When -T cookiealwaysvalid is passed to named, DNS cookie checks for
the incoming queries always pass, given they are structurally correct.
(cherry picked from commit 807ef8545d)
When answering queries, don't add data to the additional section if
the answer has more than 13 names in the RDATA. This limits the
number of lookups into the database(s) during a single client query,
reducing query processing load.
Also, don't append any additional data to type=ANY queries. The
answer to ANY is already big enough.
(cherry picked from commit a1982cf1bb)
implement, document, and test the 'max-query-restarts' option
which specifies the query restart limit - the number of times
we can follow CNAMEs before terminating resolution.
(cherry picked from commit 104f3b82fb)
(cherry picked from commit 2e04f0380c)
the number of iterative queries that can be sent to resolve a
name now defaults to 32 rather than 100.
(cherry picked from commit 7e3b425dc2)
(cherry picked from commit a11367ade3)
MAX_RESTARTS is no longer hard-coded; ns_server_setmaxrestarts()
and dns_client_setmaxrestarts() can now be used to modify the
max-restarts value at runtime. in both cases, the default is 11.
(cherry picked from commit c5588babaf)
(cherry picked from commit bfbc6a6c84)
the number of steps that can be followed in a CNAME chain
before terminating the lookup has been reduced from 16 to 11.
(this is a hard-coded value, but will be made configurable later.)
(cherry picked from commit 05d78671bb)
(cherry picked from commit dd88a4cdfc)
The lame-ttl cache is implemented in ADB as per-server locked
linked-list "indexed" with <qname,qtype>. This list has to be walked
every time there's a new query or new record added into the lame cache.
Determined attacker can use this to degrade performance of the resolver.
Resolver testing has shown that disabling the lame cache has little
impact on the resolver performance and it's a minimal viable defense
against this kind of attack.
When answering a query, named should never attempt to add the same RRset
to the ANSWER section more than once. However, such a situation may
arise when chasing DNAME records: one of the DNAME records placed in the
ANSWER section may turn out to be the final answer to a client query,
but there is no way to know that in advance. Tweak the relevant INSIST
assertion in query_find() so that it handles this case properly. The
rdataset is freed later anyway, so there is no need to clean it up
immediately.
The DNS Flag Day 2020 reduced all the EDNS buffer sizes to 1232. In
this commit, we revert the default value for nocookie-udp-size back to
4096 because the option is too obscure and most people don't realize
that they also need to change this configuration option in addition to
max-udp-size.
(cherry picked from commit 79c196fc77)
Since the queries sent towards root and TLD servers are now included in
the count (as a result of the fix for CVE-2020-8616),
"max-recursion-queries" has a higher chance of being exceeded by
non-attack queries. Increase its default value from 75 to 100.
(cherry picked from commit ab0bf49203)
Every DocBook source document can be namespaced (DocBook 5) or
non-namespaced (DocBook 4). The set of XSL stylesheets used for
producing an output document can also be namespaced or non-namespaced.
Namespaced source documents should be used with namespaced stylesheets
and non-namespaced source documents should be used with non-namespaced
stylesheets. However, both stylesheet flavors contain processing rules
which allow them to be used interchangeably for any type of source
document.
Unfortunately, these processing rules became broken in version 1.79.1 of
the stylesheets, which means that non-namespaced source documents can no
longer be correctly transformed into man pages using namespaced
stylesheets and vice versa. This problem was fixed upstream [1], but no
released version of the XSL stylesheets contains that fix yet.
Back in 2016, this problem was reported as RT #43831 and allegedly fixed
in commit 1b8ce3b330. However, that fix
only helped for the non-namespaced version of the stylesheets - while
also breaking man page generation for the namespaced flavor.
Since using namespaced DocBook sources is the current best practice
(DocBook 5), make BIND DocBook sources namespaced again. When using
version 1.79.1 or 1.79.2 of the XSL stylesheets, care must be taken to
ensure namespaced stylesheets are used for generating BIND
documentation.
[1] https://github.com/docbook/xslt10-stylesheets/issues/109
The DNS Flag Day 2020 aims to remove the IP fragmentation problem from
the UDP DNS communication. In this commit, we implement the minimal
required changes by changing the defaults for `edns-udp-size`,
`max-udp-size` and `nocookie-udp-size` to `1232` (the value picked by
DNS Flag Day 2020).
(cherry picked from commit bb990030d3)
(cherry picked from commit df06773f9a73e4a4bb0b5e7f4bd8e40394e64032)
This commit fix the problems that arose when moving the dns_message_t
object from fetchctx_t to the query structure.
Since the lifetime of query objects are different than that of a
fetchctx and the dns_message_t object held by the query may be being
used by some external module, e.g. validator, even after the query may
have been destroyed, propery handling of the references to the message
were added in this commit to avoid accessing an already destroyed
object.
Specifically, in resquery_response(), a reference to the message is
attached at the beginning of the function and detached at the end, since
a possible call to fctx_cancelquery() would release the dns_message_t
object, and in the next lines of code a call to add_bad() would require
a valid pointer to the same object.
In valcreate() a new reference is attached to the message object, this
ensures that if the corresponding query object is destroyed before the
validator attempts to access it, no invalid pointer access occurs.
In validated() we have to attach a new reference to the message, since
we destroy the validator object at the beginning of the function, and we
need access to the message in the next lines of the same function.
The dotat() function has been changed to send the TAT
query asynchronously, so there's no lock order loop
because we initialize the data first and then we schedule
the TAT send to happen asynchronously.
This breaks following lock-order loops:
zone->lock (dns_zone_setviewcommit) while holding view->lock
(dns_view_setviewcommit)
keytable->lock (dns_keytable_find) while holding zone->lock
(zone_asyncload)
view->lock (dns_view_findzonecut) while holding keytable->lock
(dns_keytable_forall)
(cherry picked from commit 3c4b68af7c)
ret->link.next is tested to see if it is NULL unlocked
to avoid obtaining taillock when it is not nexessary then
retested once the taillock is obtained.
Now that the log message has been printed set the result code to
DNS_R_FORMERR. We don't do this via dns_result_torcode() as we
don't want upstream errors to produce FORMERR if that processing
end with DNS_R_BADTSIG.
(cherry picked from commit 20488d6ad3)
server might be created, but not yet fully initialized, when fatal
function is called. Check both server and task before attaching
exclusive task.
(cherry picked from commit c5e7152cf0)
it was possible for the count_newzones() function to try to
unlock view->new_zone_lock on return before locking it, which
caused a crash on shutdown.
(cherry picked from commit ed37c63e2b)
When "rndc reconfig" is run, named first configures a fresh set of views
and then tears down the old views. Consider what happens for a single
view with LMDB enabled; "envA" is the pointer to the LMDB environment
used by the original/old version of the view, "envB" is the pointer to
the same LMDB environment used by the new version of that view:
1. mdb_env_open(envA) is called when the view is first created.
2. "rndc reconfig" is called.
3. mdb_env_open(envB) is called for the new instance of the view.
4. mdb_env_close(envA) is called for the old instance of the view.
This seems to have worked so far. However, an upstream change [1] in
LMDB which will be part of its 0.9.26 release prevents the above
sequence of calls from working as intended because the locktable mutexes
will now get destroyed by the mdb_env_close() call in step 4 above,
causing any subsequent mdb_txn_begin() calls to fail (because all of the
above steps are happening within a single named process).
Preventing the above scenario from happening would require either
redesigning the way we use LMDB in BIND, which is not something we can
easily backport, or redesigning the way BIND carries out its
reconfiguration process, which would be an even more severe change.
To work around the problem, set MDB_NOLOCK when calling mdb_env_open()
to stop LMDB from controlling concurrent access to the database and do
the necessary locking in named instead. Reuse the view->new_zone_lock
mutex for this purpose to prevent the need for modifying struct dns_view
(which would necessitate library API version bumps). Drop use of
MDB_NOTLS as it is made redundant by MDB_NOLOCK: MDB_NOTLS only affects
where LMDB reader locktable slots are stored while MDB_NOLOCK prevents
the reader locktable from being used altogether.
[1] https://git.openldap.org/openldap/openldap/-/commit/2fd44e325195ae81664eb5dc36e7d265927c5ebc
(cherry picked from commit 53120279b5)
enough buffer space. Also change named_os_uname() prototype so
that it is now returning (const char *) rather than (char *). If
uname() is not supported on a UNIX build prepopulate unamebuf[]
with "unknown architecture".
(cherry picked from commit 4bc3de070f)
named_os_openfile was being called with switch_user set to true
unconditionally leading to log messages about being unable to
switch user identity from named when regenerating the key.
(cherry picked from commit 071bc29962)
Aram Sargsyan
Ondřej SurýandNicki Křížek
Evan HuntandOndřej Surý
Petr ŠpačekandNicki Křížek
Tinderbox User
Michal Nowak
Ondřej SurýandMichał Kępień
Mark Andrews
Diego Fronza
Petr MenšíkandMark Andrews