Dmitrij Tejblum [Thu, 18 Aug 2011 16:22:17 +0000 (20:22 +0400)]
lib: simplify interface of setsockopt_multicast_ipv4().
* sockopt.[ch] (setsockopt_ipv4_multicast): ifindex is now mandatory (all
non-ancient OSes can use it anyway), and if_addr parameter (the address
of the interface) is now gone. (setsockopt_ipv4_multicast_if):
IP_MULTICAST_IF processing moved to this new function
* ospf_network.c (ospf_if_add_allspfrouters, ospf_if_drop_allspfrouters,
ospf_if_add_alldrouters, ospf_if_drop_alldrouters, ospf_if_ipmulticast),
rip_interface.c (ipv4_multicast_join, ipv4_multicast_leave,
rip_interface_new): adapt to the new interface
* bgp_attr.c
* bgp_attr_atomic(): accept extra argument, add checks for
"optional", "transitive" and "partial" bits, log each error
condition independently
* bgp_attr_parse(): provide extra argument
* bgp_attr.c
* bgp_attr_local_pref(): accept extra argument, add checks for
"optional" and "transitive" bits, log each error condition
independently
* bgp_attr_parse(): provide extra argument
This change is based on Xavier Beaudouin's patch (which fixes detection
of 3 config.h macros on FreeBSD without any impact to Linux build of
Quagga) and FreeBSD port patch (which fixes 5 config.h macros, but
breaks the Linux build), it fixes 5 macros and works for both FreeBSD 8
and Linux.
Fritz Reichmann [Wed, 14 Sep 2011 15:31:51 +0000 (19:31 +0400)]
isisd: fix crash on "no router isis" (BZ#536)
The crash is due to threads accessing data that gets destroyed
during the removal of the configuration.
* isis_circuit.c: Destroy adjacencies to stop adjacency expiry thread.
Stop PSNP threads.
* isisd.c: Change state of circuit back to INIT and reassign the
circuit structure to isis->init_circ_list rather than destroying
the circuit data structure. Stop SPF threads. Stop LSP generation
threads.
* isisd.h: Add pointers to LSP threads into area structure in order to
stop them in isisd.c
* isis_lsp.c: Store pointer to LSP thread in area structure.
* isis_pdu.c: Stop PDU generation for a circuit with a removed area.
* isis_pfpacket.c: Stop processing received PDUs for a circuit with a
removed area.
* ospf_route.c: Function ospf_asbr_route_cmp is called uniquely from
ospf_route_cmp() when the flag OSPF_RFC1583_COMPATIBLE is not set.
Therefore, the check that the flag is set doesn't make sense at all
and it can consequently be removed without doing any harm.
Signed-off-by: Alexandre Chappuis <alc@open.ch> Signed-off-by: Roman Hoog Antink <rha@open.ch>
This vulnerability (CERT-FI #514840) was reported by CROSS project.
ospf6d processes IPv6 prefix structures in incoming packets without
verifying that the declared prefix length is valid. This leads to a
crash
caused by out of bounds memory access.
* ospf6_abr.h: new macros for size/alignment validation
* ospf6_asbr.h: idem
* ospf6_intra.h: idem
* ospf6_lsa.h: idem
* ospf6_message.h: idem
* ospf6_proto.h: idem
* ospf6_message.c
* ospf6_packet_minlen: helper array for ospf6_packet_examin()
* ospf6_lsa_minlen: helper array for ospf6_lsa_examin()
* ospf6_hello_recv(): do not call ospf6_header_examin(), let upper
layer verify the input data
* ospf6_dbdesc_recv(): idem
* ospf6_lsreq_recv(): idem
* ospf6_lsupdate_recv(): idem
* ospf6_lsack_recv(): idem
* ospf6_prefixes_examin(): new function, implements A.4.1
* ospf6_lsa_examin(): new function, implements A.4
* ospf6_lsaseq_examin(): new function, an interface to above
* ospf6_packet_examin(): new function, implements A.3
* ospf6_rxpacket_examin(): new function, replaces
ospf6_header_examin()
* ospf6_header_examin(): sayonara
* ospf6_receive(): perform passive interface check earliest possible,
employ ospf6_rxpacket_examin()
This vulnerability (CERT-FI #514839) was reported by CROSS project.
When Database Description LSA header list contains trailing zero octets,
ospf6d tries to process this data as an LSA header. This triggers an
assertion in the code and ospf6d shuts down.
* ospf6_lsa.c
* ospf6_lsa_is_changed(): handle header-only argument(s)
appropriately, do not treat LSA length underrun as a fatal error.
This vulnerability (CERT-FI #514838) was reported by CROSS project.
The error is reproducible only when ospfd debugging is enabled:
* debug ospf packet all
* debug ospf zebra
When incoming packet header type field is set to 0x0a, ospfd will crash.
* ospf_packet.c
* ospf_verify_header(): add type field check
* ospf_read(): perform input checks early
This vulnerability (CERT-FI #514837) was reported by CROSS project.
They have also suggested a fix to the problem, which was found
acceptable.
Quagga ospfd does not seem to handle unknown LSA types in a Link State
Update message correctly. If LSA type is something else than one
supported
by Quagga, the default handling of unknown types leads to an error.
* ospf_flood.c
* ospf_flood(): check return value of ospf_lsa_install()
This vulnerability (CERT-FI #513254) was reported by CROSS project.
They have also suggested a fix to the problem, which was found
acceptable.
The problem occurs when bgpd receives an UPDATE message containing
255 unknown AS_PATH attributes in Path Attribute Extended Communities.
This causes a buffer overlow in bgpd.
Denis Ovsienko [Sat, 20 Aug 2011 19:12:00 +0000 (23:12 +0400)]
ospf6d: add missing include
Two extern declarations in ospf6_abr.h are based on struct ospf6_route,
which may not be available at the time ospf6_abr.h is included. This may
lead to warnings after including ospf6_abr.h just for the structures
defined in it.
Paul Jakma [Fri, 29 Jul 2011 17:16:25 +0000 (18:16 +0100)]
Merge branch 'attr-errors'
Contains BGP fixes:
- set extcommunity crash: tihs patch tries to make the refcounting more robust
but does not fully solve the problem, sadly.
- BGP attribute error handling: Little testing.
"mtu-ignore" is an option ospfd used to mimic from the vendor's
implementation, now ospf6d will also implement it.
* ospf6_interface.h: extend ospf6_interface structure by one flag
* ospf6_interface.c: (ipv6_ospf6_mtu_ignore, no_ipv6_ospf6_mtu_ignore):
new declarations; (ospf6_interface_create): show initial value for
consistency; (ospf6_interface_show): print flag status
* ospf6_message.c: (ospf6_dbdesc_recv): consider interface-specific flag
when checking MTU
Paul Jakma [Mon, 11 Apr 2011 15:31:43 +0000 (16:31 +0100)]
bgpd: trivial format / warning fixes
* bgp_aspath.c: (assegments_parse) just bail early if length doesn't match
and fix the formatting.
* bgp_network.c: add include needed for set_nonblocking
* bgp_packet.c: formatting
Michael Lambert [Mon, 13 Sep 2010 15:48:11 +0000 (11:48 -0400)]
bgpd: fix memory leak when deleting a community
* bgp_routemap.c: (route_set_community_delete) When deleting a
community in a route-map the old community was being orphaned. Like
the description of the same code in route_set_community, this is a
hack, not a true fix.
Paul Jakma [Tue, 29 Mar 2011 13:18:49 +0000 (14:18 +0100)]
bgpd: Fix merge error in jhash commit
* bgp_attr.c: (attrhash_key_make) 98e30f should have changed jhash2 to jhash.
These kinds of merge errors would be reduced and life would be easier if
people would submit fully-formed fixes that could be chucked directly into
git-am.
* bgpd.h: Add error code for setting GTSM on iBGP
* bgpd.c: (peer_ttl_security_hops_set) use previous error code and signal
incompatibility of GTSM+iBGP to vty.
Consider the session state when setting GTSM, and reset Open/Active peers
to let them pick up new TTL from start.
Greg Troxel [Thu, 24 Mar 2011 14:19:43 +0000 (10:19 -0400)]
build: treat opaque-lsa default correctly for ospfapi
In deciding whether to default ospfapi to on or off, use the same rule
for opaque-lsa as earlier: != no rather than = yes, so that not having
set it implies yes.
* bgp_vty.c: (peer_ebgp_multihop_{un,}set_vty) tail-call cleanup.
({no_,}neighbor_ttl_security) ditto.
* bgpd.c: (peer_ttl_security_hops_set) Peer group checks and TTL set only
need to be done on transition.
* sockunion.c: (sockopt_minttl) remove always-on debug and improve readability.
Nick Hilliard [Wed, 23 Mar 2011 15:33:17 +0000 (15:33 +0000)]
bgpd: RFC 5082 Generalized TTL Security Mechanism support
* bgpd: Add support for RFC 5082 GTSM, which allows the TTL field to be used
to verify that incoming packets have been sent from neighbours no more
than X IP hops away. In other words, this allows packets that were sent from
further away (i.e. not by the neighbour with known distance, and so possibly
a miscreant) to be filtered out.
* lib/sockunion.{c,h}: (sockopt_minttl) new function, to set a minimum TTL
using the IP_MINTTL socket opt.
* bgpd.h: (BGP_ERR_NO_EBGP_MULTIHOP_WITH_TTLHACK) define for command
error for minttl.
(struct peer) add a config variable, to store the configured minttl.
(peer_ttl_security_hops_{set,unset}) configuration handlers
* bgpd.c: (peer_group_get) init gtsm_hops
(peer_ebgp_multihop_{un,}set) check for conflicts with GTSM. Multihop and
GTSM can't both be active for a peer at the same time.
(peer_ttl_security_hops_set) set minttl, taking care to avoid conflicts with
ebgp_multihop.
(bgp_config_write_peer) write out minttl as "neighbor .. ttl-security hops X".
* bgp_vty.c: (bgp_vty_return) message for
BGP_ERR_NO_EBGP_MULTIHOP_WITH_TTLHACK
(peer_ebgp_multihop_{un,}set_vty)
* bgp_network.c: (bgp_accept) set minttl on accepted sockets if appropriate.
(bgp_connect) ditto for outbound.
Paul Jakma [Wed, 23 Mar 2011 10:30:30 +0000 (10:30 +0000)]
bgpd: Fix compile failure if IPv6 build was disabled.
* bgp_route.c: ({no_,}ipv6_bgp_network_ttl_cmd) depends on ipv6_bgp_network
which is HAVE_IPV6, so these should be too.
(bgp_route_init) and the installs should be similarly ifdefed
bgpd: attribute jhash call should use a standard interface to in6_addr data
* bgp_attr.c; (attrhash_key_make) s6_addr is only member of in6_addr
guaranteed to be available - s6_addr32 isn't. Fix to be more portable, and
thus allow compilation on BSD again.
Paul Jakma [Tue, 22 Mar 2011 15:23:55 +0000 (15:23 +0000)]
ospfd: Compile fix for opaque support
* ospfd: Refresher logic cleanup broke OSPF opaque, which does its own thing
with regard to refresher logic and which also, in the protocol, requires
implementations to keep state of which OI an LSA is received on (rather
than providing information in the LSA to allow it to be looked up - as
other LSAs requiring such assocation were careful to do).
* ospf_lsa.h: (struct ospf_interface) Add back the pointer to oi, but only
for type-9 now.
* ospf_nsm.c: (ospf_db_summary_add) check the oi actually exists first -
doesn't obviate the need for opaque to ensure oi pointers get cleaned up
when ospf_interfaces disappear.
* ospf_opaque.{c,h}: (ospf_opaque_functab,ospf_opaque_lsa_refresh) Refresher
LSA functions now need to return the LSA to the general refresh logic,
to indicate whether the LSA was refreshed.
Paul Jakma [Tue, 22 Mar 2011 10:18:05 +0000 (10:18 +0000)]
build: change sense of opaque-{lsa,te} enable args to enable by default
* configure.ac: (AC_ARG_ENABLE({ospf-te,opaque-lsa})) reverse the sense to
--disable
(enable_{opaque_lsa,ospf_te}) treat as enabled unless explicitly disabled.
* bgp_packet.c: (bgp_write) On BGP write, use TCP_CORK to provide hints to
kernel about TCP buffering. This will cause BGP packets to occur in
bigger chunks (full size MTU), improving performance and getting rid of
one of the problems reported in the UNH BGP conformance test.
Michael Lambert [Sat, 24 Jul 2010 16:44:07 +0000 (12:44 -0400)]
bgpd: "Intern" communities in route maps
* bgp_community.[ch]: (community_lookup) New helper function to look
up a community list in the hash table.
* bgp_routemap.c: A new community structure was being allocated for
every BGP update which matched a route map which set a community.
This behavior led to rapid growth in the memory consumed by bgpd.
Adding the communities to the hash table addresses the memory
growth, but may introduce a problem in modifying or deleting the
'set community' statement in the route map.
Michael Lambert [Fri, 23 Jul 2010 18:43:04 +0000 (14:43 -0400)]
bgpd: New show commands for improved view and address family support
Many show commands do not have support for multiple views and do not
treat different address families uniformly. The following changes add
a number of commands with support for views and rationalized treatment
of IPv4 v IPv6 and unicast v multicast (such as in JUNOS, IOS XR and
more recent versions of IOS).
* bgp_route.c: (bgp_show_community) Inserted a new second argument (the
name of the view) and the code to look up that name in the BGP structure.
The NULL argument in the call to bgp_show (indicating the default view)
was replaced by the specified view. The existing calls to
bgp_show_community had a NULL second argument inserted to make clear
that they refer to the default view.
(top level) Added new functions via the DEFUN and/or ALIAS macros (and
the associated command table entries) to add the commands
show bgp ipv4 (unicast|multicast)
show bgp ipv4 (unicast|multicast) A.B.C.D
show bgp ipv4 (unicast|multicast) A.B.C.D/M
show bgp ipv6 (unicast|multicast)
show bgp ipv6 (unicast|multicast) X:X::X:X
show bgp ipv6 (unicast|multicast) X:X::X:X/M
These show either the full BGP table or the specified route or
prefix for the given address family.
show bgp view WORD (ipv4|ipv6) (unicast|multicast) community
show bgp view WORD (ipv4|ipv6) (unicast|multicast) community \
(AA:NN|local-AS|no-advertise|no-export){1,4}
For the specified view and address family, these show entries
matching any community or the specified communit(y)(ies).
show bgp view WORD (ipv4|ipv6) (unicast|multicast) neighbors \
(A.B.C.D|X:X::X:X) (advertised-routes|received-routes)
For the specified view and address family, show the routes
advertised to or received from the given BGP neighbor.
show bgp [view WORD] ipv4 (unicast|multicast) \
rsclient (A.B.C.D|X:X::X:X)
show bgp [view WORD] ipv4 (unicast|multicast) \
rsclient (A.B.C.D|X:X::X:X) A.B.C.D
show bgp [view WORD] ipv4 (unicast|multicast) \
rsclient (A.B.C.D|X:X::X:X) A.B.C.D/M
show bgp [view WORD] ipv6 (unicast|multicast) \
rsclient (A.B.C.D|X:X::X:X)
show bgp [view WORD] ipv6 (unicast|multicast) \
rsclient (A.B.C.D|X:X::X:X) X:X::X:X
show bgp [view WORD] ipv6 (unicast|multicast) \
rsclient (A.B.C.D|X:X::X:X) X:X::X:X/M
For the specifed (optional) view and address family, show either
the full BGP table or the specified route or prefix for the given
route server client peer.
* bgp_vty.c: (top level) Added new functions via the DEFUN and/or ALIAS
macros (and the associated command table entries) to add the commands
show bgp [view WORD] (ipv4|ipv6) (unicast|multicast) summary
show bgp [view WORD] (ipv4|ipv6) (unicast|multicast) rsclient summary
For the specified (optional) view and address family, display
either the normal summary table for BGP peers, or the route server
client table showing the import and export policies.
Paul Jakma [Tue, 23 Nov 2010 16:35:42 +0000 (16:35 +0000)]
bgpd: Implement revised error handling for partial optional/trans. attributes
* BGP error handling generally boils down to "reset session". This was fine
when all BGP speakers pretty much understood all BGP messages. However
the increasing deployment of new attribute types has shown this approach
to cause problems, in particular where a new attribute type is "tunneled"
over some speakers which do not understand it, and then arrives at a speaker
which does but considers it malformed (e.g. corruption along the way, or
because of early implementation bugs/interop issues).
To mitigate this drafts before the IDR (likely to be adopted) propose to
treat errors in partial (i.e. not understood by neighbour), optional
transitive attributes, when received from eBGP peers, as withdrawing only
the NLRIs in the affected UPDATE, rather than causing the entire session
to be reset. See:
* bgp_aspath.c: (assegments_parse) Replace the "NULL means valid, 0-length
OR an error" return value with an error code - instead taking
pointer to result structure as arg.
(aspath_parse) adjust to suit previous change, but here NULL really
does mean error in the external interface.
* bgp_attr.h (bgp_attr_parse) use an explictly typed and enumerated
value to indicate return result.
(bgp_attr_unintern_sub) cleans up just the members of an attr, but not the
attr itself, for benefit of those who use a stack-local attr.
* bgp_attr.c: (bgp_attr_unintern_sub) split out from bgp_attr_unintern
(bgp_attr_unintern) as previous.
(bgp_attr_malformed) helper function to centralise decisions on how to
handle errors in attributes.
(bgp_attr_{aspathlimit,origin,etc..}) Use bgp_attr_malformed.
(bgp_attr_aspathlimit) Subcode for error specifc to this attr should be
BGP_NOTIFY_UPDATE_OPT_ATTR_ERR.
(bgp_attr_as4_path) be more rigorous about checks, ala bgp_attr_as_path.
(bgp_attr_parse) Adjust to deal with the additional error level that
bgp_attr_ parsers can raise, and also similarly return appropriate
error back up to (bgp_update_receive). Try to avoid leaking as4_path.
* bgp_packet.c: (bgp_update_receive) Adjust to deal with BGP_ATTR_PARSE_WITHDRAW
error level from bgp_attr_parse, which should lead to a withdraw, by
making the attribute parameter in call to (bgp_nlri_parse) conditional
on the error, so the update case morphs also into a withdraw.
Use bgp_attr_unintern_sub from above, instead of doing this itself.
Fix error case returns which were not calling bgp_attr_unintern_sub
and probably leaking memory.
* tests/aspath_test.c: Fix to work for null return with bad segments
Paul Jakma [Sun, 5 Dec 2010 20:28:02 +0000 (20:28 +0000)]
bgpd: Remove AS Path limit/TTL functionality
* draft-ietf-idr-as-pathlimit doesn't seem to have gone anywhere, and its
author does not think it will make progress in IDR. Remove all support
introduced for it, but leave stubs for the commands to avoid breaking
any configurations.
* Extended communities has some kind of resource allocation problem which
causes a double-free if the 'set extcommunity ...' command is used.
Try fix by properly interning extcommunities.
Also, more generally, make unintern functions take a double pointer
so they can NULL out callers references - a usefully defensive programming
pattern for functions which make refs invalid.
Sadly, this patch doesn't fix the problem entirely - crashes still
occur on session clear.
* bgp_ecommunity.h: (ecommunity_{free,unintern}) take double pointer
args.
* bgp_community.h: (community_unintern) ditto
* bgp_attr.h: (bgp_attr_intern) ditto
* bgp_aspath.h: (bgp_aspath.h) ditto
* (general) update all callers of above
* bgp_routemap.c: (route_set_ecommunity_{rt,soo}) intern the new extcom added
to the attr, and unintern any old one.
(route_set_ecommunity_{rt,soo}_compile) intern the extcom to be used
for the route-map set.
(route_set_ecommunity_*_free) unintern to match, instead of free
(route_set_ecommunity_soo) Do as _rt does and don't just leak
any pre-existing community, add to it (is additive right though?)
Paul Jakma [Sat, 27 Nov 2010 23:14:02 +0000 (23:14 +0000)]
tests: Extend aspath_test.c with cases for invalid segments & attributes
* aspath_test.c: Add more test cases. In particular ones to cover the
last invalid-segment problem. Also add ability to specify aspath attribute
headers and test them somewhat.
NB: It's obvious this test has not been run for a year by anyone, despite
2 non-trivial commits to bgpd aspath code.
Paul Jakma [Sat, 27 Nov 2010 22:48:34 +0000 (22:48 +0000)]
bgpd: Rollback some of the changes made for invalid AS_PATH segment fix
Some of the changes made in commit cddb8112b80fa9867156c637d63e6e79eeac67bb
don't work particularly well for other changes that need to be made to
address BGP attribute error handling problems. In particular, returning
a pointer from complex attribute data parsing functions will not suffice
to express the require range of return status conditions.
* bgp_aspath.c: (assegments_parse) Rollback to a more minimal set of
changes to fix the original problem.
(aspath_parse) Slightly needless pushing around of code, and taking
2 parameters to say whether ot use 2 or 4 byte encoding seems unnecessary.
* bgp_attr.c: (bgp_attr_as{,4}path) Rollback, in preparation for BGP
attribute error handling update.
Paul Jakma [Sun, 5 Dec 2010 17:17:26 +0000 (17:17 +0000)]
bgpd/security: CVE-2010-1674 Fix crash due to extended-community parser error
* bgp_attr.c: (bgp_attr_ext_communities) Certain extended-community attrs
can leave attr->flag indicating ext-community is present, even though no
extended-community object has been attached to the attr structure. Thus a
null-pointer dereference can occur later.
(bgp_attr_community) No bug fixed here, but tidy up flow so it has same
form as previous.
Tom Goff [Wed, 10 Nov 2010 21:02:11 +0000 (13:02 -0800)]
ospf6d: Extend the "[no] debug ospf6 route" vty commands
* ospf6_route.c ([no_]debug_ospf6_route) Include memory as a debug
option. This allows ospf6 route memory debugging to be enabled or
disabled interactively or from a config file.
Tom Goff [Wed, 10 Nov 2010 21:03:02 +0000 (13:03 -0800)]
ospf6d: Route locking (memory) cleanup
* ospf6_route.c: (ospf6_route_best_next) Allows unlock route, even
when there's no next route. This is consistent with how
ospf6_route_next() behaves.
* ospf6_intra.c: (ospf6_intra_prefix_lsa_remove) Make sure the last
route considered is always unlocked. This is needed when the for
loop terminates because ospf6_route_is_prefix() returns zero.
Tom Goff [Wed, 10 Nov 2010 21:01:41 +0000 (13:01 -0800)]
ospf6d: Have ospf6d cleanup when it terminates normally
A clean exit makes it easier to use memory debuggers.
* ospf6_asbr.c: (ospf6_asbr_terminate) Add a function to do route map
cleanup.
* ospf6_lsa.c: (ospf6_lsa_terminate) Add a function to cleanup the lsa
handler vector.
* ospf6_main.c: (ospf6_exit) Add an function that causes ospf6d to
gracefully exit.
* ospf6_message.c: (ospf6_message_terminate) Add a function that frees
the send and receive buffers.
* ospf6_top.c: (ospf6_delete) Enable the ospf6_delete() function.
Disable ospf6 before freeing everything.
Tom Goff [Wed, 10 Nov 2010 21:02:38 +0000 (13:02 -0800)]
ospf6d: Fix memory allocation issues in SPF
* ospf6_area.c: Call ospf6_spf_table_finish() before deleting the spf
table. This ensures that the associated ospf6_vertex structures
are also freed.
* ospf6_spf.c: Only allocate a priority queue when a spf calculation
is actually performed. Also defer calling ospf6_spf_table_finish().
bgpd: use Jenkins hash for BGP transit, cluster and attr hashes
* bgp_attr.c: I observed while doing some debugging that even for simple
tests there was a lot of hash collisions for BGP attributes. Switch to
using Jhash rather than additive hashing. Probably overkill, but the
function is fast and available.
({attrhash,cluster,transit}_hask_key_make) convert to Jenkins hash,
instead of additive hash.
lib: Better hashing of string values using Bernstein hash
* hash.{h,c}: (string_hash_make) Hash optimised for strings, current
implementation using Bernstein hash, which offers a good compromise
between distribution and performance.
* distribute.c: (distribute_hash_make) use previous instead of additive
string hash.
* if_rmap.c: (if_rmap_hash_make) ditto
Barry Friedman [Thu, 3 Feb 2011 22:02:53 +0000 (14:02 -0800)]
bgpd: Remove extra lock on interior table node
If the radix tree creates an extra interior node in bgp_node_get(),
it locks the interior node even though this node is not returned to
the caller, so it may never be unlocked. The lock prevents this node
from being deleted.
* bgpd/bgp_table.c: (bgp_node_get) Remove lock on interior node which
prevents proper node deletion
* bgp_route.c: (route_vty_out*) The local prefix, metric and weight values
are all stored as uint32_t. Change the format to %u so that large values
are not displayed as negative integers.
* bgp_route.c: (bgp_static_update_rsclient) BGP sometimes crashes when
removing route server client because of use after free.
The code to update rsclient created a local static copy of bgp attributes
but neglected to handle the extra information pointer. The extra
information was getting freed by bgp_attr_unintern() and reused later when
the copy was passed to bgp_attr_intern().
The fix is to use the attr_dup function to create a copy of the extra
information, then clean it up.
* bgpd: (bgp_damp_parameter_set) The BGP reuse_index is not initialized
properly. This would cause sporadic crash when disabling dampening. Use
XCALLOC correctly and the right size array is initialized and no memset is
needed.
Chris Caputo [Tue, 27 Jul 2010 16:28:55 +0000 (16:28 +0000)]
bgpd: fix bgp_node locking issues
* bgpd: Connected table locks were being locked but not unlocked, such that
eventually a lock would exceed 2^31 and become negative, thus triggering
an assert later on.
* bgp_main.c: (bgp_exit) delete connected elements along with ifp's.
* bgp_nexthop.c: (bgp_nexthop_lookup{,_ipv6}) add missing unlocks
(bgp_multiaccess_check_v4) ditto
(bgp_connected_{add,delete}) Use a distinct memtype for bgp_connected_ref.
(bgp_scan_finish) reset the nexthop cache to clean it up when bgpd exits
* bgp_route.c: fix missing bgp_node unlocks
* lib/memtype.c: (memory_list_bgp) add MTYPE_BGP_CONN
* testing: has been tested for almost 2 months now.
John Kemp [Fri, 18 Mar 2011 14:52:18 +0000 (17:52 +0300)]
bgpd: improve "monotonic" uptime correction
Older versions of Quagga/Zebra would output a value in MRT table
dump files for "uptime" aka "ORIGINATED" that was a WALL clock
value. Given that uptime is now internally a bgp_clock MONOTONIC
value, the output in the MRT files is showing up as monotonic.
Note: time of MRT dump is still recorded correctly as a
time() based value, so we haven't lost that value.
Proposal is to correct the uptime output on the vty and in the
MRT files to again display something more akin to WALL time.
* bgp_dump.c: (bgp_dump_routes_func) add conditional correction
* bgp_route.c: (route_vty_out_detail) make correction conditional, move
variable declaration to beginning of the function
Andrew J. Schorr [Thu, 24 Feb 2011 10:52:14 +0000 (13:52 +0300)]
ripd: resolve debug statements issue (bug 442)
...A nasty bug, if you forgot to disable debugging, stored the config
and reboot your machine - if you really depend on ripd, then the machine
will not fully come back on the network, because ripd fails.
Dmitrij Tejblum [Fri, 14 Jan 2011 15:27:05 +0000 (18:27 +0300)]
bgpd: fix handling of "Unsupported Capability"
* bgp_packet.c: (bgp_notify_receive) justify the difference between
BGP_NOTIFY_OPEN_UNSUP_PARAM and BGP_NOTIFY_OPEN_UNSUP_CAPBL cases, as
it is explained in RFC5492, page 3, paragraph 1.
"Unsupported Capability" error does not mean, that the peer doesn't
support capabilities advertisement -- quite the opposite (if the peer
would not support capabilities advertisement, the code would be
"Unsupported Optional Parameter"). Thus there is no reason to mark
the peer as one non-supporting capabilities advertisement.
Example: suppose the peer is in fact IPv6-only, but we didn't configure
anything address-family specific for it. Then, the peer would refuse
the session with "Unsupported Capability" code. If we internally set
the peer as non-supporting capabilities advertisement after that, we
will not be able to establish the session with it ever, even with a
fixed configuration -- IPv6-only BGP session cannot be established
without capabilities.
In practice an edge case would be seen as the same IPv6 peer working
with its "neighbor" block read from bgpd.conf, but not working, when
slowly input in "conf t" mode.
Dmitrij Tejblum [Thu, 13 Jan 2011 15:25:40 +0000 (18:25 +0300)]
ospf6d: fix crash in SPF calculation
* ospf6_spf.c: Don't replace a node with another node with a lower
number of hops, instead get them from the queue in the correct
order. (Actually, the replacement crashed the ospf6d daemon
rather than worked.)
Paul Jakma [Thu, 15 Apr 2010 07:11:51 +0000 (08:11 +0100)]
ospfd: Remove oi field from LSA, have network_lsa_refresh look up when needed
* ospf_lsa.h: (struct ospf_lsa) remove oi pointer
* ospf_lsa.c: (ospf_network_lsa_refresh) instead of keeping a pointer, just
lookup the oi when it's needed. This decouples network LSA from oi lifetime
and avoids having to invalidate pointers in LSAs when an oi changes,
simplifying the code.