Help!!! Dual monitor display issue shifted | pressku.com

Trending 2 months ago

Ubuntu 20.04 patched up to date. I person a problem pinch nan video output from a Radeon paper (dual monitor) into 2 monitors. The paper is:

lspci –v 01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] RV620 LE [Radeon HD 3450] (prog-if 00 [VGA controller]) Subsystem: Dell OptiPlex 980 Flags: autobus master, accelerated devsel, latency 0, IRQ 33 Memory astatine e0000000 (64-bit, prefetchable) [size=256M] Memory astatine f7df0000 (64-bit, non-prefetchable) [size=64K] I/O ports astatine dc00 [size=256] Expansion ROM astatine 000c0000 [disabled] [size=128K] Capabilities: <access denied> Kernel driver successful use: radeon Kernel modules: radeon"

OR

sudo lshw -c display *-display description: VGA compatible controller product: RV620 LE [Radeon HD 3450] vendor: Advanced Micro Devices, Inc. [AMD/ATI] beingness id: 0 autobus info: pci@0000:01:00.0 version: 00 width: 64 bits clock: 33MHz capabilities: p.m. pciexpress msi vga_controller bus_master cap_list rom configuration: driver=radeon latency=0 resources: irq:33 memory:e0000000-efffffff memory:f7df0000-f7dfffff ioport:dc00(size=256) memory:c0000-dffff

With a "radeon" driver:

modinfo radeon filename: /lib/modules/5.15.0-94-generic/kernel/drivers/gpu/drm/radeon/radeon.ko license: GPL and further rights description: ATI Radeon author: Gareth Hughes, Keith Whitwell, others.

Both monitors spell done different KVMs. They are connected a “Join Displays” configuration successful Ubuntu (aka extended desktop). The setup is arsenic follows:

PC -> Radeon paper (dual output) -> KVM #1 -> DVI-0 -> KVM #2 -> DVI-1

DVI-0 goes done nan “good” KVM #1 and provides Ubuntu pinch nan correct characteristics for nan monitor. Consequently, edid tin publication its characteristics. This is nan “primary” monitor.

sudo get-edid | parse-edid This is read-edid type 3.0.2. Prepare for immoderate fun. Attempting to usage i2c interface Problem requesting enslaved address: Device aliases assets busy No EDID connected autobus 1 No EDID connected autobus 2 No EDID connected autobus 4 No EDID connected autobus 5 1 imaginable busses found: 3 256-byte EDID successfully retrieved from i2c autobus 3 Looks for illustration i2c was successful. Have a bully day. Checksum Correct Section "Monitor" Identifier "ASUS VA24E" ModelName "ASUS VA24E" VendorName "AUS" # Monitor Manufactured week 49 of 2020 # EDID type 1.3 # Analog Display DisplaySize 530 300 Gamma 2.20 Option "DPMS" "true" Horizsync 24-84 VertRefresh 48-75 # Maximum pixel timepiece is 180MHz #Not giving modular mode: 1920x1080, 60Hz #Not giving modular mode: 1680x1050, 60Hz #Not giving modular mode: 1440x900, 60Hz #Not giving modular mode: 1280x1024, 60Hz #Not giving modular mode: 1280x960, 60Hz #Not giving modular mode: 1280x720, 60Hz #Not giving modular mode: 1152x864, 75Hz Modeline "Mode 0" 148.50 1920 2008 2052 2200 1080 1084 1089 1125 +hsync +vsync EndSection

DVI-1 goes done nan “old” KVM#2 which does not supply Ubuntu pinch immoderate show characteristics. From nan documentation, I cognize that nan 2nd show is an ASUS VP249 which is highly akin to nan first one. This is nan “secondary” monitor. Edid cannot publication its characteristics. I get zilch. Even if I run:

ls /sys/class/drm/*/edid | xargs -i{} sh -c "echo {}; parse-edid < {}" /sys/class/drm/card0-DIN-1/edid Partial Read... Try again /sys/class/drm/card0-DVI-I-1/edid Checksum Correct Section "Monitor" Identifier "ASUS VA24E" ModelName "ASUS VA24E" VendorName "AUS" # Monitor Manufactured week 49 of 2020 # EDID type 1.3 # Analog Display DisplaySize 530 300 Gamma 2.20 Option "DPMS" "true" Horizsync 24-84 VertRefresh 48-75 # Maximum pixel timepiece is 180MHz #Not giving modular mode: 1920x1080, 60Hz #Not giving modular mode: 1680x1050, 60Hz #Not giving modular mode: 1440x900, 60Hz #Not giving modular mode: 1280x1024, 60Hz #Not giving modular mode: 1280x960, 60Hz #Not giving modular mode: 1280x720, 60Hz #Not giving modular mode: 1152x864, 75Hz Modeline "Mode 0" 148.50 1920 2008 2052 2200 1080 1084 1089 1125 +hsync +vsync EndSection /sys/class/drm/card0-DVI-I-2/edid Partial Read... Try again

Consequently, xrander switches to a “generic” group of resolutions for this monitor. This is what I get:

xrandr -q Screen 0: minimum 320 x 200, existent 2944 x 1080, maximum 8192 x 8192 DIN disconnected (normal near inverted correct x axis y axis) DVI-0 connected superior 1920x1080+0+0 (normal near inverted correct x axis y axis) 527mm x 296mm 1920x1080 60.00*+ 1680x1050 59.95 1280x1024 75.02 60.02 1440x900 59.89 1280x960 60.00 1152x864 75.00 1280x720 60.00 1024x768 75.03 70.07 60.00 832x624 74.55 800x600 72.19 75.00 60.32 56.25 640x480 75.00 72.81 66.67 59.94 720x400 70.08 DVI-1 connected 1024x768+1920+0 (normal near inverted correct x axis y axis) 0mm x 0mm 1024x768 60.00* 800x600 60.32 56.25 848x480 60.00 640x480 59.94

So, arsenic nan solution I request is not listed for DVI-1, I went nan modular route:

cvt 1920 1080 60 # 1920x1080 59.96 Hz (CVT 2.07M9) hsync: 67.16 kHz; pclk: 173.00 MHz Modeline "1920x1080_60.00" 173.00 1920 2048 2248 2576 1080 1083 1088 1120 -hsync +vsync xrandr --newmode "1920x1080_60.00" 173.00 1920 2048 2248 2576 1080 1083 1088 1120 -hsync +vsync xrandr --addmode DVI-1 "1920x1080_60.00"

And past I selected nan caller 1920x1080 solution successful Settings and changed nan DVI-1 output. This works. Just to make judge that this gets automatically setup during boot, I added nan past 2 commands into nan ~/.profile file. So, each clip it boots, nan DVI-1 solution gets group to 1920x1080.

BUT, location is simply a problem which I thought I could lick easily. It turns retired that nan DVI-1 show is shifted to nan near and down nan DVI-0 (the primary), thus, leaving a “dead band” connected nan near and a achromatic file to nan right.

Now, I thought that I could lick this because if I spell to settings and move to immoderate of nan different modes xrandr auto-generates (i.e. 1024x768 aliases 800x600), nan DVI-1 show IS centered correctly without overlaps of immoderate kind. So, theoretically speaking, if I could “clone” nan nonstop bid xrandr is utilizing for a little solution but swapping successful nan 1920x1080, that should beryllium it.

However, I person been astatine this for almost a week and nary matter what I effort it simply does not work!!!

Before we continue, please statement that I person astir 7 PCs pinch each kinds of Windows OS installed successful them going done nan nonstop aforesaid KVMs and nary of them, not one, has this problem. Furthermore, successful immoderate of those PCs I usage moreover higher solution without immoderate problems. So, it’s not nan KVM’s.

Furthermore, immoderate clip ago, I had a different Ubuntu PC (20.04) connected nan aforesaid KVMs and it did make nan correct centered show pinch nan 1920x1080 resolution. Unfortunately, that container was wiped retired because of hardware failure. So, nary thought why it worked but from nan backmost of my mind I look to retrieve that erstwhile I implemented nan cvt circus, it auto-centered. And no, I did NOT usage nan Auto characteristic successful nan monitor. And no, I can’t usage nan auto-feature of nan show because a) it does not set that acold and b) moreover if it did, it would screw-up nan different PCs. So, nan Auto characteristic of nan show is not an option.

Lastly, yes, if I move nan monitors astir betwixt KVM#1 and KVM#2, I get nan aforesaid problem connected KVM#2. Obviously, location is an rumor betwixt Ubuntu and KVM#2.

This leaves maine pinch nan “soft” action of adjusting via xrandr. These are immoderate of nan things I tried:

Xvidtune: nan only solution it accepts is 1024x768 and if I effort to alteration to thing else, it generates nan error: “Sorry: You person requested a mode-line that is not possible, aliases not supported by your hardware configuration”. So, this is simply a dormant end.

HSyncStart and HSyncEnd: I tried nan solution outlined here:

Similar problem

which worked for nan OP and it consisted successful “pushing” nan surface to nan near (in his case) by modifying HSyncStart and HSyncEnd by +300 units. In my case, I did nan aforesaid but I decreased by 300 units leaving maine pinch this command:

xrandr --newmode "1920x1080_60.00" 173.00 1920 1748 1948 2576 1080 1083 1088 1120 -hsync +vsync

But it does not work. The strategy rejects nan option:

xrandr --newmode "1920x1080_60.00" 173.00 1920 1748 1948 2576 1080 1083 1088 1120 -hsync +vsync peter@DEVU:~$ xrandr --addmode DVI-1 "1920x1080_60.00" peter@DEVU:~$ xrandr --output DVI-1 --mode "1920x1080_60.00" xrandr: Configure crtc 1 failed

And yes, I did play pinch different portion numbers pinch nan aforesaid error. CRTC 1 (whatever that whitethorn be) fails each time. So, this is different dormant end.

Position: I thought that nan easiest measurement to displacement to nan correct would beryllium thing on nan Pos bid (as hinted by nan xrandr output aliases calculated):

DVI-1 connected 1024x768+1920+0 (normal near inverted correct x axis y axis) 0mm x 0mm

So, I went pinch nan modular CRV 1920 1080 60 (same arsenic above) and changed nan output for DVI-1 to nan acceptable 1920x1080 resolution, aforesaid arsenic before. This is OK but shifted. Then I played astir pinch nan Pos bid arsenic so:

xrandr --output DVI-0 --mode "1920x1080" --pos 0x0 --rotate normal --output DVI-1 --mode 1920x1080_60.00 --pos 1920x0 --rotate normal

In nan illustration above, I americium shifting DVI-1 1920 “units” to nan right, 1920 being nan width of DVI-0 (same arsenic hinted by xrandr –q output). But nothing. No effect.

I tried aggregate parameter values connected DVI-1 and DVI-0 but nary luck. Pos does nan occupation of moving nan surface correct aliases near comparative to each different BUT it does not get free of nan “permanent” displacement of DVI-1 to nan left.

Driver replacement:I tried to spot if I could switch nan driver (currently “radeon”) pinch thing proprietary. The AMDGPU does not support my existent paper (too old) arsenic acold arsenic I tin tell. The AMD Catalyst (fglrx) driver does not support my existent kernel (up to 3.4). So, nary luck here. I americium stuck pinch “radeon”. But, past again, nan full driver circus for AMD is rather confusing and I could beryllium mistaken.

Anyhow, I americium astatine nan extremity of my statement and person nary thought whatsoever what to effort next.

Any thief would beryllium immensely appreciated,

Tx!

More
close