[pkg-fso-maint] debian: issues with latest packed fso

[Luca Capello]

Hi Arne!

On Fri, 12 Jun 2009 18:08:21 +0200, arne anka wrote:
>>> - fso needs probably some better detection for recamping
>>
>> If I read correctly upstream bug
>>
>>   https://docs.openmoko.org/trac/ticket/1024#comment:55
>>
>> various workarounds should be available for it,
>
> where do you read that? (and how do you know which commenbt is #55?)

If you move your mouse next to the comment author, a paragraph sign
(pilcrow) appears and this contains the link to the comment number.
Yes, it is ugly, but we should live with that since I do not think
Openmoko or freesmartphone.org will move to debbugs :-(

I read that on Michael Lauer comments #55:

  #1024 will not be fixed by a hardware exchange. All current
  distributions have a satisfying workaround, the real issue is still
  undergoing research wrt. GSM firmware.

Still I do not know which are these workarounds.

> afaiu, the workaround is disabling deep sleep, ie using SLEEP=2, in
> fso it is achieved by
>
> ti_calypso_deep_sleep = never
>
> in frameworkd.conf.

I do not remember I have experienced similar problems, thus I have never
played with that value, sorry.

>> is this not the case for
>> Debian?  And if you do not experience the same bug with fso images, do
>> you know what is different in Debian?
>
> i am not sure, i understand, what you're saying -- but i was referring
> not to fso images (never used those) but to the older debian package.
> my understanding is, that fso tries to detect by itself if the phone
> is prone to #1024, and accordingly enables or disables deep sleep.
> from my experinece that detection does not work (reliably).
> inhowfar improvements were made in newer fso versions, i can't say.

I will ping Michael Lauer about the ETA for fso-milestone5.5: it is
maybe better to upgrade to a new frameworkd version before milestone5.5
is out...

>> However, according to
>>
>>   http://totalueberwachung.de/blog/2009/06/03/freerunner-deep-sleep-standby-time
>>
>> Openmoko bug #1024 there should be a "final" workaround.
>
> disabling deep sleep.  according to joerg(?) the standby time may
> increase by up to 25% with a hw fix, but improving power management in
> the kernel  will yield better over all results.

Do you mean that disabling deep-sleep (which I guess is something
implemented in hardware) and with proper power management (implemented
in software) you will increase stand-by time?

Thx, bye,
Gismo / Luca
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 314 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/pkg-fso-maint/attachments/20090613/3b5efa67/attachment.pgp>