Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.user > #237086 > unrolled thread

Failure to use 3840x2160 30hz with Intel 620 chipset

Back to article view | Back to linux.debian.user

Contents

  Failure to use 3840x2160 30hz with Intel 620 chipset Pierre Couderc <pierre@couderc.eu> - 2021-07-08 23:00 +0200
    Re: Failure to use 3840x2160 30hz with Intel 620 chipset Felix Miata <mrmazda@earthlink.net> - 2021-07-08 23:10 +0200
      Re: Failure to use 3840x2160 30hz with Intel 620 chipset Felix Miata <mrmazda@earthlink.net> - 2021-07-08 23:40 +0200
        Re: Failure to use 3840x2160 30hz with Intel 620 chipset Polyna-Maude Racicot-Summerside <debian@polynamaude.com> - 2021-07-09 00:40 +0200
        Re: Failure to use 3840x2160 30hz with Intel 620 chipset Pierre Couderc <pierre@couderc.eu> - 2021-07-09 09:10 +0200
          Re: Failure to use 3840x2160 30hz with Intel 620 chipset Felix Miata <mrmazda@earthlink.net> - 2021-07-10 02:20 +0200
            Re: Failure to use 3840x2160 30hz with Intel 620 chipset Pierre Couderc <pierre@couderc.eu> - 2021-07-10 23:10 +0200
              Re: Failure to use 3840x2160 30hz with Intel 620 chipset Felix Miata <mrmazda@earthlink.net> - 2021-07-11 00:40 +0200
                Re: Failure to use 3840x2160 30hz with Intel 620 chipset Pierre Couderc <pierre@couderc.eu> - 2021-07-12 07:30 +0200
                Re: Failure to use 3840x2160 30hz with Intel 620 chipset Felix Miata <mrmazda@earthlink.net> - 2021-07-12 10:20 +0200
                  Re: Failure to use 3840x2160 30hz with Intel 620 chipset Pierre Couderc <pierre@couderc.eu> - 2021-07-16 08:40 +0200
                    Re: Failure to use 3840x2160 30hz with Intel 620 chipset Pierre Couderc <pierre@couderc.eu> - 2021-07-16 08:40 +0200
                      Re: Failure to use 3840x2160 30hz with Intel 620 chipset Felix Miata <mrmazda@earthlink.net> - 2021-07-16 10:00 +0200
                        Re: Failure to use 3840x2160 30hz with Intel 620 chipset Pierre Couderc <pierre@couderc.eu> - 2021-07-16 12:10 +0200
                Re: Failure to use 3840x2160 30hz with Intel 620 chipset Pierre Couderc <pierre@couderc.eu> - 2021-07-17 08:20 +0200

#237086 — Failure to use 3840x2160 30hz with Intel 620 chipset

I have a problem when trying a 4K display with Acer swift5. They say 
that 3840x2160 30hz is supported...

When I try (under bullseye) :

gtf  3840 2160 30

xrandr --newmode  "3840x2160_30.00"  339.57  3840 4080 4496 5152 2160 
2161 2164 2197  -HSync +Vsync
xrandr --addmode HDMI-1   3840x2160_30.00
xrandr --output HDMI-1 --mode "3840x2160_30.00"

the PC crashes (all is black, need to power off).

Did anyone success in similar case..?

Thnaks for any help.

PC

[toc] | [next] | [standalone]

#237087

Pierre Couderc composed on 2021-07-08 22:42 (UTC+0200):

> I have a problem when trying a 4K display with Acer swift5. They say 
> that 3840x2160 30hz is supported...

> When I try (under bullseye) :

> gtf  3840 2160 30

> xrandr --newmode  "3840x2160_30.00"  339.57  3840 4080 4496 5152 2160 
> 2161 2164 2197  -HSync +Vsync
> xrandr --addmode HDMI-1   3840x2160_30.00
> xrandr --output HDMI-1 --mode "3840x2160_30.00"

> the PC crashes (all is black, need to power off).
											
You should have no need of gtf or mode creation with 620 graphics. Mode creation
via xrandr is an anachronism. X knows very well how to auto generate modes when
provided with needed specs. If EDID, the usual source for the data, is broken,
then provide the specs via:

	/etc/X11/xorg.conf.d/50-monitor.conf

containing the horizontal sync and vertical refresh rates for your display, e.g.:

	Section "Monitor"
	    Identifier	"Default Monitor"
		HorizSync	28-85
		VertRefresh	50-100
	EndSection

Usually you can get these two specs if not elsewhere using:

	hwinfo --monitor

As to why you're having the problem, let us see the resulting URL from:

	pastebinit /var/log/Xorg.0.log

as a start.
-- 
Evolution as taught in public schools is, like religion,
	based on faith, not based on science.

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata

[toc] | [prev] | [next] | [standalone]

#237088

Oops, left out a line.

Felix Miata composed on 2021-07-08 17:08 (UTC-0400):

> Pierre Couderc composed on 2021-07-08 22:42 (UTC+0200):

>> I have a problem when trying a 4K display with Acer swift5. They say 
>> that 3840x2160 30hz is supported...

>> When I try (under bullseye) :

>> gtf  3840 2160 30

>> xrandr --newmode  "3840x2160_30.00"  339.57  3840 4080 4496 5152 2160 
>> 2161 2164 2197  -HSync +Vsync
>> xrandr --addmode HDMI-1   3840x2160_30.00
>> xrandr --output HDMI-1 --mode "3840x2160_30.00"

>> the PC crashes (all is black, need to power off).
 											
> You should have no need of gtf or mode creation with 620 graphics. Mode creation
> via xrandr is an anachronism. X knows very well how to auto generate modes when
> provided with needed specs. If EDID, the usual source for the data, is broken,
> then provide the specs via:

> 	/etc/X11/xorg.conf.d/50-monitor.conf

> containing the horizontal sync and vertical refresh rates for your display, e.g.:
> 
> 	Section "Monitor"
> 	    Identifier	"Default Monitor"
> 		HorizSync	28-85
> 		VertRefresh	50-100
		Option	"PreferredMode"	"3840x2160"
> 	EndSection
> 
> Usually you can get these two specs if not elsewhere using:
> 
> 	hwinfo --monitor
> 
> As to why you're having the problem, let us see the resulting URL from:
> 
> 	pastebinit /var/log/Xorg.0.log
> 
> as a start.-- 
Evolution as taught in public schools is, like religion,
	based on faith, not based on science.

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata

[toc] | [prev] | [next] | [standalone]

#237090

[Multipart message — attachments visible in raw view] — view raw

Hi,
>  Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!
OS/2 all good memories !
> 
> Felix Miata
> 

-- 
Polyna-Maude R.-Summerside
-Be smart, Be wise, Support opensource development

[toc] | [prev] | [next] | [standalone]

#237095

On 7/8/21 11:34 PM, Felix Miata wrote:
>
>>> the PC crashes (all is black, need to power off).
>   											
>> You should have no need of gtf or mode creation with 620 graphics. Mode creation
>> via xrandr is an anachronism. X knows very well how to auto generate modes when
>> provided with needed specs. If EDID, the usual source for the data, is broken,
>> then provide the specs via:

Yes, EDID is wrong and does not show 3840x2160 warranted by Samsung for UR59C 32".
  

>> 	/etc/X11/xorg.conf.d/50-monitor.conf
>> containing the horizontal sync and vertical refresh rates for your display, e.g.:
>>
>> 	Section "Monitor"
>> 	    Identifier	"Default Monitor"
>> 		HorizSync	28-85
>> 		VertRefresh	50-100
> 		Option	"PreferredMode"	"3840x2160"
>> 	EndSection
>>
>
This has no effect but I suppose the problem is due to the 2 screen 
configuration. So I have tried replacing Identifier with :

  Identifier    "HDMI-1"

But it is not better...

I have put in https://paste.debian.net/1203873/
the content of  .local/share/xorg/Xorg.0.log

It seems that the Modeline 3840x2160 is recognized, but the 30hz 
limitation is not expressed.

Acer swift5 specifies 3840x2160 but with a 30hz limitation, no more 
frequency.

Thanks you very much

PC

[toc] | [prev] | [next] | [standalone]

#237144

Pierre Couderc composed on 2021-07-09 09:05 (UTC+0200):

> Felix Miata wrote:

>>>> the PC crashes (all is black, need to power off).
   											
>>> You should have no need of gtf or mode creation with 620 graphics. Mode creation
>>> via xrandr is an anachronism. X knows very well how to auto generate modes when
>>> provided with needed specs. If EDID, the usual source for the data, is broken,
>>> then provide the specs via:
 
> Yes, EDID is wrong and does not show 3840x2160 warranted by Samsung for UR59C 32".
 
>>> 	/etc/X11/xorg.conf.d/50-monitor.conf
>>> containing the horizontal sync and vertical refresh rates for your display, e.g.:
>>>
>>> 	Section "Monitor"
>>> 	    Identifier	"Default Monitor"
>>> 		HorizSync	28-85
>>> 		VertRefresh	50-100
>> 		Option	"PreferredMode"	"3840x2160"
>>> 	EndSection

> This has no effect but I suppose the problem is due to the 2 screen 
> configuration. So I have tried replacing Identifier with :
 
>   Identifier    "HDMI-1"
 
> But it is not better...
 
> I have put in https://paste.debian.net/1203873/
> the content of  .local/share/xorg/Xorg.0.log
 
> It seems that the Modeline 3840x2160 is recognized, but the 30hz 
> limitation is not expressed.
 
> Acer swift5 specifies 3840x2160 but with a 30hz limitation, no more 
> frequency.
											
With dual displays, setup is more complicated, instead of /etc/X11/xorg.conf.d/50-monitor.conf
try /etc/X11/xorg.conf with the following content:

Section "Device"
    Identifier	"UHD620"
	Driver	"modesetting"
	Option	"monitor-eDP-1"		"eDPcon"
	Option	"monitor-HDMI-1"	"HDMIcon"
EndSection

Section "Monitor"
    Identifier	"HDMIcon"
	HorizSync	30-135
	VertRefresh	29-31
	Option	"PreferredMode"	"3840x2160"
EndSection

Section "Monitor"
    Identifier	"eDPcon"
	Option	"Primary" "true"
EndSection

Section "Screen"
    Identifier "extScreen"
	Device	"UHD620"
	Monitor	"HDMIcon"
EndSection

Section "Screen"
    Identifier "intScreen"
	Device	"UHD620"
	Monitor	"eDPcon"
EndSection

Also, make sure your HDMI cable is up to specs for 4K.
-- 
Evolution as taught in public schools is, like religion,
	based on faith, not based on science.

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata

[toc] | [prev] | [next] | [standalone]

#237175

[Multipart message — attachments visible in raw view] — view raw

On 7/10/21 2:17 AM, Felix Miata wrote:

> Pierre Couderc composed on 2021-07-09 09:05 (UTC+0200):  > > >> Yes, EDID is wrong and does not show 3840x2160 warranted by 
Samsung >> for UR59C 32". > > > > With dual displays, setup is more 
complicated, instead of > /etc/X11/xorg.conf.d/50-monitor.conf try 
/etc/X11/xorg.conf with the > following content: > > Section "Device" 
Identifier "UHD620" Driver "modesetting" Option > "monitor-eDP-1" 
"eDPcon" Option "monitor-HDMI-1" "HDMIcon" > EndSection > > Section 
"Monitor" Identifier "HDMIcon" HorizSync 30-135 VertRefresh > 29-31 
Option "PreferredMode" "3840x2160" EndSection > > Section "Monitor" 
Identifier "eDPcon"Section "Monitor" Identifier > "HDMIcon" HorizSync 
30-135 VertRefresh 29-31 ModeLine > "3840x2160_30.00" 339.57 3840 4080 
4496 5152 2160 2161 2164 2197 > -HSync +Vsync Option "PreferredMode" 
"3840x2160_30.00" EndSection > > Section "Monitor" > > Option "Primary" 
"true" EndSection > > Section "Screen" Identifier "extScreen" Device 
"UHD620" Monitor > "HDMIcon" EndSection > > Section "Screen" Identifier 
"intScreen" Device "UHD620" Monitor > "eDPcon" EndSection > > Also, make 
sure your HDMI cable is up to specs for 4K.

The HDMI cable was provided with the Samsung display.

I have tried your solution which starts the 2 displays in  1920x1080

see : https://paste.debian.net/1203999/

for( .local/share/xorg/Xorg.0.log)


Then I have tried  with : (please remind that the 4K display is 
specified 30hz for 3840x2160).

Section "Monitor"
     Identifier  "HDMIcon"
         HorizSync       30-135
         VertRefresh     29-31
         ModeLine "3840x2160_30.00"  339.57  3840 4080 4496 5152 2160 
2161 2164 2197  -HSync +Vsync
         Option  "PreferredMode" "3840x2160_30.00"
EndSection

The result is  : https://paste.debian.net/1204000

It does not strictly crashes but only the first monitor is displayed in 
X mode, the second monitor (HDMI) remains as in non X mode, and the 
mouse does not want to quit the high left corner of the HDMI monitor.

I feel as if  there is some driver problem...

Than you very much, Felix, I progress !

[toc] | [prev] | [next] | [standalone]

#237180

Pierre Couderc composed on 2021-07-10 23:07 (UTC+0200):
 
> see : https://paste.debian.net/1203999/
 
> for( .local/share/xorg/Xorg.0.log)
											
What does xrandr -q report booted this way?

https://wiki.archlinux.org/title/Backlight may warrant a read.

I have to suspect acpi_backlight=vendor on kernel command line is not taken
well by the 3840x2160 screen.

Whatever caused [ 313.732] (WW) modeset(0): Option "PreferredMode" is not used
I can't tell.

> https://paste.debian.net/plain/1204000

[ 870.193] (EE) modeset(0): failed to set mode: Invalid argument appears multiple times quite some lines after these:
[   869.149] (II) modeset(0): Using user preference for initial modes
[   869.149] (II) modeset(0): Output eDP-1 using initial mode 1920x1080 +0+0
[   869.149] (II) modeset(0): Output HDMI-1 using initial mode 3840x2160_30.00 +0+0

The last line indicates the screens overlap instead of side by side placement. It
should end +1920+0 if everything else is OK, which pretty much explains absence of
4k screen output.

Have you tried booting with a clean kernel command line, e.g. without
acpi_backlight=vendor in particular?

Likely dmesg and journal need to be explored to find a reason for the invalid
argument errors. Adding drm.debug=0x1e & log_buf_len=1M to kernel command line
may be needed to help diagnosis.

I only had a 4K display for 3 weeks 3 years ago, so it's hard for me to be very
helpful by trying to match behavior, or construct a configuration file free of
errors.
-- 
Evolution as taught in public schools is, like religion,
	based on faith, not based on science.

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata

[toc] | [prev] | [next] | [standalone]

#237226

On 7/11/21 12:29 AM, Felix Miata wrote:
> Pierre Couderc composed on 2021-07-10 23:07 (UTC+0200):
>   
>
Thank your for your precious help.

I quit this big 4k for a week of holidays without any monitor...

And come back to it nextweek !

[toc] | [prev] | [next] | [standalone]

#237233

Felix Miata composed on 2021-07-10 18:29 (UTC-0400):

> I only had a 4K display for 3 weeks 3 years ago, so it's hard for me to be very
> helpful by trying to match behavior, or construct a configuration file free of
> errors.
											
I forgot something important to this thread. I got a broken 55" 4K LG TV
for free 8 weeks ago. I spent $40US on parts to fix it.

I just tried connecting it to a Kaby Lake Intel PC (HD 620, not UHD 620)
via direct HDMI, along with an ordinary 1920x1080 Samsung HDTV via DisplayPort
to HDMI adapter, using openSUSE 15.2. With only the LG connected, X opened up
in 3840x2160@30 automatically. With both connected, both come up in 1920x1080@60.
However, xrandr --output HDMI-2 --primary --mode 3840x2160 --output DP-1 --mode 1920x1080 --above HDMI-2
worked a charm to set 3840x2160@30 on the LG and set the small Samsung above
the big LG.

Rebooting to Bullseye to try the same, I can't get the Samsung HDTV out of
unsupported mode, even though xrandr reports the Samsung is using its native
mode, or trying with any other supported mode selected. This smells like a
Bullseye bug waiting to be reported.

Next I tried openSUSE Tumbleweed. It works as expected there too:
# xrandr --output HDMI-2 --primary --mode 3840x2160 --output DP-1 --mode 1920x1080 --above HDMI-2
# xrandr | egrep 'onnect|creen|\*' | grep -v disconn | sort -r
Screen 0: minimum 320 x 200, current 3840 x 3240, maximum 16384 x 16384
HDMI-2 connected primary 3840x2160+0+1080 (normal left inverted right x axis y axis) 1600mm x 900mm
DP-1 connected 1920x1080+0+0 (normal left inverted right x axis y axis) 160mm x 90mm
   3840x2160     30.00*   25.00    24.00    29.97    23.98
   1920x1080     60.00*+  59.94    30.00    24.00    29.97    23.98
# inxi -SGay
System:
  Host: ab250 Kernel: 5.12.9-1-default x86_64 bits: 64 compiler: gcc v: 11.1.1
  parameters:...
  Desktop: Trinity R14.0.10 tk: Qt 3.5.0 info: kicker wm: Twin 3.0 vt: 7
  dm: TDM Distro: openSUSE Tumbleweed 20210611
Graphics:
  Device-1: Intel HD Graphics 630 vendor: ASUSTeK driver: i915 v: kernel
  bus-ID: 00:02.0 chip-ID: 8086:5912 class-ID: 0300
  Display: x11 server: X.Org 1.20.11 driver: loaded: modesetting
  unloaded: fbdev,vesa alternate: intel display-ID: :0 screens: 1
  Screen-1: 0 s-res: 3840x3240 s-dpi: 144 s-size: 678x572mm (26.7x22.5")
  s-diag: 887mm (34.9")
  Monitor-1: DP-1 res: 1920x1080 hz: 60 dpi: 305 size: 160x90mm (6.3x3.5")
  diag: 184mm (7.2")
  Monitor-2: HDMI-2 res: 3840x2160 hz: 30 dpi: 61
  size: 1600x900mm (63.0x35.4") diag: 1836mm (72.3")
  OpenGL: renderer: Mesa DRI Intel HD Graphics 630 (KBL GT2)
  v: 4.6 Mesa 21.1.2 compat-v: 3.0 direct render: Yes
-- 
Evolution as taught in public schools is, like religion,
	based on faith, not based on science.

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata

[toc] | [prev] | [next] | [standalone]

#237468

On 7/12/21 10:11 AM, Felix Miata wrote:
> Felix Miata composed on 2021-07-10 18:29 (UTC-0400):
>
>> I only had a 4K display for 3 weeks 3 years ago, so it's hard for me to be very
>> helpful by trying to match behavior, or construct a configuration file free of
>> errors.
> 											
> I forgot something important to this thread. I got a broken 55" 4K LG TV
> for free 8 weeks ago. I spent $40US on parts to fix it.
>
> I just tried connecting it to a Kaby Lake Intel PC (HD 620, not UHD 620)
> via direct HDMI, along with an ordinary 1920x1080 Samsung HDTV via DisplayPort
> to HDMI adapter, using openSUSE 15.2. With only the LG connected, X opened up
> in 3840x2160@30 automatically. With both connected, both come up in 1920x1080@60.
> However, xrandr --output HDMI-2 --primary --mode 3840x2160 --output DP-1 --mode 1920x1080 --above HDMI-2
> worked a charm to set 3840x2160@30 on the LG and set the small Samsung above
> the big LG.
>
> Rebooting to Bullseye to try the same, I can't get the Samsung HDTV out of
> unsupported mode, even though xrandr reports the Samsung is using its native
> mode, or trying with any other supported mode selected. This smells like a
> Bullseye bug waiting to be reported.
>
> Next I tried openSUSE Tumbleweed. It works as expected there too:
> # xrandr --output HDMI-2 --primary --mode 3840x2160 --output DP-1 --mode 1920x1080 --above HDMI-2
> # xrandr | egrep 'onnect|creen|\*' | grep -v disconn | sort -r
> Screen 0: minimum 320 x 200, current 3840 x 3240, maximum 16384 x 16384
> HDMI-2 connected primary 3840x2160+0+1080 (normal left inverted right x axis y axis) 1600mm x 900mm
> DP-1 connected 1920x1080+0+0 (normal left inverted right x axis y axis) 160mm x 90mm
>     3840x2160     30.00*   25.00    24.00    29.97    23.98
>     1920x1080     60.00*+  59.94    30.00    24.00    29.97    23.98
> # inxi -SGay
> System:
>    Host: ab250 Kernel: 5.12.9-1-default x86_64 bits: 64 compiler: gcc v: 11.1.1
>    parameters:...
>    Desktop: Trinity R14.0.10 tk: Qt 3.5.0 info: kicker wm: Twin 3.0 vt: 7
>    dm: TDM Distro: openSUSE Tumbleweed 20210611
> Graphics:
>    Device-1: Intel HD Graphics 630 vendor: ASUSTeK driver: i915 v: kernel
>    bus-ID: 00:02.0 chip-ID: 8086:5912 class-ID: 0300
>    Display: x11 server: X.Org 1.20.11 driver: loaded: modesetting
>    unloaded: fbdev,vesa alternate: intel display-ID: :0 screens: 1
>    Screen-1: 0 s-res: 3840x3240 s-dpi: 144 s-size: 678x572mm (26.7x22.5")
>    s-diag: 887mm (34.9")
>    Monitor-1: DP-1 res: 1920x1080 hz: 60 dpi: 305 size: 160x90mm (6.3x3.5")
>    diag: 184mm (7.2")
>    Monitor-2: HDMI-2 res: 3840x2160 hz: 30 dpi: 61
>    size: 1600x900mm (63.0x35.4") diag: 1836mm (72.3")
>    OpenGL: renderer: Mesa DRI Intel HD Graphics 630 (KBL GT2)
>    v: 4.6 Mesa 21.1.2 compat-v: 3.0 direct render: Yes

[toc] | [prev] | [next] | [standalone]

#237469

On 7/16/21 8:31 AM, Pierre Couderc wrote:
>
> On 7/12/21 10:11 AM, Felix Miata wrote:
>> Felix Miata composed on 2021-07-10 18:29 (UTC-0400):
>>
>>> I only had a 4K display for 3 weeks 3 years ago, so it's hard for me 
>>> to be very
>>> helpful by trying to match behavior, or construct a configuration 
>>> file free of
>>> errors.
>>
>> I forgot something important to this thread. I got a broken 55" 4K LG TV
>> for free 8 weeks ago. I spent $40US on parts to fix it.
>>
>> I just tried connecting it to a Kaby Lake Intel PC (HD 620, not UHD 620)
>> via direct HDMI, along with an ordinary 1920x1080 Samsung HDTV via 
>> DisplayPort
>> to HDMI adapter, using openSUSE 15.2. With only the LG connected, X 
>> opened up
>> in 3840x2160@30 automatically. With both connected, both come up in 
>> 1920x1080@60.
>> However, xrandr --output HDMI-2 --primary --mode 3840x2160 --output 
>> DP-1 --mode 1920x1080 --above HDMI-2
>> worked a charm to set 3840x2160@30 on the LG and set the small 
>> Samsung above
>> the big LG.
>>
>> Rebooting to Bullseye to try the same, I can't get the Samsung HDTV 
>> out of
>> unsupported mode, even though xrandr reports the Samsung is using its 
>> native
>> mode, or trying with any other supported mode selected. This smells 
>> like a
>> Bullseye bug waiting to be reported.
>>
>> Next I tried openSUSE Tumbleweed. It works as expected there too:
>> # xrandr --output HDMI-2 --primary --mode 3840x2160 --output DP-1 
>> --mode 1920x1080 --above HDMI-2
>> # xrandr | egrep 'onnect|creen|\*' | grep -v disconn | sort -r
>> Screen 0: minimum 320 x 200, current 3840 x 3240, maximum 16384 x 16384
>> HDMI-2 connected primary 3840x2160+0+1080 (normal left inverted right 
>> x axis y axis) 1600mm x 900mm
>> DP-1 connected 1920x1080+0+0 (normal left inverted right x axis y 
>> axis) 160mm x 90mm
>>     3840x2160     30.00*   25.00    24.00    29.97    23.98
>>     1920x1080     60.00*+  59.94    30.00    24.00    29.97 23.98
>> # inxi -SGay
>> System:
>>    Host: ab250 Kernel: 5.12.9-1-default x86_64 bits: 64 compiler: gcc 
>> v: 11.1.1
>>    parameters:...
>>    Desktop: Trinity R14.0.10 tk: Qt 3.5.0 info: kicker wm: Twin 3.0 
>> vt: 7
>>    dm: TDM Distro: openSUSE Tumbleweed 20210611
>> Graphics:
>>    Device-1: Intel HD Graphics 630 vendor: ASUSTeK driver: i915 v: 
>> kernel
>>    bus-ID: 00:02.0 chip-ID: 8086:5912 class-ID: 0300
>>    Display: x11 server: X.Org 1.20.11 driver: loaded: modesetting
>>    unloaded: fbdev,vesa alternate: intel display-ID: :0 screens: 1
>>    Screen-1: 0 s-res: 3840x3240 s-dpi: 144 s-size: 678x572mm 
>> (26.7x22.5")
>>    s-diag: 887mm (34.9")
>>    Monitor-1: DP-1 res: 1920x1080 hz: 60 dpi: 305 size: 160x90mm 
>> (6.3x3.5")
>>    diag: 184mm (7.2")
>>    Monitor-2: HDMI-2 res: 3840x2160 hz: 30 dpi: 61
>>    size: 1600x900mm (63.0x35.4") diag: 1836mm (72.3")
>>    OpenGL: renderer: Mesa DRI Intel HD Graphics 630 (KBL GT2)
>>    v: 4.6 Mesa 21.1.2 compat-v: 3.0 direct render: Yes
>
Before reporting a bug, maybe it could be too a question of driver...?

[toc] | [prev] | [next] | [standalone]

#237470

Pierre Couderc composed on 2021-07-16 08:34 (UTC+0200):

> Before reporting a bug, maybe it could be too a question of driver...?
											
There are two X drivers for Intel IGPs capable at least theoretically of
supporting 3840x2160:

	Intel DDX, provided by xserver-xorg-video-intel
	Modesetting DIX, provided by xserver-xorg-core

The Intel driver has been essentially in maintenance mode for the past 7 years,
its main reason for keeping it maintained being it is the only driver available
supporting ancient Intel IGPs. The only way to tell if it works better on newer
hardware is to try it. Don't expect much from anything as new as yours using the
Intel DDX instead of the Modesetting DIX.

Either of the two drivers can be specifically selected via a .conf file in
/etc/X11/xorg.conf.d/, e.g., assuming xserver-xorg-video-intel is installed:

	# file: 15-ddxdrv.conf
	Section "Device"
	  Identifier "DDX"
		Driver "intel"
#		Driver "modesetting"
	EndSection

selects the Intel DDX, blocking X from attempting use of the Modesetting DIX.
-- 
Evolution as taught in public schools is, like religion,
	based on faith, not based on science.

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata

[toc] | [prev] | [next] | [standalone]

#237471

On 7/16/21 9:52 AM, Felix Miata wrote:
> Pierre Couderc composed on 2021-07-16 08:34 (UTC+0200):
>
>> Before reporting a bug, maybe it could be too a question of driver...?
> 											
> There are two X drivers for Intel IGPs capable at least theoretically of
> supporting 3840x2160:
>
> 	Intel DDX, provided by xserver-xorg-video-intel
> 	Modesetting DIX, provided by xserver-xorg-core
>
> The Intel driver has been essentially in maintenance mode for the past 7 years,
> its main reason for keeping it maintained being it is the only driver available
> supporting ancient Intel IGPs. The only way to tell if it works better on newer
> hardware is to try it. Don't expect much from anything as new as yours using the
> Intel DDX instead of the Modesetting DIX.
>
> Either of the two drivers can be specifically selected via a .conf file in
> /etc/X11/xorg.conf.d/, e.g., assuming xserver-xorg-video-intel is installed:
>
> 	# file: 15-ddxdrv.conf
> 	Section "Device"
> 	  Identifier "DDX"
> 		Driver "intel"
> #		Driver "modesetting"
> 	EndSection
>
> selects the Intel DDX, blocking X from attempting use of the Modesetting DIX.

I have tried intel driver which does not work but does not crashes.

I come back to "modesetting" driver.

[toc] | [prev] | [next] | [standalone]

#237479

On 7/11/21 12:29 AM, Felix Miata wrote:
> Pierre Couderc composed on 2021-07-10 23:07 (UTC+0200):
>   
>> see :https://paste.debian.net/1203999/
>   
>> for( .local/share/xorg/Xorg.0.log)
> 											
> What does xrandr -q report booted this way?
Screen 0: minimum 320 x 200, current 3840 x 1080, maximum 16384 x 16384
eDP-1 connected primary 1920x1080+0+0 (normal left inverted right x axis 
y axis) 309mm x 174mm panning 1920x1080+0+0
    1920x1080     60.05 +  60.01    59.97    59.96*   59.93
    1680x1050     59.95    59.88
    1600x1024     60.17
    1400x1050     59.98
    1600x900      59.99    59.94    59.95    59.82
    1280x1024     60.02
    1440x900      59.89
    1400x900      59.96    59.88
    1280x960      60.00
    1440x810      60.00    59.97
    1368x768      59.88    59.85
    1360x768      59.80    59.96
    1280x800      59.99    59.97    59.81    59.91
    1152x864      60.00
    1280x720      60.00    59.99    59.86    59.74
    1024x768      60.04    60.00
    960x720       60.00
    928x696       60.05
    896x672       60.01
    1024x576      59.95    59.96    59.90    59.82
    960x600       59.93    60.00
    960x540       59.96    59.99    59.63    59.82
    800x600       60.00    60.32    56.25
    840x525       60.01    59.88
    864x486       59.92    59.57
    800x512       60.17
    700x525       59.98
    800x450       59.95    59.82
    640x512       60.02
    720x450       59.89
    700x450       59.96    59.88
    640x480       60.00    59.94
    720x405       59.51    58.99
    684x384       59.88    59.85
    680x384       59.80    59.96
    640x400       59.88    59.98
    576x432       60.06
    640x360       59.86    59.83    59.84    59.32
    512x384       60.00
    512x288       60.00    59.92
    480x270       59.63    59.82
    400x300       60.32    56.34
    432x243       59.92    59.57
    320x240       60.05
    360x202       59.51    59.13
    320x180       59.84    59.32
HDMI-1 connected 1920x1080+1920+0 (normal left inverted right x axis y 
axis) 697mm x 392mm panning 1920x1080+1920+0
    1920x1080     30.00*   29.97
DP-1 disconnected (normal left inverted right x axis y axis)
HDMI-2 disconnected (normal left inverted right x axis y axis)

> https://wiki.archlinux.org/title/Backlight  may warrant a read.
>
> I have to suspect acpi_backlight=vendor on kernel command line is not taken
> well by the 3840x2160 screen.

I have tried to remove acpi_backlight=vendor, but

https://paste.debian.net/1204642/

shows no significant difference with https://paste.debian.net/1203999

> Whatever caused [ 313.732] (WW) modeset(0): Option "PreferredMode" is not used
> I can't tell.

My idea is that this reason is that the limitation to 30Hz is not 
described by


Section "Monitor"
     Identifier  "HDMIcon"
         HorizSync       30-135
         VertRefresh     29-31
         Option  "PreferredMode" "3840x2160"
EndSection

So, I have used gtf to create a 3840 2160 30 Modeline. As i have read 
the cvt is more "modern", I trie now with

Modeline "3840x2160_30.00"  338.75  3840 4080 4488 5136  2160 2163 2168 
2200 -hsync +vsync
  And...

It fails too with :

https://paste.debian.net/1204643/

With the following result :

- eDP-1 is dislayed correctly

- HDMI seems to remain as before startx is executed, but a cursor, 
certainly a X cursor appears in the high left corner of the screen. This 
cursor seems to move a little when I move the mouse but returns 
immediately to the corner. I have to kill kill the session with another 
console.

I think that either I have not found hye good "Modeline" either it is a 
bug (maybe in the driver), but it is surprising that I am the 1rst to 
have this problem as this monitor is available since 2019...

[toc] | [prev] | [standalone]

Back to top | Article view | linux.debian.user

csiph-web