Red Hat Bugzilla – Bug 585728
keyboard/mouse input not accepted in graphical anaconda
Last modified: 2010-04-29 12:50:45 EDT
Description of problem:
When Anaconda switches to graphical mode the keyboard stops working and the mouse pointer won't move either.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Boot into the installer (I used preupgrade, booting from /boot/upgrade but removed the kickstart argument)
2. Choose language and keyboard (English, sv-latin1)
Keyboard is PS/2, mouse is USB.
Is this "bug 553780 - keyboard/mouse input not accepted in stage2" again?
(In reply to comment #0)
> Description of problem:
> When Anaconda switches to graphical mode the keyboard stops working and the
> mouse pointer won't move either.
> Version-Release number of selected component (if applicable):
> anaconda 13.38
> How reproducible:
> Steps to Reproduce:
> 1. Boot into the installer (I used preupgrade, booting from /boot/upgrade but
> removed the kickstart argument)
> 2. Choose language and keyboard (English, sv-latin1)
> Additional info:
> Keyboard is PS/2, mouse is USB.
> Is this "bug 553780 - keyboard/mouse input not accepted in stage2" again?
Could be, but then I would expect a lot more similar reports, I think this is xorg / the kernel locking up. Moving over to xorg.
Maybe, but not completely. I forgot to mention: With the right kickstart config it shows a text entry box with a still blinking cursor, so it's still alive.
I'll try to figure out a way to debug this.
I have the same problem. It's not a crash. Here's how I debugged it:
1. Boot with nokill. Makes life a *lot* easier.
2. Wait for the install to "hang." This is 100% reliable.
OK, now we're in business. dmesg shows no errors at all.
Looking at /tmp/X.log, I see lots of "(EE) No input driver/identifier specified (ignoring)"
I think that there are two problems with the preupgrade image:
1. /etc/X11/xorg.conf.d/10-evdev.conf is missing. That file is importable.
2. Note the above path. /etc/xorg.conf.d should be /etc/X11/xorg.conf.d
Fixing both of those from the console causes Xorg to work (although it's too late for anaconda at that point).
Created attachment 410124 [details]
/tmp/X.log from a failed anaconda run
This is X.log pulled after forcibly switching the VT.
Thanks for the ping (by mail), you're right this is not an X-server issue but a bad interaction between the X server and anaconda, which has been fixed now.
*** This bug has been marked as a duplicate of bug 585621 ***