Hi, I came across this lovely segfault when using BFD in Bird v1.4.5 and v1.5: (gdb) bt #0 bfd_send_ctl (p=0x5555557cf840, s=0x5555557e2cf8, final=0) at ../../../proto/bfd/packets.c:66 #1 0x000055555556fc86 in timers_fire (loop=loop@entry=0x5555557dfba0) at ../../../proto/bfd/io.c:437 #2 0x000055555556fe24 in birdloop_main (arg=0x5555557dfba0) at ../../../proto/bfd/io.c:759 #3 0x00007ffff7bc4182 in start_thread () from /lib/x86_64-linux-gnu/libpthread.so.0 #4 0x00007ffff78f147d in clone () from /lib/x86_64-linux-gnu/libc.so.6 Any idea if this is fixed in latest master or that it might be result of misconfiguration? Cheers Michal
On Fri, Jul 17, 2015 at 03:22:45PM +0200, Michal Humpula wrote:
Hi,
I came across this lovely segfault when using BFD in Bird v1.4.5 and v1.5:
(gdb) bt #0 bfd_send_ctl (p=0x5555557cf840, s=0x5555557e2cf8, final=0) at ../../../proto/bfd/packets.c:66 #1 0x000055555556fc86 in timers_fire (loop=loop@entry=0x5555557dfba0) at ../../../proto/bfd/io.c:437 #2 0x000055555556fe24 in birdloop_main (arg=0x5555557dfba0) at ../../../proto/bfd/io.c:759 #3 0x00007ffff7bc4182 in start_thread () from /lib/x86_64-linux-gnu/libpthread.so.0 #4 0x00007ffff78f147d in clone () from /lib/x86_64-linux-gnu/libc.so.6
Any idea if this is fixed in latest master or that it might be result of misconfiguration?
Hi There is a known bug related to crashing after not being able to allocate a socket (which would be seen in log). Well, i don't see any obvious problem. Could you send me config, logs and/or core dump? -- Elen sila lumenn' omentielvo Ondrej 'Santiago' Zajicek (email: santiago@crfreenet.org) OpenPGP encrypted e-mails preferred (KeyID 0x11DEADC3, wwwkeys.pgp.net) "To err is human -- to blame it on a computer is even more so."
Thanks for the hint! The crashing after inability to allocate socket, got me thinking that it only occurs only in one connection and only on one instance. So yeah, I've got something silly like this in my setup: protocol bgp test { neighbor 10.0.1.2 as 1; source address 10.1.1.1; bfd 1; } where source address wasn't on any interface (typo error). And sure, the log confirms that: 2015-07-17 13:31:46 <ERR> bfd1: Socket error: bind: Cannot assign requested address Funny thing is, that BGP was able to ignore the mistake in address and continue to successfully established connection with pear, so I've never noticed the problem until now:) Might be worth documenting the bug somewhere if it's not going to be fixed soon:) Thanks again. Cheers Michal On 2015-07-17 23:19, Ondrej Zajicek wrote:
On Fri, Jul 17, 2015 at 03:22:45PM +0200, Michal Humpula wrote:
Hi,
I came across this lovely segfault when using BFD in Bird v1.4.5 and v1.5:
(gdb) bt #0 bfd_send_ctl (p=0x5555557cf840, s=0x5555557e2cf8, final=0) at ../../../proto/bfd/packets.c:66 #1 0x000055555556fc86 in timers_fire (loop=loop@entry=0x5555557dfba0) at ../../../proto/bfd/io.c:437 #2 0x000055555556fe24 in birdloop_main (arg=0x5555557dfba0) at ../../../proto/bfd/io.c:759 #3 0x00007ffff7bc4182 in start_thread () from /lib/x86_64-linux-gnu/libpthread.so.0 #4 0x00007ffff78f147d in clone () from /lib/x86_64-linux-gnu/libc.so.6
Any idea if this is fixed in latest master or that it might be result of misconfiguration?
Hi
There is a known bug related to crashing after not being able to allocate a socket (which would be seen in log).
Well, i don't see any obvious problem. Could you send me config, logs and/or core dump?
On 18.07.2015 15:14, Michal Humpula wrote:
Thanks for the hint! The crashing after inability to allocate socket, got me thinking that it only occurs only in one connection and only on one instance. So yeah, I've got something silly like this in my setup:
protocol bgp test { neighbor 10.0.1.2 as 1; source address 10.1.1.1; bfd 1; }
where source address wasn't on any interface (typo error).
I have similiar config on my Linux router, but in my case source address exists until i shut the interface down. When the interface is shut down (ifdown interfacename), bird crashes. May be dangerous, i could imaginge situations when i want to shut interface down or it just happens accidently. Workaround that works in my lab: sysctl -w net.ipv4.ip_nonlocal_bind=1 -- A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? Z poważaniem Bartosz Radwan
On Tue, Dec 01, 2015 at 11:53:55AM +0100, Bartosz Radwan wrote:
I have similiar config on my Linux router, but in my case source address exists until i shut the interface down. When the interface is shut down (ifdown interfacename), bird crashes.
May be dangerous, i could imaginge situations when i want to shut interface down or it just happens accidently.
Well, it is a bug (fixed in devel branch), not intended behavior. Thanks for the sysctl note. -- Elen sila lumenn' omentielvo Ondrej 'Santiago' Zajicek (email: santiago@crfreenet.org) OpenPGP encrypted e-mails preferred (KeyID 0x11DEADC3, wwwkeys.pgp.net) "To err is human -- to blame it on a computer is even more so."
On 05.12.2015 16:01, Ondrej Zajicek wrote:
Well, it is a bug (fixed in devel branch), not intended behavior Thanks Ondriej.
When approximately next stable release including this fix is expectded? -- A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? Z poważaniem Bartosz Radwan
participants (3)
-
Bartosz Radwan -
Michal Humpula -
Ondrej Zajicek