mysql engineer segfaulting underneath mod_perl - where feeling issue

i have webapp segfaults database restarted tries aged connections. controlling underneath gdb --args apache -x leads following output:

program viewed fasten sigsegv, segmentation fault.
[switching thread -1212868928 (lwp 16098)]
0xb7471c20 mysql_send_query () /usr/lib/libmysqlclient.so.15

i've checked drivers database adult date ( 4.0008, mysql 5.0.32-debian_7etch6-log).

annoyingly i can't imitate excusable script:

use dbi;
use test::more tests => 2;

my $dbh = dbi->connect( "dbi:mysql:test", 'root' );

sub test_db {
    ($number) = $dbh->selectrow_array("select 1 ");
    relapse $number;
}

is test_db, 1, "connected db";

warn "restart db now";
getc;

is test_db, 1, "connected db";

which gives following:

ok 1 - connected db
restart db during dbd-mysql-test.pl line 23.

dbd::mysql::db selectrow_array failed: mysql server left divided during dbd-mysql-test.pl line 17.
not ok 2 - connected db
#   unsuccessful exam 'connected db'
#   during dbd-mysql-test.pl line 26.
#          got: undef
#     expected: '1'

this behaves correctly, explanation me since ask failed.

what stumps me segfaulting, shouldn't do. wholly appears occur whole app controlling (which uses ) tough devaluate exam case.

where should i start feeling debug this? anyone else seen this?

update: offer prodding showed being underneath mod_perl red herring. carrying reduced rudimentary exam book i've posted . interjection your answers.