Michael 'Mickey' Lauer | 8 Jan 2011 19:01
Picon
Gravatar

test

Are we still up and running?

Cheers,
--

-- 
:M:
Denis 'GNUtoo' Carikli | 8 Jan 2011 19:31
Favicon

Resources handling plugins

hi,
Previously, by accident, having multiple plugins for handling one
resource worked.
It doesn't work anymore.
Instead Mickeyl told me that he will come up with something but he needs
usage scenario.

Here's what needs to be done on the htcdream to activate wifi:
modprobe msm_wifi
cd /sys/class/rfkill/rfkill1/
echo 1 > state
ifconfig wlan0 up

That needs to be acompaniated with a firmware loader(be it udev or
fsodeviced when devtmpfs is used instead of udev)

iwconfig wlan0 power on doesn't need to be executed right now because
PSM doesn't work.(if it is it doesn't do anything on the current kernel)

For bluetooth loading here's what needs to be done:
echo 1 > /sys/class/rfkill/rfkill0/state
hciattach     -n -s 115200 /dev/ttyHS0 texasalt 115200 flow
echo 1 > /sys/class/rfkill/rfkill1/state #if rfkill1 is still bluetooth
class and not wifi(for instance if msm_wifi is not loaded)
hciconfig hci0 up

I'll leave the palm-pre, the nokia900 and the om-gta02 to other people to
describe them better than I could.

Denis.
(Continue reading)

Simon Busch | 8 Jan 2011 22:07
Picon

Re: Resources handling plugins

On 01/08/11 at 07:31pm, Denis 'GNUtoo' Carikli wrote:
> hi,
> Previously, by accident, having multiple plugins for handling one
> resource worked.
> It doesn't work anymore.
> Instead Mickeyl told me that he will come up with something but he needs
> usage scenario.

You mean the wifi part, right? What will we have than different to the 
current situation, as we now a have a power control plugin for each 
architecture we support. Which other way do you want to go? Implementing 
the steps mentioned below (module loading, rfkill, ifconfig, psm) in such a 
generic way that they apply's for every machine?

> I'll leave the palm-pre, the nokia900 and the om-gta02 to other people to
> describe them better than I could.

For palmpre it's the following for wifi (as already implemented in a power 
control plugin):
- load 'sirloin_wifi' module to take wifi chip out of reset
- load 'libertas' module
- load 'libertas_sdio' module
- load firmware by udev or our firmware_loader
- use wifi (even ifconfig eth1 power off/on works)

For bluetooth there is currently no process as nobody has ever looked for 
it. But I think it should be something like this:
- enable chip via sysfs
- attach the hsuart responsible for bluetooth via hci_attach
- let bluez handle the rest
(Continue reading)

Jakob | 8 Jan 2011 22:43

Re: test

yup, we are

On 1/8/11, Michael 'Mickey' Lauer <mickey <at> vanille-media.de> wrote:
> Are we still up and running?
>
> Cheers,
> --
> :M:
>
>
> _______________________________________________
> Smartphones-userland mailing list
> Smartphones-userland <at> linuxtogo.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/smartphones-userland
>
Klaus Kurzmann | 9 Jan 2011 12:42
Picon
Gravatar

Re: Resources handling plugins

Am Samstag, 8. Januar 2011, 19:31:01 schrieb Denis 'GNUtoo' Carikli:
> hi,
> Previously, by accident, having multiple plugins for handling one
> resource worked.
> It doesn't work anymore.
> Instead Mickeyl told me that he will come up with something but he needs
> usage scenario.
> 
> Here's what needs to be done on the htcdream to activate wifi:
> modprobe msm_wifi
> cd /sys/class/rfkill/rfkill1/
> echo 1 > state
> ifconfig wlan0 up
> 
> That needs to be acompaniated with a firmware loader(be it udev or
> fsodeviced when devtmpfs is used instead of udev)
> 
> iwconfig wlan0 power on doesn't need to be executed right now because
> PSM doesn't work.(if it is it doesn't do anything on the current kernel)
> 
> For bluetooth loading here's what needs to be done:
> echo 1 > /sys/class/rfkill/rfkill0/state
> hciattach     -n -s 115200 /dev/ttyHS0 texasalt 115200 flow
> echo 1 > /sys/class/rfkill/rfkill1/state #if rfkill1 is still bluetooth
> class and not wifi(for instance if msm_wifi is not loaded)
> hciconfig hci0 up
> 
> I'll leave the palm-pre, the nokia900 and the om-gta02 to other people to
> describe them better than I could.

(Continue reading)

Klaus Kurzmann | 9 Jan 2011 13:54
Picon
Gravatar

Re: Resources handling plugins

Am Samstag, 8. Januar 2011, 22:07:51 schrieb Simon Busch:
> On 01/08/11 at 07:31pm, Denis 'GNUtoo' Carikli wrote:
> > hi,
> > Previously, by accident, having multiple plugins for handling one
> > resource worked.
> > It doesn't work anymore.
> > Instead Mickeyl told me that he will come up with something but he needs
> > usage scenario.
> 
> You mean the wifi part, right? What will we have than different to the 
> current situation, as we now a have a power control plugin for each 
> architecture we support. Which other way do you want to go? Implementing 
> the steps mentioned below (module loading, rfkill, ifconfig, psm) in such
> a  generic way that they apply's for every machine?

more something like being able to use for example rfkill in a machine specific 
control plugin without code duplication I think

--

-- 
Klaus 'mrmoku' Kurzmann
Rico Rommel | 30 Jan 2011 10:16
Picon
Favicon

phoneuid crashes

Hi all,

yesterday I rebuilt cornucopia, libfso-glib and all the shr things from git 
and now phoneuid is crashing everytime i run phoneui-contacts or phoneui-
messages.

DISPLAY=":0" gdb /usr/bin/phoneuid:

Program received signal SIGSEGV, Segmentation fault.
0x403d768c in free_smartphone_pim_messages_query (self=0x0, query=0x87120, 
_callback_=0x402e7538, _user_data_=0x152ab8) at freesmartphone-pim.c:13866

13866           FREE_SMARTPHONE_PIM_MESSAGES_GET_INTERFACE (self)->query 
(self, query, _callback_, _user_data_);

Any idea, what's going wrong?

Rico
_______________________________________________
Smartphones-userland mailing list
Smartphones-userland <at> linuxtogo.org
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/smartphones-userland
Timo Juhani Lindfors | 30 Jan 2011 11:10
Picon
Picon
Favicon
Gravatar

Re: phoneuid crashes

Rico Rommel <rico <at> bierrommel.de> writes:
> 0x403d768c in free_smartphone_pim_messages_query (self=0x0, query=0x87120, 

I haven't used phoneuid but: are you using packages from debian or
some other distro or building them yourself? Can you provide out of
gdb "backtrace full" command?
Rico Rommel | 30 Jan 2011 12:58
Picon
Favicon

Re: phoneuid crashes

Am Sonntag, 30. Januar 2011, 11:10:37 schrieb Timo Juhani Lindfors:
> Rico Rommel <rico <at> bierrommel.de> writes:
> > 0x403d768c in free_smartphone_pim_messages_query (self=0x0,
> > query=0x87120,
> 
> I haven't used phoneuid but: are you using packages from debian or
> some other distro or building them yourself? Can you provide out of
> gdb "backtrace full" command?

I'm using debian but build everything related to efl, fso and shr myself.

Here's the full backtrace

http://pastebin.com/jDqp8c3k
_______________________________________________
Smartphones-userland mailing list
Smartphones-userland <at> linuxtogo.org
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/smartphones-userland
Klaus Kurzmann | 30 Jan 2011 13:36
Picon
Gravatar

Re: phoneuid crashes

Am Sonntag, 30. Januar 2011, 12:58:59 schrieb Rico Rommel:
> Am Sonntag, 30. Januar 2011, 11:10:37 schrieb Timo Juhani Lindfors:
> > Rico Rommel <rico <at> bierrommel.de> writes:
> > > 0x403d768c in free_smartphone_pim_messages_query (self=0x0,
> > > query=0x87120,
self=0x0 is bad :P

looks like opimd is not there?

> > 
> > I haven't used phoneuid but: are you using packages from debian or
> > some other distro or building them yourself? Can you provide out of
> > gdb "backtrace full" command?
> 
> I'm using debian but build everything related to efl, fso and shr myself.
> 
> Here's the full backtrace
> 
> http://pastebin.com/jDqp8c3k

--

-- 
Klaus 'mrmoku' Kurzmann

Gmane