[Openchrome-devel] VGA (crt-tube) connected to DVI port of VX900

[Kevin Brace]

Hi John,

> Sent: Fri, 1 Apr 2016 19:18:42 +0000
> From: John Friend <cepb13 at live.com>
> To: "openchrome-devel at lists.freedesktop.org"
> 	<openchrome-devel at lists.freedesktop.org>
> Subject: [Openchrome-devel] VGA (crt-tube) connected to DVI port of
> 	VX900
> Message-ID: <BLU180-W13B03A97495D9D7823E1F3B59A0 at phx.gbl>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> Hi Kevin 
> 
> 
> 
> First thanks for your efforts and congratulations for the new version. 
> 
> 
> 
> I
>  have tested it (commit 52730d1) on an HP T5550 (VX900) device but unfortunately there is
>  still no signal on non-LCD (ie. crt-tube) VGA monitors connected trough
>  DVI-VGA adapter. I have also tested it on CLE266, CN700 and VX800 
> devices. There is no problem on them. Actually I wonder if you or anyone
>  else tested VX900 with a crt-tube monitor. Because time to time I have tested some previous commits but could never have a signal. (Comment # 248)
> 
> 
> Please note that crt monitor works with via-opensource driver on VX900. But
>  it seems that its kind of faking EDID data. Because although it can 
> detect the VGA monitor, it supports only 60Hz on all resolutions unless 
> there is Modeline settings in xorg.conf file. Also there is "Monitor 
> name: S3FakeEdidCRT" message seen on xorg log file. 
> 
> 
> 
> I can send xorg 
> log files of five different tests I have performed via e-mail or open a new bug. 
> Please let me know which one you prefer. By the way, IMHO, if you do not have any, having a tube monitor for your own tests will be the best for future works as probably there is still many people using these old monitors 
> together with their old VIA devices. 
> 
> John

I apologize for VGA coming out of DVI to VGA adapter not working for your particular system.
I verified VGA, DVI, and LVDS FP to work with my Sylvania gnet 13001 netbook which has a DVI-I port, although DVI and LVDS FP cannot do handle clone mode at this point (VGA and LVDS FP can handle clone mode).
For this VGA, DVI to VGA adapter was used.
Whether or not I like it, I had to make a trade off on when the new version should go out, and due to the previous stock version (Version 0.3.3) having numerous problems with devices that were not known by OpenChrome (known device table), I decided that a few version had to be released at this point since it will solve many people's problem.
As for VX900 chipset users, solving the display detection bug affecting VX900 chipset will be one of the highest priority issue I will work on starting around this Thursday.
    Just to let you know, I did test OpenChrome Version 0.4.0 with numerous VIA chipset laptop / mainboards before the new code release (CLE266 chipset to VX855 chipset).
Unfortunately, I do not own a VX900 chipset box with VGA coming out of it (I do own ZOTAC ZBOX nano VD01, but this one has HDMI and DisplayPort only.), so the testing on VX900 chipset had to be sacrificed.
I do own several VGA input capable monitors, and they are used for testing.
As I have discussed in the lengthy announcement piece I sent out (I will not write such a long piece in the future, just to let you know.), I do care about making sure that OpenChrome will work with older generation equipment, and this will obviously include CRT based VGA monitors.
Just to show my commitment to making sure OpenChrome works with really old equipment, I tested it with CLE266 chipset prior to the release (EPIA-CL and EPIA-M mainboards).
Although, I do have a few CRT monitors for testing, I will admit that most testing is done with LCD monitors that have VGA input (this should not make a difference, however).
    If you do not mind spending some time filing a new bug report with http://bugs.freedesktop.org, that will be highly appreciated.
I do have some ideas on how to fix VX900 chipset display detection bug, and I hope to start implementing them in the next few weeks.

Regards,

Kevin Brace