| Age | Commit message (Collapse) | Author |
|---|
|
commit fe9bb64... "bgpd: CVE-2012-1820, DoS in bgp_capability_orf()"
made the length test in bgp_capability_orf_entry() stricter and is now
causing us to refuse (with CEASE) ORF capabilites carrying any excess
data. This does not conform to the robustness principle as laid out by
RFC1122 ("be liberal in what you accept").
Even worse, RFC5291 is quite unclear on how to use the ORF capability
with multiple AFI/SAFIs. It can be interpreted as either "use one
instance, stuff everything in" but also as "use multiple instances".
So, if not for applying robustness, we end up clearing sessions from
implementations going by the former interpretation. (or if anyone dares
add a byte of padding...)
Cc: Denis Ovsienko <infrastation@yandex.ru>
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
|
|
An ORF (code 3) capability TLV is defined to contain exactly one
AFI/SAFI block. Function bgp_capability_orf(), which parses ORF
capability TLV, uses do-while cycle to call its helper function
bgp_capability_orf_entry(), which actually processes the AFI/SAFI data
block. The call is made at least once and repeated as long as the input
buffer has enough data for the next call.
The helper function, bgp_capability_orf_entry(), uses "Number of ORFs"
field of the provided AFI/SAFI block to verify, if it fits the input
buffer. However, the check is made based on the total length of the ORF
TLV regardless of the data already consumed by the previous helper
function call(s). This way, the check condition is only valid for the
first AFI/SAFI block inside an ORF capability TLV.
For the subsequent calls of the helper function, if any are made, the
check condition may erroneously tell, that the current "Number of ORFs"
field fits the buffer boundary, where in fact it does not. This makes it
possible to trigger an assertion by feeding an OPEN message with a
specially-crafted malformed ORF capability TLV.
This commit fixes the vulnerability by making the implementation follow
the spec.
|
|
Use the array_size() helper macro. Replaces several instances of local
macros with the same definition.
Reviewed-by: Scott Feldman <sfeldma@cumulusnetworks.com>
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
|
|
Address problem where bgpd would reject a session if a peer sent some
capabilities in its Open message, but did not include a Multiprotocol
extensions capability. Note that the session would come up if there
were no capabilities at all in the Open message.
* Add the 'mp_capability' out parameter to
bgp_capability_parse(). Set it to '1' if a Multiprotocol
extensions capability is encountered.
* Switch on 'mp_capability' instead of 'capability' in the calling
functions to determine if the peer indicated the set of AFI/SAFIs
it supports.
The net result is that when a peer does not send an MP capability,
it is assumed to support the AFI/SAFIs configured for it locally.
|
|
* bgp_packet.c: (bgp_open_receive) Errors from bgp_open_option_parse are
detected, and the code will stop processing the OPEN and return. However
it does so without calling bgp_notify_send to send a NOTIFY - which means
the peer FSM doesn't get stopped, and bgp_read will be called again later.
Because it returns, it doesn't go through the code near the end of the
function that removes the current message from the peer input streaam.
Thus the next call to bgp_read will try to parse a half-parsed stream as
if it were a new BGP message, leading to an assert later in the code when
it tries to read stuff that isn't there. Add the required call to
bgp_notify_send before returning.
* bgp_open.c: (bgp_capability_as4) Be a bit stricter, check the length field
corresponds to the only value it can be, which is the amount we're going to
read off the stream. And make sure the capability flag gets set, so
callers can know this capability was read, regardless.
(peek_for_as4_capability) Let bgp_capability_as4 do the length check.
|
|
"While setting up a testbed, I ran across a little problem in the
parsing of the "graceful restart" BGP capability that resulted in
Quagga not actually activating it for the peer in question - when
the peer sent a single AFI/SAFI block."
* bgp_open.c
* bgp_capability_restart(): actually process the last AFI/SAFI block
|
|
- SAFI value 3 is reserved. It was assigned by RFC 2858 for a use
that was never fully implemented, so it is deprecated by this
document.
* zebra.h: rename macro
* bgp_fsm.c: (bgp_graceful_restart_timer_expire,
bgp_graceful_stale_timer_expire, bgp_stop, bgp_establish): update
* bgpd.c: (peer_nsf_stop): update
* bgp_open.c: (bgp_capability_vty_out): SAFI 3 isn't a recognized case
any more
|
|
(with resolved conflict in bgpd/bgp_packet.c)
Two macros resolving to the same integer constant broke a case block and
a more thorough merge of BGP_SAFI_VPNV4 and BGP_SAFI_VPNV6 was
performed.
* bgpd.h: MPLS-labeled VPN SAFI is AFI-independent, switch to single
* macro
* bgp_capability_test.c: update test data
* bgp_mp_attr_test.c: idem
* bgp_route.c: (bgp_maximum_prefix_overflow, bgp_table_stats_vty) update
macro and check conditions (where appropriate)
* bgp_packet.c: (bgp_route_refresh_send, bgp_capability_send,
bgp_update_receive, bgp_route_refresh_receive): idem
* bgp_open.c: (bgp_capability_vty_out, bgp_afi_safi_valid_indices,
bgp_open_capability_orf, bgp_open_capability): idem
* bgp_attr.c: (bgp_mp_reach_parse, bgp_packet_attribute,
bgp_packet_withdraw): idem
|
|
|
|
Fix printf format warning and make capability table 'const static'
|
|
2008-03-13 Paul Jakma <paul.jakma@sun.com>
* (various) Remove 0 entries from struct message's, unneeded due to
recent improvements in mes_lookup/LOOKUP.
|
|
2007-12-22 Paul Jakma <paul.jakma@sun.com>
* Fix series of vulnerabilities reported by "Mu Security
Research Team", where bgpd can be made to crash by sending
malformed packets - requires that bgpd be configured with a
session to the peer.
* bgp_attr.c: (bgp_attr_as4_path) aspath_parse may fail, only
set the attribute flag indicating AS4_PATH if we actually managed
to parse one.
(bgp_attr_munge_as4_attrs) Assert was too general, it is possible
to receive AS4_AGGREGATOR before AGGREGATOR.
(bgp_attr_parse) Check that we have actually received the extra
byte of header for Extended-Length attributes.
* bgp_attr.h: Fix BGP_ATTR_MIN_LEN to account for the length byte.
* bgp_open.c: (cap_minsizes) Fix size of CAPABILITY_CODE_RESTART,
incorrect -2 left in place from a development version of as4-path
patch.
* bgp_packet.c: (bgp_route_refresh_receive) ORF length parameter
needs to be properly sanity checked.
* tests/bgp_capability_test.c: Test for empty capabilities.
|
|
2007-10-14 Paul Jakma <paul.jakma@sun.com>
* NEWS: Note that MRT dumps are now version 2
* (general) Merge in Juergen Kammer's AS4 patch.
2007-09-27 Paul Jakma <paul.jakma@sun.com>
* bgp_aspath.c: (assegment_normalise) remove duplicates from
from sets.
(aspath_reconcile_as4) disregard a broken part of the RFC around
error handling in path reconciliation.
* aspath_test.c: Test dupe-weeding from sets.
Test that reconciliation merges AS_PATH and AS4_PATH where
former is shorter than latter.
2007-09-26 Paul Jakma <paul.jakma@sun.com>
* aspath_test.c: Test AS4_PATH reconcilation where length
of AS_PATH and AS4_PATH is same.
2007-09-25 Paul Jakma <paul.jakma@sun.com>
* bgp_open.c: (peek_for_as4_capability) Fix to work.
* bgp_packet.c: (bgp_open_receive) Fix sanity check of as4.
* tests/bgp_capability_test.c: (general) Extend tests to validate
peek_for_as4_capability.
Add test of full OPEN Option block, with multiple capabilities,
both as a series of Option, and a single option.
Add some crap to beginning of stream, to prevent code depending
on getp == 0.
2007-09-18 Paul Jakma <paul.jakma@sun.com>
* bgp_open.c: (bgp_capability_as4) debug printf inline with others.
(peek_for_as4_capability) There's no need to signal failure, as
failure is better dealt with through full capability parser -
just return the AS4, simpler.
* bgp_packet.c: (bgp_open_receive) Update to match
peek_for_as4_capability change.
Allow use of BGP_AS_TRANS by 2b speakers.
Use NOTIFY_OPEN_ERR rather than CEASE for OPEN parsing errors.
(bgp_capability_msg_parse) missing argument to debug print
(bgp_capability_receive) missing return values.
* tests/bgp_capability_test.c: (parse_test) update for changes to
peek_for_as4_capability
2007-07-25 Paul Jakma <paul.jakma@sun.com>
* Remove 2-byte size macros, just make existing macros take
argument to indicate which size to use.
Adjust all users - typically they want '1'.
* bgp_aspath.c: (aspath_has_as4) New, return 1 if there are any
as4's in a path.
(aspath_put) Return the number of bytes actually written, to
fix the bug Juergen noted: Splitting of segments will change
the number of bytes written from that already written to the
AS_PATH header.
(aspath_snmp_pathseg) Pass 2-byte flag to aspath_put. SNMP
is still defined as 2b.
(aspath_aggregate) fix latent bug.
(aspath_reconcile_as4) AS_PATH+NEW_AS_PATH reconciliation
function.
(aspath_key_make) Hash the AS_PATH string, rather than
just taking the addition of assegment ASes as the hash value,
hopefully sligthly more collision resistant.
(bgp_attr_munge_as4_attrs) Collide the NEW_ attributes
together with the OLD 2-byte forms, code Juergen
had in bgp_attr_parse but re-organised a bit.
(bgp_attr_parse) Bunch of code from Juergen moves
to previous function.
(bgp_packet_attribute) Compact significantly by
just /always/ using extended-length attr header.
Fix bug Juergen noted, by using aspath_put's
(new) returned size value for the attr header rather
than the (guesstimate) of aspath_size() - the two could
differ when aspath_put had to split large segments, unlikely
this bug was ever hit in the 'wild'.
(bgp_dump_routes_attr) Always use extended-len and
use aspath_put return for header length. Output 4b ASN
for AS_PATH and AGGREGATOR.
* bgp_ecommunity.c: (ecommunity_{hash_make,cmp}) fix
hash callback declarations to match prototypes.
(ecommunity_gettoken) Updated for ECOMMUNITY_ENCODE_AS4,
complete rewrite of Juergen's changes (no asdot support)
* bgp_open.c: (bgp_capability_as4) New, does what it says
on the tin.
(peek_for_as4_capability) Rewritten to use streams and
bgp_capability_as4.
* bgp_packet.c: (bgp_open_send) minor edit
checked (in the abstract at least) with Juergen.
Changes are to be more accepting, e.g, allow AS_TRANS on
a 2-byte session.
* (general) Update all commands to use CMD_AS_RANGE.
* bgp_vty.c: (bgp_clear) Fix return vals to use CMD_..
Remove stuff replicated by VTY_GET_LONG
(bgp_clear_vty) Return bgp_clear directly to vty.
* tests/aspath_test.c: Exercise 32bit parsing. Test reconcile
function.
* tests/ecommunity_test.c: New, test AS4 ecommunity changes,
positive test only at this time, error cases not tested yet.
2007-07-25 Juergen Kammer <j.kammer@eurodata.de>
* (general) AS4 support.
* bgpd.h: as_t changes to 4-bytes.
* bgp_aspath.h: Add BGP_AS4_MAX and BGP_AS_TRANS defines.
* bgp_aspath.c: AS_VALUE_SIZE becomes 4-byte, AS16_VALUE_SIZE
added for 2-byte.
Add AS16 versions of length calc macros.
(aspath_count_numas) New, count number of ASes.
(aspath_has_as4) New, return 1 if there are any as4's in a
path.
(assegments_parse) Interpret assegment as 4 or 2 byte,
according to how the caller instructs us, with a new
argument.
(aspath_parse) Add use32bit argument to pass to
assegments_parse. Adjust all its callers to pass 1, unless
otherwise noted.
(assegment_data_put) Adjust to be able to write 2 or 4 byte
AS, according to new use32bit argument.
(aspath_put) Adjust to write 2 or 4.
(aspath_gettoken) Use a long for passed in asno.
* bgp_attr.c: (attr_str) Add BGP_ATTR_AS4_PATH and
BGP_ATTR_AS4_AGGREGATOR.
(bgp_attr_aspath) Call aspath_parse with right 2/4 arg, as
determined by received-capability flag.
(bgp_attr_aspath_check) New, code previously in attr_aspath
but moved to new func so it can be run after NEW_AS_PATH
reconciliation.
(bgp_attr_as4_path) New, handle NEW_AS_PATH.
(bgp_attr_aggregator) Adjust to cope with 2/4 byte ASes.
(bgp_attr_as4_aggregator) New, read NEW_AGGREGATOR.
(bgp_attr_parse) Add handoffs to previous parsers for the two
new AS4 NEW_ attributes.
Various checks added for NEW/OLD reconciliation.
(bgp_packet_attribute) Support 2/4 for AS_PATH and
AGGREGATOR, detect when NEW_ attrs need to be sent.
* bgp_debug.{c,h}: Add 'debug bgp as4'.
* bgp_dump.c: MRTv2 support, unconditionally enabled, which
supports AS4. Based on patches from Erik (RIPE?).
* bgp_ecommunity.c: (ecommunity_ecom2str) ECOMMUNITY_ENCODE_AS4
support.
* bgp_open.c: (peek_for_as4_capability) New, peek for AS4
capability prior to full capability parsing, so we know which
ASN to use for struct peer lookup.
(bgp_open_capability) Always send AS4 capability.
* bgp_packet.c: (bgp_open_send) AS4 handling for AS field
(bgp_open_receive) Peek for AS4 capability first, and figure
out which AS to believe.
* bgp_vty.c: (bgp_show_peer) Print AS4 cap
* tests/aspath_test.c: Support asn32 changes, call aspath_parse
with 16 bit.
* vtysh/extract.pl: AS4 compatibility for router bgp ASNUMBER
* vtysh/extract.pl.in: AS4 compatibility for router bgp ASNUMBER
* vtysh/vtysh.c: AS4 compatibility for router bgp ASNUMBER
|
|
2007-09-17 Paul Jakma <paul.jakma@sun.com>
* bgp_open.c: (bgp_capability_mp) We were setting
afc_nego[safi][safi] rather than afc_nego[afi][safi], thus
failling to announce any non-IPv4 prefixes. Remove the extra,
typo-ed character.
* bgp_capability_test.c: Test that peer's adv_recv and adv_nego get
set correctly for MP capability and given AFI/SAFI.
Colour OK/failed result so it's easier to find them.
|
|
2007-07-26 Paul Jakma <paul.jakma@sun.com>
* (general) Clean up and compact capability parsing slightly.
Consolidate validation of length and logging of generic TLV, and
memcpy of capability data, thus removing such from cap specifc
code (not always present or correct).
* bgp_open.h: Add structures for the generic capability TLV header
and for the data formats of the various specific capabilities we
support. Hence remove the badly named, or else misdefined, struct
capability.
* bgp_open.c: (bgp_capability_vty_out) Use struct capability_mp_data.
Do the length checks *before* memcpy()'ing based on that length
(stored capability - should have been validated anyway on input,
but..).
(bgp_afi_safi_valid_indices) new function to validate (afi,safi)
which is about to be used as index into arrays, consolidates
several instances of same, at least one of which appeared to be
incomplete..
(bgp_capability_mp) Much condensed.
(bgp_capability_orf_entry) New, process one ORF entry
(bgp_capability_orf) Condensed. Fixed to process all ORF entries.
(bgp_capability_restart) Condensed, and fixed to use a
cap-specific type, rather than abusing capability_mp.
(struct message capcode_str) added to aid generic logging.
(size_t cap_minsizes[]) added to aid generic validation of
capability length field.
(bgp_capability_parse) Generic logging and validation of TLV
consolidated here. Code compacted as much as possible.
* bgp_packet.c: (bgp_open_receive) Capability parsers now use
streams, so no more need here to manually fudge the input stream
getp.
(bgp_capability_msg_parse) use struct capability_mp_data. Validate
lengths /before/ memcpy. Use bgp_afi_safi_valid_indices.
(bgp_capability_receive) Exported for use by test harness.
* bgp_vty.c: (bgp_show_summary) fix conversion warning
(bgp_show_peer) ditto
* bgp_debug.h: Fix storage 'extern' after type 'const'.
* lib/log.c: (mes_lookup) warning about code not being in
same-number array slot should be debug, not warning. E.g. BGP
has several discontigious number spaces, allocating from
different parts of a space is not uncommon (e.g. IANA
assigned versus vendor-assigned code points in some number
space).
|
|
* (global) The great bgpd extern and static'ification.
* bgp_routemap.c: remove unused ROUTE_MATCH_ASPATH_OLD code
(route_set_metric_compile) fix u_int32_t to ULONG_MAX comparison
warnings.
* bgp_route.h: (bgp_process, bgp_withdraw, bgp_update) export these
used by various files which had their own private declarations,
in the case of mplsvpn - incorrect.
|
|
* (global) Update code to match stream.h changes.
stream_get_putp effectively replaced with stream_get_endp.
stream_forward renamed to stream_forward_getp.
stream_forward_endp introduced to replace some previous
setting/manual twiddling of putp by daemons.
* lib/stream.h: Remove putp. Update reference to putp with endp.
Add stream_forward_endp, which daemons were doing manually.
Rename stream_forward to stream_forward_getp.
lib/stream.c: Remove/update references to putp.
introduce stream_forward_endp.
|
|
bgpd.[ch]: Add BGP_INFO_STALE flag and end-of-rib support. "bgp
graceful-restart" commands added. Show numbers of individual
messages in "show ip bgp neighbor" command. Final pieces of graceful
restart.
[merge from GNU Zebra]
|
|
capability route-refresh" commands. Route refresh capability is sent
anyway now. Preserve dummy deprecated commands.
[merge from GNU Zebra]
|
|
* *.c: Change level of debug messages to LOG_DEBUG.
|
|
* type mismatch fixes
|
|
repository by Rivo Nurges.
|
|
|
