https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=241698
Bug ID: 241698
Summary: pfctl segmentation fault, weird "set skip" behavior
Product: Base System
Version: 12.0-RELEASE
Hardware: Any
OS: Any
Status: New
Severity: Affects Only Me
Priority: ---
Component: kern
Assignee: b...@freebsd.org
Reporter: darks...@darkbsd.org
I was running pf with the following line :
set skip on lo
Replacing it to :
set skip on lo0
And then reloading configuration with the following command ends up causing a
segmentation fault and a core dump :
# pfctl -f /etc/pf.conf
The core dump reveals it occured in strncmp() but pfctl did not provide the
base symbols so I can't be sure yet of what is going on.
#0 0x00080048c1d5 in strncmp () from /lib/libc.so.7
(gdb) bt
#0 0x00080048c1d5 in strncmp () from /lib/libc.so.7
#1 0x0021d7f0 in ?? ()
#2 0x00220c6a in ?? ()
#3 0x00222cbb in ?? ()
#4 0x0021d11b in ?? ()
#5 0x00080026d000 in ?? ()
#6 0x in ?? ()
It should also be noted that re-trying the command right after this works, this
time. Also, for some reason, I had to run pfctl twice to have traffic on the
loopback interface processed properly. (This would prove to be nasty because
DNS queries and traffic between jails would end up blocked...)
As a workaround, I have now removed "set skip on lo" and gone for "pass quick
on lo0 no state". This has mitigated both issues (the pf segfault when
rewriting pf.conf between lo and lo0, and traffic to lo0 being blocked)
--
You are receiving this mail because:
You are the assignee for the bug.
___
freebsd-bugs@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-bugs
To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"