| Age | Commit message (Collapse) | Author |
|---|
|
|
|
|
|
|
|
|
|
as appropriate
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
handling that was re-using ipv4_input() instead of using ip4_input()
from netinet/ip_ip4.c
|
|
|
|
|
|
|
|
a perfect emulation of a "real" architecture's splsoftclock, as it
assumes it is only invoked from higher spl levels. Use splsoftnet instead.
|
|
|
|
|
|
james@oaktree.co.uk by re-working icmp embedded-packet code so that
ip_forward() m_copy()-aliased packet can be forwarded to ip_output and
icmp_error() safely, because no packet tweaking is needed before
calling icmp_error()
|
|
|
|
|
|
|
|
|
|
|
|
to an attack pointed out by David Wagner.
|
|
|
|
|
|
s remote interfaces, both ARP and local delivery had bugs
|
|
other families than inet.
|
|
|
|
|
|
|
|
location, as attempts to do it differently have caused too many problems.
|
|
|
|
hopefully without random kernel data corruption.
|
|
host on a different interface. Also add interface name to other overwrite
diagnostic messages.
|
|
|
|
data corruptions and panics that people have experienced. by reverting
we loose tcp signatures and ipv6 cleanups, the code looked correct to me.
|
|
of running out of memory when adding SPIs.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
