hi there,
is there any other possible reason for segmentation faults
with squid 2.3.stable5 than the known ones (gcc -O2) ?
i have compiled it using gcc 2.95.3 without the '-O' flags, but
squid seems to insist on quick segfault nearly each time i
use 'squid -k reconfigure'.
+++
the machine is a p3 866 (dual board, 1 cpu)
suse 7.2
kernel 2.4.13 smp
1 gb ram
2x 18gb uscsi3 disks
squid 2.3.stable5 (built from daily tarball 2001/10/29)
my compile options:
--enable-snmp \
--enable-cache-digests \
--disable-ident-lookups \
--enable-gnuregex \
--enable-time-hack \
--enable-dlmalloc
i have also tried building squid without dlmalloc and time-hack
but it wouldn't spare me from the segfaults.
i have also tried downloading a newer version of 'dlmalloc.c'
(v2.7.0) and putting it into the 'lib' directory of squid's source tree.
i have also placed the corresponding 'malloc.h' in the 'include'
directory - it compiles fine and everything works ok - except from
the sporadic segfaults.
the error msg in cache.log is:
FATAL: Received Segment Violation...dying.
2001/10/30 12:09:58| storeDirWriteCleanLogs: Starting...
2001/10/30 12:09:59| WARNING: Closing open FD 9
2001/10/30 12:09:59| Finished. Wrote 65464 entries.
2001/10/30 12:09:59| Took 0.3 seconds (192720.3 entries/sec).
CPU Usage: 3901.260 seconds = 1460.060 user + 2441.200 sys
Maximum Resident Size: 0 KB
Page faults with physical i/o: 362
Memory usage for squid via mallinfo():
total space in arena: 180415 KB
Ordinary blocks: 166703 KB 5923 blks
Small blocks: 0 KB 0 blks
Holding blocks: 824 KB 2 blks
Free Small blocks: 0 KB
Free Ordinary blocks: 13711 KB
Total in use: 167527 KB 93%
Total free: 13711 KB 8%
2001/10/30 12:10:10| Starting Squid Cache version 2.3.STABLE5 for i686-pc-linux
any ideas?
ps: perhaps s.o. can help me out doing a debug trace, i never
did that before.
Received on Tue Oct 30 2001 - 03:32:07 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:03:12 MST