[Intel HD 4400] strongly irritating artefacts on 2560x1440 laptop display

[Intel HD 4400] strongly irritating artefacts on 2560x1440 laptop display

[Martin Wilck]

Dear list members,

I suffer from highly disturbing artefacts with the Intel HD 4400 chip
set on thw 2560x1440 display (Sharp LQ133T1JW19) of my Fujitsu Lifebook
S904. The effects are hard to describe, therefore I have uploaded a
video to
https://www.dropbox.com/sh/9wh0nzvp0w1phxz/AAAsvAqHN8ApXzor58FmGbwLa?dl=0.
 I can distinguish 2 effects that are possibly related to each other:

 1. Maximized windows under GNOME3 (fully maximized or half-maximized on
the right edge of the screen) fill the screen vertically with a single
color (usually grey). Half-maximzed windows on the left edge are not
affected. The effect is only seen if the Laptop screen is the main
monitor in GNOME (showing the panel on top). Some applications, which
start in maximized mode and can't be resized, are unusable on this screen.

 2. Artefacts appear when windows are placed on the right and bottom
areas of the screen. The artefacts look as if pixels extend vertically
from their horizontal position to the bottom edge of the screen
(resulting in vertical lines). The strength of the effect depends on the
content of the windows that are visible. The effect "extends" from the
window causing it to other windows and the background. The effect is not
seen in screenshots (screenshot looks fine although the screen was
unreadable at the time I took it).

1.) is observed only under GNOME, but I see 2.) on KDE and LXDE, too. I
also tried various distributions. Most of my experiments have been done
on OpenSUSE 13.2, kernel-vanilla-3.19.rc5, xorg 7.6_1.16.1-5.1, intel
2.99.916, libdrm_intel1-2.4.58-1.1, Mesa 10.3.0-91.3.2. But I see the
artefacts also on Fedora 21 and Ubuntu 14.04.

As a matter of fact, I tried the modesetting driver and even fbdev and
observed similar effects, too. Thus i915 may actually not be the culprit
here, but could I figure no other place to ask for help than this
mailing list.

I have also experimented with resolutions and refresh rates. Only at
1024x768 and lower I was unable to reproduce the problem.

Any help would be highly appreciated. I am highly willing to do any
testing and/or debugging that may be necessary to find the problem.

Best regards
Martin


martin@artemis:~> xrandr --verbose
Screen 0: minimum 8 x 8, current 2560 x 1440, maximum 32767 x 32767
eDP1 connected primary 2560x1440+0+0 (0x49) normal (normal left inverted
right x axis y axis) 294mm x 165mm
	Identifier: 0x43
	Timestamp:  19447
	Subpixel:   unknown
	Gamma:      1.0:1.0:1.0
	Brightness: 1.0
	Clones:
	CRTC:       0
	CRTCs:      0 1 2
	Transform:  1.000000 0.000000 0.000000
	            0.000000 1.000000 0.000000
	            0.000000 0.000000 1.000000
	           filter:
	_MUTTER_PRESENTATION_OUTPUT: 0
	EDID:
		00ffffffffffff004d10121400000000
		24170104a51d1178068800a4564d9a26
		0d505400000001010101010101010101
		010101010101735f00a0a0a03a503020
		e50026a510000018454c00a0a0a03a50
		3020e50026a510000018aa3f00a0a0a0
		3a503020e50026a510000018000000fc
		004c5131333354314a5731390a2000b6
	BACKLIGHT: 892
		range: (0, 892)
	Backlight: 892
		range: (0, 892)
	scaling mode: Full aspect
		supported: None, Full, Center, Full aspect
	Broadcast RGB: Automatic
		supported: Automatic, Full, Limited 16:235
	audio: auto
		supported: force-dvi, off, auto, on
  2560x1440 (0x49) 244.350MHz -HSync -VSync *current +preferred
        h: width  2560 start 2608 end 2640 total 2720 skew    0 clock
89.83KHz
        v: height 1440 start 1454 end 1459 total 1498           clock
59.97Hz
  2560x1440 (0xae) 195.250MHz -HSync -VSync
        h: width  2560 start 2608 end 2640 total 2720 skew    0 clock
71.78KHz
        v: height 1440 start 1454 end 1459 total 1498           clock
47.92Hz
  2560x1440 (0xaf) 162.980MHz -HSync -VSync
        h: width  2560 start 2608 end 2640 total 2720 skew    0 clock
59.92KHz
        v: height 1440 start 1454 end 1459 total 1498           clock
40.00Hz
  1920x1440 (0xb0) 234.000MHz -HSync +VSync
        h: width  1920 start 2048 end 2256 total 2600 skew    0 clock
90.00KHz
        v: height 1440 start 1441 end 1444 total 1500           clock
60.00Hz
  1856x1392 (0xb1) 218.300MHz -HSync +VSync
        h: width  1856 start 1952 end 2176 total 2528 skew    0 clock
86.35KHz
        v: height 1392 start 1393 end 1396 total 1439           clock
60.01Hz
  1792x1344 (0xb2) 204.800MHz -HSync +VSync
        h: width  1792 start 1920 end 2120 total 2448 skew    0 clock
83.66KHz
        v: height 1344 start 1345 end 1348 total 1394           clock
60.01Hz
  1600x1200 (0xb3) 162.000MHz +HSync +VSync
        h: width  1600 start 1664 end 1856 total 2160 skew    0 clock
75.00KHz
        v: height 1200 start 1201 end 1204 total 1250           clock
60.00Hz
  1400x1050 (0xb4) 122.000MHz +HSync +VSync
        h: width  1400 start 1488 end 1640 total 1880 skew    0 clock
64.89KHz
        v: height 1050 start 1052 end 1064 total 1082           clock
59.98Hz
  1280x1024 (0xb5) 108.000MHz +HSync +VSync
        h: width  1280 start 1328 end 1440 total 1688 skew    0 clock
63.98KHz
        v: height 1024 start 1025 end 1028 total 1066           clock
60.02Hz
  1280x960 (0xb6) 108.000MHz +HSync +VSync
        h: width  1280 start 1376 end 1488 total 1800 skew    0 clock
60.00KHz
        v: height  960 start  961 end  964 total 1000           clock
60.00Hz
  1024x768 (0xb7) 65.000MHz -HSync -VSync
        h: width  1024 start 1048 end 1184 total 1344 skew    0 clock
48.36KHz
        v: height  768 start  771 end  777 total  806           clock
60.00Hz
  800x600 (0xb8) 40.000MHz +HSync +VSync
        h: width   800 start  840 end  968 total 1056 skew    0 clock
37.88KHz
        v: height  600 start  601 end  605 total  628           clock
60.32Hz
  800x600 (0xb9) 36.000MHz +HSync +VSync
        h: width   800 start  824 end  896 total 1024 skew    0 clock
35.16KHz
        v: height  600 start  601 end  603 total  625           clock
56.25Hz
  640x480 (0xba) 25.175MHz -HSync -VSync
        h: width   640 start  656 end  752 total  800 skew    0 clock
31.47KHz
        v: height  480 start  490 end  492 total  525           clock
59.94Hz
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel HD 4400] strongly irritating artefacts on 2560x1440 laptop display

[Chris Wilson]

On Tue, Jan 27, 2015 at 10:49:43PM +0100, Martin Wilck wrote:
> Dear list members,
> 
> I suffer from highly disturbing artefacts with the Intel HD 4400 chip
> set on thw 2560x1440 display (Sharp LQ133T1JW19) of my Fujitsu Lifebook
> S904. The effects are hard to describe, therefore I have uploaded a
> video to
> https://www.dropbox.com/sh/9wh0nzvp0w1phxz/AAAsvAqHN8ApXzor58FmGbwLa?dl=0.
>  I can distinguish 2 effects that are possibly related to each other:

[snip]
 
> 1.) is observed only under GNOME, but I see 2.) on KDE and LXDE, too. I
> also tried various distributions. Most of my experiments have been done
> on OpenSUSE 13.2, kernel-vanilla-3.19.rc5, xorg 7.6_1.16.1-5.1, intel
> 2.99.916, libdrm_intel1-2.4.58-1.1, Mesa 10.3.0-91.3.2. But I see the
> artefacts also on Fedora 21 and Ubuntu 14.04.
> 
> As a matter of fact, I tried the modesetting driver and even fbdev and
> observed similar effects, too. Thus i915 may actually not be the culprit
> here, but could I figure no other place to ask for help than this
> mailing list.

From watching the video, and your comments here, it is clear this is not
a driver bug (neither in the ddx nor opengl compositor). That leaves us
with hardware misconfiguration, aka kernel bug. Some of the artefacts
could be a display underrun. It would be good to check a drm.debug=0xe
dmesg for any warnings (and the info itself will be useful). Another
thing worth playing with is i915.enable_fbc=1 (though that is really
just swapping one problem with another).
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel HD 4400] strongly irritating artefacts on 2560x1440 laptop display

[Jani Nikula]

On Tue, 27 Jan 2015, Martin Wilck <mwilck@arcor.de> wrote:
> Dear list members,
>
> I suffer from highly disturbing artefacts with the Intel HD 4400 chip
> set on thw 2560x1440 display (Sharp LQ133T1JW19) of my Fujitsu Lifebook
> S904. The effects are hard to describe, therefore I have uploaded a
> video to
> https://www.dropbox.com/sh/9wh0nzvp0w1phxz/AAAsvAqHN8ApXzor58FmGbwLa?dl=0.
>  I can distinguish 2 effects that are possibly related to each other:

Just a quick guess, what does this say:

# cat /sys/module/i915/parameters/enable_{fbc,ips,psr}

Please try these module parameters for the non-zero ones:

i915.enable_fbc=0
i915.enable_ips=0
i915.enable_psr=0

BR,
Jani.

>
>  1. Maximized windows under GNOME3 (fully maximized or half-maximized on
> the right edge of the screen) fill the screen vertically with a single
> color (usually grey). Half-maximzed windows on the left edge are not
> affected. The effect is only seen if the Laptop screen is the main
> monitor in GNOME (showing the panel on top). Some applications, which
> start in maximized mode and can't be resized, are unusable on this screen.
>
>  2. Artefacts appear when windows are placed on the right and bottom
> areas of the screen. The artefacts look as if pixels extend vertically
> from their horizontal position to the bottom edge of the screen
> (resulting in vertical lines). The strength of the effect depends on the
> content of the windows that are visible. The effect "extends" from the
> window causing it to other windows and the background. The effect is not
> seen in screenshots (screenshot looks fine although the screen was
> unreadable at the time I took it).
>
> 1.) is observed only under GNOME, but I see 2.) on KDE and LXDE, too. I
> also tried various distributions. Most of my experiments have been done
> on OpenSUSE 13.2, kernel-vanilla-3.19.rc5, xorg 7.6_1.16.1-5.1, intel
> 2.99.916, libdrm_intel1-2.4.58-1.1, Mesa 10.3.0-91.3.2. But I see the
> artefacts also on Fedora 21 and Ubuntu 14.04.
>
> As a matter of fact, I tried the modesetting driver and even fbdev and
> observed similar effects, too. Thus i915 may actually not be the culprit
> here, but could I figure no other place to ask for help than this
> mailing list.
>
> I have also experimented with resolutions and refresh rates. Only at
> 1024x768 and lower I was unable to reproduce the problem.
>
> Any help would be highly appreciated. I am highly willing to do any
> testing and/or debugging that may be necessary to find the problem.
>
> Best regards
> Martin
>
>
> martin@artemis:~> xrandr --verbose
> Screen 0: minimum 8 x 8, current 2560 x 1440, maximum 32767 x 32767
> eDP1 connected primary 2560x1440+0+0 (0x49) normal (normal left inverted
> right x axis y axis) 294mm x 165mm
> 	Identifier: 0x43
> 	Timestamp:  19447
> 	Subpixel:   unknown
> 	Gamma:      1.0:1.0:1.0
> 	Brightness: 1.0
> 	Clones:
> 	CRTC:       0
> 	CRTCs:      0 1 2
> 	Transform:  1.000000 0.000000 0.000000
> 	            0.000000 1.000000 0.000000
> 	            0.000000 0.000000 1.000000
> 	           filter:
> 	_MUTTER_PRESENTATION_OUTPUT: 0
> 	EDID:
> 		00ffffffffffff004d10121400000000
> 		24170104a51d1178068800a4564d9a26
> 		0d505400000001010101010101010101
> 		010101010101735f00a0a0a03a503020
> 		e50026a510000018454c00a0a0a03a50
> 		3020e50026a510000018aa3f00a0a0a0
> 		3a503020e50026a510000018000000fc
> 		004c5131333354314a5731390a2000b6
> 	BACKLIGHT: 892
> 		range: (0, 892)
> 	Backlight: 892
> 		range: (0, 892)
> 	scaling mode: Full aspect
> 		supported: None, Full, Center, Full aspect
> 	Broadcast RGB: Automatic
> 		supported: Automatic, Full, Limited 16:235
> 	audio: auto
> 		supported: force-dvi, off, auto, on
>   2560x1440 (0x49) 244.350MHz -HSync -VSync *current +preferred
>         h: width  2560 start 2608 end 2640 total 2720 skew    0 clock
> 89.83KHz
>         v: height 1440 start 1454 end 1459 total 1498           clock
> 59.97Hz
>   2560x1440 (0xae) 195.250MHz -HSync -VSync
>         h: width  2560 start 2608 end 2640 total 2720 skew    0 clock
> 71.78KHz
>         v: height 1440 start 1454 end 1459 total 1498           clock
> 47.92Hz
>   2560x1440 (0xaf) 162.980MHz -HSync -VSync
>         h: width  2560 start 2608 end 2640 total 2720 skew    0 clock
> 59.92KHz
>         v: height 1440 start 1454 end 1459 total 1498           clock
> 40.00Hz
>   1920x1440 (0xb0) 234.000MHz -HSync +VSync
>         h: width  1920 start 2048 end 2256 total 2600 skew    0 clock
> 90.00KHz
>         v: height 1440 start 1441 end 1444 total 1500           clock
> 60.00Hz
>   1856x1392 (0xb1) 218.300MHz -HSync +VSync
>         h: width  1856 start 1952 end 2176 total 2528 skew    0 clock
> 86.35KHz
>         v: height 1392 start 1393 end 1396 total 1439           clock
> 60.01Hz
>   1792x1344 (0xb2) 204.800MHz -HSync +VSync
>         h: width  1792 start 1920 end 2120 total 2448 skew    0 clock
> 83.66KHz
>         v: height 1344 start 1345 end 1348 total 1394           clock
> 60.01Hz
>   1600x1200 (0xb3) 162.000MHz +HSync +VSync
>         h: width  1600 start 1664 end 1856 total 2160 skew    0 clock
> 75.00KHz
>         v: height 1200 start 1201 end 1204 total 1250           clock
> 60.00Hz
>   1400x1050 (0xb4) 122.000MHz +HSync +VSync
>         h: width  1400 start 1488 end 1640 total 1880 skew    0 clock
> 64.89KHz
>         v: height 1050 start 1052 end 1064 total 1082           clock
> 59.98Hz
>   1280x1024 (0xb5) 108.000MHz +HSync +VSync
>         h: width  1280 start 1328 end 1440 total 1688 skew    0 clock
> 63.98KHz
>         v: height 1024 start 1025 end 1028 total 1066           clock
> 60.02Hz
>   1280x960 (0xb6) 108.000MHz +HSync +VSync
>         h: width  1280 start 1376 end 1488 total 1800 skew    0 clock
> 60.00KHz
>         v: height  960 start  961 end  964 total 1000           clock
> 60.00Hz
>   1024x768 (0xb7) 65.000MHz -HSync -VSync
>         h: width  1024 start 1048 end 1184 total 1344 skew    0 clock
> 48.36KHz
>         v: height  768 start  771 end  777 total  806           clock
> 60.00Hz
>   800x600 (0xb8) 40.000MHz +HSync +VSync
>         h: width   800 start  840 end  968 total 1056 skew    0 clock
> 37.88KHz
>         v: height  600 start  601 end  605 total  628           clock
> 60.32Hz
>   800x600 (0xb9) 36.000MHz +HSync +VSync
>         h: width   800 start  824 end  896 total 1024 skew    0 clock
> 35.16KHz
>         v: height  600 start  601 end  603 total  625           clock
> 56.25Hz
>   640x480 (0xba) 25.175MHz -HSync -VSync
>         h: width   640 start  656 end  752 total  800 skew    0 clock
> 31.47KHz
>         v: height  480 start  490 end  492 total  525           clock
> 59.94Hz
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx

-- 
Jani Nikula, Intel Open Source Technology Center
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel HD 4400] strongly irritating artefacts on 2560x1440 laptop display

[Martin Wilck]

On 01/28/2015 10:45 AM, Chris Wilson wrote:

> From watching the video, and your comments here, it is clear this is not
> a driver bug (neither in the ddx nor opengl compositor). That leaves us
> with hardware misconfiguration, aka kernel bug. Some of the artefacts
> could be a display underrun. It would be good to check a drm.debug=0xe
> dmesg for any warnings (and the info itself will be useful). Another
> thing worth playing with is i915.enable_fbc=1 (though that is really
> just swapping one problem with another).
> -Chris
	
Thanks a lot for looking into this. I uploaded the dmesg to the dropbox
area on
https://www.dropbox.com/sh/9wh0nzvp0w1phxz/AAAsvAqHN8ApXzor58FmGbwLa?dl=0.

I will report the results of playing with the various parameters later.

The defaults (what I've tested so far) are:

enable_cmd_parser: 1
enable_execlists: 0
enable_fbc: -1
enable_hangcheck: Y
enable_ips: 1
enable_ppgtt: 1
enable_psr: 0
enable_rc6: 1

2 more questions to the experts: 1. Would it make sense to check out and
try the latest drm-intel-fixes or drm-intel-next branch? 2. How likely
is it that this phenomenon indicates faulty hardware?

Martin

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel HD 4400] strongly irritating artefacts on 2560x1440 laptop display

[Martin Wilck]

On 01/28/2015 10:45 AM, Jani Nikula wrote:

> Just a quick guess, what does this say:
> 
> # cat /sys/module/i915/parameters/enable_{fbc,ips,psr}

fbc: -1, ips: 1, psr: 0

> 
> Please try these module parameters for the non-zero ones:
> 
> i915.enable_fbc=0
> i915.enable_ips=0
> i915.enable_psr=0

These settings had no effect.
Thanks a lot anyway.

Martin

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel HD 4400] strongly irritating artefacts on 2560x1440 laptop display

[Martin Wilck]

On 01/28/2015 10:45 AM, Chris Wilson wrote:

> From watching the video, and your comments here, it is clear this is not
> a driver bug (neither in the ddx nor opengl compositor). That leaves us
> with hardware misconfiguration, aka kernel bug. Some of the artefacts
> could be a display underrun. It would be good to check a drm.debug=0xe
> dmesg for any warnings (and the info itself will be useful). Another
> thing worth playing with is i915.enable_fbc=1 (though that is really
> just swapping one problem with another).

i915.enable_fbc=1 had no effect.

I agree with you that this is probably a low-level problem, although the
fact that the artefacts strongly depend on Window content doesn't fit
that theory.

Martin

PS: I have just done a thing I thought I'd never do - I am downloading
an evaluation copy of Windows in order to check whether my HW is faulty.

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel HD 4400] strongly irritating artefacts on 2560x1440 laptop display

[Martin Wilck]

On 01/29/2015 10:47 PM, Martin Wilck wrote:> PS: I have just done a
> thing I thought I'd never do - I am downloading
> an evaluation copy of Windows in order to check whether my HW is faulty.

Just did that - the problem occurs on Windows, too. So this was a case
of a broken display.

I am sorry for having bothered, and would like to thank Chris and Jani
for replying.

Martin



_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel HD 4400] strongly irritating artefacts on 2560x1440 laptop display

[Jani Nikula]

On Sat, 31 Jan 2015, Martin Wilck <mwilck@arcor.de> wrote:
> On 01/29/2015 10:47 PM, Martin Wilck wrote:> PS: I have just done a
>> thing I thought I'd never do - I am downloading
>> an evaluation copy of Windows in order to check whether my HW is faulty.
>
> Just did that - the problem occurs on Windows, too. So this was a case
> of a broken display.
>
> I am sorry for having bothered, and would like to thank Chris and Jani
> for replying.

No problem, thanks for following up with this. Sorry to hear about the
display.

BR,
Jani.



-- 
Jani Nikula, Intel Open Source Technology Center
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel HD 4400] strongly irritating artefacts on 2560x1440 laptop display

[Martin Wilck]

Am 31. Januar 2015 16:36:47 MEZ, schrieb Jani Nikula <jani.nikula@linux.intel.com>:

>> I am sorry for having bothered, and would like to thank Chris and
>Jani
>> for replying.
>
>No problem, thanks for following up with this. Sorry to hear about the
>display.

It'll be a warranty issue. Actually I'm relieved because if the display had worked well under Windows, I wouldn't have known what to do.

Martin

>
>BR,
>Jani.


-- 
Diese Nachricht wurde von meinem Android-Mobiltelefon mit K-9 Mail gesendet.
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx