Bug 22162

I also tried DRI off without success
Monitor is SAMSUNG 2333HD, opensuse recognised the info from vendor's cd
I'll attach xorg.conf dmesg, etc  as soon as I figure out how to do it here ....

here xrandr output:
Screen 0: minimum 320 x 200, current 1920 x 1080, maximum 3200 x 1080
VGA disconnected (normal left inverted right x axis y axis)
HDMI-1 connected 1920x1080+0+0 (normal left inverted right x axis y axis) 518mm x 32mm
   1920x1080      59.9*+
   1680x1050      59.9  
   1280x1024      75.0     60.0  
   1440x900       59.9  
   1280x960       60.0  
   1280x800       59.9  
   1152x864       75.0  
   1024x768       75.0     70.1     60.0  
   832x624        74.6  
   800x600        72.2     75.0     60.3     56.2  
   640x480        75.0     72.8     66.7     59.9  
   720x400        70.1  
HDMI-2 disconnected (normal left inverted right x axis y axis)

and with --verbose:

Screen 0: minimum 320 x 200, current 1920 x 1080, maximum 3200 x 1080
VGA disconnected (normal left inverted right x axis y axis)
	Identifier: 0x3b
	Timestamp:  31443
	Subpixel:   unknown
	Clones:    
	CRTCs:      0 1
	Transform:  1.000000 0.000000 0.000000
	            0.000000 1.000000 0.000000
	            0.000000 0.000000 1.000000
	           filter: 
HDMI-1 connected 1920x1080+0+0 (0x3e) normal (normal left inverted right x axis 
y axis) 518mm x 32mm
	Identifier: 0x3c
	Timestamp:  31443
	Subpixel:   unknown
	Clones:     HDMI-2
	CRTC:       0
	CRTCs:      0 1
	Panning:    0x0+0+0
	Tracking:   0x0+0+0
	Border:     0/0/0/0
	Transform:  1.000000 0.000000 0.000000
	            0.000000 1.000000 0.000000
	            0.000000 0.000000 1.000000
	           filter: 
	EDID_DATA:
		00ffffffffffff004c2d850433324d43
		1a120103803420782aee91a3544c9926
		0f5054bfef80714f8100814081809500
		a940b30001011a3680a070381f403020
		350006202000001a000000fd00384b1e
		5111000a202020202020000000fc0053
		796e634d61737465720a2020000000ff
		004831414b3530303030300a202000e9
  1920x1080 (0x3e)  138.5MHz +HSync -VSync *current +preferred
        h: width  1920 start 1968 end 2000 total 2080 skew    0 clock   66.6KHz
        v: height 1080 start 1083 end 1088 total 1111           clock   59.9Hz
  1680x1050 (0x3f)  119.0MHz +HSync -VSync
        h: width  1680 start 1728 end 1760 total 1840 skew    0 clock   64.7KHz
        v: height 1050 start 1053 end 1059 total 1080           clock   59.9Hz
  1280x1024 (0x40)  135.0MHz +HSync +VSync
        h: width  1280 start 1296 end 1440 total 1688 skew    0 clock   80.0KHz
        v: height 1024 start 1025 end 1028 total 1066           clock   75.0Hz
  1280x1024 (0x41)  108.0MHz +HSync +VSync
        h: width  1280 start 1328 end 1440 total 1688 skew    0 clock   64.0KHz
        v: height 1024 start 1025 end 1028 total 1066           clock   60.0Hz
  1440x900 (0x42)   88.8MHz +HSync -VSync
        h: width  1440 start 1488 end 1520 total 1600 skew    0 clock   55.5KHz
        v: height  900 start  903 end  909 total  926           clock   59.9Hz

  1280x960 (0x43)  108.0MHz +HSync +VSync
        h: width  1280 start 1376 end 1488 total 1800 skew    0 clock   60.0KHz
        v: height  960 start  961 end  964 total 1000           clock   60.0Hz
  1280x800 (0x44)   71.0MHz +HSync -VSync
        h: width  1280 start 1328 end 1360 total 1440 skew    0 clock   49.3KHz
        v: height  800 start  803 end  809 total  823           clock   59.9Hz
  1152x864 (0x45)  108.0MHz +HSync +VSync
        h: width  1152 start 1216 end 1344 total 1600 skew    0 clock   67.5KHz
        v: height  864 start  865 end  868 total  900           clock   75.0Hz
  1024x768 (0x46)   78.8MHz +HSync +VSync
        h: width  1024 start 1040 end 1136 total 1312 skew    0 clock   60.0KHz
        v: height  768 start  769 end  772 total  800           clock   75.0Hz
  1024x768 (0x47)   75.0MHz -HSync -VSync
        h: width  1024 start 1048 end 1184 total 1328 skew    0 clock   56.5KHz
        v: height  768 start  771 end  777 total  806           clock   70.1Hz
  1024x768 (0x48)   65.0MHz -HSync -VSync
        h: width  1024 start 1048 end 1184 total 1344 skew    0 clock   48.4KHz
        v: height  768 start  771 end  777 total  806           clock   60.0Hz
  832x624 (0x49)   57.3MHz -HSync -VSync
        h: width   832 start  864 end  928 total 1152 skew    0 clock   49.7KHz
        v: height  624 start  625 end  628 total  667           clock   74.6Hz
  800x600 (0x4a)   50.0MHz +HSync +VSync
        h: width   800 start  856 end  976 total 1040 skew    0 clock   48.1KHz
        v: height  600 start  637 end  643 total  666           clock   72.2Hz
  800x600 (0x4b)   49.5MHz +HSync +VSync
        h: width   800 start  816 end  896 total 1056 skew    0 clock   46.9KHz
        v: height  600 start  601 end  604 total  625           clock   75.0Hz
  800x600 (0x4c)   40.0MHz +HSync +VSync
        h: width   800 start  840 end  968 total 1056 skew    0 clock   37.9KHz
        v: height  600 start  601 end  605 total  628           clock   60.3Hz
  800x600 (0x4d)   36.0MHz +HSync +VSync
        h: width   800 start  824 end  896 total 1024 skew    0 clock   35.2KHz
        v: height  600 start  601 end  603 total  625           clock   56.2Hz
  640x480 (0x4e)   31.5MHz -HSync -VSync
        h: width   640 start  656 end  720 total  840 skew    0 clock   37.5KHz
        v: height  480 start  481 end  484 total  500           clock   75.0Hz
  640x480 (0x4f)   31.5MHz -HSync -VSync
        h: width   640 start  664 end  704 total  832 skew    0 clock   37.9KHz
        v: height  480 start  489 end  492 total  520           clock   72.8Hz
  640x480 (0x50)   30.2MHz -HSync -VSync
        h: width   640 start  704 end  768 total  864 skew    0 clock   35.0KHz
        v: height  480 start  483 end  486 total  525           clock   66.7Hz
  640x480 (0x51)   25.2MHz -HSync -VSync
        h: width   640 start  656 end  752 total  800 skew    0 clock   31.5KHz
        v: height  480 start  490 end  492 total  525           clock   59.9Hz
  720x400 (0x52)   28.3MHz -HSync +VSync
        h: width   720 start  738 end  846 total  900 skew    0 clock   31.5KHz
        v: height  400 start  412 end  414 total  449           clock   70.1Hz
HDMI-2 disconnected (normal left inverted right x axis y axis)
	Identifier: 0x3d
	Timestamp:  31443
	Subpixel:   unknown
	Clones:     HDMI-1
	CRTCs:      0 1
	Transform:  1.000000 0.000000 0.000000
	            0.000000 1.000000 0.000000
	            0.000000 0.000000 1.000000
	           filter: 


XORG.CONF:
# xorg.conf (X.Org X Window System server configuration file)
#
# This file was generated by dexconf, the Debian X Configuration tool, using
# values from the debconf database.
#
# Edit this file with caution, and see the xorg.conf manual page.
# (Type "man xorg.conf" at the shell prompt.)
#
# This file is automatically updated on xserver-xorg package upgrades *only*
# if it has not been modified since the last upgrade of the xserver-xorg
# package.
#
# Note that some configuration settings that could be done previously
# in this file, now are automatically configured by the server and settings
# here are ignored.
#
# If you have edited this file but would like it to be automatically updated
# again, run the following command:
#   sudo dpkg-reconfigure -phigh xserver-xorg

Section "Monitor"
	Identifier	"Configured Monitor"
EndSection

Section "Screen"
	Identifier	"Default Screen"
	Monitor		"Configured Monitor"
	Device		"Configured Video Device"
	SubSection "Display"
		Virtual	3200 1080
	EndSubSection
EndSection

Section "Device"
	Identifier	"Configured Video Device"
	Option	"ModeDebug" "yes"
EndSection

Section "ServerFlags"
	Option	"DontZap"	"False"
EndSection

Ok, when I say VGA output I mean DVI-I to VGA conversion, like that it works ok. Using any of the two DVI outputs as DVI-D will reproduce the problem (blank screens randomly, some times one in 5 minutes, some times 10 in 2 minutes ...)

Created attachment 26557 [details]
Xorg.0.log

Created attachment 26559 [details]
dmesg

Created attachment 26562 [details]
xrandr --verbose output

Created attachment 26563 [details]
xorg.conf

This should have been fixed by Ling's "G4x DPLL setting" patch. Could you try xf86-video-intel master branch?

(In reply to comment #7)
> This should have been fixed by Ling's "G4x DPLL setting" patch. Could you try
> xf86-video-intel master branch?
> 

 Ok, I am new here, I guess I should download last release of xf86-video-intel master branch from http://intellinuxgraphics.org/download.html and build it for my distribution and test it.  right?  ok, I'll do it as soon as I have some spare time to do it correctly following all the instructions (this weekend I hope) 

 My only doubt is if I need to build the four components: 2D driver, 3D GL driver, DRM and AGPGART kernel module and finally LIBDRM userland DRM module.

 I'll let you know the results as soon as I finish.

Thanks a lot for your reply. Regards,

Javier.

Ok, I found a repository (ppa.launchpad.net) for my ubuntu with 2.7.1 intel driver, now I have:

javi@pc-javi:~$ dpkg -l | grep -i drm
ii  libdrm-intel1                              2.4.9-1ubuntu1~xup~1                 Userspace interface to intel-specific kernel
ii  libdrm2                                    2.4.5-0ubuntu4                       Userspace interface to kernel DRM services -

javi@pc-javi:~$ dpkg -l | grep -i xorg-video-intel
ii  xserver-xorg-video-intel                   2:2.7.1-0ubuntu1~xup~1               X.Org X server -- Intel i8xx, i9xx display d


 and I did not get the trouble during last 10 minutes,  oops !! I got right now another 1 second blank screen :-(   I was streesing the machine: 3d chess, virtual machines, web browsing ....

 it seems to happen less often but the problem seems to remain there. Anyhow I will try more and report.


Thanks for your help and sorry for my English :-)

(In reply to comment #9)
> 
>  it seems to happen less often but the problem seems to remain there. Anyhow I
> will try more and report.
> 
> 

pls keep us posted and attach reg_dump when the issue happens. thanks.

Created attachment 26719 [details]
reg_dump probably before the blank screen

Problem persists, I set up a cron every minute with the reg_dump, but it is difficult to do the reg_dump during the very same second when problem arises, I join two reg_dumps, probably from a bit before and a bit after the 1 second (aprox) black screen ...

 I do not see any interesting thing in dmesg, any other place where to look for traces??

thanks again. Regards,

javier.


(In reply to comment #10)
> (In reply to comment #9)
> > 
> >  it seems to happen less often but the problem seems to remain there. Anyhow I
> > will try more and report.
> > 
> > 
> 
> pls keep us posted and attach reg_dump when the issue happens. thanks.
>

Created attachment 26720 [details]
reg_dump a few secons after the black screen

Created attachment 26721 [details]
xorg.0.log after several black screens

 I put Driver "vesa" in xorg.conf and I  also had black screens (although I could not get my native screens resolution 1920x2080) , so after all it might not be an intel's video driver issue  ??  

 I am going to test installing windows and putting the screeen to another computer to discard a moterboard or screen issue. It is weird because through VGA (DVI-I to VGA) everything works ok, but I get worse quality  and ubuntu has problems to recognize native resolution ... (1920x1080) 

 I keep trying ....

Created attachment 26844 [details]
please try the debug patch on your machine, thanks.

from reg_dump, we can see clock is set correctly, this is only debug patch, please try it on your machine.

Thanks for your help
Ma Ling

Ok, I'll report as soon as I get to build video driver with the debug patch.

 I noticed in X logs and setting refresh rates manually, that it seems to take into account only 2 decimals, while my monitor's manual specifies refresh rates (vertical and horizontal and global) with 3 decimals. Could this have something to do with the problem?

Thanks for your help. Regards,

Javier.

>  I noticed in X logs and setting refresh rates manually, that it seems to take
> into account only 2 decimals, while my monitor's manual specifies refresh rates
> (vertical and horizontal and global) with 3 decimals. Could this have something
> to do with the problem?
Sorry I 'm not clear about what's your meaning, could you please specify it?

Thanks
Ma Ling

I tried writing in xorg.conf refresh rates as specified by screens manufacture, so I put in monitor section:

 VertRefresh  59.934
 HorizSync    66.587

 and then, I saw somewhere (xorg.0.log?) that it took only 59.93 and 66.58 ( and total clock was rounded to 139 instead of 138,500. Anyhow, now I do not see it in my new Xorg.0.log after I applied your patch .....

 So yes, I built the driver with your patch and  ... problem remains there :-(  indeed I get poor glxgears performance and no compiz effects, I guess it is due to the following error from Xorg.0.log:

 [drm] failed to load kernel module "i915"
(EE) [drm] drmOpen failed.
(EE) intel(0): [dri] DRIScreenInit failed. Disabling DRI.

 I will try building everything again, as I might have missed some step in building the modules. How can I check the version of the driver that is running ??

 Anyhow, I checked that I have the problem even when running X with driver "vesa"  in xorg.conf  . could it mean that the problem is not necessary in the intel driver??

thanks again for your help. Regards,

Javier.

would you please try the mode of 1680x1050 to see if the problem gets better or even disappear? thanks.

 I switched to 1680x1050 and I worked for some 5 hours without getting the problem  (yes, it dissapeared :-)  ). It is quite amazing, because when I tried Driver "vesa" it switched automatically to 1600x1200 but the 1 second black screens continued to be there ...

 I would like to switch to 1920x1080 as it is my screen's native resolution ... with 1680x1050 I get worse quality and a bit enlarged image ...

thanks anyhow for your help. 


(In reply to comment #19)
> would you please try the mode of 1680x1050 to see if the problem gets better or
> even disappear? thanks.
> 

(In reply to comment #20)
>  I switched to 1680x1050 and I worked for some 5 hours without getting the
> problem  (yes, it dissapeared :-)  ). It is quite amazing, because when I tried
> Driver "vesa" it switched automatically to 1600x1200 but the 1 second black
> screens continued to be there ...
>  I would like to switch to 1920x1080 as it is my screen's native resolution ...
> with 1680x1050 I get worse quality and a bit enlarged image ...
> thanks anyhow for your help. 
> (In reply to comment #19)
> > would you please try the mode of 1680x1050 to see if the problem gets better or
> > even disappear? thanks.
> > 

How about the patch in comments #15 when you set resolution as 1920x1080 ?

thanks
Ma Ling

(In reply to comment #20)
>  I switched to 1680x1050 and I worked for some 5 hours without getting the
> problem  (yes, it dissapeared :-)  ). It is quite amazing, because when I tried
> Driver "vesa" it switched automatically to 1600x1200 but the 1 second black
> screens continued to be there ...
> 
>  I would like to switch to 1920x1080 as it is my screen's native resolution ...
> with 1680x1050 I get worse quality and a bit enlarged image ...
> 
> thanks anyhow for your help. 
> 
> 
> 

the reason I ask you to try with 1680x1050 mode is because the device can generate a very close pixel clock to it, compared with 1920x1080. In other way, for the 1920x1080 mode your monitor suggested in its EDID, the required clock is 138.5Mhz, but the closest one we can generate is 138Mhz. Usually it's close enough for monitor to tolerate...  but unfortunately, it seems your monitor is very strict on its input pixel clock.. that also explains why you often see the flicker in vesa mode as well using the G45 mobo.

Pls give it a try this official CVT mode of 1080p (see man page for Option PreferredMode )

Modeline "1920x1080_CVT"  148.50  1920 2008 2052 2200  1080 1084 1089 1125 +hsync +vsync (67.5 kHz)

thanks.

> 
> How about the patch in comments #15 when you set resolution as 1920x1080 ?
> 
> thanks
> Ma Ling
> 

see comment# 18.

Because of hardware reason, we can not provide 100% persent 138.5M clock.
Could you please try the modeline by xrandr command, at least the tolerance derived from hardware is smaller and your EDID indicates monitor supports it.

Modeline "1920x1200R"  154.00  1920 1968 2000 2080  1200 1203 1209 1235 +hsync -vsync.

Thanks
Ma Ling

 
> Pls give it a try this official CVT mode of 1080p (see man page for Option
> PreferredMode )
> 
> Modeline "1920x1080_CVT"  148.50  1920 2008 2052 2200  1080 1084 1089 1125
> +hsync +vsync (67.5 kHz)
> 

 I tried that modeline with xrandr but I got the flicker after a while. I am now trying the other modeline "1920x1200"  ... no flicker for the last 15 minutes ... I'll keep updated, but again, it is not a good resolution for my screen :-(

 One question, when you say "the closest one we can generate is 138Mhz" talking about pixel clock, is it a hardware limitation or a software one (operating system, Xorg, driver arquitecture ....)  ??

 I read in the 2.6.30 kernel release notes something about flicker free graphics , could that have something to do here ??

Thank you again.

(In reply to comment #25)
> > Pls give it a try this official CVT mode of 1080p (see man page for Option
> > PreferredMode )
> > 
> > Modeline "1920x1080_CVT"  148.50  1920 2008 2052 2200  1080 1084 1089 1125
> > +hsync +vsync (67.5 kHz)
> > 
> 
>  I tried that modeline with xrandr but I got the flicker after a while. I am
> now trying the other modeline "1920x1200"  ... no flicker for the last 15
> minutes ... I'll keep updated, but again, it is not a good resolution for my
> screen :-(
> 

anyway, please keep us updated about your testing.

>  One question, when you say "the closest one we can generate is 138Mhz" talking
> about pixel clock, is it a hardware limitation or a software one (operating
> system, Xorg, driver arquitecture ....)  ??
> 

HW. but that should be acceptable to monitor.

>  I read in the 2.6.30 kernel release notes something about flicker free
> graphics , could that have something to do here ??

no, that's not related with your case.

> 
> Thank you again.
> 

Hi again,

> anyway, please keep us updated about your testing.
 After some 4 hours working I got only 1 flicker with 1920x1200 mode, and it was a few minutes after waking up from hibernation.
 I only see two possible tests more to do done: check what happens with other operating systems (windows) and test screen in another computer. I'll do them as soon as I can and I'll post results. After that, I guess I could only try to blame samsung for such monitor intolerability to signals and try to change screen ... but I guess that is almost impossible. ok, let's wait the last tests :-)

Some questions, anyhow: 
 I guess it works ok for VGA output (DVI-I) because circuit for analog output is more capable of producing the exact clock signal. right?  Anyway, I would need some reading to understand why a 5ms response time monitor needs a 138.5 Mhz clock input for the digital input ... can you recomend me any good reading?  thanks in advance, anyhow.  

 And an off-topic question, as you guys must have intel mobos and you must run linux: do you get this mobos to suspend properly ???  I tried everything and there is no way to get my DQ45CB wake up from suspend :-( ...  (reported at http://communities.intel.com/message/29707#29707 but I got no answers yet ...)

thanks for any help, again. Regards,

Javier.

(In reply to comment #27)
> Hi again,
> > anyway, please keep us updated about your testing.
>  After some 4 hours working I got only 1 flicker with 1920x1200 mode, and it
> was a few minutes after waking up from hibernation.
>  I only see two possible tests more to do done: check what happens with other
> operating systems (windows) and test screen in another computer. I'll do them
> as soon as I can and I'll post results. After that, I guess I could only try to
It is good news, I also think the monitor is too sensitive about clock, you know it is very hard for hardware to support 100 percent accurate clock :).
So could you use the modeline as preferred mode, we close this issue ?
 
> Some questions, anyhow: 
>  I guess it works ok for VGA output (DVI-I) because circuit for analog output
> is more capable of producing the exact clock signal. right?  Anyway, I would
> need some reading to understand why a 5ms response time monitor needs a 138.5
> Mhz clock input for the digital input ... can you recomend me any good reading?
could you search related message from wikipedia?
 
>  And an off-topic question, as you guys must have intel mobos and you must run
> linux: do you get this mobos to suspend properly ???  I tried everything and
> there is no way to get my DQ45CB wake up from suspend :-( ...  (reported at
> http://communities.intel.com/message/29707#29707 but I got no answers yet ...)
It should belong to ACPI issues, could you file your issue at http://bugzilla.kernel.org/enter_bug.cgi?product=Power management

Thanks
Ma Ling
 

 Hi again,
Could you please try Windows environment to check whether the same issue will occur?

Thanks for your help.
Ma Ling

Created attachment 27000 [details]
please try the patch on your machine, thanks.

(In reply to comment #28)
> It is good news, I also think the monitor is too sensitive about clock, you
> know it is very hard for hardware to support 100 percent accurate clock :).
> So could you use the modeline as preferred mode, we close this issue ?
 No, I cannot use that modeline (1920x1200) as it is not my monitor's native resolution, image format is different (16:10 vs. 16:9) and thus image is enlarged :-(

 I'll try the last patch as soon as I have some spare time, as well as trying windows (I have to install it) I would prefer not to close the ticket untill I do these tests.

 One last question, would the same issue arise with other similar motherboards as the intel DG45ID  (G45 chipset with hdmi output) ??

(In reply to comment #31)
> (In reply to comment #28)
> > It is good news, I also think the monitor is too sensitive about clock, you
> > know it is very hard for hardware to support 100 percent accurate clock :).
> > So could you use the modeline as preferred mode, we close this issue ?
>  No, I cannot use that modeline (1920x1200) as it is not my monitor's native
> resolution, image format is different (16:10 vs. 16:9) and thus image is
> enlarged :-(
> 
>  I'll try the last patch as soon as I have some spare time, as well as trying
> windows (I have to install it) I would prefer not to close the ticket untill I
> do these tests.
> 
>  One last question, would the same issue arise with other similar motherboards
> as the intel DG45ID  (G45 chipset with hdmi output) ??
> 

we seldom see monitor has such issue on latest driver. likely this is a monitor specific issue rather than motherboard. thanks.

 Ok, I tried with windows XP (new install + last drivers for intel video and for monitor) and flicker was there and more often.

We can close the ticket.

I'll try the last patch anyhow, and the screen in another computer. I'll talk to Samsung also to see if it is a defective screen or the model has that little toleration to pixel clock.

 Thanks a lot for your help. Regards,

Javier.

By the way, should I care about this dmesg message:

[  105.363676] [drm:i915_get_vblank_counter] *ERROR* trying to get vblank count for disabled pipe 1
[  105.368092] mtrr: no MTRR for c0000000,10000000 found

Thanks anyway.

(In reply to comment #34)
> By the way, should I care about this dmesg message:
> [  105.363676] [drm:i915_get_vblank_counter] *ERROR* trying to get vblank count
> for disabled pipe 1
> [  105.368092] mtrr: no MTRR for c0000000,10000000 found
> Thanks anyway.

No, root cause is sensitive monitor :)
Thanks
Ma Ling

Well, I feel a bit stupid .... finally the problem was the cable .. I changed DVI-D cable and now it works ok. I am sorry I wasted your time ...  anyway thanks to you now I know how to have the latest intel video drivers in my machine and I am quite happy to have an intel board with free and powerful drivers for my linux ... indeed I did also solve the suspend issue :-)

 From now on I will always suspect of cables .... even if it works for some resolutions and not for others (I guess it got worse from a certain bandwidth on)