Xorg-7.4 evdev (keybaord and mouse config)

[DJ Lucas]

Need to add information on using evdev driver. Right now, it works with legacy hardware by honoring the legacy entries in xorg.conf. When this ticket is addressed, the workaround should be removed from the xorg.conf file (not from the book as we will not be able to test xorg otherwise). My previous suggestion was to address one common mouse (ImPS/2 with side scroll) and one common ps2 keyboard in the book completely (preferably not a US 101/105 keyboard....unless setup US-International)...the rest can go to the wiki.

[dnicholson@…]

The nice thing about evdev is that it will usually do the right thing on its own without any additional setup. So, that makes the example configuration really simple. BTW, if you're using xf86-input-evdev-2.x and you find that your mouse buttons aren't mapped correctly, this is a bug in the kernel.

However, if you've enabled input hotplugging with HAL (--enable-config-hal), then having xorg.conf entries for input devices can actually be harmful. Peter Hutterer has a great blog entry on how this works:

​http://who-t.blogspot.com/2008/07/input-configuration-in-nutshell.html

[DJ Lucas]

So basically, get rid of mouse and keyboard drivers altogether, then drop the workaround. It just seems really sloppy to me that the wrong driver is referenced in the automatically created xorg.conf.new (even if it doesn't exist) and then it is ignored when going over the config.

[DJ Lucas]

Actually, this does not work either on my old box. Though replacing mouse and kbd with evdev pointing at the correct devices in the default xorg.conf.new does solve the no input devices issue without the workaround. Now just need a way identify the correct devices in the book. I have three old boxes to test with, two P4 1.8s and one AthlonXP 1800+. On all, the keyboard device is /dev/input/by-path/platform-i8042-serio-0-event-kbd and the mouse device is /dev/input/by-path/platform-i8042-serio-1-event-mouse. I'm not sure yet how the by-id paths were generated in the article that Dan referenced....have to dig into it a bit more. In the mean time, are these devices expected to be consistent across different hardware, MPD excluded?

[ken@…]

On this box with a KVM (which probably simulates a usb connection for keyboard and mouse) and no hal in sight, I had to add

Section "ServerFlags"
        Option  "AllowEmptyInput" "false"
EndSection

to xorg.conf with xorg-server-1.5.3, otherwise I have no keyboard and no mouse although MagicSysRQ allowed me to umount and reboot.

[Randy McMurchy]

Modified milestone from 6.4 to 6.5

[DJ Lucas]

This issue has simply been superseded by the newer xorg packages. Current xorg with evdev does the right thing with everything I've thrown at it thus far without even bothering to create an xorg.conf file. Even my multimedia keyboard is mapped to the correct shortcuts for Gnome without manual configuration.

dj [ ~ ]$ ls /etc/X11/
app-defaults  mwm  twm  xinit
dj [ ~ ]$ 

Closing invalid.

[(none)]

Milestone 6.5 deleted