Hi all,
... another small step done - I just hope my direction is right :-) ...
After reading about printk, klogd & syslogd I managed to squeeze some
information out of the freshly compiled h3600-uda1341 module.
A small change in h3600_audio_init() changed the behaviour in that way, that
opening & closing /dev/dsp did not produce clicks anymore ... but that can´t
be the solution - as the other problems remain:
-> gain (somethings) not ok after opening the device
-> samplerate changes only @ the 2nd attempt
-> bass/treble missing (in opie mixer)
In familiar 0.7.2 all this worked ok ... therefore I wonder what has changed
since then in respect to the iPAQ38xx´s audio functions? The driver code
looks unchanged since quite some time - therefore I expect the solution
somewhere outside the driver´s code. Maybe the HAL, or the configuration of
the kernel ...?
-> is there a 2.6 kernel for the iPAQ38xx in development?
-> from where can I get the kernel .config that was used to build familiar
0.7.2
-> was another module used in 0.7.2 - as there are @ lest 2 others which
implement uda1341 in conjunction with sa11xx.
@ the moment I´m not sure how to continue - therefore advice would be
appreciated.
with kind regards,
Patrik Gfeller
On Thursday 08 September 2005 20.26, Patrik Gfeller wrote:
> Hi all,
> ... I need a little help/advice, as programming & especially the linux
> kernel is new ground to me. Well - I´ve managed to configure the kernel
> K2-4-19-rmk6-pxa1-hh37 & to x-compile it for my iPAQ 3870. With make
> xconfig I´ve enabled all the kernel hacking options (except spinlock &
> waitqueue) in the hope the sound modules will become verbouse ...
>
> The size of the so compiled modules increase significantly (h3600-uda1341
> from 7576 to 154352 Bytes) and sound works with those new modules.
> But dmesg nor /var/log/messages (if I start syslogd) provide any
> information what is happening in those modules.
>
> My questions therefore are:
> - is there something else I´ve to do to activate the debug
> - where can I grap those messages
>
> any advice and/or literary/www references are very welcome.
>
> with kind regards,
> Patrik
>
> On Thursday 08 September 2005 18.04, Patrik Gfeller wrote:
> > Hi Paul (et all),
> >
> > it seems that iPAQ-3870 is equipped with a philips uda1341 sound chip.
> > Therefore sa1100-audio, uda1341 & h3600-uda1341 modules are involved ...
> > I was able to compile/build them myself (with the click/pop problem).
> >
> > Now I´ll try to understand this printk thing - & how to build them with
> > the debug stuff enable ... maybe this will throw some more light on the
> > issue.
> >
> > with kind regards,
> > Patrik
> >
> > On Tuesday 30 August 2005 03.39, Paul Eggleton wrote:
> > > Hi Patrik,
> > >
> > > > The main symptoms are clicks when opening & closing /dev/dsp. As I
> > > > listen to e-books with opie-reader & flite I've those clicks after
> > > > every sentence (there's a work around - but it's still anoying).
> > >
> > > Yes I agree this is very annoying. There are other problems too - on
> > > the h3800 the audio device does not always work after resume, either.
> > > But that's a separate issue.
> > >
> > > > Well - I decided to try to fix this, but I've to confess that I'll
> > > > need some help/tips.
> > > > As the bug was not present in 0.7.2 I thought with diff'ing some
> > > > driver source files I should find the track - but I was wrong.
> > > > Bugzilla #722 has the same symptoms - but I don't know if 39xx uses
> > > > the same driver as the 38xx family.
> > >
> > > It doesn't, they have different sound chips. I can't recall how it
> > > works on h3800, but the h3900 uses a uda1380.
> > >
> > > > What is "soft-start" - & where can it be configured (Google was not
> > > > able to answer ...)?
> > >
> > > Soft-start I believe means ramping the volume up and down rather than
> > > just turning the output on/off. I imagine you'd have to do it quite
> > > quickly so that it's not noticable. At this stage I don't believe the
> > > driver supports this so you would have to modify it. I can't imagine
> > > why this is more of a problem now than it was before, though.
> > >
> > > Cheers,
> > > Paul
> >
> > _______________________________________________
> > Familiar-dev mailing list
> > Familiar-dev_at_handhelds.org
> > https://handhelds.org/mailman/listinfo/familiar-dev
>
> _______________________________________________
> Familiar-dev mailing list
> Familiar-dev_at_handhelds.org
> https://handhelds.org/mailman/listinfo/familiar-dev
This archive was generated by hypermail 2.2.0 : Sun Sep 18 2005 - 12:51:32 EDT