Debian Bug report logs - #778598
atop sometimes fails with a floating point exception or a trap exception

Toggle useless messages

Message #5 received at submit@bugs.debian.org (full text, mbox, reply):

From: Klaus Ethgen <Klaus@Ethgen.de>

To: Debian Bug Tracking System <submit@bugs.debian.org>

Subject: atop sometimes fails with a floating point exception or a trap exception

Date: Tue, 17 Feb 2015 09:50:31 +0100

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

Package: atop
Version: 1.27.3-1
Severity: normal

I do not know since when it started but since some time I get more and
more floating point exceptions when starting it on the console or "trap
divide error" when rotated via cron.

The trap error from log is:
   traps: atop[28297] trap divide error ip:407b9a sp:7fff857869b8 error:0 in atop[400000+29000]

... and on the console just a "floating point exception".

The error do not happen every time and often atop TUI start with the
second or subsequent start. But it happen that often that I believe in a
more important bug.

- -- System Information:
Debian Release: 8.0
  APT prefers unstable
  APT policy: (800, 'unstable'), (110, 'experimental')
Architecture: amd64 (x86_64)
Foreign Architectures: i386

Kernel: Linux 3.18.6 (SMP w/8 CPU cores)
Locale: LANG=de_DE, LC_CTYPE=de_DE (charmap=ISO-8859-1) (ignored: LC_ALL set to de_DE)
Shell: /bin/sh linked to /bin/dash
Init: sysvinit (via /sbin/init)

Versions of packages atop depends on:
ii  libc6        2.19-15
ii  libncurses5  5.9+20140913-1+b1
ii  libtinfo5    5.9+20140913-1+b1
ii  lsb-base     4.1+Debian13+nmu1
ii  zlib1g       1:1.2.8.dfsg-2+b1

Versions of packages atop recommends:
ii  cron  3.0pl1-127

atop suggests no packages.

- -- Configuration Files:
/etc/cron.d/atop changed:
0 0 * * * root /usr/sbin/invoke-rc.d atop _cron

/etc/default/atop changed:
INTERVAL=600
LOGPATH="/var/log/atop"
OUTFILE="$LOGPATH/daily.log"
START_DAEMON="no"

/etc/init.d/atop changed:
PATH=/sbin:/usr/sbin:/bin:/usr/bin
DESC="atop system monitor"
NAME=atop
DAEMON=/usr/bin/atop
WRAPPER=/usr/share/atop/atop.wrapper
INTERVAL=600            # interval 10 minutes
LOGPATH="/var/log/atop"
OUTFILE=$LOGPATH/daily.log
PIDFILE=/var/run/$NAME.pid
SCRIPTNAME=/etc/init.d/$NAME
[ -x $DAEMON ] || exit 0
START_DAEMON="yes"
[ -r /etc/default/$NAME ] && . /etc/default/$NAME
. /lib/init/vars.sh
. /lib/lsb/init-functions
CURDAY=$(date +%Y%m%d)
DAEMON_ARGS="-a -w $LOGPATH/atop_$CURDAY $INTERVAL"
do_start()
{
	# Return
	#   0 if daemon has been started
	#   1 if daemon was already running
	#   2 if daemon could not be started
        start-stop-daemon --start --background --quiet \
		--pidfile $PIDFILE \
		--test --startas $WRAPPER > /dev/null \
		|| return 1
        start-stop-daemon --start --background --quiet \
	        --pidfile $PIDFILE --make-pidfile \
		--startas $WRAPPER -- $DAEMON $OUTFILE \
		$DAEMON_ARGS \
		|| return 2
}
do_stop()
{
	# Return
	#   0 if daemon has been stopped
	#   1 if daemon was already stopped
	#   2 if daemon could not be stopped
	#   other if a failure occurred
        start-stop-daemon --stop --quiet --retry=USR2/30/KILL/5 --pidfile $PIDFILE --name $NAME
	RETVAL="$?"
	[ "$RETVAL" = 2 ] && return 2
	rm -f $PIDFILE
	return "$RETVAL"
}
case "$1" in
   start)
      if [ "x$START_DAEMON" = "xyes" ]
      then
	 [ "$VERBOSE" != no ] && log_daemon_msg "Starting $DESC " "$NAME"
	 do_start
	 case "$?" in
	    0|1) [ "$VERBOSE" != no ] && log_end_msg 0 ;;
	    2) [ "$VERBOSE" != no ] && log_end_msg 1 ;;
	 esac
      fi
   ;;
   stop)
      [ "$VERBOSE" != no ] && log_daemon_msg "Stopping $DESC" "$NAME"
      do_stop
      case "$?" in
	 0|1) [ "$VERBOSE" != no ] && log_end_msg 0 ;;
	 2) [ "$VERBOSE" != no ] && log_end_msg 1 ;;
      esac
   ;;
   status)
       status_of_proc "$DAEMON" "$NAME" && exit 0 || exit $?
   ;;
  restart|force-reload|_cron)
  	[ "$1" = "_cron" ] && VERBOSE="no"
      [ "$VERBOSE" != no ] && log_daemon_msg "Restarting $DESC" "$NAME"
      do_stop
      case "$?" in
	 0|1)
	    do_start
	    case "$?" in
                        0) [ "$VERBOSE" != no ] && log_end_msg 0
			   [ "$1" = "_cron" ] && sleep 3 && find $LOGPATH -name 'atop_*' -mtime +28 -exec rm {} \;
	 ;;
	       1) [ "$VERBOSE" != no ] && log_end_msg 1 ;; # Old process is still running
	       *) [ "$VERBOSE" != no ] && log_end_msg 1 ;; # Failed to start
	    esac
	 ;;
	 *)
	    # Failed to stop
	    [ "$VERBOSE" != no ] && log_end_msg 1
	 ;;
      esac
   ;;
   *)
	echo "Usage: $SCRIPTNAME {start|stop|status|restart|force-reload}" >&2
	exit 1
   ;;
esac
exit 0


- -- no debconf information

- -- 
Klaus Ethgen                              http://www.ethgen.ch/
pub  4096R/4E20AF1C 2011-05-16   Klaus Ethgen <Klaus@Ethgen.de>
Fingerprint: 85D4 CA42 952C 949B 1753  62B3 79D0 B06F 4E20 AF1C
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1

iQGcBAEBCgAGBQJU4wDNAAoJEKZ8CrGAGfasPJ0L+wUNJSw8XlJdy0b3ZZe6f7+v
3Sv3Uy+zELAB2C9GdAYvz5tWdKybBgM+/YUFj5k63Jgpmpj0hhObLBGxV2Ifcele
RhHiAEpOO8raIkB33skATNruRl21NdyDj1l+YBjnpmHqWy5RVMiVbeyvPXyHt59/
udNfADzrkatRPfuLG3UiaHSI4ghc5+jwRCfsaBUYMCj0m069+JQ9TjALGmPouI9H
jwLjGM5WNUAUJ5t9fmvXH1F5GdYI6hhSX2BhUfTfzjlCWaQrNQrzPUIn0k6RZL0W
RS/N3TRyd8zdxPjUNfrvbvn9Y7NkO/48cfyLnj9YIRdlrtmE3VXt/aZse0GhWgKN
DRlujNCzalwde0jFUGn8dN2nHAFgilrYljvIvNxHz5GE/kPI7XAd/p3w6BI5GhVj
iBdlJdxVfJj0r3ux4h5HKF5S5dr7hUZFO2fvHKbns1miESWsejVQn+NiJcD6FYVe
GX9Swq5ud+/N+5+erHwsx4g5hW1vQ0jDGdqZJJYZzQ==
=Ertj
-----END PGP SIGNATURE-----

Message #10 received at 778598@bugs.debian.org (full text, mbox, reply):

From: Jakob Haufe <sur5r@sur5r.net>

To: Klaus Ethgen <Klaus@Ethgen.de>, 778598@bugs.debian.org

Subject: Re: Bug#778598: atop sometimes fails with a floating point exception or a trap exception

Date: Sun, 23 Aug 2015 13:36:52 +0200

[Message part 1 (text/plain, inline)]

Hi,

I stumbled across this as well and I rebuild atop with nostrip which gave me
this backtrace:

----SNIP----
/tmp/atop-dbg/atop-1.27.3$ gdb /usr/bin/atop /var/lib/coredumps/atop.1440280803/core
Reading symbols from /usr/bin/atop...done.
[New LWP 27858]
Core was generated by `/usr/bin/atop -a -w /var/log/atop/atop_20150823 600'.
Program terminated with signal SIGFPE, Arithmetic exception.
#0  0x0000000000407662 in acctprocnt () at acctproc.c:571
      
warning: Source file is more recent than executable.
571             return (statacc.st_size - acctsize) / acctrecsz;
gdb $ bt full
#0  0x0000000000407662 in acctprocnt () at acctproc.c:571
        statacc = {
          st_dev = 18,
          st_ino = 996486,
          st_nlink = 1,
          st_mode = 33152,
          st_uid = 0,
          st_gid = 0,
          __pad0 = 0,
          st_rdev = 0,
          st_size = 0,
          st_blksize = 4096,
          st_blocks = 0,
          st_atim = {
            tv_sec = 1440280803,
            tv_nsec = 90171947
          },
          st_mtim = {
            tv_sec = 1440280803,
            tv_nsec = 89171920
          },
          st_ctim = {
            tv_sec = 1440280803,
            tv_nsec = 89171920
          },
          __glibc_reserved = {0, 0, 0}
        }
#1  0x0000000000403242 in engine () at atop.c:829
        lastcmd = <optimized out>
        nactproc = 4285065470
        totslpi = -1613224016
        devsstat = 0x7fb47489d010
        curpact = 0x7fb474867010
        curpexit = <optimized out>
        devtstat = <optimized out>
        ntask = 280
        sigact = {
          __sigaction_handler = {
            sa_handler = 0x403890 <getalarm>,
            sa_sigaction = 0x403890 <getalarm>
          },
          sa_mask = {
            __val = {0 <repeats 16 times>}
          },
          sa_flags = 0,
          sa_restorer = 0x0
        }
        devpstat = <optimized out>
        nexit = <optimized out>
        ndeviat = <optimized out>
        totproc = 1
        totrun = 0
        totslpu = 32767
        totzombie = 1954458716
        presstat = 0x7fb474981010
        curplen = 298
        i = <optimized out>
        noverflow = <optimized out>
        hlpsstat = 0x7fb474981010
        j = <optimized out>
        timelimit = 0
        cursstat = 0x7fb47490f010
#2  main (argc=<optimized out>, argv=<optimized out>) at atop.c:659
        i = <optimized out>
        c = <optimized out>
        p = <optimized out>
        rlim = {
          rlim_cur = 18446744073709551615,
          rlim_max = 18446744073709551615
        }
gdb $ print acctrecsz
$1 = 0
----SNIP----

This is a global variable in acctproc.c. Unfortunately, I didn't have the
time to investigate this yet.

Cheers,
sur5r

-- 
ceterum censeo microsoftem esse delendam.

[Message part 2 (application/pgp-signature, inline)]

Message #15 received at 778598@bugs.debian.org (full text, mbox, reply):

From: Marc Haber <mh+debian-packages@zugschlus.de>

To: Jakob Haufe <sur5r@sur5r.net>, 778598@bugs.debian.org, 778598-submitter@bugs.debian.org

Cc: Klaus Ethgen <Klaus@Ethgen.de>, Marc Haber <mh+debian-packages@zugschlus.de>

Subject: Re: Bug#778598: atop sometimes fails with a floating point exception or a trap exception

Date: Thu, 11 Aug 2016 12:02:31 +0200

tags #778598 moreinfo
thanks

Can you please re-check with 2.2.3 from experimental?

Greetings
Marc

On Sun, Aug 23, 2015 at 01:36:52PM +0200, Jakob Haufe wrote:
> From: Jakob Haufe <sur5r@sur5r.net>
> Subject: Bug#778598: atop sometimes fails with a floating point exception
>  or a trap exception
> To: Klaus Ethgen <Klaus@Ethgen.de>, 778598@bugs.debian.org
> Reply-To: Jakob Haufe <sur5r@sur5r.net>, 778598@bugs.debian.org
> Date: Sun, 23 Aug 2015 13:36:52 +0200
> X-Mailer: Claws Mail 3.11.1 (GTK+ 2.24.28; x86_64-pc-linux-gnu)
> 
> Hi,
> 
> I stumbled across this as well and I rebuild atop with nostrip which gave me
> this backtrace:
> 
> ----SNIP----
> /tmp/atop-dbg/atop-1.27.3$ gdb /usr/bin/atop /var/lib/coredumps/atop.1440280803/core
> Reading symbols from /usr/bin/atop...done.
> [New LWP 27858]
> Core was generated by `/usr/bin/atop -a -w /var/log/atop/atop_20150823 600'.
> Program terminated with signal SIGFPE, Arithmetic exception.
> #0  0x0000000000407662 in acctprocnt () at acctproc.c:571
>       
> warning: Source file is more recent than executable.
> 571             return (statacc.st_size - acctsize) / acctrecsz;
> gdb $ bt full
> #0  0x0000000000407662 in acctprocnt () at acctproc.c:571
>         statacc = {
>           st_dev = 18,
>           st_ino = 996486,
>           st_nlink = 1,
>           st_mode = 33152,
>           st_uid = 0,
>           st_gid = 0,
>           __pad0 = 0,
>           st_rdev = 0,
>           st_size = 0,
>           st_blksize = 4096,
>           st_blocks = 0,
>           st_atim = {
>             tv_sec = 1440280803,
>             tv_nsec = 90171947
>           },
>           st_mtim = {
>             tv_sec = 1440280803,
>             tv_nsec = 89171920
>           },
>           st_ctim = {
>             tv_sec = 1440280803,
>             tv_nsec = 89171920
>           },
>           __glibc_reserved = {0, 0, 0}
>         }
> #1  0x0000000000403242 in engine () at atop.c:829
>         lastcmd = <optimized out>
>         nactproc = 4285065470
>         totslpi = -1613224016
>         devsstat = 0x7fb47489d010
>         curpact = 0x7fb474867010
>         curpexit = <optimized out>
>         devtstat = <optimized out>
>         ntask = 280
>         sigact = {
>           __sigaction_handler = {
>             sa_handler = 0x403890 <getalarm>,
>             sa_sigaction = 0x403890 <getalarm>
>           },
>           sa_mask = {
>             __val = {0 <repeats 16 times>}
>           },
>           sa_flags = 0,
>           sa_restorer = 0x0
>         }
>         devpstat = <optimized out>
>         nexit = <optimized out>
>         ndeviat = <optimized out>
>         totproc = 1
>         totrun = 0
>         totslpu = 32767
>         totzombie = 1954458716
>         presstat = 0x7fb474981010
>         curplen = 298
>         i = <optimized out>
>         noverflow = <optimized out>
>         hlpsstat = 0x7fb474981010
>         j = <optimized out>
>         timelimit = 0
>         cursstat = 0x7fb47490f010
> #2  main (argc=<optimized out>, argv=<optimized out>) at atop.c:659
>         i = <optimized out>
>         c = <optimized out>
>         p = <optimized out>
>         rlim = {
>           rlim_cur = 18446744073709551615,
>           rlim_max = 18446744073709551615
>         }
> gdb $ print acctrecsz
> $1 = 0
> ----SNIP----
> 
> This is a global variable in acctproc.c. Unfortunately, I didn't have the
> time to investigate this yet.
> 
> Cheers,
> sur5r
> 
> -- 
> ceterum censeo microsoftem esse delendam.


-- 
-----------------------------------------------------------------------------
Marc Haber         | "I don't trust Computers. They | Mailadresse im Header
Leimen, Germany    |  lose things."    Winona Ryder | Fon: *49 6224 1600402
Nordisch by Nature |  How to make an American Quilt | Fax: *49 6224 1600421

Message #25 received at 778598@bugs.debian.org (full text, mbox, reply):

From: Matt <matt@mattkeenan.net>

To: 778598@bugs.debian.org, 778598-submitter@bugs.debian.org

Cc: Jakob Haufe <sur5r@sur5r.net>, Marc Haber <mh+debian-packages@zugschlus.de>, Klaus Ethgen <Klaus@Ethgen.de>

Subject: atop: SIGFPE

Date: Sat, 29 Oct 2016 19:36:18 +0100

Hi all,

So, I've tracked down this bug and what is causing it; in accprocnt()
you'll find the following,


int

acctprocnt(void)

{

  struct stat     statacc;

  /*

  ** if accounting not supported, skip call

  */

  if (acctfd == -1)

    return 0;

  /*

  ** determine how many processes exited last interval

  */

  (void) fstat(acctfd, &statacc);

  if (acctsize > statacc.st_size) /* somebody did reset accounting? */

  {

    /*

    ** reposition to start of file

    */

    lseek(acctfd, 0, SEEK_SET);

    acctsize = 0;

  }

  return (statacc.st_size - acctsize) / acctrecsz;

}


As you can see the last line does a division, and this is the source of
the SIGFPE; when acctrecsz is zero obviously.

Now here comes the tricky bit as far as a bug fix is concerned;
acctrecsz is a static variable, i.e. a file local global variable.

And it gets worse, there are times when it might be used uninitialised,
or used after the accounting file is considered corrupt or otherwise
inaccessible (xref functions acctvers(), acctswon() (acctrecsz may not
be set after call to acctvers()) x3, and acctswon() acctrecsz may not be
set after call to acctvers() and return code of acctvers() is ignored.

So, the best way to fix this would require an extensive refactor and
re-write to fix these use of globals. This would mean using one struct
to encapsulate the accounting info (the accounting file fd, accounting
file version, account file record sizes, etc), and then allocate this
struct at runtime (in a suitable init function, quite possibly the one
that opens and initialises the accounting file), pass the accounting
file struct to functions that handle this accounting file data.

The other fix would be a quick and dirty fix to make sure that the
division never happens when acctrecsz is zero.

There could be a middle ground as well; but half hearted measures aren't
so appealing to me.

So I can write either, but which would you prefer? I'm guess the big
refactor / re-write would be harder to push upstream, but since I'm not
the maintainer I couldn't say.

Matt

Message #33 received at 778598@bugs.debian.org (full text, mbox, reply):

From: Marc Haber <mh+debian-packages@zugschlus.de>

To: Matt <matt@mattkeenan.net>

Cc: 778598@bugs.debian.org, 778598-submitter@bugs.debian.org, Jakob Haufe <sur5r@sur5r.net>, Klaus Ethgen <Klaus@Ethgen.de>

Subject: Re: atop: SIGFPE

Date: Sun, 30 Oct 2016 19:45:54 +0100

On Sat, Oct 29, 2016 at 07:36:18PM +0100, Matt wrote:
> So I can write either, but which would you prefer? I'm guess the big
> refactor / re-write would be harder to push upstream, but since I'm not
> the maintainer I couldn't say.

I apologize for any wasted work, but there is no record of the issue
happening with the new upstream versions in experimental.

Did you work against experimental atop 2.2.x, or against the outdated
and soon-to-be-replaced atop 1.27 in jessie, stretch or sid?

Greetings
Marc

-- 
-----------------------------------------------------------------------------
Marc Haber         | "I don't trust Computers. They | Mailadresse im Header
Leimen, Germany    |  lose things."    Winona Ryder | Fon: *49 6224 1600402
Nordisch by Nature |  How to make an American Quilt | Fax: *49 6224 1600421

Message #41 received at 778598@bugs.debian.org (full text, mbox, reply):

From: Marc Haber <mh+debian-packages@zugschlus.de>

To: Marc Haber <mh+debian-packages@zugschlus.de>, 778598@bugs.debian.org, 778598-submitter@bugs.debian.org

Cc: Matt <matt@mattkeenan.net>, Jakob Haufe <sur5r@sur5r.net>, Klaus Ethgen <Klaus@Ethgen.de>

Subject: Re: Bug#778598: atop: SIGFPE

Date: Sat, 17 Dec 2016 08:49:35 +0100

On Sun, Oct 30, 2016 at 07:45:54PM +0100, Marc Haber wrote:
> Did you work against experimental atop 2.2.x, or against the outdated
> and soon-to-be-replaced atop 1.27 in jessie, stretch or sid?

This is probably fixed in experimental for months. Please try again
and report back. I will close this bug by the end of Januar 2017 if no
report is received.

Greetings
Marc

Message #49 received at 778598@bugs.debian.org (full text, mbox, reply):

From: Jakob Haufe <sur5r@sur5r.net>

To: Marc Haber <mh+debian-packages@zugschlus.de>

Cc: 778598@bugs.debian.org, 778598-submitter@bugs.debian.org

Subject: Re: Bug#778598: atop: SIGFPE

Date: Sun, 18 Dec 2016 22:43:07 +0100

[Message part 1 (text/plain, inline)]

On Sat, 17 Dec 2016 08:49:35 +0100
Marc Haber <mh+debian-packages@zugschlus.de> wrote:

> This is probably fixed in experimental for months. Please try again
> and report back. I will close this bug by the end of Januar 2017 if no
> report is received.

I installed 2.2.3-1~exp1 back in August and never experienced this bug since
then.

Cheers,
sur5r

-- 
ceterum censeo microsoftem esse delendam.

[Message part 2 (application/pgp-signature, inline)]

Message #57 received at 778598@bugs.debian.org (full text, mbox, reply):

From: Marc Haber <mh+debian-packages@zugschlus.de>

To: 778598@bugs.debian.org, 778598-submitter@bugs.debian.org

Cc: Marc Haber <mh+debian-packages@zugschlus.de>

Subject: Re: Bug#778598: atop: SIGFPE

Date: Wed, 18 Jan 2017 20:16:53 +0100

tags #778598 unreproducible
thanks

as this is probably fixed but still missing feedback from the original
reporter, I am marking this as unreproducible.

Greetings
Marc

Message #67 received at 778598-done@bugs.debian.org (full text, mbox, reply):

From: Marc Haber <mh+debian-packages@zugschlus.de>

To: 778598-done@bugs.debian.org

Subject: Re: Bug#778598: atop: SIGFPE

Date: Sat, 4 Feb 2017 11:11:10 +0100

Version: 2.2.6-2

On Sat, Dec 17, 2016 at 08:49:35AM +0100, Marc Haber wrote:
> This is probably fixed in experimental for months. Please try again
> and report back. I will close this bug by the end of Januar 2017 if no
> report is received.

Closing.

Greetings
Marc

-- 
-----------------------------------------------------------------------------
Marc Haber         | "I don't trust Computers. They | Mailadresse im Header
Leimen, Germany    |  lose things."    Winona Ryder | Fon: *49 6224 1600402
Nordisch by Nature |  How to make an American Quilt | Fax: *49 6224 1600421

Send a report that this bug log contains spam.