ntpd 4.2.4 crash on NetBSD/i386 - NTP

This is a discussion on ntpd 4.2.4 crash on NetBSD/i386 - NTP ; -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I just got a new NetBSD box running. It's an old 486 with 16MB RAM and 515MB hard drive. The NetBSD 3.1 GENERIC kernel takes up more than half the RAM, and launching the ...

+ Reply to Thread
Results 1 to 14 of 14

Thread: ntpd 4.2.4 crash on NetBSD/i386

  1. ntpd 4.2.4 crash on NetBSD/i386

    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    I just got a new NetBSD box running. It's an old 486 with 16MB RAM and
    515MB hard drive. The NetBSD 3.1 GENERIC kernel takes up more than half
    the RAM, and launching the stock ntpd (4.2.0) that's part of the
    distribution failed with a message saying "mlockall(): Cannot allocate
    memory". There is a 64MB swap partition on the machine.

    I figured things were too tight, so I built a custom kernel, without USB,
    SCSI, IPV6, parallel port, and so on. Basically, the only options left
    are those for devices I want to use and IPV4. A few file systems are
    disabled as well, but the NTP option is still present. The machine boots
    fine, with no errors that I notice. I've got more RAM available for ntpd.

    From that point, if I try again to run the stock ntpd (4.2.0), it crashes
    on startup with a ntpd.core in the root. I figured I'd try to reduce the
    size of the ntpd object by building the current distribution from source
    (4.2.4). It worked well on my NetBSD/macppc before, so I decided to do
    the same for the 486.

    I ran ./configure --disable-all-clocks --disable-debug
    --enable-local-clock --disable-ipv6

    The make and installation were successful. However, I still get a crash
    on ntpd startup, again with a core dump. I'm out of ideas.


    Here's my ntpd.conf

    pidfile /var/run/ntpd.pid
    driftfile /etc/ntp.drift
    keys /etc/ntp.keys
    statsdir /var/NTP

    trustedkey 1
    controlkey 1
    requestkey 1

    server 192.168.3.14 iburst # That's the NetBSD/macppc machine

    statistics loopstats
    enable stats

    restrict localhost
    restrict 192.168.3.0 mask 255.255.255.0
    restrict default nomodify


    Looking at /var/log/messages doesn't say much... (I have manually put in
    0.000 in the drift file as a startup value, since I don't know yet what
    value will end up there).

    Feb 24 14:39:49 localhost ntpd[28722]: ntpd 4.2.4@1.1437-o Sat Feb 24
    18:04:36 GMT 2007 (1)
    Feb 24 14:39:51 localhost ntpd[28576]: precision = 1250.000 usec
    Feb 24 14:39:51 localhost ntpd[28576]: Listening on interface #0 wildcard,
    0.0.0.0#123 Disabled
    Feb 24 14:39:51 localhost ntpd[28576]: Listening on interface #1 we1,
    192.168.3.41#123 Enabled
    Feb 24 14:39:51 localhost ntpd[28576]: Listening on interface #2 lo0,
    127.0.0.1#123 Enabled
    Feb 24 14:39:51 localhost ntpd[28576]: Listening on routing socket on fd
    #23 for interface updates
    Feb 24 14:39:51 localhost ntpd[28576]: kernel time sync status 0040
    Feb 24 14:39:52 localhost ntpd[28576]: frequency initialized 0.000 PPM
    from /etc/ntp.drift

    That's all she wrote.


    Here's the dmesg, in case it may help

    NetBSD 3.1 (I486) #2: Sat Feb 24 09:14:17 EST 2007
    pldubuc@localhost.localhost:/usr/src/sys/arch/i386/compile/I486
    total memory = 15616 KB
    avail memory = 13180 KB
    mainbus0 (root)
    cpu0 at mainbus0: (uniprocessor)
    cpu0: Intel 486DX (486-class)
    isa0 at mainbus0
    com0 at isa0 port 0x3f8-0x3ff irq 4: ns8250 or ns16450, no fifo
    com1 at isa0 port 0x2f8-0x2ff irq 3: ns8250 or ns16450, no fifo
    pckbc0 at isa0 port 0x60-0x64
    pckbd0 at pckbc0 (kbd slot)
    pckbc0: using irq 1 for kbd slot
    wskbd0 at pckbd0: console keyboard
    wdc0 at isa0 port 0x1f0-0x1f7 irq 14
    atabus0 at wdc0 channel 0
    we1 at isa0 port 0x300-0x31f iomem 0xcc000-0xcdfff irq 10
    we1: SMC8416T Ethernet (16-bit)
    we1: Ethernet address 00:e0:29:16:74:09
    vga0 at isa0 port 0x3b0-0x3df iomem 0xa0000-0xbffff
    wsdisplay0 at vga0 kbdmux 1: console (80x25, vt100 emulation), using
    wskbd0
    wsmux1: connecting to wsdisplay0
    pcppi0 at isa0 port 0x61
    sysbeep0 at pcppi0
    fdc0 at isa0 port 0x3f0-0x3f7 irq 6 drq 2
    fd0 at fdc0 drive 0: 1.44MB, 80 cyl, 2 head, 18 sec
    wd0 at atabus0 drive 0:
    wd0: drive supports 16-sector PIO transfers, chs addressing
    wd0: 515 MB, 1048 cyl, 16 head, 63 sec, 512 bytes/sect x 1056384 sectors
    wd0: drive supports PIO mode 3
    boot device: wd0
    root on wd0a dumps on wd0b
    root file system type: ffs
    wsdisplay0: screen 1 added (80x25, vt100 emulation)
    wsdisplay0: screen 2 added (80x25, vt100 emulation)
    wsdisplay0: screen 3 added (80x25, vt100 emulation)
    wsdisplay0: screen 4 added (80x25, vt100 emulation)


    I'm open to suggestions at this point. I also question whether there are
    specific kernel options that must be turned on in order for ntpd to work
    (other than basic networking, I mean).

    Thanks a bunch.

    - --
    Pierre Dubuc
    pldubuc@yahoo.ca
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.6 (NetBSD)

    iEYEARECAAYFAkXgnZYACgkQaMA1iJ2qYmLMwgCg2q4PuJinIl gNKYet7VQiijAc
    uXkAnjlVZ27o/1QQon1AGLQQvhgkdE+w
    =IYBa
    -----END PGP SIGNATURE-----

  2. Re: ntpd 4.2.4 crash on NetBSD/i386


    >Looking at /var/log/messages doesn't say much... (I have manually put in
    >0.000 in the drift file as a startup value, since I don't know yet what
    >value will end up there).


    Leaving the file non-existant works better. NTP knows about that
    case and goes through a special startup dance to figure out what
    the drift should be.

    --
    These are my opinions, not necessarily my employer's. I hate spam.


  3. Re: ntpd 4.2.4 crash on NetBSD/i386

    >I just got a new NetBSD box running. It's an old 486 with 16MB RAM and
    >515MB hard drive. The NetBSD 3.1 GENERIC kernel takes up more than half
    >the RAM, and launching the stock ntpd (4.2.0) that's part of the
    >distribution failed with a message saying "mlockall(): Cannot allocate
    >memory". There is a 64MB swap partition on the machine.


    I don't think swap space helps. You are trying to lock things
    in main memory so it will respond without the delays of swapping
    things in.


    >I ran ./configure --disable-all-clocks --disable-debug
    > --enable-local-clock --disable-ipv6


    I think "local-clock" refers to a special last-resort
    refclock. Your conf file doesn't include one so you
    don't need it.


    >I'm open to suggestions at this point. I also question whether there are
    >specific kernel options that must be turned on in order for ntpd to work
    >(other than basic networking, I mean).


    I don't have much experience at running things it boxes without
    much memory.

    How much memory do you have left and/or how much does it need?

    ntpd.html says there is a -N option to run at high priority.
    I think that includes locking things in memory. You might
    try turning that switch off (or commenting out that line in
    the source code).

    --
    These are my opinions, not necessarily my employer's. I hate spam.


  4. Re: ntpd 4.2.4 crash on NetBSD/i386

    Pierre Dubuc wrote:
    > -----BEGIN PGP SIGNED MESSAGE-----
    > Hash: SHA1
    >
    > I just got a new NetBSD box running. It's an old 486 with 16MB RAM and
    > 515MB hard drive. The NetBSD 3.1 GENERIC kernel takes up more than half
    > the RAM, and launching the stock ntpd (4.2.0) that's part of the
    > distribution failed with a message saying "mlockall(): Cannot allocate
    > memory". There is a 64MB swap partition on the machine.


    > I'm open to suggestions at this point. I also question whether there are
    > specific kernel options that must be turned on in order for ntpd to work
    > (other than basic networking, I mean).
    >


    If this is an exercise in mascohism I'll leave you to your enjoyment
    thereof. Otherwise it's time to swallow your pride and go trash
    picking!! Better machines than your old 486 are being curbsided every
    day! You not only get more processing power you also get more memory
    and more disk. You get PCI bus slots instead of ISA or EISA. Not only
    that, but you can still buy memory for them!

    If you happen to be in the Philadelphia area I could probably give you a
    better machine!


  5. Re: ntpd 4.2.4 crash on NetBSD/i386

    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    Replying to message <45E0C229.9070006@comcast.net>

    > Date: Sat, 24 Feb 2007 17:54:33 -0500
    > From: Richard B. gilbert
    > Newsgroups: comp.protocols.time.ntp
    > Subject: Re: ntpd 4.2.4 crash on NetBSD/i386
    >
    >
    > If this is an exercise in mascohism I'll leave you to your enjoyment thereof.
    > Otherwise it's time to swallow your pride and go trash picking!! Better
    > machines than your old 486 are being curbsided every day! You not only get
    > more processing power you also get more memory and more disk. You get PCI
    > bus slots instead of ISA or EISA. Not only that, but you can still buy
    > memory for them!
    >


    Actually, it's more an exercise than anything. I hate to waste perfectly
    good hardware, and I thought it would be interesting to play with this old
    box and see if I could eventually set it up as a CHU receiver. Oh well...

    - --
    Pierre Dubuc
    pldubuc@yahoo.ca
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.6 (NetBSD)

    iEYEARECAAYFAkXgy/UACgkQaMA1iJ2qYmJJ8ACg+3r+EzRIhVtfdZjRqW1ie4Z9
    L7cAn2iOftgKO2ESGl6UoUCKAir2Yi5w
    =2YuU
    -----END PGP SIGNATURE-----

  6. Re: ntpd 4.2.4 crash on NetBSD/i386

    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    Replying to message

    > Date: Sat, 24 Feb 2007 15:18:19 -0500
    > From: Pierre Dubuc
    > Newsgroups: comp.protocols.time.ntp
    > Subject: ntpd 4.2.4 crash on NetBSD/i386
    >
    > [big snip]
    >
    >
    > I'm open to suggestions at this point. I also question whether there are
    > specific kernel options that must be turned on in order for ntpd to work
    > (other than basic networking, I mean).


    As an interesting follow-up, I noticed that the kernel doesn't seem to
    handle math very well. Commands "ping" and "top" crash as well, and ping
    yielded an interesting message: "math_emulate: 0xd9fa not implemented
    Illegal instruction", which points to the kernel option that is supposed
    to handle this. I did enable "options math_emulate", but old posts on the
    netbsd lists seem to indicate that the emulation is or was not complete
    at one point. This may be my problem since ntpd probably does a lot of
    statistical calculations. I'll post there and see what people say.

    - --
    Pierre Dubuc
    pldubuc@yahoo.ca
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.6 (NetBSD)

    iEYEARECAAYFAkXgzX8ACgkQaMA1iJ2qYmKMggCgsjvLbJVany cXyuQfm7cqV3VA
    /ygAoI8uiW2QS6JKuxNiuMGSlbP8CRsB
    =9NlZ
    -----END PGP SIGNATURE-----

  7. Re: ntpd 4.2.4 crash on NetBSD/i386

    Pierre,

    Please see http://bugs.ntp.isc.org/387 . There is a link on that bug to a
    couple of pages at http://ntp.isc.org that may also help.

    H

  8. Re: ntpd 4.2.4 crash on NetBSD/i386

    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    Replying to message

    > Date: Sun, 25 Feb 2007 08:10:11 +0000
    > From: Harlan Stenn
    > Newsgroups: comp.protocols.time.ntp
    > Subject: Re: ntpd 4.2.4 crash on NetBSD/i386
    >
    > Pierre,
    >
    > Please see http://bugs.ntp.isc.org/387 . There is a link on that bug to a
    > couple of pages at http://ntp.isc.org that may also help.
    >
    > H
    >


    That's an interesting idea. Turning on debug messages on the stock 4.2.0
    gives me the following output.

    Debug1: 2 -> 2 = 2
    ntpd 4.2.0-r Tue Oct 31 03:50:01 UTC 2006 (1)
    Debug1: 2 -> 2 = 2
    addto_syslog: ntpd 4.2.0-r Tue Oct 31 03:50:01 UTC 2006 (1)
    addto_syslog: set_process_priority: Leave priority alone: priority_done is
    <2>
    addto_syslog: precision = 1260.000 usec
    create_sockets(123)
    addto_syslog: no IPv6 interfaces found
    bind() fd 4, family 2, port 123, addr 0.0.0.0, flags=8
    flags for fd 4: 0106
    bind() fd 5, family 2, port 123, addr 192.168.3.41, flags=8
    flags for fd 5: 0106
    bind() fd 6, family 2, port 123, addr 127.0.0.1, flags=0
    flags for fd 6: 0106
    create_sockets: ninterfaces=3
    interface 0: fd=4, bfd=-1, name=wildcard, flags=0x8
    sin=0.0.0.0 bcast=0.0.0.0, mask=255.255.255.255
    interface 1: fd=5, bfd=-1, name=we1, flags=0x9
    sin=192.168.3.41 bcast=192.168.3.255, mask=255.255.255.0
    interface 2: fd=6, bfd=-1, name=lo0, flags=0x5
    sin=127.0.0.1 mask=255.0.0.0
    init_io: maxactivefd 6
    local_clock: time 0 offset 0.000000 state 0
    Debug2: 2 -> 2 = 2
    addto_syslog: frequency initialized 0.000 PPM from /etc/ntp.drift
    local_clock: time 0 offset 0.000000 state 1
    getnetnum given 192.168.3.14, got 192.168.3.14
    key_expire: at 0
    peer_clear: at 0 assoc ID 44148 refid INIT
    newpeer: 192.168.3.41->192.168.3.14 mode 3 vers 4 poll 6 10 flags 0x201
    0x1 ttl 0 key 00000000
    getnetnum given localhost, got ::1
    getnetnum given 192.168.3.0, got 192.168.3.0
    getnetnum given 255.255.255.0, got 255.255.255.0
    report_event: system event 'event_restart' (0x01) status 'sync_alarm,
    sync_unspec, 1 event, event_unspec' (0xc010)
    MCAST *****sendpkt(fd=5 dst=192.168.3.14, src=192.168.3.41,
    ttl=0, len=48)
    transmit: at 1 192.168.3.41->192.168.3.14 mode 3
    poll_update: at 1 192.168.3.14 flags 0201 poll 6 burst 8 last 1 next 3
    auth_agekeys: at 1 keys 1 expired 0
    timer: refresh ts 0
    receive: at 1 192.168.3.41<-192.168.3.14 restrict 000
    receive: at 1 192.168.3.41<-192.168.3.14 mode 4 code 1
    peer 192.168.3.14 event 'event_reach' (0x84) status 'unreach, conf, 1
    event, event_reach' (0xa014)
    math_emulate: 0xd9fa not implemented
    [1] Illegal instruction (core dumped) /usr/sbin/ntpd -...


    Thie problem your link was about seemed more about interface binding, but
    what I get instead from running -D2 is exactly the same error message that
    I get when running ping. I guess floating point emulation is not completely
    implemented in NetBSD/i386. I'll follow-up with those guys.

    Thanks.

    - --
    Pierre Dubuc
    pldubuc@yahoo.ca
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.6 (NetBSD)

    iEYEARECAAYFAkXhf1IACgkQaMA1iJ2qYmJDsACgr9QQ7oijHn 8CuNZvaIWsNuxN
    7agAn3E01NkNtLCrz5AiA/eaXNa+oXWa
    =3Ylh
    -----END PGP SIGNATURE-----

  9. ***SPAM*** Re: ntpd 4.2.4 crash on NetBSD/i386

    Pierre Dubuc wrote:
    >
    > Debug1: 2 -> 2 = 2
    > ntpd 4.2.0-r Tue Oct 31 03:50:01 UTC 2006 (1)
    > Debug1: 2 -> 2 = 2
    > addto_syslog: ntpd 4.2.0-r Tue Oct 31 03:50:01 UTC 2006 (1)
    > addto_syslog: set_process_priority: Leave priority alone: priority_done is
    > <2>
    > addto_syslog: precision = 1260.000 usec
    > create_sockets(123)
    > addto_syslog: no IPv6 interfaces found
    > bind() fd 4, family 2, port 123, addr 0.0.0.0, flags=8
    > flags for fd 4: 0106
    > bind() fd 5, family 2, port 123, addr 192.168.3.41, flags=8
    > flags for fd 5: 0106
    > bind() fd 6, family 2, port 123, addr 127.0.0.1, flags=0
    > flags for fd 6: 0106
    > create_sockets: ninterfaces=3
    > interface 0: fd=4, bfd=-1, name=wildcard, flags=0x8
    > sin=0.0.0.0 bcast=0.0.0.0, mask=255.255.255.255
    > interface 1: fd=5, bfd=-1, name=we1, flags=0x9
    > sin=192.168.3.41 bcast=192.168.3.255, mask=255.255.255.0
    > interface 2: fd=6, bfd=-1, name=lo0, flags=0x5
    > sin=127.0.0.1 mask=255.0.0.0
    > init_io: maxactivefd 6
    > local_clock: time 0 offset 0.000000 state 0
    > Debug2: 2 -> 2 = 2
    > addto_syslog: frequency initialized 0.000 PPM from /etc/ntp.drift
    > local_clock: time 0 offset 0.000000 state 1
    > getnetnum given 192.168.3.14, got 192.168.3.14
    > key_expire: at 0
    > peer_clear: at 0 assoc ID 44148 refid INIT
    > newpeer: 192.168.3.41->192.168.3.14 mode 3 vers 4 poll 6 10 flags 0x201
    > 0x1 ttl 0 key 00000000
    > getnetnum given localhost, got ::1


    You shouldn't be getting this since you are not running IPv6. This may
    be the cause of your crash. Please upgrade to 4.2.4 at least. Check if
    you have localhost defined in your hosts file or if your local DNS has
    it defined.

    What does the line look like in your config file that uses localhost?
    Why are you using it at all?


    Danny
    _______________________________________________
    questions mailing list
    questions@lists.ntp.isc.org
    https://lists.ntp.isc.org/mailman/listinfo/questions


  10. Re: ntpd 4.2.4 crash on NetBSD/i386

    -----BEGIN PGP SIGNED MESSAGE-----
    Hash: SHA1

    Replying to message <45E26AB3.6000407@ntp.isc.org>

    > Date: Mon, 26 Feb 2007 05:05:55 GMT
    > From: Danny Mayer
    > Newsgroups: comp.protocols.time.ntp
    > Subject: ***SPAM*** Re: ntpd 4.2.4 crash on NetBSD/i386
    >
    > Pierre Dubuc wrote:
    >>
    >> Debug1: 2 -> 2 = 2
    >> ntpd 4.2.0-r Tue Oct 31 03:50:01 UTC 2006 (1)
    >> Debug1: 2 -> 2 = 2
    >> addto_syslog: ntpd 4.2.0-r Tue Oct 31 03:50:01 UTC 2006 (1)
    >> addto_syslog: set_process_priority: Leave priority alone: priority_done is
    >> <2>
    >> addto_syslog: precision = 1260.000 usec
    >> create_sockets(123)
    >> addto_syslog: no IPv6 interfaces found
    >> bind() fd 4, family 2, port 123, addr 0.0.0.0, flags=8
    >> flags for fd 4: 0106
    >> bind() fd 5, family 2, port 123, addr 192.168.3.41, flags=8
    >> flags for fd 5: 0106
    >> bind() fd 6, family 2, port 123, addr 127.0.0.1, flags=0
    >> flags for fd 6: 0106
    >> create_sockets: ninterfaces=3
    >> interface 0: fd=4, bfd=-1, name=wildcard, flags=0x8
    >> sin=0.0.0.0 bcast=0.0.0.0, mask=255.255.255.255
    >> interface 1: fd=5, bfd=-1, name=we1, flags=0x9
    >> sin=192.168.3.41 bcast=192.168.3.255, mask=255.255.255.0
    >> interface 2: fd=6, bfd=-1, name=lo0, flags=0x5
    >> sin=127.0.0.1 mask=255.0.0.0
    >> init_io: maxactivefd 6
    >> local_clock: time 0 offset 0.000000 state 0
    >> Debug2: 2 -> 2 = 2
    >> addto_syslog: frequency initialized 0.000 PPM from /etc/ntp.drift
    >> local_clock: time 0 offset 0.000000 state 1
    >> getnetnum given 192.168.3.14, got 192.168.3.14
    >> key_expire: at 0
    >> peer_clear: at 0 assoc ID 44148 refid INIT
    >> newpeer: 192.168.3.41->192.168.3.14 mode 3 vers 4 poll 6 10 flags 0x201
    >> 0x1 ttl 0 key 00000000
    >> getnetnum given localhost, got ::1

    >
    > You shouldn't be getting this since you are not running IPv6. This may
    > be the cause of your crash. Please upgrade to 4.2.4 at least. Check if
    > you have localhost defined in your hosts file or if your local DNS has
    > it defined.
    >
    > What does the line look like in your config file that uses localhost?
    > Why are you using it at all?
    >
    >
    > Danny


    You are right that I compiled my version of 4.2.4 with debugging and IPv6
    disabled, and I also built my streamlined kernel without IPv6 support.

    However, what you see above was actually the debug output from
    starting the stock ntpd 4.2.0 from a GENERIC kernel, which does enable
    IPv6. I would therefore expect the daemon to attempt to use IPv6 on
    startup.

    - --
    Pierre Dubuc
    pldubuc@yahoo.ca
    -----BEGIN PGP SIGNATURE-----
    Version: GnuPG v1.4.6 (NetBSD)

    iEYEARECAAYFAkXiwPsACgkQaMA1iJ2qYmKHxgCeM5qdDjHQIt 69U8957JV+RFuu
    rk8An2C04dWHoskjrVj8BHhKrFu81ggf
    =Z6do
    -----END PGP SIGNATURE-----

  11. Re: ntpd 4.2.4 crash on NetBSD/i386

    Pierre Dubuc wrote:
    > Replying to message <45E26AB3.6000407@ntp.isc.org>
    >
    >> Date: Mon, 26 Feb 2007 05:05:55 GMT
    >> From: Danny Mayer
    >> Newsgroups: comp.protocols.time.ntp
    >> Subject: ***SPAM*** Re: ntpd 4.2.4 crash on NetBSD/i386

    >
    >> Pierre Dubuc wrote:
    >>> Debug1: 2 -> 2 = 2
    >>> ntpd 4.2.0-r Tue Oct 31 03:50:01 UTC 2006 (1)
    >>> Debug1: 2 -> 2 = 2
    >>> addto_syslog: ntpd 4.2.0-r Tue Oct 31 03:50:01 UTC 2006 (1)
    >>> addto_syslog: set_process_priority: Leave priority alone: priority_done is
    >>> <2>
    >>> addto_syslog: precision = 1260.000 usec
    >>> create_sockets(123)
    >>> addto_syslog: no IPv6 interfaces found
    >>> bind() fd 4, family 2, port 123, addr 0.0.0.0, flags=8
    >>> flags for fd 4: 0106
    >>> bind() fd 5, family 2, port 123, addr 192.168.3.41, flags=8
    >>> flags for fd 5: 0106
    >>> bind() fd 6, family 2, port 123, addr 127.0.0.1, flags=0
    >>> flags for fd 6: 0106
    >>> create_sockets: ninterfaces=3
    >>> interface 0: fd=4, bfd=-1, name=wildcard, flags=0x8
    >>> sin=0.0.0.0 bcast=0.0.0.0, mask=255.255.255.255
    >>> interface 1: fd=5, bfd=-1, name=we1, flags=0x9
    >>> sin=192.168.3.41 bcast=192.168.3.255, mask=255.255.255.0
    >>> interface 2: fd=6, bfd=-1, name=lo0, flags=0x5
    >>> sin=127.0.0.1 mask=255.0.0.0
    >>> init_io: maxactivefd 6
    >>> local_clock: time 0 offset 0.000000 state 0
    >>> Debug2: 2 -> 2 = 2
    >>> addto_syslog: frequency initialized 0.000 PPM from /etc/ntp.drift
    >>> local_clock: time 0 offset 0.000000 state 1
    >>> getnetnum given 192.168.3.14, got 192.168.3.14
    >>> key_expire: at 0
    >>> peer_clear: at 0 assoc ID 44148 refid INIT
    >>> newpeer: 192.168.3.41->192.168.3.14 mode 3 vers 4 poll 6 10 flags 0x201
    >>> 0x1 ttl 0 key 00000000
    >>> getnetnum given localhost, got ::1

    >> You shouldn't be getting this since you are not running IPv6. This may
    >> be the cause of your crash. Please upgrade to 4.2.4 at least. Check if
    >> you have localhost defined in your hosts file or if your local DNS has
    >> it defined.

    >
    >> What does the line look like in your config file that uses localhost?
    >> Why are you using it at all?

    >
    >
    >> Danny

    >
    > You are right that I compiled my version of 4.2.4 with debugging and IPv6
    > disabled, and I also built my streamlined kernel without IPv6 support.
    >
    > However, what you see above was actually the debug output from
    > starting the stock ntpd 4.2.0 from a GENERIC kernel, which does enable
    > IPv6. I would therefore expect the daemon to attempt to use IPv6 on
    > startup.
    >


    4.2.0 is very old and it's possible you ran into the same problem as Bug
    #393( https://ntp.isc.org/bugs/show_bug.cgi?id=393) where it got an IPv6
    address as a result of a DNS lookup and tried to use it and then ntpq
    was used and that caused the crash as the IPv6 address wasn't valid (I
    forget the exact reason). In that case it wasn't carefully ignoring IPv6
    address as it should have. It's fixed.

    Danny
    _______________________________________________
    questions mailing list
    questions@lists.ntp.isc.org
    https://lists.ntp.isc.org/mailman/listinfo/questions


  12. Re: ntpd 4.2.4 crash on NetBSD/i386

    In article ,
    Pierre Dubuc writes:

    >to handle this. I did enable "options math_emulate", but old posts on the
    >netbsd lists seem to indicate that the emulation is or was not complete
    >at one point. This may be my problem since ntpd probably does a lot of
    >statistical calculations. I'll post there and see what people say.


    486's have math co-processors built in. Try turning off the emulation.

    Chris

  13. Re: ntpd 4.2.4 crash on NetBSD/i386

    wrote in message
    newsrOdnYmd64QckXjYnZ2dnUVZ_tTinZ2d@comcast.com...
    > In article ,
    > Pierre Dubuc writes:


    >> [...] I did enable "options math_emulate", but old posts on the
    >> netbsd lists seem to indicate that the emulation is or was not
    >> complete at one point. ...

    >
    > 486's have math co-processors built in. Try turning off the emulation.


    486DXs do. 486SXs do not.

    Groetjes,
    Maarten Wiltink



  14. Re: ntpd 4.2.4 crash on NetBSD/i386

    Maarten Wiltink wrote:
    > wrote in message

    []
    >> 486's have math co-processors built in. Try turning off the
    >> emulation.

    >
    > 486DXs do. 486SXs do not.
    >
    > Groetjes,
    > Maarten Wiltink


    Pedantically, 486SXs have the co-processor built-in, but it's either
    disabled or not connected (I can't remember which).



    David



+ Reply to Thread