Thanks Luc, I had already tracked down the correct place in via_mode.c
just after I wrote this mail, and this now works perfectly. If anyone
might be interested, we could probably hack up a patch to configure this
from the config file, although given that it is so easy to force I
assume there may be some underlying reason why this may not be a good
way to do it ? This could certainly be a useful option for people using
epia-M for embedded applications, set top boxes etc, where we don't want
autoconfiguration of tv output. We haven't had time to properly debug
why the xv issues were happening, but we've worked around it by using a
single form in our app.
I also noticed in poking around the source for the driver, that cle266
appears to be dual head capable, so I have been playing around with
X86Config and using the cle both as the control surface via the vga, and
using the tv output for our video playback output. This works to a
certain degree, when X starts I have two different X displays, but the
secondary one has the colors wrong ( blue becomes green ), and our
xine-lib based player starts and then immediately segfaults, I'm
assuming related to xv again. Our programmer who is working on the
player is on holiday this week, so I cant really get anything meaningful
from debugging at the moment.
Just thought I would send it to the list and see if that makes sense to
anyone. To be honest, now we have a working setup using a pci card for
primary vga and just using cle26/unichrome for secondary playback, we
will probably stick with that for the shipping product, but I was
interested to see if the cle266 could handle dual head correctly.
Luc Verhaegen wrot ve:
> On Thu, Jul 15, 2004 at 09:11:44AM +0100, Matt Jarvis wrote:
>>I've just noticed in testing, with cle266 as secondary vga, that if you
>>do not have a connection present when XFree86 starts then the tv out is
>>turned off. Is there an option to force this even if a connection is not
>>present ? Other cards we have tested ( eg nvidia geforce 2mx ) will
>>still enable the tv out without a connection, if you have specified the
>>active device in the XF86Config.
> Again, most of the options we inherited from via are either not reasonable
> or not properly implemented (why would people want to force the tv-encoder
> type when you need to detect and configure it through i2c, if the detect
> fails, the configure will fail too).
> Anyway, the easy way would be to force this would be the route ECS G320 or
> Averatec 3200 users currently need to take: force the return value of
> VIAGetDeviceDetect (via_mode.c:~700) to return the value they need.
> Most of this code is already deprecated in the devel-branch-bios_cleanup
> branch (which is hopelessly out of tune with HEAD).
> So did you manage to track down the XV problem you had?
> Luc Verhaegen.
> This SF.Net email is sponsored by BEA Weblogic Workshop
> FREE Java Enterprise J2EE developer tools!
> Get your free copy of BEA WebLogic Workshop 8.1 today.
> Unichrome-devel mailing list
Technical Development Manager
Rolec Music Ltd
210 Belgravia Works, Marlborough Road, London, UK N19 4NF
Tel: +44 207 281 4776 Fax : +44 207 281 4565
This email is strictly confidential and intended solely for the
addressee(s). It may contain personal and confidential information and
as such may be protected by the Data Protection Act 1998. If this email
has come to you in error you must take no action based on it, nor must
you copy or show it to anyone; please reply to this email and highlight
Any views or opinions expressed within this email are those of the
author, and do not necessarily represent those of the company.
Although we have taken steps to ensure that this email and attachments
are free from any virus, we advise that in keeping with good computer
practice the recipient should ensure they are actually virus free.