Just did a zypper up and now have one new problem and one problem that's been around awhile that I was hoping a fix would show up for.
During boot, I get the following errors:
getpt: could not getpty for boot.loadmodules
boot.devicemapper
boot.localnet
boot.md
then after I login, when I attempt to launch an xterm, it fails with a could not get pty error.
the long outstanding error is that kdeinit4 is seg faulting during login.
Any one have any ideas of what’s wrong?
Thanks
So I did have a couple of problems with openSUSE 11.4 M5, but not the ones you list above. I am writing this message to you right now using openSUSE 11.4 M5. I did not use the zypper up command but rather I did a clean install from a DVD. If you can’t get it to work enough to test it out as you have loaded it now, I then suggest you reinstall from a DVD or forget about it until M6 comes out and try again.
Please htumblin, when posting about non released versions, do this in Prereleased/Beta. Then people who are not testing these releases are not confused reading problems which are of no connection to their stable release.
I would go for Factory, which seems quite stable just now and will be very close to 11.4-ms6 which is due this week. This is especially true for KDE, which is presently 4.6-RC2 in Factory, whereas the 4.6-beta2 in MS5 was problematic. Also you would get the final kernel-2.6.37 rather than 2.6.37-rc5.
Do you have the correct versions of packages “sysvinit” and “sysvinit-tools” installed?
Did you dup to Factory as suggested? You have given no information about what you are actually doing, starting position, repositories enabled, the exact commands you are using, etc.
I been upgrading as new milestones become available, so was fine with milestone 4(but had problem with kdeinit crashing)
when I upgraded to M5, I started to see the getpty errors.
The cleanest solution is to do a clean install rather than upgrading. It’s the only way that you can be realy confident about testing and submitting bug reports. Something has gone wrong somewhere with the installation/configuration of udev or mtab.
/etc/mtab could be a text file, but is better as a symbolic link to /proc/self/mounts
You could also try, from run level 3, trashing /etc/udev and /lib/udev/ then reinstalling udev (zypper install --force udev). But only when you are prepared to make a new install, as I have never tried this.
seems a bit odd that this is happening so early during startup. Must be a missing link and/or wrong config for the ptys. Looks like the filesystem gets mounted early on, but cannot allocate ptys. Could it be permissions or a missing directory in /dev/pts? Any other ideas?
seems a bit odd that this is happening so early during startup
The console(terminal) has to be available at the front otherwise you could not monitor/interact with the boot process.
Is this related to /dev/pts ? If so, it is mounted but empty.
Yes. If it is empty it is either not mounted or udev has not created the device files.
/etc/mtab should show what is actualy mounted. But if it is a file (because it was inherited from a pre- kernel-2.6 system or has been copied during an update etc.) then it could be wrong/corrupted. If it is a normal file remove/rename it and
There should be a line like:
devpts /dev/pts devpts mode=0620,gid=5 0 0
as per /etc/fstab.
It should be the third thing mounted, after the root and temporary filesystems.
/dev is empty on a udev (kernel-2.6) system until the system starts, when it is dynamically populated by udwv. Hence my suggestion to remove /rename the udev configuration directories
/etc/udev/ and /lib/udev/
and then reinstalling udev.