While "rndc reload" causes dns_zone_asyncload() to be called for the
signed version of an inline-signed zone, the subsequent zone_load() call
causes the raw version to be reloaded from storage. This means that
DNS_ZONEFLG_LOADPENDING gets set for the signed version of the zone by
dns_zone_asyncload() before the reload is attempted, but zone_postload()
is only called for the raw version and thus DNS_ZONEFLG_LOADPENDING is
cleared for the raw version, but not for the signed version. This in
turn prevents zone maintenance from happening for the signed version of
the zone.
Until commit 29b7efdd9f, this problem
remained dormant because DNS_ZONEFLG_LOADPENDING was previously
immediately, unconditionally cleared after zone loading was started
(whereas it should only be cleared when zone loading is finished or an
error occurs). This behavior caused other issues [1] and thus had to be
changed.
Fix reloading inline-signed zones by clearing DNS_ZONEFLG_LOADPENDING
for the signed version of the zone once the raw version reload
completes. Take care not to clear it prematurely during initial zone
load. Also make sure that DNS_ZONEFLG_LOADPENDING gets cleared when
zone_postload() encounters an error or returns early, to prevent other
scenarios from resulting in the same problem. Add comments aiming to
help explain code flow.
[1] see RT #47076
Trying to resolve a trust anchor telemetry query for a locally served
zone does not cause upstream queries to be sent as the response is
determined just by consulting local data. Work around this issue by
calling dns_view_findzonecut() first in order to determine the NS RRset
for a given domain name and then passing the zone cut found to
dns_resolver_createfetch().
Note that this change only applies to TAT queries generated by the
resolver itself, not to ones received from downstream resolvers.
If transferring or loading a mirror zone fails, resolution should still
succeed by means of falling back to regular recursive queries.
Currently, though, if a slave zone is present in the zone table and not
loaded, a SERVFAIL response is generated. Thus, mirror zones need
special handling in this regard.
Add a new dns_zt_find() flag, DNS_ZTFIND_MIRROR, and set it every time a
domain name is looked up rather than a zone itself. Handle that flag in
dns_zt_find() in such a way that a mirror zone which is expired or not
yet loaded is ignored when looking up domain names, but still possible
to find when the caller wants to know whether the zone is configured.
This causes a fallback to recursion when mirror zone data is unavailable
without making unloaded mirror zones invisible to code checking a zone's
existence.
When a resolver is a regular slave (i.e. not a mirror) for some zone,
non-recursive queries for names below that slaved zone will return a
delegation sourced from it. This behavior is suboptimal for mirror
zones as their contents should rather be treated as validated, cached
DNS responses. Modify query_delegation() and query_zone_delegation() to
permit clients allowed cache access to check its contents for a better
answer when responding to non-recursive queries.
Make ns3 mirror the "root" zone from ns1 and query the former for a
properly signed record below the root. Ensure ns1 is not queried during
resolution and that the AD bit is set in the response.
Update axfr_commit() so that all incoming versions of a mirror zone
transferred using AXFR are verified before being used. If zone
verification fails, discard the received version of the zone, wait until
the next refresh and retry.
This commit only moves code around, with the following exceptions:
- the check_dns_dbiterator_current() macro and functions
is_delegation() and has_dname() were removed from
bin/dnssec/dnssectool.{c,h} and duplicated in two locations:
bin/dnssec/dnssec-signzone.c and lib/dns/zoneverify.c; these
functions are used both by the code in bin/dnssec/dnssec-signzone.c
and verifyzone(), but are not a good fit for being exported by a
code module responsible for zone verification,
- fatal() and check_result() were duplicated in lib/dns/zoneverify.c
as static functions which do not use the "program" variable any more
(as it is only set by the tools in bin/dnssec/); this is a temporary
step which only aims to prevent compilation from breaking - these
duplicate functions will be removed once lib/dns/zoneverify.c is
refactored not to use them,
- the list of header files included by lib/dns/zoneverify.c was
expanded to encompass all header files that are actually used by the
code in that file,
- a description of the purpose of the commented out "fields" inside
struct nsec3_chain_fixed was added.
- make qname-minimization option tristate {strict,relaxed,disabled}
- go straight for the record if we hit NXDOMAIN in relaxed mode
- go straight for the record after 3 labels without new delegation or 7 labels total
- use start of fetch (and not time of response) as 'now' time for querying cache for
zonecut when following delegation.
This commit reverts the previous change to use system provided
entropy, as (SYS_)getrandom is very slow on Linux because it is
a syscall.
The change introduced in this commit adds a new call isc_nonce_buf
that uses CSPRNG from cryptographic library provider to generate
secure data that can be and must be used for generating nonces.
Example usage would be DNS cookies.
The isc_random() API has been changed to use fast PRNG that is not
cryptographically secure, but runs entirely in user space. Two
contestants have been considered xoroshiro family of the functions
by Villa&Blackman and PCG by O'Neill. After a consideration the
xoshiro128starstar function has been used as uint32_t random number
provider because it is very fast and has good enough properties
for our usage pattern.
The other change introduced in the commit is the more extensive usage
of isc_random_uniform in places where the usage pattern was
isc_random() % n to prevent modulo bias. For usage patterns where
only 16 or 8 bits are needed (DNS Message ID), the isc_random()
functions has been renamed to isc_random32(), and isc_random16() and
isc_random8() functions have been introduced by &-ing the
isc_random32() output with 0xffff and 0xff. Please note that the
functions that uses stripped down bit count doesn't pass our
NIST SP 800-22 based random test.
The three functions has been modeled after the arc4random family of
functions, and they will always return random bytes.
The isc_random family of functions internally use these CSPRNG (if available):
1. getrandom() libc call (might be available on Linux and Solaris)
2. SYS_getrandom syscall (might be available on Linux, detected at runtime)
3. arc4random(), arc4random_buf() and arc4random_uniform() (available on BSDs and Mac OS X)
4. crypto library function:
4a. RAND_bytes in case OpenSSL
4b. pkcs_C_GenerateRandom() in case PKCS#11 library
Add a new ATF test, sigs_test, containing everything required to start
defining test cases for dns__zone_updatesigs(). The framework is
written in a way which ensures that changes to zone database applied by
any dns__zone_updatesigs() invocation are preserved between subsequent
checks.
Add a new private header file, lib/dns/zone_p.h, which will hold type
definitions and function prototypes not meant to be exported by libdns,
but required by zone-related unit tests.
- added tests to the dnssec system test that duplicate the ones
from bin/tests/dnssec-signzone
- changed cleanall.sh so it doesn't automatically remove all
key files, because there are now some of those that are part of the
distribution
- add CHANGES note
- update copyrights and license headers
- add -j to the make commands in .gitlab-ci.yml to take
advantage of parallelization in the gitlab CI process
