Bug 30050

Summary: [[GM45] HDMI] bogus resolutions which are unuseable
Product: xorg Reporter: Julius Schwartzenberg <julius.schwartzenberg>
Component: Driver/intelAssignee: Carl Worth <cworth>
Status: RESOLVED DUPLICATE QA Contact: Xorg Project Team <xorg-team>
Severity: normal    
Priority: medium    
Version: 7.5 (2009.10)   
Hardware: x86-64 (AMD64)   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
output of xrandr --verbose
none
Xorg.0.log none

Description Julius Schwartzenberg 2010-09-06 14:19:53 UTC
I tried to use my X4500HD's HDMI port of my notebook with two devices and had some strange issues. I'm using Ubuntu Lucid with the lastest drivers from the xorg-edgers ppa.

I first connected it to a cheap brand HDTV. A number of resolutions showed up, with the maximum resolution being 1280x720 and the lowest 640x480. None of these resolutions however gave a useable picture, in each case the edges were just off the screen. This resolution list didn't seem to match properly what the TV supports.

Expected behaviour: show the complete image on the connected HDTV.


The second device was an amplifier by Sony. It has no video output, only audio.
The first problem was that for the audio output to function, I needed to switch the HDMI port on with xrandr. I would say xrandr is only for video & HDMI audio should always just work without it.
After enabling the HDMI port, more problems appear. The notebook display can be disabled, leaving the user without a working display and a need to reset the whole system. Also the possible resolution list sometimes changes and instead of showing the 1280x800 resolution of my notebook it shows again the exact same list of resolutions that I got when I had the HDTV connected.

This shows that when HDMI is used, the driver just makes up some resolutions, with the highest one being 1280x720, irregardless of what is actually supported by the connected device. The driver should be capable of detecting 'no video output' and not offer any resolution at all (just turn on the audio) or it should offer the resolutions that can be displayed properly by the connected monitor instead of making up its own.

When using my amplifier, my notebook screen sometimes also got switched to 1280x720 instead of 1280x800.

I can perform more tests and provide more details. Please let me know if I need to provide more information.
Comment 1 Julien Cristau 2010-09-06 14:30:15 UTC
On Mon, Sep  6, 2010 at 14:19:54 -0700, bugzilla-daemon@freedesktop.org wrote:

> I can perform more tests and provide more details. Please let me know if I need
> to provide more information.

http://intellinuxgraphics.org/how_to_report_bug.html has good
guidelines.
Comment 2 Chris Wilson 2010-09-07 05:08:37 UTC
Julius, thanks for the bug report and *thankyou* for testing with xorg-edgers first.

TVs overscan. That is they lie about their resolution and part of the image will fall outside of the visible area. The reason is simply compatibility with historical broadcasts. Good TVs with have 1:1 pixel mode with zero overscan.

On the integrated TV output we support defining the margin and so compensate, however we do not yet have anything like that for HDMI. So we have to rely on the TV's EDID for the list of valid modes, and you need to attach both the Xorg.0.log and xrandr --verbose.

The second set of problems with the amplifier is interesting. You are definitely the first person to have tried it. ;-) Can you please file a separate bug report for it as it is a distinct problem that will need to be addressed separately.
Comment 3 Julius Schwartzenberg 2010-09-07 12:38:09 UTC
Thanks a lot for your advice! My apologies for merging the issues at first. I created bug #30067 for the issues with my amplifier.

Unfortunately I have no other good devices to test the TV with, so I cannot verify its behavior. Overscan seems likely though, it seemed every border was missing the same amount of pixels.

I do not have access to the TV right now, I'll be sure to upload my Xorg.0.log and the verbose xrandr output when I soon have access again. (I *do* have access to my amp.)
Comment 4 Julius Schwartzenberg 2010-09-12 04:23:11 UTC
Created attachment 38641 [details]
output of xrandr --verbose
Comment 5 Julius Schwartzenberg 2010-09-12 04:24:06 UTC
Created attachment 38642 [details]
Xorg.0.log
Comment 6 Chris Wilson 2010-09-12 04:46:45 UTC
Thanks Julius, afaics we are interpreting the capabilities of the connection/monitor correctly just we not taking the visible area of the monitor into account.

*** This bug has been marked as a duplicate of bug 29723 ***

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.