[Dnsmasq-discuss] segmentation fault after upgrade to 2.77

[AW]

Hi!
Is there anything new in 2.77 (compared to 2.76), that might cause a segmentation fault directly after a file has been sent?Jul 16 09:24:01 neo dnsmasq-tftp[1967]: sent /var/dnsmasq/tftpboot/pxe/x86_64-efi/core.efi to 192.168.1.2Jul 16 09:24:01 neo systemd[1]: dnsmasq.service: Main process exited, code=killed, status=11/SEGV
https://bugs.archlinux.org/task/54834
Thx.
ByeArne

[Simon Kelley]

Post by AW
Hi!
Is there anything new in 2.77 (compared to 2.76), that might cause a
segmentation fault directly after a file has been sent?
Jul 16 09:24:01 neo dnsmasq-tftp[1967]: sent
/var/dnsmasq/tftpboot/pxe/x86_64-efi/core.efi to 192.168.1.2
Jul 16 09:24:01 neo systemd[1]: dnsmasq.service: Main process exited,
code=killed, status=11/SEGV
https://bugs.archlinux.org/task/54834

Juts tried with a similar config to yours and I can't reproduce this, so
as the only known possessor of a system showing this bug, we'll need you
to do some more tests, if that's possible.

Can you build a binary with debugging symbols and run it under gdb?
Failing that, running under strace would be useful.


Cheers,

Simon.

[AW]

Hi!
I removed that dhcp-script line from my config file.And now it works fine (with and without strace) (I did 3 suspend+thaw cycles with no segv).
Thanks.
ByeArne

Post by AW
Hi!
Is there anything new in 2.77 (compared to 2.76), that might cause a
segmentation fault directly after a file has been sent?
Jul 16 09:24:01 neo dnsmasq-tftp[1967]: sent
/var/dnsmasq/tftpboot/pxe/x86_64-efi/core.efi to 192.168.1.2
Jul 16 09:24:01 neo systemd[1]: dnsmasq.service: Main process exited,
code=killed, status=11/SEGV
https://bugs.archlinux.org/task/54834

Juts tried with a similar config to yours and I can't reproduce this, so
as the only known possessor of a system showing this bug, we'll need you
to do some more tests, if that's possible.

Can you build a binary with debugging symbols and run it under gdb?
Failing that, running under strace  would be useful.


Cheers,

Simon.

[Simon Kelley]

OK, I'm guessing that the problem lies in this commit,

http://thekelleys.org.uk/gitweb/?p=dnsmasq.git;a=commit;h=c77fb9d8f09d136fa71bde2469c4fd11cefa6f4a

but I still can't reproduce the problem here.

Are you able to build dnsmasq with debug symbols, run it under gdb and
trigger the crash? If you're not sure how to do that, I can help, but
for the build part, it might be better to go back to the Archlinux people.


Cheers,

Simon.

Post by AW
Hi!
I removed that dhcp-script line from my config file.
And now it works fine (with and without strace) (I did 3 suspend+thaw
cycles with no segv).
Thanks.
Bye
Arne

Post by AW
Hi!
Is there anything new in 2.77 (compared to 2.76), that might cause a
segmentation fault directly after a file has been sent?
Jul 16 09:24:01 neo dnsmasq-tftp[1967]: sent
/var/dnsmasq/tftpboot/pxe/x86_64-efi/core.efi to 192.168.1.2
Jul 16 09:24:01 neo systemd[1]: dnsmasq.service: Main process exited,
code=killed, status=11/SEGV
https://bugs.archlinux.org/task/54834

Juts tried with a similar config to yours and I can't reproduce this, so
as the only known possessor of a system showing this bug, we'll need you
to do some more tests, if that's possible.
Can you build a binary with debugging symbols and run it under gdb?
Failing that, running under strace would be useful.
Cheers,
Simon.
_______________________________________________
Dnsmasq-discuss mailing list
http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss
_______________________________________________
Dnsmasq-discuss mailing list
http://lists.thekelleys.org.uk/mailman/listinfo/dnsmasq-discuss