Bug#1038644: nfdump: segfault if started with -R option
> Could you please try the just uploaded 1.7.1-3 to verify the fix for > both bugs? Verified: in 1.7.1-3 no segfault and no false "failed!" message from /etc/init.d/nfdump start. Thank you! On Wed, Jun 21, 2023 at 09:42:38PM +0200, Bernhard Schmidt wrote: > On 19/06/23 05:25 PM, Yury Shevchuk wrote: > > > # /usr/bin/nfcapd -D -P /var/run/nfcapd/default.pid -w /var/cache/nfdump > > -S1 -b 120.0.1 -p 2055 -R 127.0.0.2 2055 > > Segmentation fault > > The patch (trivial) is attached. > > Thanks. For the record, this is included in the much larger > > https://github.com/phaag/nfdump/commit/abfab42419117add44e1ea15ad9559d265642219#diff-c95665baa1999e70e29344d1dc05f3282cd1cf7f31b47341581cd1cf81b7d062R593 > > in v1.7.2 > > > A minor change in /etc/init.d/nfdump conffile (added return 0) fixes false > > "failed!" message from "/etc/init.d/nfdump start" which appears on systems > > using sysvinit-core rather than systemd. > > I really don't get what this code is supposed to do though. And I don't > want to invest much time into sysvinit. From my understanding > > start-stop-daemon --start --quiet \ > --pidfile "$PIDFILE" --exec "$NFCAPD" --test > /dev/null \ > || return 1 > > First we run with --test. If start-stop-daemon returns zero (process not > already running) we continue, else we return 1. So far so good. > > start-stop-daemon --start --quiet \ > --pidfile "$PIDFILE" \ > --exec "$NFCAPD" -- \ > -D -P "$PIDFILE" \ > $options \ > || return 2 > > Now we really start it. If we can do it we continue, if we can't we > return 2 (could not be started) > > sleep 1 > start-stop-daemon --start --quiet \ > --pidfile "$PIDFILE" --exec "$NFCAPD" --test > /dev/null \ > && return 2 > > Now we basically test again if the daemon is already running. If it > isn't, we return 2, > > At this point we have checked that 1 second after the start the process > is still running, and can return 0. > > Could you please try the just uploaded 1.7.1-3 to verify the fix for > both bugs? > > Bernhard
Bug#1038644: nfdump: segfault if started with -R option
On 19/06/23 05:25 PM, Yury Shevchuk wrote: > # /usr/bin/nfcapd -D -P /var/run/nfcapd/default.pid -w /var/cache/nfdump -S1 > -b 120.0.1 -p 2055 -R 127.0.0.2 2055 > Segmentation fault > The patch (trivial) is attached. Thanks. For the record, this is included in the much larger https://github.com/phaag/nfdump/commit/abfab42419117add44e1ea15ad9559d265642219#diff-c95665baa1999e70e29344d1dc05f3282cd1cf7f31b47341581cd1cf81b7d062R593 in v1.7.2 > A minor change in /etc/init.d/nfdump conffile (added return 0) fixes false > "failed!" message from "/etc/init.d/nfdump start" which appears on systems > using sysvinit-core rather than systemd. I really don't get what this code is supposed to do though. And I don't want to invest much time into sysvinit. From my understanding start-stop-daemon --start --quiet \ --pidfile "$PIDFILE" --exec "$NFCAPD" --test > /dev/null \ || return 1 First we run with --test. If start-stop-daemon returns zero (process not already running) we continue, else we return 1. So far so good. start-stop-daemon --start --quiet \ --pidfile "$PIDFILE" \ --exec "$NFCAPD" -- \ -D -P "$PIDFILE" \ $options \ || return 2 Now we really start it. If we can do it we continue, if we can't we return 2 (could not be started) sleep 1 start-stop-daemon --start --quiet \ --pidfile "$PIDFILE" --exec "$NFCAPD" --test > /dev/null \ && return 2 Now we basically test again if the daemon is already running. If it isn't, we return 2, At this point we have checked that 1 second after the start the process is still running, and can return 0. Could you please try the just uploaded 1.7.1-3 to verify the fix for both bugs? Bernhard
Bug#1038644: nfdump: segfault if started with -R option
Package: nfdump
Version: 1.7.1-2
Severity: normal
Tags: patch
X-Debbugs-Cc: si...@botik.ru
Dear Maintainer,
-R option is broken in nfdump-1.7.1-2. It used to work in infdump-1.6.22-2.
How to reproduce:
# /usr/bin/nfcapd -D -P /var/run/nfcapd/default.pid -w /var/cache/nfdump -S1 -b
120.0.1 -p 2055 -R 127.0.0.2 2055
Segmentation fault
The patch (trivial) is attached.
A minor change in /etc/init.d/nfdump conffile (added return 0) fixes false
"failed!" message from "/etc/init.d/nfdump start" which appears on systems
using sysvinit-core rather than systemd.
Thank you!
-- System Information:
Debian Release: 12.0
APT prefers stable-security
APT policy: (500, 'stable-security'), (500, 'stable')
Architecture: amd64 (x86_64)
Kernel: Linux 6.1.0-9-amd64 (SMP w/4 CPU threads; PREEMPT)
Kernel taint flags: TAINT_OOT_MODULE, TAINT_UNSIGNED_MODULE
Locale: LANG=C.UTF-8, LC_CTYPE=ru_RU.UTF-8 (charmap=UTF-8), LANGUAGE not set
Shell: /bin/sh linked to /usr/bin/dash
Init: sysvinit (via /sbin/init)
LSM: AppArmor: enabled
Versions of packages nfdump depends on:
ii init-system-helpers1.65.2
ii libbz2-1.0 1.0.8-5+b1
ii libc6 2.36-9
ii libpcap0.8 1.10.3-1
ii librrd81.7.2-4+b8
ii lsb-base 11.6
ii sysvinit-utils [lsb-base] 3.06-4
nfdump recommends no packages.
nfdump suggests no packages.
-- Configuration Files:
/etc/init.d/nfdump changed:
PATH=/sbin:/usr/sbin:/bin:/usr/bin
NAME=nfcapd
DESC='Netflow capture daemon'
NFCAPD='/usr/bin/nfcapd'
PIDDIR="/var/run/$NAME/"
[ -x "$NFCAPD" ] || exit 0
mkdir -p "$PIDDIR"
. /lib/init/vars.sh
. /lib/lsb/init-functions
do_start () {
local INSTANCE="$1"
local CONFIG="$2"
sh -n "$CONFIG" 2>/dev/null || return 2
cache_dir=
user=
group=
options=
. "$CONFIG"
[ "$options" ] || return 2
if [ "$cache_dir" ] ; then
mkdir -p "$cache_dir"
if [ "$user" ] && [ "$group" ] ; then
chown "$user:$group" "$cache_dir"
elif [ "$user" ] ; then
chown "$user" "$cache_dir"
fi
fi
local PIDFILE="$PIDDIR$INSTANCE.pid"
start-stop-daemon --start --quiet \
--pidfile "$PIDFILE" --exec "$NFCAPD" --test > /dev/null \
|| return 1
start-stop-daemon --start --quiet \
--pidfile "$PIDFILE" \
--exec "$NFCAPD" -- \
-D -P "$PIDFILE" \
$options \
|| return 2
sleep 1
start-stop-daemon --start --quiet \
--pidfile "$PIDFILE" --exec "$NFCAPD" --test > /dev/null \
&& return 2
return 0 # 20230617, sizif: shut up "failed!" complaint
}
do_stop () {
local INSTANCE="$1"
local PIDFILE="$PIDDIR/$INSTANCE.pid"
start-stop-daemon --stop --quiet \
--retry=TERM/30/KILL/5 --pidfile "$PIDFILE" --name "$NAME"
RETVAL="$?"
[ "$RETVAL" = 2 ] && return 2
# Many daemons don't delete their pidfiles when they exit.
rm -f "$PIDFILE"
return "$RETVAL"
}
EXIT=0
do_action () {
local CONFIG="$1"
INSTANCE="$(basename "${CONFIG%%.conf}")"
case "$ACTION" in
start)
[ "$VERBOSE" != no ] && log_daemon_msg "Starting $DESC" "$INSTANCE"
do_start "$INSTANCE" "$CONFIG"
case "$?" in
0|1)[ "$VERBOSE" != no ] && log_end_msg 0 ;;
2)
[ "$VERBOSE" != no ] && log_end_msg 1
EXIT=1
;;
esac
;;
stop)
[ "$VERBOSE" != no ] && log_daemon_msg "Stopping $DESC" "$INSTANCE"
do_stop "$INSTANCE"
case "$?" in
0|1)[ "$VERBOSE" != no ] && log_end_msg 0 ;;
2)
[ "$VERBOSE" != no ] && log_end_msg 1
EXIT=1
;;
esac
;;
status)
status_of_proc -p "$PIDDIR/$INSTANCE.pid" "$NFCAPD" "nfcapd
$INSTANCE instance" || EXIT=$?
;;
restart|force-reload)
log_daemon_msg "Restarting $DESC" "$INSTANCE"
do_stop "$INSTANCE"
case "$?" in
0|1)
do_start "$INSTANCE" "$CONFIG"
case "$?" in
0) log_end_msg 0 ;;
*)
# Old process is still running or
# failed to start
log_end_msg 1 ;;
esac
;;
*)
# Failed to stop
log_end_msg 1
;;
esac
;;
*)
echo "Usage: /etc/init.d/nfdump
{start|stop|status|restart|force-reload} [ ...]" >&2
exit 3
;;
esac
}
ACTION="$1"
shift
if [ "$1" ] ; then
while [ "$1" ] ; do
CONFIG="/etc/nfdump/$1.conf"
if [ -f "$CONFIG" ] ; then
do_action
