MySQL (All 4.x), Segmentation Faults, and ... (Fixed!)

[Xystance]

Hello all, this is my first post to the forums as everything else I've been able to solve through bugs.gentoo.org or via #gentoo or #gentoo-amd64 on freenode.

But I'm out of ideas.

MySQL 4.0.21 (Tried 4.0.20-r1, 4.0.18 ebuilds as well) builds successfully, but when running the ebuild config, manually running mysql_install_db or even if just attempting to start mysqld, I get a segmentation fault.

The mysql.err log file gives nothing other than "mysqld started, mysqld stopped".

However, in /var/log/everything/current, I do get the following:
"Oct 18 16:41:32 [kernel] Unable to handle kernel paging request at virtual address 00016b54"

I get the same message at the same address everytime, whether it's different versions of MySQL or if I have other servers running.


Summary:

I -am- using the 2004.3 profile with GCC 3.4.2.

I am on an Athlon XP 2600+, 256MB DDR333, multiple hard drives. Root is reiser3.x, all other drives (which are just under /mnt and used for storage) are reiser4. I run an ssh server, a samba server, an apache2 server, and use netfilter as a firewall which is also allowing me to run stateful firewalling, network address translation (it is the firewall/gateway to a few computers), dhcp server, and dnsmasq. Webmin is also installed.

I did stop sshd, samba, webmin, and apache2 just in case something was misbehaving. No other daemons complain and I get great performance out of this machine otherwise.



System information so you all know:

emerge --info, with extraneous information pulled:

Portage 2.0.51_rc9 (default-linux/x86/2004.3, gcc-3.4.2, glibc-2.3.4.20041006-r0, 2.6.9-rc4-mm1 i686)
=================================================================
System uname: 2.6.9-rc4-mm1 i686 AMD Athlon(TM) XP 2600+
Gentoo Base System version 1.5.3
Autoconf: sys-devel/autoconf-2.59-r5
Automake: sys-devel/automake-1.8.5-r1
Binutils: sys-devel/binutils-2.15.92.0.2-r1
Headers: sys-kernel/linux26-headers-2.6.8.1-r1
Libtools: sys-devel/libtool-1.5.2-r5
ACCEPT_KEYWORDS="x86 ~x86"
AUTOCLEAN="yes"
CFLAGS="-march=athlon-xp -O2 -pipe -fomit-frame-pointer"
CHOST="i686-pc-linux-gnu"
COMPILER=""
CXXFLAGS="-march=athlon-xp -O2 -pipe -fomit-frame-pointer"
DISTDIR="/usr/portage/distfiles"
FEATURES="autoaddcvs ccache distlocks sandbox sfperms"
MAKEOPTS="-j2"
USE="acpi apache2 apm arts avi berkdb bitmap-fonts crypt cups curses dba encode f77 foomaticdb gd gdbm gif gpm gtk2 imlib ithreads jpeg jpg kde libg++ libwww mad mikmod motif mpeg mysql ncurses nls nptl oggvorbis opengl oss pam pdflib perl php png python qt quicktime readline samba sdl slang spell ssl svga tcpd threads tiff truetype x86 xml2 xmms xprint xv zlib"

Any ideas? need any more information?

[r00tzz]

Try to use another kernel and sse if it works.

If it works, try reversing the optimize-profile-path-slightly of the mm kernel.

I used to have segfaults in gecko browsers 'till I reversed the patch.

You can find it here:

http://kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.9-rc4/2.6.9-rc4-mm1/broken-out/optimize-profile-path-slightly.patch

HTH

[Xystance]

I'll give a couple of other kernels a try, and I'll use that patch if they don't work. Thanks much!

[pludodog]

Still getting the same error with the latest gentoo-dev-sources... anyone have any ideas?

[Xystance]

Works : 2.6.9-mm1
Works : 2.6.9-nitro1
Broken: 2.6.9-rc4-mm1

But I can deal with that.

Thanks for your help. (Just a note: that patch you asked me to reverse didn't make a difference unfortunately... but fortunately 2.6.9 became stable)