user release.debian....@packages.debian.org usertag 587948 squeeze-can-defer tag 587948 squeeze-ignore kthxbye
On Thu, Jan 6, 2011 at 02:01:37 +0100, Mirco Bauer wrote: > Ok, I can finally reproduce and confirm this issue if shared memory > (/dev/shm) is exhausted. The mono runtime needs to deal with it though. > (I had to create a chroot btw, as the already mounted /dev/shm can not be > changed as long as it's used by some process) > > redbull:/# mount -t tmpfs -o size=1 tmpfs /dev/shm > redbull:/# mount > proc on /proc type proc (rw) > tmpfs on /dev/shm type tmpfs (rw,size=1) > redbull:/# gacutil > > ** (/usr/lib/mono/2.0/gacutil.exe:6145): WARNING **: Thread (nil) may have > been prematurely finalized > > ** (/usr/lib/mono/2.0/gacutil.exe:6145): WARNING **: Thread (nil) may have > been prematurely finalized > > ** (/usr/lib/mono/2.0/gacutil.exe:6145): WARNING **: Thread (nil) may have > been prematurely finalized > > ** (/usr/lib/mono/2.0/gacutil.exe:6145): WARNING **: Thread (nil) may have > been prematurely finalized > Segmentation fault > redbull:/# > > > As temporarily workaround MONO_DISABLE_SHM=1 as environment variable can be > used: > redbull:/# MONO_DISABLE_SHM=1 gacutil > The following assemblies are installed into the GAC: > redbull:/# > Considering that 1) there is a workaround 2) this seems hard to fix reliably, according to https://bugzilla.novell.com/show_bug.cgi?id=662646#c1 I'm tagging this bug as not a blocker for the squeeze release. Cheers, Julien
signature.asc
Description: Digital signature
