segmentation violation

[BillT]

Getting segmentation violations all over the place, to the point where Linux version is pretty much unusable.

(Been using Mac version happily for quite a while now, but now trying to use BC to help organize my Linux machines).

I'm running CentOS 6.5, and have installed bcompare 4.0.5 (build 19480) using rpm.

Attaching stack trace from gdb, along with zip file.

Any help would be appreciated -- thanks!

[BillT]

P.S. Typically I install Linux w/o 32-bit support -- all my development is 64-bit only, and I prefer not to have to deal with potentially picking up wrong libs etc.

Since I installed from rpm, I'm assuming that all dependencies were installed correctly, but perhaps not.

It would be nice to have a pure 64-bit version on Linux that did not require 32-bit support libs.

[BillT]

Here's some more info -- edited script to echo output and get the following:

+ /bin/bash -c 'exec -a /usr/bin/bcompare /usr/lib/beyondcompare/BCompare ' /usr/bin/bcompare
/usr/bin/bcompare: line 45: 19695 Segmentation fault (core dumped) /bin/bash -c "exec -a $0 $EXEC $ARGS" $0

########################################
# set exit code / wait on existing instance
EXIT=$?
+ EXIT=139
if [ $EXIT = 99 ]; then
# detected existing instance of Beyond Compare
EXIT=`cat /tmp/BC4_.$$`
rm -f /tmp/BC4_.$$
fi
+ '[' 139 = 99 ']'
exit $EXIT
+ exit 139
/home/btorpey $

[Aaron]

Hello,

Thanks for the report. We haven't seen an issue like this crop up, yet. Normally, if there are missing libraries, BC4 fails to boot entirely until we are able to determine which are missing. The .rpm should install them, but there may be something different about your current configuration vs. an assumption we've made. The rpm is the method we recommend for systems like Centos that support it.

Would it be possible for you to email into [email protected] with a link back to this forum thread for our reference, and also the export of your installed libraries using /sbin/ldconfig -p ?

[BillT]

just sent info you requested....

x

[BillT]

Here's a clue: I'm *pretty* sure at this point that the crash is triggered by doing a move (not copy) of a symlink to a directory on a different filesystem.

For now, I'm doing a copy followed by delete, and that seems to work fine...

HTH ...

[Aaron]

Hello BillT,

Thanks for the follow-up. I searched through our email and we had a few questions for you. I just resent that email.