interesting, run
ulimit -c number_of_blocks
to set core dump limit size, by default it is set to 0 so nothing is dumpped, just remember that 1 block is 4k by default for ext2/3 filesystems, and 2k for reiserFS, so calculate how many blocks you are willing to allocate for core dumps, then when core is created use strings or a hex editor to get an idea what could be causing the rpm to segfault. You know there is nothing we can help you out if the message reads segmentation fault - it could be anything from a corrupted rpm to filesystem corruption
|