- Do not calculate constant length values at run time,
CTASSERT() their sanity.
- Remove superfluous cleaning of mbuf fields after allocation.
- Replace compat macros with function calls.
Sponsored by: Nginx, Inc.
Currently we use interface indeces as zone IDs for link-local and
interface-local scopes, and since we don't have any tool to configure
zone IDs, there is no need to acquire the afdata lock several times per
packet only to read if_index value.
So, now in6_setscope reads zone IDs for interface-local, link-local and
global scopes without a lock.
Sponsored by: Yandex LLC
MFC after: 2 weeks
It stops treating the address on the interface as special by source
address selection rule even when the interface is outgoing interface.
This is desired in some situation.
Requested by: hrs
Reviewed by: IHANet folks including hrs
MFC after: 1 week
and embeds it into address. Inside the kernel we keep addresses with
embedded zone id only for two scopes: link-local and interface-local.
For other scopes this function is nop in most cases. To reduce an
overhead of locking, first check that address is capable for embedding.
Also, handle the loopback address before acquire the lock.
Sponsored by: Yandex LLC
MFC after: 1 week
all interested parties in case if interface flag IFF_UP has changed.
However, not only SIOCSIFFLAGS can raise the flag, but SIOCAIFADDR
and SIOCAIFADDR_IN6 can, too. The actual |= is done not in the protocol
code, but in code of interface drivers. To fix this historical layering
violation, we will check whether ifp->if_ioctl(SIOCSIFADDR) raised the
IFF_UP flag, and if it did, run the if_up() handler.
This fixes configuring an address under CARP control on an interface
that was initially !IFF_UP.
P.S. I intentionally omitted handling the IFF_SMART flag. This flag was
never ever used in any driver since it was introduced, and since it
means another layering violation, it should be garbage collected instead
of pretended to be supported.
For now use 256 buckets and fnv_hash function. Use xor'ed 32-bit
s6_addr32 parts of in6_addr structure as a hash key. Update
in6_localip and in6_is_addr_deprecated to use hash table for fastest
lookup.
Sponsored by: Yandex LLC
Discussed with: dwmalone, glebius, bz
entering llentry_free(), and in case if we lose the race, we should simply
perform LLE_FREE_LOCKED(). Otherwise, if the race is lost by the thread
performing arptimer(), it will remove two references from the lle instead
of one.
Reported by: Ian FREISLICH <ianf clue.co.za>
- Deembed scope id in L3 address in in6_lltable_dump().
- Simplify scope id recovery in rtsock routines.
- Remove embedded scope id handling in ndp(8) and route(8) completely.
Instead, add protocol specific mbuf flags M_IP_NEXTHOP and
M_IP6_NEXTHOP. Use them to indicate that the mbuf's chain
contains the PACKET_TAG_IPFORWARD tag. And do a tag lookup
only when this flag is set.
Suggested by: andre
on the related functionality in the runtime via the sysctl variable
net.pfil.forward. It is turned off by default.
Sponsored by: Yandex LLC
Discussed with: net@
MFC after: 2 weeks
in network byte order. Any host byte order processing is
done in local variables and host byte order values are
never[1] written to a packet.
After this change a packet processed by the stack isn't
modified at all[2] except for TTL.
After this change a network stack hacker doesn't need to
scratch his head trying to figure out what is the byte order
at the given place in the stack.
[1] One exception still remains. The raw sockets convert host
byte order before pass a packet to an application. Probably
this would remain for ages for compatibility.
[2] The ip_input() still subtructs header len from ip->ip_len,
but this is planned to be fixed soon.
Reviewed by: luigi, Maxim Dounin <mdounin mdounin.ru>
Tested by: ray, Olivier Cochard-Labbe <olivier cochard.me>
are using (different) ND-based approach described in RFC 4861. This change
is similar to r241406 which conditionally skips the same check in IPv4.
This change is part of bigger patch eliminating rte locking.
Sponsored by: Yandex LLC.
OK'd by: hrs
MFC after: 2 weeks
host byte order, was sometimes called with net byte order. Since we are
moving towards net byte order throughout the stack, the function was
converted to expect net byte order, and its consumers fixed appropriately:
- ip_output(), ipfilter(4) not changed, since already call
in_delayed_cksum() with header in net byte order.
- divert(4), ng_nat(4), ipfw_nat(4) now don't need to swap byte order
there and back.
- mrouting code and IPv6 ipsec now need to switch byte order there and
back, but I hope, this is temporary solution.
- In ipsec(4) shifted switch to net byte order prior to in_delayed_cksum().
- pf_route() catches up on r241245 changes to ip_output().
llentry_free() and arptimer():
o Use callout_init_rw() for lle timeout, this allows us safely
disestablish them.
- This allows us to simplify the arptimer() and make it
race safe.
o Consistently use ifp->if_afdata_lock to lock access to
linked lists in the lle hashes.
o Introduce new lle flag LLE_LINKED, which marks an entry that
is attached to the hash.
- Use LLE_LINKED to avoid double unlinking via consequent
calls to llentry_free().
- Mark lle with LLE_DELETED via |= operation istead of =,
so that other flags won't be lost.
o Make LLE_ADDREF(), LLE_REMREF() and LLE_FREE_LOCKED() more
consistent and provide more informative KASSERTs.
The patch is a collaborative work of all submitters and myself.
PR: kern/165863
Submitted by: Andrey Zonov <andrey zonov.org>
Submitted by: Ryan Stone <rysto32 gmail.com>
Submitted by: Eric van Gyzen <eric_van_gyzen dell.com>
adding any extension header, or rather before calling into IPsec
processing as we may send the packet and not return to IPv6 output
processing here.
PR: kern/170116
MFC After: 3 days
implementation of RFC 3484 for this purpose for a long time and "prefer_source"
was never implemented actually. ND6_IFF_PREFER_SOURCE macro is left intact.
it skips FLOWTABLE lookup. However, the non-NULL ro has dual meaning
here: it may be supplied to provide route, and it may be supplied to
store and return to caller the route that ip_output()/ip6_output()
finds. In the latter case skipping FLOWTABLE lookup is pessimisation.
The difference between struct route filled by FLOWTABLE and filled
by rtalloc() family is that the former doesn't hold a reference on
its rtentry. Reference is hold by flow entry, and it is about to
be released in future. Thus, route filled by FLOWTABLE shouldn't
be passed to RTFREE() macro.
- Introduce new flag for struct route/route_in6, that marks route
not holding a reference on rtentry.
- Introduce new macro RO_RTFREE() that cleans up a struct route
depending on its kind.
- All callers to ip_output()/ip6_output() that do supply non-NULL
but empty route should use RO_RTFREE() to free results of
lookup.
- ip_output()/ip6_output() now do FLOWTABLE lookup always when
ro->ro_rt == NULL.
Tested by: tuexen (SCTP part)
Possibly do some entra work in case we would not get into the
ifa0 != NULL paths later as we already do for the mltaddr before.
XXX We should possibly error in case in6_setscope fails.
Reference: http://lists.freebsd.org/pipermail/freebsd-net/2011-September/029829.html
Submitted by: bz
MFC after: 1 week
Interface routes are refcounted as packets move through the stack,
and there's garbage collection tied to it so that route changes can
safely propagate while traffic is flowing. In our setup, we weren't
changing or deleting any routes, but the refcounting logic in
ip6_input() was wrong and caused a reference leak on every inbound
V6 packet. This eventually caused a 32bit overflow, and the resulting
0 value caused the garbage collection to run on the active route.
That then snowballed into the panic.
Reviewed by: scottl
MFC after: 3 days
headers for TSO but also for generic checksum offloading. Ideally we
would only have one common function shared amongst all drivers, and
perhaps when updating them for IPv6 we should introduce that.
Eventually we should provide the meta information along with mbufs to
avoid (re-)parsing entirely.
To not break IPv6 (checksums and offload) and to be able to MFC the
changes without risking to hurt 3rd party drivers, duplicate the v4
framework, as other OSes have done as well.
Introduce interface capability flags for TX/RX checksum offload with
IPv6, to allow independent toggling (where possible). Add CSUM_*_IPV6
flags for UDP/TCP over IPv6, and reserve further for SCTP, and IPv6
fragmentation. Define CSUM_DELAY_DATA_IPV6 as we do for legacy IP and
add an alias for CSUM_DATA_VALID_IPV6.
This pretty much brings IPv6 handling in line with IPv4.
TSO is still handled in a different way and not via if_hwassist.
Update ifconfig to allow (un)setting of the new capability flags.
Update loopback to announce the new capabilities and if_hwassist flags.
Individual driver updates will have to follow, as will SCTP.
Reported by: gallatin, dim, ..
Reviewed by: gallatin (glanced at?)
MFC after: 3 days
X-MFC with: r235961,235959,235958
already plan to support >64k payload here, the IPv6 header payload
length obviously is only 16 bit and the calculations need to be right.
Reported by: dim
Tested by: dim
MFC after: 1 day
X-MFC: with r235958
Use M_ZERO with malloc rather than calling bzero() ourselves.
Change if () panic() checks to KASSERT()s as they are only
catching invariants in code flow but not dependent on network
input/output.
Move initial assigments indirecting pointers after the lock
has been aquired.
Passing layer boundries, reset M_PROTOFLAGS.
Remove a NULL assignment before free.
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
Factor out Hop-By-Hop option processing. It's still not heavily used,
it reduces the footprint of ip6_input() and makes ip6_input() more
readable.
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
Defer checksum calulations on UDP6 output and respect the mbuf
flags set by NICs having done checksum validation for us already,
thus saving the computing time in the input path as well.
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
Add support for delayed checksum calculations in the IPv6
output path. We currently cannot offload to the card if we
add extension headers (which incl. fragmentation).
Fix two SCTP offload support copy&paste bugs: calculate
checksums if fragmenting and no need to flag IPv4 header
checksums in the IPv6 forwarding path.
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
Hide the ip6aux functions. The only one referenced outside ip6_input.c
is not compiled in yet (__notyet__) in route6.c (r235954). We do have
accessor functions that should be used.
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
X-MFC: KPI?
Simplify the code removing a return from an earlier else case,
not differing from the default function return called now.
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
We currently nowhere set IP6A_SWAP making the entire check useless
with the current code. Keep around but do not compile in.
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
No need to hold the (expensive) rt lock over (expensive) logging.
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
Introduce a (for now copied stripped down) in6_cksum_pseudo()
function. We should be able to use this from in6_cksum() but
we should also ponder possible MD specific improvements.
It takes an extra csum argument to allow for easy checks as
will be done by the upper layer protocol input paths.
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
Optimize in6_cksum(), re-ordering work and limiting variable
initialization, removing a bzero() for mostly re-initialized
struct values, making use of the newly introduced in6_getscope(),
as well as converting an if/panic to a KASSERT().
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
Introduce in6_getscope() to allow more effective checksum
computations without the need to copy the address to clear the
scope.
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
Reviewed by: gnn (as part of the whole)
MFC After: 3 days
casted to structs by getting rid of these buffers entirely. In r169832, it
was tried to paper over this issue by 32-bit aligning the buffers. Depending
on compiler optimizations that still was insufficient for 64-bit architectures
with strong alignment requirements though.
While at it, add comments regarding the total lack of locking in this area.
Tested by: bz
Reviewed by: bz (slightly earlier version), yongari (earlier version)
MFC after: 1 week
inp_socket. To avoid panic, do not dereference inp_socket,
but obtain reuse port option from inp_flags2, like this
is done after next call to in_pcblookup_local() a few lines
down below.
Submitted by: rwatson
call in an #if 0 section.
In in6_selecthlim() optimize a case where in6p cannot be NULL due to an
earlier check.
More consistently use u_int instead of int for fibnum function arguments.
Sponsored by: Cisco Systems, Inc.
MFC after: 3 days
at which the lle_tbl pointer points to freed memory and the llt_free pointer is no longer
valid.
Move the free pointer in to the llentry itself and update the initalization sites.
MFC after: 2 weeks
return ifnet double pointer.
Pass that hint down to in6_selectif() to be used when i) the default FIB
is queried and ii) route lookup fails because the network is not present
(i.e. someone deleted the connected subnet).
This hint should not be generally used from anywhere outside the neighbor
discovery code. We just make use of it from nd6_ns_output().
Extend the nd6_na_output() interface by a nd6_na_output_fib() version
and pass the FIB number from the NS mbuf on to NA to allow the new mbuf
to inherit the FIB tag and a later lookup from ip6_output() to succeed
in the aformentioned example case.
Provide a wrapper function for the old public interface also used from
CARP but mark it with BURN_BRIDGES to cleanup in HEAD after MFC.
Sponsored by: Cisco Systems, Inc.
the original IPv4 implementation from r178888:
- Use RT_DEFAULT_FIB in the IPv4 implementation where noticed.
- Use rt*fib() KPI with explicit RT_DEFAULT_FIB where applicable in
the NFS code.
- Use the new in6_rt* KPI in TCP, gif(4), and the IPv6 network stack
where applicable.
- Split in6_rtqtimo() and in6_mtutimo() as done in IPv4 and equally
prevent multiple initializations of callouts in in6_inithead().
- Use wrapper functions where needed to preserve the current KPI to
ease MFCs. Use BURN_BRIDGES to indicate expected future cleanup.
- Fix (related) comments (both technical or style).
- Convert to rtinit() where applicable and only use custom loops where
currently not possible otherwise.
- Multicast group, most neighbor discovery address actions and faith(4)
are locked to the default FIB. Individual IPv6 addresses will only
appear in the default FIB, however redirect information and prefixes
of connected subnets are automatically propagated to all FIBs by
default (mimicking IPv4 behavior as closely as possible).
Sponsored by: Cisco Systems, Inc.
The actual ia6->ia6_lifetime access is hidden in
IFA6_IS_INVALID/IFA6_IS_DEPRECATED macros since a long time ago
(see netinet6/nd6.c, r1.104 of KAME for the reference).
MFC after: 3 days
comments to longer, also refining strange ones.
Properly use #ifdef rather than #if defined() where possible. Four
#if defined(PCBGROUP) occurances (netinet and netinet6) were ignored to
avoid conflicts with eventually upcoming changes for RSS.
Reported by: bde (most)
Reviewed by: bde
MFC after: 3 days
If set to 1, no ABORT is sent back in response to an incoming
INIT. If set to 2, no ABORT is sent back in response to
an out of the blue packet. If set to 0 (the default), ABORTs
are sent.
Discussed with rrs@.
MFC after: 1 month.
in6m_release_locked() to defer calls to mld_v1_transmit_report() until
after the IF_ADDR_LOCK is dropped. This removes a race where the lock
is dropped and reacquired while attempting to walk an interface's
address list.
Reviewed by: bz
MFC after: 1 week
reference on a group in the leaving state while iterating over the loop.
Instead, use the same approach used in igmp_ifdetach() and mld_ifdetach()
of placing the groups to free on pending release list and then releasing
the references after dropping the IF_ADDR_LOCK. This closes an ugly race
where the code was dropping the lock in the middle of iterating over the
list. It also fixes some additional potential use-after-free bugs since
the cancellation routine also applied other changes to the group after
dropping the reference. Now those changes are performed before the
reference is dropped and the group is potentially freed.
Prodded to fix by: glebius
Reviewed by: bz
MFC after: 1 week
of the SIOC[DG]LIFADDR icotls before dropping the IF_ADDR_LOCK() and
release the reference after using it. This prevents the address from
being potentially freed out from under the ioctl handler.
Reviewed by: bz
MFC after: 1 week
from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.
The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.
ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.
To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]
The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.
Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!
PR: kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by: bz
Submitted by: bz [1]
after its installation. This removal may be accidental and can
prevent the default route from being installed in the future if
the associated default router has the best preference. The cause
is the lack of status update in the default router on the state
of its route installation in the kernel FIB. This patch fixes
the described problem.
Reviewed by: hrs, discussed with hrs
MFC after: 5 days
The SYSCTL_NODE macro defines a list that stores all child-elements of
that node. If there's no SYSCTL_DECL macro anywhere else, there's no
reason why it shouldn't be static.
inp_socket->so_options dereference when we may not acquire the lock on
the inpcb.
This fixes the crash due to NULL pointer dereference in
in_pcbbind_setup() when inp_socket->so_options in a pcb returned by
in_pcblookup_local() was checked.
Reported by: dave jones <s.dave.jones@gmail.com>, Arnaud Lacombe <lacombar@gmail.com>
Suggested by: rwatson
Glanced by: rwatson
Tested by: dave jones <s.dave.jones@gmail.com>
separation rewrite changes. r196865 was committed to fix a scope
violation problem in the following test scenario:
box-1# ifconfig em0 inet6 2001:db8:1:: prefixlen 64 anycast
box-1# ifconfig em1 inet6 2001:db8:2::1 prefixlen 64
box-2# ifconfig re0 inet6 2001:db8:1::6 prefixlen 64
em0 and re0 are on the same link.
box-2# ping6 2001:db8:1::
PING6(56=40+8+8 bytes) 2001:db8:1::6 --> 2001:db8:1::
the ICMPv6 response should have a source address of em1, which
is 2001:db8:2::1, not the link-local address of em0.
That code is no longer necessary and breaks the IPv6-Ready logo
testing, so revert it now.
Reviewed by: hrs
MFC after: 3 days
inlined by Qing Li in his big new-ARP commit. I am going to utilize
them in my newcarp work, and also these functions left declared
in in6_var.h for all the time they were absent.
Reviewed by: bz
determine if a loopback route should be installed for an interface
IPv6 address. Another condition is the address must not belong to a
looopback interface.
Reviewed by: hrs
MFC after: 3 days
appending the new mbuf to the chain reference but possibly causing an mbuf
nextpkt loop leading to a memory used after handoff (or having been freed)
and leaking an mbuf here.
Reviewed by: rwatson, brooks
MFC after: 3 days
(r225485). When setting an interface name to it, the following
configurations will be enabled:
1. "no_radr" is set to all IPv6 interfaces automatically.
2. "-no_radr accept_rtadv" will be set only for $ipv6_cpe_wanif. This is
done just before evaluating $ifconfig_IF_ipv6 in the rc.d scripts (this
means you can manually supersede this configuration if necessary).
3. The node will add RA-sending routers to the default router list
even if net.inet6.ip6.forwarding=1.
This mode is added to conform to RFC 6204 (a router which connects
the end-user network to a service provider network). To enable
packet forwarding, you still need to set ipv6_gateway_enable=YES.
Note that accepting router entries into the default router list when
packet forwarding capability and a routing daemon are enabled can
result in messing up the routing table. To minimize such unexpected
behaviors, "no_radr" is set on all interfaces but $ipv6_cpe_wanif.
Approved by: re (bz)
mld_set_version() is called only from mld_v1_input_query() and
mld_v2_input_query() both holding the if_addr_mtx lock, and then calling
into mld_v2_cancel_link_timers() acquires it the second time, which results
in mtx recursion. To avoid that, delay if_addr_mtx acquisition until after
mld_set_version() is called; while here, further reduce locking scope
to protect only the needed pieces: if_multiaddrs, in6m_lookup_locked().
PR: kern/158426
Reported by: Thomas <tps vr-web.de>,
Tom Vijlbrief <tom.vijlbrief xs4all.nl>
Tested by: Tom Vijlbrief
Reviewed by: bz
Approved by: re (kib)
Distinguish IPv4 and IPv6 addresses and optional port numbers in
user space to set the option for the correct protocol family.
Add support in the kernel for carrying the new IPv6 destination
address and port.
Add support to TCP and UDP for IPv6 and fix UDP IPv4 to not change
the address in the IP header.
Add support for IPv6 forwarding to a non-local destination.
Add a regession test uitilizing VIMAGE to check all 20 possible
combinations I could think of.
Obtained from: David Dolson at Sandvine Incorporated
(original version for ipfw fwd IPv6 support)
Sponsored by: Sandvine Incorporated
PR: bin/117214
MFC after: 4 weeks
Approved by: re (kib)
people think: returning true for an address in any connected subnet, not
necessarily on the local machine.
Sponsored by: Sandvine Incorporated
MFC after: 2 weeks
Approved by: re (kib)
* Decouple the path supervision using a separate HB timer per path.
* Add support for potentially failed state.
* Bring back RTO.min to 1 second.
* Accept packets on IP-addresses already announced via an ASCONF
* While there: do some cleanups.
Approved by: re@
MFC after: 2 months.
same as the host address. This already works fine for INET6 and ND6.
While here, remove two function pointers from struct lltable which are
only initialized but never used.
MFC after: 3 days
whether decapsulated IPsec packets will be passed to pfil again depending
on the setting of the net.ip6.ipsec6.filtertunnel sysctl.
PR: kern/157670
Submitted by: Manuel Kasper (mk neon1.net)
MFC after: 2 weeks
struct inpcbgroup. pcbgroups, or "connection groups", supplement the
existing inpcbinfo connection hash table, which when pcbgroups are
enabled, might now be thought of more usefully as a per-protocol
4-tuple reservation table.
Connections are assigned to connection groups base on a hash of their
4-tuple; wildcard sockets require special handling, and are members
of all connection groups. During a connection lookup, a
per-connection group lock is employed rather than the global pcbinfo
lock. By aligning connection groups with input path processing,
connection groups take on an effective CPU affinity, especially when
aligned with RSS work placement (see a forthcoming commit for
details). This eliminates cache line migration associated with
global, protocol-layer data structures in steady state TCP and UDP
processing (with the exception of protocol-layer statistics; further
commit to follow).
Elements of this approach were inspired by Willman, Rixner, and Cox's
2006 USENIX paper, "An Evaluation of Network Stack Parallelization
Strategies in Modern Operating Systems". However, there are also
significant differences: we maintain the inpcb lock, rather than using
the connection group lock for per-connection state.
Likewise, the focus of this implementation is alignment with NIC
packet distribution strategies such as RSS, rather than pure software
strategies. Despite that focus, software distribution is supported
through the parallel netisr implementation, and works well in
configurations where the number of hardware threads is greater than
the number of NIC input queues, such as in the RMI XLR threaded MIPS
architecture.
Another important difference is the continued maintenance of existing
hash tables as "reservation tables" -- these are useful both to
distinguish the resource allocation aspect of protocol name management
and the more common-case lookup aspect. In configurations where
connection tables are aligned with hardware hashes, it is desirable to
use the traditional lookup tables for loopback or encapsulated traffic
rather than take the expense of hardware hashes that are hard to
implement efficiently in software (such as RSS Toeplitz).
Connection group support is enabled by compiling "options PCBGROUP"
into your kernel configuration; for the time being, this is an
experimental feature, and hence is not enabled by default.
Subject to the limited MFCability of change dependencies in inpcb,
and its change to the inpcbinfo init function signature, this change
in principle could be merged to FreeBSD 8.x.
Reviewed by: bz
Sponsored by: Juniper Networks, Inc.
an explicit action for INET6 configuration happens. The changes are:
1. When an ND6 flag is changed via SIOCSIFINFO_FLAGS ioctl,
setting ND6_IFF_ACCEPT_RTADV and/or ND6_IFF_AUTO_LINKLOCAL now triggers
an attempt to clear the ND6_IFF_IFDISABLED flag.
2. When an AF_INET6 address is added successfully to an interface and
it is marked as ND6_IFF_IFDISABLED, an attempt to clear the
ND6_IFF_IFDISABLED happens.
This simplifies ND6_IFF_IFDISABLED flag manipulation by users via ifconfig(8);
in most cases manual configuration is no longer needed.
- When ND6_IFF_AUTO_LINKLOCAL is set and no link-local address is assigned to
an interface, SIOCSIFINFO_FLAGS ioctl now calls in6_ifattach() to configure
a link-local address.
This change ensures link-local address configuration when "ifconfig IF inet6"
command is invoked. For example, "ifconfig IF inet6 auto_linklocal" now
always try to configure an LL addr even if ND6_IFF_AUTO_LINKLOCAL is already
set to 1 (i.e. down/up cycle is no longer needed).
Reviewed by: bz
- A new per-interface knob IFF_ND6_NO_RADR and sysctl IPV6CTL_NO_RADR.
This controls if accepting a route in an RA message as the default route.
The default value for each interface can be set by net.inet6.ip6.no_radr.
The system wide default value is 0.
- A new sysctl: net.inet6.ip6.norbit_raif. This controls if setting R-bit in
NA on RA accepting interfaces. The default is 0 (R-bit is set based on
net.inet6.ip6.forwarding).
Background:
IPv6 host/router model suggests a router sends an RA and a host accepts it for
router discovery. Because of that, KAME implementation does not allow
accepting RAs when net.inet6.ip6.forwarding=1. Accepting RAs on a router can
make the routing table confused since it can change the default router
unintentionally.
However, in practice there are cases where we cannot distinguish a host from
a router clearly. For example, a customer edge router often works as a host
against the ISP, and as a router against the LAN at the same time. Another
example is a complex network configurations like an L2TP tunnel for IPv6
connection to Internet over an Ethernet link with another native IPv6 subnet.
In this case, the physical interface for the native IPv6 subnet works as a
host, and the pseudo-interface for L2TP works as the default IP forwarding
route.
Problem:
Disabling processing RA messages when net.inet6.ip6.forwarding=1 and
accepting them when net.inet6.ip6.forward=0 cause the following practical
issues:
- A router cannot perform SLAAC. It becomes a problem if a box has
multiple interfaces and you want to use SLAAC on some of them, for
example. A customer edge router for IPv6 Internet access service
using an IPv6-over-IPv6 tunnel sometimes needs SLAAC on the
physical interface for administration purpose; updating firmware
and so on (link-local addresses can be used there, but GUAs by
SLAAC are often used for scalability).
- When a host has multiple IPv6 interfaces and it receives multiple RAs on
them, controlling the default route is difficult. Router preferences
defined in RFC 4191 works only when the routers on the links are
under your control.
Details of Implementation Changes:
Router Advertisement messages will be accepted even when
net.inet6.ip6.forwarding=1. More precisely, the conditions are as
follow:
(ACCEPT_RTADV && !NO_RADR && !ip6.forwarding)
=> Normal RA processing on that interface. (as IPv6 host)
(ACCEPT_RTADV && (NO_RADR || ip6.forwarding))
=> Accept RA but add the router to the defroute list with
rtlifetime=0 unconditionally. This effectively prevents
from setting the received router address as the box's
default route.
(!ACCEPT_RTADV)
=> No RA processing on that interface.
ACCEPT_RTADV and NO_RADR are per-interface knob. In short, all interface
are classified as "RA-accepting" or not. An RA-accepting interface always
processes RA messages regardless of ip6.forwarding. The difference caused by
NO_RADR or ip6.forwarding is whether the RA source address is considered as
the default router or not.
R-bit in NA on the RA accepting interfaces is set based on
net.inet6.ip6.forwarding. While RFC 6204 W-1 rule (for CPE case) suggests
a router should disable the R-bit completely even when the box has
net.inet6.ip6.forwarding=1, I believe there is no technical reason with
doing so. This behavior can be set by a new sysctl net.inet6.ip6.norbit_raif
(the default is 0).
Usage:
# ifconfig fxp0 inet6 accept_rtadv
=> accept RA on fxp0
# ifconfig fxp0 inet6 accept_rtadv no_radr
=> accept RA on fxp0 but ignore default route information in it.
# sysctl net.inet6.ip6.norbit_no_radr=1
=> R-bit in NAs on RA accepting interfaces will always be set to 0.
hash install, etc. For now, these are arguments are unused, but as we add
RSS support, we will want to use hashes extracted from mbufs, rather than
manually calculated hashes of header fields, due to the expensive of the
software version of Toeplitz (and similar hashes).
Add notes that it would be nice to be able to pass mbufs into lookup
routines in pf(4), optimising firewall lookup in the same way, but the
code structure there doesn't facilitate that currently.
(In principle there is no reason this couldn't be MFCed -- the change
extends rather than modifies the KBI. However, it won't be useful without
other previous possibly less MFCable changes.)
Reviewed by: bz
Sponsored by: Juniper Networks, Inc.
- The existing ipi_lock continues to protect the global inpcb list and
inpcb counter. This lock is now relegated to a small number of
allocation and free operations, and occasional operations that walk
all connections (including, awkwardly, certain UDP multicast receive
operations -- something to revisit).
- A new ipi_hash_lock protects the two inpcbinfo hash tables for
looking up connections and bound sockets, manipulated using new
INP_HASH_*() macros. This lock, combined with inpcb locks, protects
the 4-tuple address space.
Unlike the current ipi_lock, ipi_hash_lock follows the individual inpcb
connection locks, so may be acquired while manipulating a connection on
which a lock is already held, avoiding the need to acquire the inpcbinfo
lock preemptively when a binding change might later be required. As a
result, however, lookup operations necessarily go through a reference
acquire while holding the lookup lock, later acquiring an inpcb lock --
if required.
A new function in_pcblookup() looks up connections, and accepts flags
indicating how to return the inpcb. Due to lock order changes, callers
no longer need acquire locks before performing a lookup: the lookup
routine will acquire the ipi_hash_lock as needed. In the future, it will
also be able to use alternative lookup and locking strategies
transparently to callers, such as pcbgroup lookup. New lookup flags are,
supplementing the existing INPLOOKUP_WILDCARD flag:
INPLOOKUP_RLOCKPCB - Acquire a read lock on the returned inpcb
INPLOOKUP_WLOCKPCB - Acquire a write lock on the returned inpcb
Callers must pass exactly one of these flags (for the time being).
Some notes:
- All protocols are updated to work within the new regime; especially,
TCP, UDPv4, and UDPv6. pcbinfo ipi_lock acquisitions are largely
eliminated, and global hash lock hold times are dramatically reduced
compared to previous locking.
- The TCP syncache still relies on the pcbinfo lock, something that we
may want to revisit.
- Support for reverting to the FreeBSD 7.x locking strategy in TCP input
is no longer available -- hash lookup locks are now held only very
briefly during inpcb lookup, rather than for potentially extended
periods. However, the pcbinfo ipi_lock will still be acquired if a
connection state might change such that a connection is added or
removed.
- Raw IP sockets continue to use the pcbinfo ipi_lock for protection,
due to maintaining their own hash tables.
- The interface in6_pcblookup_hash_locked() is maintained, which allows
callers to acquire hash locks and perform one or more lookups atomically
with 4-tuple allocation: this is required only for TCPv6, as there is no
in6_pcbconnect_setup(), which there should be.
- UDPv6 locking remains significantly more conservative than UDPv4
locking, which relates to source address selection. This needs
attention, as it likely significantly reduces parallelism in this code
for multithreaded socket use (such as in BIND).
- In the UDPv4 and UDPv6 multicast cases, we need to revisit locking
somewhat, as they relied on ipi_lock to stablise 4-tuple matches, which
is no longer sufficient. A second check once the inpcb lock is held
should do the trick, keeping the general case from requiring the inpcb
lock for every inpcb visited.
- This work reminds us that we need to revisit locking of the v4/v6 flags,
which may be accessed lock-free both before and after this change.
- Right now, a single lock name is used for the pcbhash lock -- this is
undesirable, and probably another argument is required to take care of
this (or a char array name field in the pcbinfo?).
This is not an MFC candidate for 8.x due to its impact on lookup and
locking semantics. It's possible some of these issues could be worked
around with compatibility wrappers, if necessary.
Reviewed by: bz
Sponsored by: Juniper Networks, Inc.
feature_present(3) to dynamically decide whether to use one or the
other family.
Reviewed by: gnn
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
MFC after: 10 days
in_pcb_lport(), in_pcblookup_local(), and in_pcblookup_hash(), and similarly
for IPv6 functions. In the future, we would like to support other flags
relating to locking strategy.
This change doesn't appear to modify the KBI in practice, as callers already
passed in INPLOOKUP_WILDCARD rather than a simple boolean.
MFC after: 3 weeks
Reviewed by: bz
Sponsored by: Juniper Networks, Inc.
interface is brought down, even though the interface address is still
valid. This patch maintains the permanent ARP entries as long as the
interface address (having the same prefix as that of the ARP entries)
is valid.
Reviewed by: delphij
MFC after: 5 days
Some bugs where fixed while doing this:
* ASCONF-ACK messages might use wrong port number when using
IPv6.
* Checking for additional addresses takes the correct address
into account and also does not do more comparisons than
necessary.
This patch is based on one received from bz@ who was
sponsored by The FreeBSD Foundation and iXsystems.
MFC after: 1 week
as well compiling out most functions adding or extending #ifdef INET
coverage.
Reviewed by: gnn
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
MFC after: 4 days
Unfold the IPSEC_COMMON_INPUT_CB() macro in xform_{ah,esp,ipcomp}.c
to not need three different versions depending on INET, INET6 or both.
Mark two places preparing for not yet supported functionality with IPv6.
Reviewed by: gnn
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
MFC after: 4 days
Not compiling in and not initializing from inetsw from in_proto.c for
IPv6 only, we need to initialize upper layer protocols from inet6sw.
Make sure to not initialize them twice in a Dual-Stack
environment but only conditionally on no INET as we have done for
TCP for a long time. Otherwise we would leak resources.
Reviewed by: gnn
Sponsored by: The FreeBSD Foundation
Sponsored by: iXsystems
MFC after: 3 days
passing the cached proxydl reference (sockaddr_dl initialized or not) to
nd6_na_output(). nd6_na_output() will thus assume a proxy NA. Revert to
conditionally passing either &proxydl or NULL if no proxy case desired.
Tested by: ipv6gw and ref9-i386
Reported by: Pete French (petefrench ingresso.co.uk on stable)
Reported by: bz, simon on Y! cluster
Reported by: kib
PR: kern/151908
MFC after: 3 days
working. We store v4 and v6 addresses as a union but for v4-mapped
addresses only store the 32bits w/o the ::ffff: word. That failed the
check as for example 127.0.0.1 would be ::7f00:1 rather than ::ffff:7f00:1
and the IN6_IS_ADDR_V4MAPPED() never worked here. Given we can hardly get
here with an unbound local address or invalid inp_vflags remove the check.
Reported by: tuexen
Reviewed by: tuexen
MFC after: 3 days
some more. Similar to what we do for TCP check for v4-mapped
addresses and then handle them or the normal v6 address case.
For either set inp_vflags before calling into the pcb connect
function so that we have an unambiguous view in case we need to
set the local address or port.
Looked at: tuexen (as part of more)
MFC after: 3 days
callers. This also fixes a problem when the prison call could set
the inp->in6p_laddr (laddr) and a following priv_check_cred() call
would return an error and will allow us to merge the IPv4 and IPv6
implementation.
MFC after: 2 weeks
even after dropping the reference and unlocking. Previously we
have dereferenced a NULL pointer (after r121765).
Simply unlocking after the block does not work either because of
lock ordering (see r121765) and in addition we would still hold
a pointer to something that might be gone by the time we access it.
Thus take a copy of the value rather than just caching the pointer.
PR: kern/151908
Submitted by: chenyl (netstar2008 126.com) (initial version)
MFC after: 2 weeks
* Store the flowid when receiving an SCTP/IPv6 packet.
* Store the flowid when receiving an SCTP packet with wrong CRC.
* Initilize flowid correctly.
* Put test code under INVARIANTS.
MFC after: 3 months.
Call the handler function with the lock held, return unlocked as we
might free the entry. Rework functions later in the call graph to be
either called with the lock held or, only if needed, unlocked.
Place asserts to document and tighten assumptions on various lle locking,
which were not always true before.
We call nd6_ns_output() unlocked and the assignment of ip6->ip6_src was
decentralized to minimize possible complexity introduced with the formerly
missing locking there. This also resulted in a push down of local
variable scopes into smaller blocks.
Reported by: many
PR: kern/148857
Submitted by: Dmitrij Tejblum (tejblum yandex-team.ru) (original version)
MFC After: 4 days
DPCPU_DEFINE and VNET_DEFINE macros, as these cause problems for various
people working on the affected files. A better long-term solution is
still being considered. This reversal may give some modules empty
set_pcpu or set_vnet sections, but these are harmless.
Changes reverted:
------------------------------------------------------------------------
r215318 | dim | 2010-11-14 21:40:55 +0100 (Sun, 14 Nov 2010) | 4 lines
Instead of unconditionally emitting .globl's for the __start_set_xxx and
__stop_set_xxx symbols, only emit them when the set_vnet or set_pcpu
sections are actually defined.
------------------------------------------------------------------------
r215317 | dim | 2010-11-14 21:38:11 +0100 (Sun, 14 Nov 2010) | 3 lines
Apply the STATIC_VNET_DEFINE and STATIC_DPCPU_DEFINE macros throughout
the tree.
------------------------------------------------------------------------
r215316 | dim | 2010-11-14 21:23:02 +0100 (Sun, 14 Nov 2010) | 2 lines
Add macros to define static instances of VNET_DEFINE and DPCPU_DEFINE.
Consistently use the LLE_ prefix for lla_lookup() and the ND6_ prefix
for nd6_lookup() even though both are defined the same. Use the right
flag variable when checking each.
No real functional change.
MFC after: 4 days
