ZONEMD needs to be able to digest SIG and RRSIG records. The signer
field can be compressed in SIG so we need to call dns_name_digest().
While for RRSIG the records the signer field is not compressed the
canonical form has the signer field downcased (RFC 4034, 6.2). This
also implies that compare_rrsig needs to downcase the signer field
during comparison.
The qpcache_findzonecut() accepts two "foundnames": 'foundname' and
'dcname' could be NULL. Originally, when 'dcname' would be NULL, the
'dcname' would be set to 'foundname'. Then code like this was present:
result = find_deepest_zonecut(&search, node, nodep, foundname,
rdataset,
sigrdataset DNS__DB_FLARG_PASS);
dns_name_copy(foundname, dcname);
Which basically means that we are copying the .ndata over itself for no
apparent reason. Cleanup the dcname vs foundname usage.
Co-authored-by: Evan Hunt <each@isc.org>
Co-authored-by: Ondřej Surý <ondrej@isc.org>
This commit bumps the total number of active streams (= the opened
streams for which a request is received, but response is not ready) to
60% of the total streams limit.
The previous limit turned out to be too tight as revealed by
longer (≥1h) runs of "stress:long:rpz:doh+udp:linux:*" tests.
Previously, the code would try to avoid sending any data regardless of
what it is unless:
a) The flush limit is reached;
b) There are no sends in flight.
This strategy is used to avoid too numerous send requests with little
amount of data. However, it has been proven to be too aggressive and,
in fact, harms performance in some cases (e.g., on longer (≥1h) runs
of "stress:long:rpz:doh+udp:linux:*").
Now, additionally to the listed cases, we also:
c) Flush the buffer and perform a send operation when there is an
outgoing DNS message passed to the code (which is indicated by the
presence of a send callback).
That helps improve performance for "stress:long:rpz:doh+udp:linux:*"
tests.
Previously, a function for continuing IO processing on the next UV
tick was introduced (http_do_bio_async()). The intention behind this
function was to ensure that http_do_bio() is eventually called at
least once in the future. However, the current implementation allows
queueing multiple such delayed requests needlessly. There is currently
no need for these excessive requests as http_do_bio() can requeue them
if needed. At the same time, each such request can lead to a memory
allocation, particularly in BIND 9.18.
This commit ensures that the number of enqueued delayed IO processing
requests never exceeds one in order to avoid potentially bombarding IO
threads with the delayed requests needlessly.
This commit significantly simplifies the code flow in the
http_do_bio() function, which is responsible for processing incoming
and outgoing HTTP/2 data. It seems that the way it was structured
before was indirectly caused by the presence of the missing callback
calls bug, fixed in 8b8f4d500d.
The change introduced by this commit is known to remove a bottleneck
and allows reproducible and measurable performance improvement for
long runs (>= 1h) of "stress:long:rpz:doh+udp:linux:*" tests.
Additionally, it fixes a similar issue with potentially missing send
callback calls processing and hardens the code against use-after-free
errors related to the session object (they can potentially occur).
Previously, a gcc < 4.6 shim for _Static_assert() was included. Such an
old compiler is not supported now anyway, so the macro variant has been
removed in favor of a single definition using _Static_assert().
Previously, the LOCK()/UNLOCK() and friends macros were defined in the
isc/util.h header. Those macros were moved to their respective headers
as those would have to be included anyway if that particular lock was in
use.
Formerly, isc/util.h would pull a few extra headers (isc/list.h,
isc/attributes.h, isc/result.h and errno.h). These includes were
removed in favor of including them directly when used.
The short convenience list macros were used very sparingly and
inconsistenly in the code base. As the consistency is prefered over
the convenience, all shortened list macro were removed in favor of
their ISC_LIST API targets.
When all the addresses were already iterated over, the
dns_remote_curraddr() function asserts. So before calling it,
dns_zone_getprimaryaddr() now checks the address list using the
dns_remote_done() function. This also means that instead of
returning 'isc_sockaddr_t' it now returns 'isc_result_t' and
writes the primary's address into the provided pointer only when
returning success.
Since algorithm fetching is handled purely in libisc, FIPS mode toggling
can be purely done in within the library instead of provider fetching in
the binary for OpenSSL >=3.0.
Disabling FIPS mode isn't a realistic requirement and isn't done
anywhere in the codebase. Make the FIPS mode toggle enable-only to
reflect the situation.
Answers to an "ANY" query which are processed by the RPZ "passthru"
policy have the response-policy's 'max-policy-ttl' value unexpectedly
applied. Do not change the records' TTL when RPZ uses a policy which
does not alter the answer.
when the caching of a negative record failed because of the
presence of a positive one, ncache_adderesult() could override
this to ISC_R_SUCCESS. this could cause CNAME and DNAME responses
to be handled incorrectly. ncache_adderesult() now sets the result
code correctly in such cases.
the target name parameter to dns_adb_createfind() was always passed as
NULL, so we can safely remove it.
relatedly, the 'target' field in the dns_adbname structure was never
referenced after being set. the 'expire_target' field was used, but
only as a way to check whether an ADB name represents a CNAME or DNAME,
and that information can be stored as a single flag.
Named was stopping nameserver address resolution attempts too soon
when dual stack servers are configured. Dual stack servers are
used when there are *not* addresses for the server in a particular
address family so find->status == DNS_ADB_NOMOREADDRESSES is not a
sufficient stopping condition when dual stack servers are available.
Call fctx_try to see if the alternate servers can be used.
DNSKEY, KEY, RRSIG and SIG constraints have been relaxed to allow
empty key and signature material after the algorithm identifier for
PRIVATEOID and PRIVATEDNS. It is arguable whether this falls within
the expected use of these types as no key material is shared and
the signatures are ineffective but these are private algorithms and
they can be totally insecure.
if the NS_QUERY_DONE_BEGIN or NS_QUERY_DONE_SEND hook is
used in a plugin and returns NS_HOOK_RETURN, some of the
cleanup in ns_query_done() can be skipped over, leading
to reference leaks that can cause named to hang on shut
down.
this has been addressed by adding more housekeeping
code after the cleanup: tag in ns_query_done().
previously, dns_name_fromtext() took both a target name and an
optional target buffer parameter, which could override the name's
dedicated buffer. this interface is unnecessarily complex.
we now have two functions, dns_name_fromtext() to convert text
into a dns_name that has a dedicated buffer, and dns_name_wirefromtext()
to convert text into uncompressed DNS wire format and append it to a
target buffer.
in cases where it really is necessary to have both, we can use
dns_name_fromtext() to load the dns_name, then dns_name_towire()
to append the wire format to the target buffer.
the target buffer passed to dns_name_concatenate() was never
used (except for one place in dig, where it wasn't actually
needed, and has already been removed in a prior commit).
we can safely remove the parameter.
this parameter was added as a (minor) optimization for
cases where dns_name_towire() is run repeatedly with the
same compression context, as when rendering all of the rdatas
in an rdataset. it is currently only used in one place.
we now simplify the interface by removing the extra parameter.
the compression offset value is now part of the compression
context, and can be activated when needed by calling
dns_compress_setmultiuse(). multiuse mode is automatically
deactivated by any subsequent call to dns_compress_permitted().
the dns_rdataslab_fromrdataset() function creates a slab
from an rdataset. if the source rdataset already uses a slab,
then no processing is necessary; we can just copy the existing
slab to a new location.
Previously, the hashmap iterator for fetches-per-zone was destroy
outside the rwlock. This could lead to an assertion failure due to a
timing race with the internal rehashing of the hashmap table as the
rehashing process requires no iterators to be running when rehashing the
hashmap table. This has been fixed by moving the destruction of the
iterator inside the read locked section.
The third argument to set_offsets() was only used in
dns_name_fromregion() and not really needed. We can remove the third
argument and then manually check whether the last label is root label.
There was just a single use of passing an extra buffer to
dns_name_downcase() which have been replaced by simple call to
isc_ascii_lowercase() and the 'target' argument from dns_name_downcase()
function has been removed.
The MAKE_EMPTY() macro was clearing up the output variable in case of
the failure. However, this was breaking the usual design pattern that
the output variables are left in indeterminate state or we don't touch
them at all when a failure occurs. Remove the macro and change the
dns_name_downcase() to not touch the name contents until success.
There was a back-and-forth between static arrays and the pointers to the
offsets. Since we are now only using the static arrays, we can cleanup
the usage of the pointers that would previously point either to the
static array or name->offsets if available.
The offsets were meant to speed-up the repeated dns_name operations, but
it was experimentally proven that there's actually no real-world
benefit. Remove the offsets and labels fields from the dns_name and the
static offsets fields to save 128 bytes from the fixedname in favor of
calculating labels and offsets only when needed.
The DNS header shows if a message has multiple questions or invalid
NOTIFY sections. We can drop these messages early, right after parsing
the question. This matches RFC 9619 for multi-question messages and
Unbound's handling of NOTIFY.
To further add further robustness, we include an additional check for
unknown opcodes, and also drop those messages early.
Add early_sanity_check() function to check for these conditions:
- Messages with more than one question, as required by RFC 9619
- NOTIFY query messages containing answer sections (like Unbound)
- NOTIFY messages containing authority sections (like Unbound)
- Unknown opcodes.
A change in 6aba56ae8 (checking whether a rejected RRset was identical
to the data it would have replaced, so that we could still cache a
signature) inadvertently introduced cases where processing of a
response would continue when previously it would have been skipped.
Acquire the database refernce in the detachnode() to prevent the last
reference to be release while the NODE_LOCK being locked. The NODE_LOCK
is locked/unlocked inside the RCU critical section, thus it is most
probably this should not pose a problem as the database uses call_rcu
memory reclamation, but this it is still safer to acquire the reference
before releasing the node.
As the default_call_rcu_thread can't be forced to flush all the work
during the executable shutdown, create one call_rcu_thread explicitly
and assign it to the all created threads.
This allows this explicit call_rcu_thread to be unassociated from the
main thread and freed before the executable destructor exits.
Instead of relying on unreliable order of execution of the library
constructors and destructors, move them to individual binaries. The
advantage is that the execution time and order will remain constant and
will not depend on the dynamic load dependency solver.
This requires more work, but that was mitigated by a simple requirement,
any executable using libisc and libdns, must include <isc/lib.h> and
<dns/lib.h> respectively (in this particular order). In turn, these two
headers must not be included from within any library as they contain
inlined functions marked with constructor/destructor attributes.
Previously, the dns_resolver_dumpfetches() would go over the fetch
counters. Alas, because of the earlier optimization, the fetch counters
would be increased only when fetches-per-zone was not 0, otherwise the
whole counting was skipped for performance reasons.
Instead of using the auxiliary fetch counters hash table, use the real
hash table that stores the fetch contexts to dump the ongoing fetches to
the recursing file.
Additionally print more information about the fetch context like start
and expiry times, number of fetch responses, number of queries and count
of allowed and dropped fetches.
The order of the fetch context hash table rwlock and the individual
fetch context was reversed when calling the release_fctx() function.
This was causing a problem when iterating the hash table, and thus the
ordering has been corrected in a way that the hash table rwlock is now
always locked on the outside and the fctx lock is the interior lock.
In the next commit, we need to know whether the timer has been started
or stopped. Add isc_timer_running() function that returns true if the
timer has been started.
After a reconfiguration the old view can be left without a valid
'rpzs' member, because when the RPZ is not changed during the named
reconfiguration 'rpzs' "migrate" from the old view into the new
view, so when a query resumes it can find that 'qctx->view->rpzs'
is NULL which query_resume() currently doesn't expect to happen if
it's recursing and 'qctx->rpz_st' is not NULL.
Fix the issue by adding a NULL-check. In order to not split the log
message to two different log messages depending on whether
'qctx->view->rpzs' is NULL or not, change the message to not log
the RPZ policy's "version" which is just a runtime counter and is
most likely not very useful for the users.
The isc_counter_create() doesn't need the return value (it was always
ISC_R_SUCCESS), use the macros to implement the reference counting,
little style cleanup, and expand the unit test.
Checking whether the authority section is properly signed should
be left to the validator. Checking in getsection (dns_message_parse)
was way too early and resulted in resolution failures of lookups
that should have otherwise succeeded.
Previously a hard-coded limitation of maximum two key or message
verification checks were introduced when checking the message's
SIG(0) signature. It was done in order to protect against possible
DoS attacks. The logic behind choosing the number two was that more
than one key should only be required only during key rotations, and
in that case two keys are enough. But later it became apparent that
there are other use cases too where even more keys are required, see
issue number #5050 in GitLab.
This change introduces two new configuration options for the views,
sig0key-checks-limit and sig0message-checks-limit, which define how
many keys are allowed to be checked to find a matching key, and how
many message verifications are allowed to take place once a matching
key has been found. The latter protects against expensive cryptographic
operations when there are keys with colliding tags and algorithm
numbers, with default being 2, and the former protects against a bit
less expensive key parsing operations and defaults to 16.
Running jobs which were entered into the isc_quota queue is the
responsibility of the isc_quota_release() function, which, when
releasing a previously acquired quota, checks whether the queue
is empty, and if it's not, it runs a job from the queue without touching
the 'quota->used' counter. This mechanism is susceptible to a possible
hangup of a newly queued job in case when between the time a decision
has been made to queue it (because used >= max) and the time it was
actually queued, the last quota was released. Since there is no more
quotas to be released (unless arriving in the future), the newly
entered job will be stuck in the queue.
Fix the wrong memory ordering for 'quota->used', as the relaxed
ordering doesn't ensure that data modifications made by one thread
are visible in other threads.
Add checks in both isc_quota_release() and isc_quota_acquire_cb()
to make sure that the described hangup does not happen. Also see
code comments.
