xrandr rotate affecting mouse cursor on all monitors.

clanger · 2010-01-05 15:41:21

Hi All,

I have a small problem with my multihead setup. I have two identical monitors, one horizontal and one vertical (like this |-- ).
I am using xrandr to rotate position and rotate the vertical monitor in my .xinitrc. Everything works as expected except the mouse cursor on the horizontal monitor has been rotated and the "click point" is no longer beneath the cursor, making it hard to browse the web etc.

Has anyone anyelse had this problem? Any luck in fixing it? Any help would be much appreciated.

Details:

I am using a Geforce9500 (2 dvi ports).
I am using the nouveau driver from the official repos (nvidia driver doesn't support xrandr and was causing me constant problems).
I am using xmonad as a WM (from the repos).

======= .xinitrc ===========
xrandr --output DVI-I-1 --rotate left
xrandr --output DVI-I-0 --right-of DVI-I-1
exec xmonad

======= xorg.conf =========

Section "ServerLayout"
Identifier "Tfellover"
Screen "Screen0"
InputDevice "Mouse0" "CorePointer"
InputDevice "Keyboard0" "CoreKeyboard"
EndSection

Section "Files"
ModulePath "/usr/lib/xorg/modules"
FontPath "/usr/share/fonts/misc"
FontPath "/usr/share/fonts/100dpi:unscaled"
FontPath "/usr/share/fonts/75dpi:unscaled"
FontPath "/usr/share/fonts/TTF"
FontPath "/usr/share/fonts/Type1"
EndSection

Section "Module"
Load "dbe"
Load "dri2"
Load "glx"
Load "extmod"
EndSection

Section "ServerFlags"
Option "AllowEmptyInput" "False"
Option "AutoAddDevices" "False"
Option "AutoEnableDevices" "False"
Option "DontZap" "False"
EndSection

Section "InputDevice"
Identifier "Keyboard0"
Driver "kbd"
Option "XkbLayout" "gb"
EndSection

Section "InputDevice"
Identifier "Mouse0"
Driver "mouse"
Option "Protocol" "auto"
Option "Device" "/dev/input/mice"
Option "ZAxisMapping" "4 5 6 7"
EndSection

Section "Monitor"
Identifier "horizontal"
VendorName "Samsung"
ModelName "Samsung SyncMaster P2050"
EndSection

Section "Monitor"
Identifier "vertical"
VendorName "Unknown"
ModelName "Samsung SyncMaster"
EndSection

Section "Device"
Identifier "GF9"
Driver "nouveau"
VendorName "NVIDIA Corporation"
BoardName "GeForce 9500 GT"
BusID "PCI:1:0:0"
Option "Monitor-DVI-I-0" "horizontal"
Option "Monitor-DVI-I-1" "vertical"
EndSection

Section "Screen"
Identifier "Screen0"
Device "GF9"
Monitor "horizontal"
DefaultDepth 24
SubSection "Display"
Depth 24
Virtual 2500 1600
EndSubSection
Option "AIGLX" "false"
EndSection

=================

Thanks

shining · 2010-01-09 11:34:54

Heh, I was going to say "yes, it's known", but it is probably thanks to you
http://lists.freedesktop.org/archives/n … 04611.html
But maybe you did not see a part of that mail that came a few hours later :
http://lists.freedesktop.org/archives/n … 04613.html

I'm quite sure that, without UMS, this function doesn't make sense
anymore, you could leave this fun for the X server (we can use the
load_cursor_argb hook even on the cards we don't advertise ARGB
support). As a bonus that would probably solve Craig's rotated cursor
issue.

If there isn't any news in 1 week, also make a bug report with the same info than in your mail. It will have less chances to be forgotten.
http://nouveau.freedesktop.org/wiki/FrontPage#Bugs

goodgrue · 2010-01-13 19:34:57

Same issue here! nouveau driver from the official repos, xmonad... My card is a GeForce 9500 GT (one DVI output and one VGA).

I read through the posts shining referenced, but 1) it didn't seem like there was concensus on whether the patch would be accepted, and 2) Apparently it breaks some other functionality, so I'm not too keen to try it.

Have there been any further developments here? Any luck on your end, clanger?

Here's my xorg.conf, for reference:

Section "Files"
    ModulePath    "/usr/lib/xorg/modules"
    
    FontPath   "/usr/share/fonts/misc"
    FontPath   "/usr/share/fonts/100dpi:unscaled"
    FontPath   "/usr/share/fonts/75dpi:unscaled"
    FontPath   "/usr/share/fonts/TTF"
    FontPath   "/usr/share/fonts/Type1"
    FontPath   "/usr/share/fonts/local"
EndSection

Section "Device"
        Identifier  "GeForce 9500 GT"
        Driver      "nouveau"
        Option      "Monitor-VGA-0" "VGA-0"
        Option      "Monitor-DVI-I-0" "DVI-I-0"
EndSection

Section "ServerLayout"
    Identifier  "Layout0"
    Screen      "Screen0" 0 420
    Option      "AIGLX" "false"
EndSection

Section "Monitor"
    Identifier "VGA-0"
    Option "DPMS" "true"
    Option "PreferredMode" "1920x1080"
EndSection

Section "Monitor"
    Identifier "DVI-I-0"
    Option "DPMS" "true"
    Option "PreferredMode" "1920x1080"
    Option "RightOf" "VGA-0"
EndSection

Section "Screen"
    Identifier "Screen0"
    Device      "GeForce 9500 GT"
    DefaultDepth 24
    SubSection "Display"
        Depth   24
        Virtual 3000 1920
    EndSubSection
EndSection

Section "ServerFlags"
    Option "AllowClosedownGrabs" "1"
EndSection

Thanks!

shining · 2010-01-13 19:55:01

goodgrue wrote:

Same issue here! nouveau driver from the official repos, xmonad... My card is a GeForce 9500 GT (one DVI output and one VGA).
I read through the posts shining referenced, but 1) it didn't seem like there was concensus on whether the patch would be accepted, and 2) Apparently it breaks some other functionality, so I'm not too keen to try it.

The guy who posted the patch is the main developer, so it was applied eventually. It mostly breaks functionality that you shouldn't use anyway (non-kms). You really should use kms if you don't already.
And then you can try the latest ddx git : http://aur.archlinux.org/packages.php?ID=11891
Please report if that helps.

goodgrue · 2010-01-13 20:16:55

I just installed and am now running the latest git version of nouveau (installed from the AUR, xf86-video-nouveau-git package). The same behavior is occurring. I am using KMS late start, set up according to the nouveau arch wiki article: http://wiki.archlinux.org/index.php/Nou … ouveau-git.

Could this be an xrandr 1.2 issue rather than a nouveau issue? I don't really understand the X stack well enough to know how to debug this, but I'm happy to try any suggestions.

edit:
Just for fun I added
Option "HWCursor" "false"
to the Device section of my xorg.conf file. Then when I used xrandr to rotate monitor 2, monitor 2 went to sleep and monitor 1 because garbled. No amount of button mashing or Ctrl-Alt-Backspacing could fix things, so I had to do a hard restart.

Last edited by goodgrue (2010-01-14 20:06:31)

shining · 2010-01-18 13:57:28

Please open two bug reports, one per issue (rotate cursor one + hwcursor false one) : http://nouveau.freedesktop.org/wiki/FrontPage#Bugs

goodgrue · 2010-01-18 18:29:24

Bug 26102 -- Rotating one monitor of a dual-head setup causes cursor problems https://bugs.freedesktop.org/show_bug.cgi?id=26102 ***
Bug 26108 -- xrandr rotate causes a segfault with HWCursor set to false. https://bugs.freedesktop.org/show_bug.cgi?id=26108

edit: both bug reports posted now.
*** edit: This issue appears to be resolved in the most recent git code.

Last edited by goodgrue (2010-02-11 01:49:48)

Arch Linux

#1 2010-01-05 15:41:21

xrandr rotate affecting mouse cursor on all monitors.

#2 2010-01-09 11:34:54

Re: xrandr rotate affecting mouse cursor on all monitors.

#3 2010-01-13 19:34:57

Re: xrandr rotate affecting mouse cursor on all monitors.

#4 2010-01-13 19:55:01

Re: xrandr rotate affecting mouse cursor on all monitors.

#5 2010-01-13 20:16:55

Re: xrandr rotate affecting mouse cursor on all monitors.

#6 2010-01-18 13:57:28

Re: xrandr rotate affecting mouse cursor on all monitors.

#7 2010-01-18 18:29:24

Re: xrandr rotate affecting mouse cursor on all monitors.

Board footer