Re: Xen & suspend

Re: Xen & suspend

[Konrad Rzeszutek Wilk]

On Sun, Apr 15, 2012 at 04:57:21PM +0200, Alessandro Di Federico wrote:
> Hello, I want to use Xen on my notebook so it's fundamental to have the
> possibility to suspend the machine.
> 
> I know there's some ongoing work to add S3 suspend support to Xen, in
> particular I've tried this:
> 
> https://git.kernel.org/?p=linux/kernel/git/konrad/xen.git;a=shortlog;h=refs/heads/devel/acpi-s3.v9
> 
> But when I try to suspend the system simply freezes.
> 
> Can you give me some hints to make it work or try to make a better
> diagnosis of the problem?

There is a new version of it. But before I point you to it - what version
of Linux are you running and want to try the patches on?

> 
> Thanks!
> Ale

RE: Xen suspend

[Neugebauer, Rolf]

I don't think anyone has looked into power management and Xen. Xen
doesn't have any special hooks for this at all and some things like APM
might work because dom0 is privileged enough. However, give that even in
Linux power management is still quite messy and unstable with Xen YMMV
even more ;(

With the plan for 2.1 to move more of the platform init code into dom0
there is a chance that Xen will be more compatible with Linux for things
like power management as well.

Rolf

> -----Original Message-----
> From: xen-devel-admin@lists.sourceforge.net [mailto:xen-devel-
> admin@lists.sourceforge.net] On Behalf Of Kurt Garloff
> Sent: 30 November 2004 09:47
> To: xen-devel@lists.sourceforge.net
> Subject: [Xen-devel] Xen suspend
> 
> Hi,
> 
> I'm running Xen-2.0.1 and a self-compiled 2.6.9-xen0 on my IBM T40.
> 
> When I close the lid the following happens:
> (a) running natively (a SL92 kernel, 2.6.9-rc2 with some
improvements):
>     The powersaved is configured to do a suspend-to-ram. The OS
suspends
>     the devices and goes to sleep. Waking up again, the devices get
>     reinitialized and everything Works like a charm.
>     Only downside is the power consumption in suspended mode.
>     It's still about 4W.
> (b) Running Xen
>     The laptop goes into suspend mode immediately.
>     It also wakes up again.
>     The OS does not seem involved in suspending at all. dmesg does not
>     show any signs of this event not even xm dmesg does.
>     After waking up, however, the touchpad (synaptics) in domain 0 is
>     unusable and the kernel spits
>     TouchPad at isa0060/serio1/input0 lost sync at byte 1
>     I'm unsure whether other devices are in a sane state.
>     The next xm create killed the machine :-(
> 
> Is this supposed to work?
> Shouldn't Xen ask the domain0 kernel to suspend the devices?
> 
> Regards,
> --
> Kurt Garloff                   <kurt@garloff.de>             [Koeln,
DE]


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://productguide.itmanagersjournal.com/

Xen suspend

[Kurt Garloff]

[-- Attachment #1: Type: text/plain, Size: 1165 bytes --]

Hi,

I'm running Xen-2.0.1 and a self-compiled 2.6.9-xen0 on my IBM T40.

When I close the lid the following happens:
(a) running natively (a SL92 kernel, 2.6.9-rc2 with some improvements):
    The powersaved is configured to do a suspend-to-ram. The OS suspends
    the devices and goes to sleep. Waking up again, the devices get 
    reinitialized and everything Works like a charm.
    Only downside is the power consumption in suspended mode.
    It's still about 4W.
(b) Running Xen
    The laptop goes into suspend mode immediately.
    It also wakes up again.
    The OS does not seem involved in suspending at all. dmesg does not
    show any signs of this event not even xm dmesg does.
    After waking up, however, the touchpad (synaptics) in domain 0 is 
    unusable and the kernel spits
    TouchPad at isa0060/serio1/input0 lost sync at byte 1
    I'm unsure whether other devices are in a sane state.
    The next xm create killed the machine :-(

Is this supposed to work?
Shouldn't Xen ask the domain0 kernel to suspend the devices?

Regards,
-- 
Kurt Garloff                   <kurt@garloff.de>             [Koeln, DE]

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Xen suspend

[Ian Pratt]

> (b) Running Xen
>     The laptop goes into suspend mode immediately.
>     It also wakes up again.
>     The OS does not seem involved in suspending at all. dmesg does not
>     show any signs of this event not even xm dmesg does.
>     After waking up, however, the touchpad (synaptics) in domain 0 is 
>     unusable and the kernel spits
>     TouchPad at isa0060/serio1/input0 lost sync at byte 1
>     I'm unsure whether other devices are in a sane state.
>     The next xm create killed the machine :-(
> 
> Is this supposed to work?
> Shouldn't Xen ask the domain0 kernel to suspend the devices?

I'm afraid 2.0 currently has no support for either APM or ACPI
power management. All bets are off when you shut the lid :-(

The roadmap for 2.1/3.0 has us heading in the right direction
with making more of this stuff visible to domain 0 so we can use
the existing Linux code. 

For the moment, it mightn't be too hard to hack software
suspend/resume to do the right thing...

Ian


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now. 
http://productguide.itmanagersjournal.com/