You are not logged in.
Pages: 1
# pacman -Syu
:: Synchronizing package databases...
:: current is up to date
extra [################] 100% 196K 24.6K/s 00:00:07
Segmentation fault
doesn't give any other information but the segfault so i don't even know where to start.
Offline
Start by running memtest86.
Offline
uhmm...
i'm not quite getting why i would want to run memtest.
i've been running arch for quite some time now.
i've been dual booting even longer.
believe me i don't have bad ram.
it is some other problem with pacman and maybe some conflict with the newest pacman i downloaded after a system upgrade and some other package maybe...
Offline
Ram can go bad it doesn't have to start that way. Segmentation faults are when either an application trys to access memory it does not have access or trys to do thing it can't with that memory. Poorly written applications (pacman is not) and bad ram are the primary caues of segfaults. Like you said yourself you've been running arch for quite some time which leds to beleive there is not a problem with pacman but with your ram going bad. Run memtest86.
Offline
You could run strace on pacman to see what calls are being made just before the segfault. Might indicate something.
Dusty
Offline
Hmm.. got a segfault too today, after upgrading loads of packages (KDE, kernel, nvidia & others)...
... and I've got 999 months guarantee for my RAM.
I got the segfault exactly after upgrading the last package in the list (which was wine), most likely on exit pacman tried to free some pointers. I believe it installed fine. Running pacman afterwards didn't create a segfault.
I usually run programs that segfault with debug symbols in valgrind to get more information. Unfortunately, I can't find this segfault reproduceable.
:: / my web presence
Offline
The same segfault happened to me. I had to restart pacman and the second time it went through.
Offline
Hmm, I recall a similar event happening a few releases back. Nobody's had any trouble in recent weeks, however. The question is, Is this the same problem or a new one?
If no bug report exists, one should be filed. I haven't had any trouble with it.
The work around, as suggested, is to run pacman a second time to complete the update.
Dusty
Offline
I just had a few more segfaults with pacman tonight. It usually happens when scrollkeeper updates its index.
Offline
Definately file a bug report then. If its happening more than once a day its a bug. :-D
Offline
yup no matter how many times i run pacman -Syu i keep gettin a segfault.
since my last upgrade jus seems to be permanently broke for me.
Offline
you can try # pacman.static -Syu
arch + gentoo + initng + python = enlisy
Offline
intrinsic: you could install valgrind (not using pacman) and run
valgrind pacman <stuff>
it is a very good debugger for such problems. It will show at which position pacman (if it does) happen to violate something (and how). This often shows problems that occur some steps _before_ the segfault, which is indirectly caused.
In your situation I would use the fact that you can reproduce it to find out the reason.
I had some segfaults my self which all were unreproducable
cu
Ford Prefect
Offline
To help troubleshooting such crashes, we need:
- pacman version
- the core file resulting of the crash
- a tarball of the directory /var/lib/pacman
- a copy of /etc/pacman.conf
To get a useful core dump, pacman needs to be compiled in debug mode:
- rebuild pacman package with the --nostrip option
- run "ulimit -c unlimited" to allow the core file creation
- run pacman to reproduce the crash (better with pacman.static)
- put the expected files somewhere where we can retrieve them, or try attaching them to a bug report (think to "gzip -9"!)
Thanks!
Offline
Pages: 1