Bug 96673 - VGA monitor goes to sleep but DVI doesn't
Summary: VGA monitor goes to sleep but DVI doesn't
Status: RESOLVED FIXED
Alias: None
Product: xorg
Classification: Unclassified
Component: Driver/openchrome (show other bugs)
Version: git
Hardware: Other All
: low minor
Assignee: Openchrome development list
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2016-06-25 01:23 UTC by Eric
Modified: 2016-07-01 21:02 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
Fix for VT1632A not turning off the screen (2.20 KB, patch)
2016-06-28 16:50 UTC, Kevin Brace
no flags Details | Splinter Review
Proposed fix for VT1632A not turning off the screen (2.20 KB, patch)
2016-07-01 04:31 UTC, Kevin Brace
no flags Details | Splinter Review

Description Eric 2016-06-25 01:23:41 UTC
Current Setup: Wyse Cx0
OpenChrome 0.4.902

Also observed previously on Wyse Vx0.

Any combination of output (DVI only, VGA only, VGA/DVI) when the Lubuntu screensaver's display power management is turned on, the VGA monitor will go into sleep mode, but the DVI will just fade to black and the monitor (backlight) stays on.
Comment 1 Kevin Brace 2016-06-28 16:18:23 UTC
(In reply to Eric from comment #0)

Hi Eric,

> Current Setup: Wyse Cx0
> OpenChrome 0.4.902
> 
> Also observed previously on Wyse Vx0.
> 
> Any combination of output (DVI only, VGA only, VGA/DVI) when the Lubuntu
> screensaver's display power management is turned on, the VGA monitor will go
> into sleep mode, but the DVI will just fade to black and the monitor
> (backlight) stays on.

Just to clarify, in order to change screen off setting,

1) "Lubuntu launch button" -> Preferences -> Screensaver
2) Within Screensaver Preferences, click "Advanced" tab
3) Check "Power Management Enabled"
4) Set appropriate minutes you want monitor to turn off


I think what is confusing is that you can also change the screen off setting from "Power Manager" under "Preferences."
Based on my testing, it appears that changing the time screen turn off time from "Power Manager" appears very unreliable. (i.e., it works from time to time, but not always)
    Assuming you have done all of this, and still seeing the problem only affecting DVI, then I will try to see what I can do.
Comment 2 Kevin Brace 2016-06-28 16:29:28 UTC
Hi Eric,

Okay, I do see a problem in the code.
I hope to get a patch to fix this bug in about 30 minutes.
Comment 3 Kevin Brace 2016-06-28 16:50:30 UTC
Created attachment 124762 [details] [review]
Fix for VT1632A not turning off the screen

Due to lack of testing when commit a8c2f04 got committed, a monitor 
with DVI input was not turning on or off when the OS tells it to turn 
the monitor on or off. This fix solves this bug by calling the correct 
function that powers up or down VT1632A.

Signed-off-by: Eric Kudzin <"Eric's e-mail address">
Signed-off-by: Kevin Brace <"Kevin's e-mail address">
Comment 4 Kevin Brace 2016-07-01 04:31:55 UTC
Created attachment 124820 [details] [review]
Proposed fix for VT1632A not turning off the screen

Due to lack of testing when commit a8c2f04 got committed, a monitor 
with DVI input was not turning on or off when the OS tells it to turn 
the monitor on or off. This fix solves this bug by calling the correct 
function that powers up or down VT1632A.

Signed-off-by: Eric Kudzin <"Eric's e-mail address">
Signed-off-by: Kevin Brace <"Kevin's e-mail address">
Comment 5 Kevin Brace 2016-07-01 04:48:41 UTC
Hi Eric,

Please try the revised patch.
This patch should be applied against OpenChrome Version 0.5 RC6 I just released.
The older patch I obsoleted should not be used since it contains a "death trap" bug in the patch that freezes the system if VT1632A is detected.
The so called 'death trap" bug is an issue of the compiler / make generating the code when pointer types aren't compatible (this should be flagged as an error, but it doesn't), and this is why when you tested the patch, it was freezing.
I found the error in the code today (sometime Thursday afternoon), but due to equipment problems at my place (i.e., the computer I commit the code to the repository was having problems) and other family commitments, I was not able to upload the patch until right now.
The computer is working fine now, so I am able to upload the revised patch.
Also, please test this patch only with Wyse Cx0 and not Vx0 at this point.
Wyse Vx0 will need an additional patch to manipulate DVP0 (Digital Video Port 0) register so that DVP0 source is IGA2 (display controller 2), not IGA1 Vx0 seems to  set DVP0 source to.
I will upload the patch to get DVI working with Vx0 over at Bug 94259 after you confirm that this particular bug (Bug 96673) is fixed and closed.
This fix will go into Version 0.5 RC7, hopefully by tomorrow (Friday).
Comment 6 Eric 2016-07-01 05:52:28 UTC
With the patch, both monitors go into sleep mode on the Cx0.
Comment 7 Kevin Brace 2016-07-01 05:59:30 UTC
(In reply to Eric from comment #6)

Hi Eric,

> With the patch, both monitors go into sleep mode on the Cx0.

Thanks for the good news.
I also saw that the fix works with my Wyse Cx0.
I will get a patch to set Wyse Vx0's DVP0 (Digital Video Port 0) to IGA2 sometime tomorrow so that you can get Vx0's DVI to work outright.
I will make the code commit to the repository shortly.
As for RC7, I will like to incorporate the DVP0 fix for Wyse Vx0 and other devices that use DVP0.
Comment 9 Kevin Brace 2016-07-01 21:02:39 UTC
Hi Eric,

I posted a patch for you to try over at Bug 94259.

https://bugs.freedesktop.org/show_bug.cgi?id=94259#c47

Apply it against the current head (Version 0.5 RC6 + 1 commit).


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.