Downloaded the 3-CD isos, burned them with k3b, and I was ready to go.
To my experience, the installer hasn't been improved any since beta 1. It was still incredibly slow. Froze for a few minutes trying to determine the package list. And the estimated time for upgrading was 4 hours! While the packages were installed, I needed to switch the CDs dozens of times - guess things would be easier if I had a DVD burner, but really not everyone has one.
Finally, after about 1 hour and a half, installation finished. I was taken to the configuration screen. Having installed Linux many times, it didn't take long for me to finish setting everything up and ready for the first boot of the new system.
On the first boot, I wasn't surprised to see the lilo bootloader in text mode instead of the nice graphical one. It has been removed since 2007.1, and there's no chance for it to come back in 2008.0, I guess :P
I was then taken to the login screen. Well, it was quite nice with a new fancy theme. I logged into KDE, my favourite desktop environment. And... things go black and I was taken back to the login screen. Wtf???
Good thing I have fluxbox installed as an alternative desktop. Logged in without any problem, I started wandering the wilderness to find the solution to my problems. Found the following lines in my kdm.log:
Backtrace:While roaming around finding the solution, I noticed that X crashed even when I start other gtk-based applications like rpmdrake, harddrake, GIMP, ekiga. Totally lost, I posted my problem at http://forum.mandriva.com, hoping that anyone there could provide any idea that could help me solve the problem.
0: /etc/X11/X(xf86SigHandler+0x85) [0x80c46e5]
1: [0xffffe420]
2: /usr/lib/xorg/modules//libexa.so(ExaOffscreenMarkUsed+0x69) [0xb7aa70b9]
3: /usr/lib/xorg/modules//libexa.so(exaDoMigration+0x597) [0xb7aa6ff7]
4: /usr/lib/xorg/modules//libexa.so [0xb7aa46d6]
5: /etc/X11/X(miFillGeneralPoly+0x4f5) [0x812ad85]
6: /etc/X11/X(miFillPolygon+0x62) [0x812a322]
7: /etc/X11/X [0x8174f82]
8: /etc/X11/X(ProcFillPoly+0x10a) [0x808666a]
9: /etc/X11/X [0x81532bd]
10: /etc/X11/X(Dispatch+0x1af) [0x8089d6f]
11: /etc/X11/X(main+0x465) [0x8071025]
12: /lib/i686/libc.so.6(__libc_start_main+0xe0) [0xb7c9bf90]
13: /etc/X11/X(FontFileCompleteXLFD+0x1e5) [0x80703a1]
Fatal server error:
Caught signal 11. Server aborting
Life without everyday applications was a real pain. While trying to fix the problem, I ran into another nasty bug with rpm:
error: Runnning db->verify ...This nasty bug is being tracked down by pixel at http://qa.mandriva.com/show_bug.cgi?id=32547. If you stumble upon it, be sure to report the bug there and help us identify the problem. In the meantime, you can use the following commands to fix the rpm database temporarily:
rpmdb: PANIC: fatal region error detected; run recovery
error: db4 error(-30975) from dbenv->open: DB_RUNRECOVERY: Fatal error, run database recovery
rm -f /var/lib/rpm/__db*Have been messing with my penguin box for so long, I didn't have any trouble getting into sleep that night, and luckily without any nightmares :P
rpm --rebuilddb
Waking up the next morning, I decided to jump into the official mandriva irc channel at irc://irc.freenode.net/#mandriva and had a nice chat with the guys there. Finally, after gathering all the relevant ideas, I decided to replace the xorg.conf file with the old one I backed up months ago. Fortunately, this completely solved the problem!
So it turned out that x have been misconfigured somehow. I wonder if it was me who picked the wrong video card/monitor/keyboard/mouse..., or the hardware detection wasn't good enough - never had this problem before. Anyway, I had enough problems with x that I wouldn't want to compare the 2 config files to find out which line was causing the all the troubles.
Time to give the damn penguin a real torture now :P
No comments:
Post a Comment