Bug 12262

Summary: External LCD (connected to laptop as Dual Head) flashes when mode set high
Product: xorg Reporter: he jiang <jiang.he>
Component: Driver/intelAssignee: Hong Liu <hong.liu>
Status: VERIFIED FIXED QA Contact: Xorg Project Team <xorg-team>
Severity: normal    
Priority: high CC: brice.goglin, daniel.zhu, fengming.pi, gordon.jin, hong.liu, keithp, mtanski, nian.wu
Version: unspecified   
Hardware: All   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:
Bug Depends on:    
Bug Blocks: 13027    
Attachments:
Description Flags
Xorg's log none

Description he jiang 2007-09-02 20:25:15 UTC 
Using Sumsung 205BW(Wide-Screen) as an external display or a VGA display. Following are the modes it can support by xrandr output:
   1680x1050      60.0 +
   1280x1024      75.0     59.9
   1280x960       59.9*
   1152x864       75.0     74.8
   1024x768       75.1     70.1     60.0
   832x624        74.6
   800x600        72.2     75.0     60.3     56.2
   640x480        75.0     72.8     66.7     60.0
   720x400        70.1
When set to 1280x1024 or 1680x1050 using xrandr, it flashes.
Comment 1 he jiang 2007-09-02 20:28:36 UTC 
Created attachment 11394 [details]
Xorg's log
Comment 2 Keith Packard 2007-09-06 18:20:42 UTC 
It sounds like the mode presented by the X server isn't compatible with the monitor. Debugging this can be tricky.

If you have a Windows disk for this monitor, looking at the configuration file that should be present may help figure out what modes are supported under Windows.

Also, if this is a LCD monitor, using a reduced blanking mode as output by

$ cvt -r 1680 1050

may work better than the EDID provided normal mode.

You can also try fiddling with the mode as returned by EDID to see if you can't get it to work.
Comment 3 Hong Liu 2007-09-10 22:18:34 UTC 
(In reply to comment #2)
> Also, if this is a LCD monitor, using a reduced blanking mode as output by
> 
> $ cvt -r 1680 1050
> 
> may work better than the EDID provided normal mode.

Using the modeline calculated by cvt is working.

Directly connecting the monitor (not dual-head) is OK with the modes probed by X.
But when dual-headed, it seems that modes (requiring dotclock > 120MHz) don't work (1280x1024@75Hz doesn't work while 1280x1024@60Hz is OK).

Is there any hardware limit (on dotclock) when using two pipes for display?
Comment 4 Gordon Jin 2007-09-10 22:44:04 UTC 
Keith, any further guide?
Comment 5 Michael Fu 2007-09-20 18:52:29 UTC 
from the log, EDID info seems broken for this monitor. vertical refresh is 0hz.
Comment 6 Hong Liu 2007-10-08 23:17:47 UTC 
*** Bug 11490 has been marked as a duplicate of this bug. ***
Comment 7 Michael Fu 2007-10-11 19:06:23 UTC 
*** Bug 11028 has been marked as a duplicate of this bug. ***
Comment 8 Michael Fu 2007-10-23 00:52:07 UTC 
*** Bug 11642 has been marked as a duplicate of this bug. ***
Comment 9 Hong Liu 2007-11-07 22:10:13 UTC 
*** Bug 10784 has been marked as a duplicate of this bug. ***
Comment 10 Hong Liu 2007-11-07 22:16:04 UTC 
Current PLL programming in our driver has problem with some modes.
for example, 1280x1024@75Hz, 1680x1050@59.9 with reduced blanking, We can't find a close enough dotclock for these kinds of modes which leads to incorrect display.
Comment 11 Gordon Jin 2007-11-14 01:17:04 UTC 
The PLL issue is fixed in xf86-video-intel 26194e19e1c80615697016e25640d4c8c244353f today by Keith.

For those CCed people (as your bug is marked as dup with this one), please retest your observed issue. If it still exists, that means it's a seperate issue so please reopen your own bug tracker.
Comment 12 Pi, Fengming 2007-11-14 17:57:31 UTC 
With the latest git source,the problem did gone.

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.