk just churns away,
> the snooze light never comes on. (Except for the first
> closing after a reboot)
>
> 2) snooze will work once, after the machine has been
> rebooted, every time after "snooze" appears to do
> nothing and "snooze -f" will report: "PMU_IOC_SLEEP:
> Device or resource busy"
Seems like the kernel never fully recovered from the first sleep. Does
switching to a 2.4 kernel, or perhaps another 2.6 kernel, improve things?
>
> 3) "pmud -d" reports before snoozing:
> pmud[3835]: pmud [treshold = 420, margin = 15]
> started
> pmud[3835]: opening /dev/adb
> pmud[3835]: opening /dev/pmu
> pmud[3835]: bind socket
> pmud[3835]: daemon stopped (bind socket)
> pretty much all the time.
>
> 4) "pmud -d" reports after snoozing:
> pmud[3441]: pmud [treshold = 420, margin = 15]
> started
> pmud[3441]: opening /dev/adb
> pmud[3441]: opening /dev/pmu
> pmud[3441]: pmu_op(0xea, ..., version req)
> PMU version 12: iBook/G3 Pismo/G4 Titanium
> pmud[3441]: PMU version 12: iBook/G3 Pismo/G4
> Titanium
> pmud[3441]: starting monitoring loop
> pmud[3441]: pmu_op(0xdc, ..., switch req)
> pmud[3441]: pmu_op(0x6f, ..., smart battery req)
> pmud[3441]: pmu_op(0x6f, ..., smart battery req)
> pmud[3441]: timeleft = -1
> pmud[3441]: running /etc/power/pwrctl maximum ac
> pmud[3441]: PMU interrupt:
> pmud[3441]: 40
> pmud[3441]: 14
> pmud[3441]: pmu_op(0xdc, ..., switch req)
> pmud[3441]: pmu_op(0x6f, ..., smart battery req)
> pmud[3441]: pmu_op(0x6f, ..., smart battery req)
> pmud[3441]: timeleft = -1
> pmud[3441]: PMU interrupt:
> pmud[3441]: 40
> pmud[3441]: 14
No idea what this is about...
>
> 5) /etc/init.d/pmud status reports after snoozing:
> pmud dead but pid file exists
pmud did exit after encountering the device busy error.
> 6) "strace pmud" reports
After it once crashed, I suppose? Otherwise the message below is OK:
> many things then...
> bind(8, {sa_family=AF_INET, sin_port=htons(879),
> sin_addr=inet_addr("127.0.0.1")}, 16) = -1 EADDRINUSE
> (Address already in use)
Only one instance of pmud may be running at any time.
I'm sorry I cannot investigate this myself - my Lombard was stolen only
last Friday and I haven't yet got myself a new Powerbook.
Michael
