| Age | Commit message (Collapse) | Author |
|---|
|
virtual MAC address is set. Among other things, this makes route6d work
correctly on systems with carp interfaces.
In order to ensure backwards compatibility, we do not include IPv6
link-local addresses in generating the HMAC, but we accept HMACs with AND
without the link-local addresses. They will be added to the HMAC in a future
release.
In short: this change should only affect backwards compatibility for
IPv6 users who are manually adding link-local addresses on carp interfaces.
testing mtu@ todd@
ok mpf@ henning@ deraadt@
|
|
ok mpf@ henning@
|
|
we've won the MASTER election.
This seems to be necessary w/ some wacky L3 switches,
that only learn by looking at ARP packets.
Fixes PR 5148.
OK mcbride@
|
|
parameter so they can work on alternate tables. table 0 hardcoded for
many callers yet, that will be adapted step by step.
input + ok claudio norby hshoexer
|
|
As a first user, move the global carp(4) demotion counter
into the interface group. Thus we have the possibility
to define which carp interfaces are demoted together.
Put the demotion counter into the reserved field of the carp header.
With this, we can have carp act smarter if multiple errors occur.
It now always takes over other carp peers, that are advertising
with a higher demote count. As a side effect, we can also have
group failovers without the need of running in preempt mode.
The protocol change does not break compability with older
implementations.
Collaborative work with mcbride@
OK mcbride@, henning@
|
|
the establish calls were bypassed or failed.
ok mpf@
|
|
for dumb non simplex interfaces, that receive packets they've just
sent. Fixes bug 5121.
OK deraadt@, mcbride@
|
|
ok henning mcbride
|
|
outbound), using a new BIOCSDIRFILT ioctl;
guidance, feedback and ok canacar@
|
|
ok henning@
|
|
directly. rather provide a rt_lookup function for regular lookups,
and a rt_gettable for those that need access to the head for some reason.
the latter cases should be revisted later probably so that nothing outside
the routing core code accesses the heads at all...
tested claudio jolan me, ok claudio markus
|
|
Add some f's; do comparison in host byte order.
OK mcbride@
|
|
OK mcbride@
|
|
router so back out the routing stuff to pre-eurobsdcon where my machine
doesn't crash immediately.
i am happy to test diffs and report success/failures but i am not happy
to have instantaneous crashes when i reboot with a new kernel that was
compiled from pristine sources.
if you are going to be an elitist asshole then you could at least make
sure your code works.
ok and "be crass towards them" deraadt@
|
|
but go through a provided wrapper.
also provide rt_lookup() instead of doing the lookup manually in many places.
ryan ok
|
|
A bug introduced in -r 1.4 led lower priotorized hosts
switching to MASTER state for a short time at bootup,
if preemption was enabled.
OK mcbride@
|
|
alias address order on the interface irrelevant.
prodding by henning@, ok mpf@
|
|
the result to determine which carp interface should answer rather than
simply using the ip address.
Fixes breakage debugged by Matt Bradford <m.bradford@isrc.qut.edu.au>
'just commit' deraadt@
|
|
freed.
ok canacar deraadt
|
|
ok mpf@
|
|
The previous code could wrongly delete multicast groups
on the parent interface. Now we forward only remembered
delete requests to the carpdev.
OK mcbride, mickey. "Get it in" deraadt.
|
|
the multicast list elsewhere, it can occur under certain conditions (PR4475).
ok deraadt@ mpf@
|
|
to bpf with either an address family or other header added.
These helpers only allocate a much smaller struct m_hdr on the stack when
needed, rather than leaving 256 byte struct mbufs on the stack in deep
call paths. Also removes a fair bit of duplicated code.
commit now, tune after deraadt@
|
|
IFF_RUNNING instead. Also, do not blindly restore flags onto an interface when
returning to the running state.
This eliminates the possibility that the kernel will bring the interface back
up after a link state change on the physical interface, even though the user
may have done an "ifconfig carpN down" in the meantime. Similarly two link
state down events on the physical interface in succession could result in us
never coming out of the INIT state.
The master/backup state is still reported via link state information, so
bgpd/ospfd, etc, should function as before.
This also addresses PR4238.
ok mpf@, ok with the idea etc henning@
|
|
with pascoe@
|
|
This converts if_link_state_change() to a generic usable
callback with dohooks().
OK henning@, camield@
Tested by camield@ and Alexey E. Suslikov
|
|
carpdev case; block outgoing packets in carp_output() instead.
- unbreak the "bow out on 'ifconfig down'" behaviour, allows instant
failover on graceful halt/reboot to work again.
ok mpf@ henning@
|
|
ok mcbride
|
|
that carp shares the IP addresse with its physical interface.
It's neither supported nor needed, since we have carpdev now.
ok mcbride@
|
|
which was removed in rev 1.57. This makes arpbalance work again, and
ensures that only one carp host will reply to a given arp request.
ok mpf@ pascoe@ dhartmei@ deraadt@
|
|
|
|
|
|
with an id10t error by yours truly.
|
|
ifp->if_flags & IFF_MULTICAST checks no longer protect against.
ok pascoe@ mpf@
|
|
use our carp_addr_updated callback to detect deletion and reconfigure
appropriately.
ok mcbride@
|
|
with an interface down avoid preempting after a reboot.
Pointed out by Stephen Marley <stephen.marley@catwoman.cl-is.com>
ok pascoe@ mpf@
|
|
Advertisements run through the carp interface first.
So we just take the address from ifp0.
While we're there,
also remove carp_macmatch6, which isn't used anymore.
Proposed by mcbride@
ok mcbride@, pascoe@
|
|
|
|
IPv4, and will allow us to get rid of carp_fix_lladdr() and the carp mbuf
tag later).
ok pascoe@ mpf@
|
|
unable to tag a packet, and free mbuf chain on all failures.
ok mcbride@
|
|
substitution.
With this change and the previous ARP changes, all Ethernet packets sent
via a CARP interface should now have the virtual MAC address as the source.
ok mcbride@ niklas@
|
|
instead of processing them in the parent interface. This causes us to send
ARP packets on a carpdev with its virtual MAC address as the source, rather
than the parent's physical address.
Permits us to fail over cleanly when connected to devices that use the
source MAC address on a received ARP request to generate an ARP entry for
the requesting host, rather than going through the who-has/is-at procedure.
ok niklas@ mcbride@
|
|
callback after in_ifinit has run. This lets us correct any routes that
in_ifinit has incorrectly added for our interface.
Also be more explicit with the routing checks/changes that we make. This
makes it possible to have different prefix lengths on a CARP interface and
the physical interface it is bound to.
ok mcbride@
|
|
kills a couple of ugly line wraps. From Max Laier.
ok pascoe@
|
|
ok pascoe@ mpf@
|
|
been experiencing output errors.
ok mcbride@
|
|
v6 advertisements are still received the old-fashioned way.
ok mcbride@
|
|
matching in the bridge receive path to make CARP operate correctly
on physical interfaces that are participating in a bridge.
ok mcbride@ henning@ dlg@
|
|
ok pascoe@
|
|
carp_iamatch() on the physical interface.
ok pascoe@ mpf@
|
