Another F11 build for the XO-1 has been released

[Yioryos Asprobounitis]

--- On Thu, 8/6/09, Paul Fox <pgf at laptop.org> wrote:

> From: Paul Fox <pgf at laptop.org>
> Subject: Re: Another F11 build for the XO-1 has been released
> To: "OLPC Development" <devel at lists.laptop.org>, "Fedora OLPC List" <fedora-olpc-list at redhat.com>
> Date: Thursday, August 6, 2009, 3:59 PM
> yioryos wrote:
>  > 
>  > > 2009/8/6 Steven M. Parrish <smparrish at gmail.com>:
>  > > > I have released a new build of F11 for the
> XO-1 named OS4
>  > > >
>  > > > You can find it at.... http://dev.laptop.org/~smparrish/xo-1/builds/
>  ...
>  > Olpc-powerd works up to sleeping but fails to blank
> and shutdown.OK on lid and 
>  > power button suspend/shutdown though. rtcwake still
> does not work, so I guess 
>  > this must be the reason.
> 
> i've now installed a fresh os4, and rtcwake works just
> fine. [1]
> 
> i'm becoming suspicious that the rtcwake issue is either a
> bug in
> rtcwake itself, or a separate system misconfig
> problem.  i did
> see a case earlier today on a different machine where
> rtcwake was
> not waking properly.  an strace showed that the value
> being
> written to the RTC was exactly 1 hour too large. 
> playing with
> the timezone and hwclock eventually got it working, but
> it's
> arguable that since the command in question is supposed to
> set a
> relative alarm ("rtcwake -m mem -a -s 5"), the timezone
> and
> system time shouldn't enter into the calculation at all
> (and now
> that i think of if, i'm wondering why the "-a" option is
> even
> there).  i'm thinking there may be a DST issue in
> there
> somewhere.
> 
> paul
> 
> [1] on the other hand, someone has just informed me that:
> 
>   > On the latest SoaS3 build ... the command 
>   > completes immediately, with the message
> "rtcwake:  /dev/rtc0 not 
>   > enabled for wakeup events".
> 
> could someone with SoaS3 insider info please verify that
> the XO-1
> kernel is being built from the olpc-2.6.30 branch of
> olpc-2.6
> repository?  i've seen that message before, but only
> on misconfigured
> kernels.
> 

I set my timezone to hwclock time (GMT). In this setting trcwake in 10% of the cases wakes up in 10.5-12.5 sec. in ~20% does nothing after I hit <return> ( I have no problems with the return key) and in the rest of the cases wakes up immediately. In all cases wakes up with the message
rtcwake: wakeup from "mem" using /dev/rtc0 at <date>
My os4 just has ntp, ntpdate, wget and yumex and runs with the 2.6.30_xo1-20090724.1605.1.olpc.d56ad9b kernel. Other than that (and the font fixing) is pristine.
I have the filling that timezone change did not improve much the situation. I was getting occasional correct wapeups though not so often. 


> paul
> =---------------------
>  paul fox, pgf at laptop.org
> 
> _______________________________________________
> Fedora-olpc-list mailing list
> Fedora-olpc-list at redhat.com
> https://www.redhat.com/mailman/listinfo/fedora-olpc-list
>