* pmac_halt() bug
@ 2002-11-20 21:03 Tim Seufert
2003-11-10 16:33 ` Igal Baevsky
0 siblings, 1 reply; 8+ messages in thread
From: Tim Seufert @ 2002-11-20 21:03 UTC (permalink / raw)
To: linuxppc-dev
In platforms/pmac_setup.c, pmac_halt() is defined as follows:
pmac_halt(void)
{
pmac_power_off();
}
Arguably, this is a bug because there should be a distinction between
halt and poweroff. I've seen posts on user mailing lists from at least
two different people who want to halt without actually powering off,
but cannot do so because of this bug. (I think the reasons have
something to do with attempting to get their machines to turn back on
automatically when power is restored after a UPS-initiated shutdown.)
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: pmac_halt() bug
2002-11-20 21:03 pmac_halt() bug Tim Seufert
@ 2003-11-10 16:33 ` Igal Baevsky
2003-11-10 23:59 ` Benjamin Herrenschmidt
2003-11-11 7:28 ` Tim Seufert
0 siblings, 2 replies; 8+ messages in thread
From: Igal Baevsky @ 2003-11-10 16:33 UTC (permalink / raw)
To: linuxppc-dev
Tim Seufert wrote:
>
> In platforms/pmac_setup.c, pmac_halt() is defined as follows:
>
> pmac_halt(void)
> {
> pmac_power_off();
> }
>
> Arguably, this is a bug because there should be a distinction between
> halt and poweroff. I've seen posts on user mailing lists from at least
> two different people who want to halt without actually powering off,
> but cannot do so because of this bug. (I think the reasons have
> something to do with attempting to get their machines to turn back on
> automatically when power is restored after a UPS-initiated shutdown.)
>
>
>
>
Tim,
Is there any way to halt PMAC machine with no powering it off ?
I will realy appretiate your answer - the problem you mentioned above is
the exact problem I'm dealing with .
Thanks.
Igal
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: pmac_halt() bug
2003-11-10 16:33 ` Igal Baevsky
@ 2003-11-10 23:59 ` Benjamin Herrenschmidt
2003-11-11 7:37 ` Tim Seufert
2003-11-11 7:28 ` Tim Seufert
1 sibling, 1 reply; 8+ messages in thread
From: Benjamin Herrenschmidt @ 2003-11-10 23:59 UTC (permalink / raw)
To: Igal Baevsky; +Cc: linuxppc-dev list
On Tue, 2003-11-11 at 03:33, Igal Baevsky wrote:
> Tim,
> Is there any way to halt PMAC machine with no powering it off ?
> I will realy appretiate your answer - the problem you mentioned above is
> the exact problem I'm dealing with .
> Thanks.
> Igal
I see no good reason why you want to do such a halt that just spins and
do nothing but suck power...
If the problem is to have the machine turn back on, I fail to see how
halting it this way would help. The only thing I can imagine is enabling
the server mode option. I added that feature to my very latest trees
for Core99 machines and later (AGP machines), and there used to be a
tool doing that for CUDA based machines.
Ben.
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: pmac_halt() bug
2003-11-10 16:33 ` Igal Baevsky
2003-11-10 23:59 ` Benjamin Herrenschmidt
@ 2003-11-11 7:28 ` Tim Seufert
1 sibling, 0 replies; 8+ messages in thread
From: Tim Seufert @ 2003-11-11 7:28 UTC (permalink / raw)
To: Igal Baevsky; +Cc: linuxppc-dev
On Nov 10, 2003, at 8:33 AM, Igal Baevsky wrote:
> Tim Seufert wrote:
>>
>> In platforms/pmac_setup.c, pmac_halt() is defined as follows:
>>
>> pmac_halt(void)
>> {
>> pmac_power_off();
>> }
> Tim,
> Is there any way to halt PMAC machine with no powering it off ?
> I will realy appretiate your answer - the problem you mentioned above
> is
> the exact problem I'm dealing with .
Assuming the above is still the implementation of pmac_halt(), there's
no way, short of modifying the kernel.
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: pmac_halt() bug
2003-11-10 23:59 ` Benjamin Herrenschmidt
@ 2003-11-11 7:37 ` Tim Seufert
2003-11-11 8:25 ` Benjamin Herrenschmidt
0 siblings, 1 reply; 8+ messages in thread
From: Tim Seufert @ 2003-11-11 7:37 UTC (permalink / raw)
To: Benjamin Herrenschmidt; +Cc: linuxppc-dev list, Igal Baevsky
On Nov 10, 2003, at 3:59 PM, Benjamin Herrenschmidt wrote:
> On Tue, 2003-11-11 at 03:33, Igal Baevsky wrote:
>
>> Tim,
>> Is there any way to halt PMAC machine with no powering it off ?
>> I will realy appretiate your answer - the problem you mentioned above
>> is
>> the exact problem I'm dealing with .
>> Thanks.
>> Igal
>
> I see no good reason why you want to do such a halt that just spins and
> do nothing but suck power...
>
> If the problem is to have the machine turn back on, I fail to see how
> halting it this way would help. The only thing I can imagine is
> enabling
> the server mode option.
That's precisely it. Some (all?) pmacs with server mode don't
automatically turn on after an intentional shutdown, only after power
is lost. So the idea is to have the UPS signal the computer to halt,
and then have the UPS cut power once it's safe. Later the UPS switches
power back on and the computer will automatically boot.
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: pmac_halt() bug
2003-11-11 7:37 ` Tim Seufert
@ 2003-11-11 8:25 ` Benjamin Herrenschmidt
2003-11-11 8:53 ` Ethan Benson
0 siblings, 1 reply; 8+ messages in thread
From: Benjamin Herrenschmidt @ 2003-11-11 8:25 UTC (permalink / raw)
To: Tim Seufert; +Cc: linuxppc-dev list, Igal Baevsky
>
> That's precisely it. Some (all?) pmacs with server mode don't
> automatically turn on after an intentional shutdown, only after power
> is lost. So the idea is to have the UPS signal the computer to halt,
> and then have the UPS cut power once it's safe. Later the UPS switches
> power back on and the computer will automatically boot.
Ok, 2 things:
- How does the UPS knows it's "safe" ?
- And right, I need to disable server mode to be able to shut down :) So
I indeed to that on those machines
Ben.
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: pmac_halt() bug
2003-11-11 8:25 ` Benjamin Herrenschmidt
@ 2003-11-11 8:53 ` Ethan Benson
0 siblings, 0 replies; 8+ messages in thread
From: Ethan Benson @ 2003-11-11 8:53 UTC (permalink / raw)
To: linuxppc-dev list
On Tue, Nov 11, 2003 at 07:25:25PM +1100, Benjamin Herrenschmidt wrote:
>
> >
> > That's precisely it. Some (all?) pmacs with server mode don't
> > automatically turn on after an intentional shutdown, only after power
> > is lost. So the idea is to have the UPS signal the computer to halt,
> > and then have the UPS cut power once it's safe. Later the UPS switches
> > power back on and the computer will automatically boot.
>
> Ok, 2 things:
>
> - How does the UPS knows it's "safe" ?
varies, the UPS i have has a smart monitoring program run over a
serial line, the UPS informs the daemon when it believes there is only
5 minutes of power remaining, and the daemon initiates a halt, the UPS
shuts down when there is only 2 minutes of power remaining, so your
shutdown process better not take longer then 3 minutes.
> - And right, I need to disable server mode to be able to shut down :) So
> I indeed to that on those machines
and there is the problem. on x86 many bioses provide an option which
lets you choose behavior post powerfail, on an x86 i have one option
is to bootup whenever AC power is lost and then restored, regardless
of previous shutdown state, for others the only way is to just halt
the machine without a powerdown.
pmac seems to have no way to accomplish either atm.
--
Ethan Benson
http://www.alaska.net/~erbenson/
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
^ permalink raw reply [flat|nested] 8+ messages in thread
* RE: pmac_halt() bug
@ 2003-11-11 12:03 Baevsky, Igal
0 siblings, 0 replies; 8+ messages in thread
From: Baevsky, Igal @ 2003-11-11 12:03 UTC (permalink / raw)
To: 'Benjamin Herrenschmidt', Tim Seufert
Cc: linuxppc-dev list, Baevsky, Igal
When in MacOS I can just unplug the power cable and put it back and the
system will power on. On linux it's not working - after power cable is back,
the system is not powering on.
After playing with /etc/init.d/halt script I managed to bring system to the
state where all processes are stoped and all FS are unmounted but the system
is still up.
As I understand, I need some module or tool to write something to the MAC
bios or eeprom whatever, during linux startup. Can somebody tell me what can
I use ?
Thanks,
Igal.
-----Original Message-----
From: Benjamin Herrenschmidt [mailto:benh@kernel.crashing.org]
Sent: Tue, November 11, 2003 10:25
Subject: Re: pmac_halt() bug
> That's precisely it. Some (all?) pmacs with server mode don't
> automatically turn on after an intentional shutdown, only after power
> is lost. So the idea is to have the UPS signal the computer to halt,
> and then have the UPS cut power once it's safe. Later the UPS switches
> power back on and the computer will automatically boot.
Ok, 2 things:
- How does the UPS knows it's "safe" ?
- And right, I need to disable server mode to be able to shut down :) So
I indeed to that on those machines
** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2003-11-11 12:03 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-11-20 21:03 pmac_halt() bug Tim Seufert
2003-11-10 16:33 ` Igal Baevsky
2003-11-10 23:59 ` Benjamin Herrenschmidt
2003-11-11 7:37 ` Tim Seufert
2003-11-11 8:25 ` Benjamin Herrenschmidt
2003-11-11 8:53 ` Ethan Benson
2003-11-11 7:28 ` Tim Seufert
-- strict thread matches above, loose matches on Subject: below --
2003-11-11 12:03 Baevsky, Igal
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).