UDP is connectionless, but systems use ICMP unreachable messages to
indicate there is no ability to reach the host or port, which can result
in a `send()` returning an error like `ECONNREFUSED`. We need to handle
non-retryable codes like that to treat it as a connection failure so we
requeue any queries on that connection to another connection/server
immediately. Otherwise what happens is we just wait on the timeout to
expire which can greatly increase the time required to get a definitive
message.
This also adds a test case to verify the behavior.
Fixes#819
Fix By: Brad Houes (@bradh352)
c-ares utilizes recursion for some operations, and some of these
processes can have unintended side effects, such as if a callback
is called that then recurses into the same function. This can cause
strange cleanup conditions that lead to crashes.
Try to disassociate queries with connections as early as possible and
move cleaning up unneeded connections to its own scan rather than
trying to detect each time a query is disassociated from a connection.
Fix By: Brad House (@bradh352)
On some broken MacOS SDK versions, if HAVE_UNISTD_H is defined, it
must be defined to 1, otherwise it will cause build errors. Autotools
always defines to 1, but CMake doesn't.
Fixes Issue: #787
Fix By: Brad House (@bradh352)
The query cache wasn't properly flushing on server list changes
and was attempting to flush even when the server list didn't
actually change.
Fix By: Brad House (@bradh352)
Due to a logic flaw dns name compression writing was not properly implemented
which would result in the name prefix not being written for a partial match.
Fixes Bug: #757
Fix By: Brad House (@bradh352)
The DNS message protocol when using non-binary strings needs to be in
the ASCII printable range. The function prototype does elude to this
but it was not actually validating the string was in anyway valid and
could be used. DNS parsing will now fail if an expected string isn't
an ASCII string.
Fixes Issue: #769
Fix By: Brad House (@bradh352)
salen validation should be greater than or equal to the required
storage size. Its not uncommon to use `struct sockaddr_storage` in
modern code which is definitely larger than `struct sockaddr_in` and
on some systems even larger than `struct sockaddr_in6`.
Fixes Issue: #752
Fix By: Brad House (@bradh352)
If initializing using default settings fails, there may be a memory leak of
search domains that were set by system configuration.
Fixes Issue: #724
Fix By: Brad House (@bradh352)
ares__strsplit provides a newly allocated buffer, so suffix list in
line variable isn't referenced anymore. Related ares_free seems to
have gone missing during refactoring made in #594
Fix By: Vojtěch Vobr (@vojtechvobr)
CMake was unconditionally setting output locations globally, but
it should not do that if it is not the top-level project (e.g.
during chain building). Detect this fact and only set the output
location when top level.
Fixes Issue: #708
Fix By: Anthony Alayo (@anthonyalayo)
It appears as though we should never sanity check the RR name vs the question name as some DNS servers may return results for alias records.
Fixes Bug: #683
Fix By: Brad House (@bradh352)
In order to be sure a different server is chosen on the next query,
a read error should result in the failure count being updated
first before requeing the request to a different server.
Fix By: Brad House (@bradh352)
There is a reported build issue where getrandom() is detected
but compile fails due to a missing prototype. This commit attempts
to resolve that issue.
Fixes Bug: #665
Fix By: Brad House (@bradh352)
New test cases depend on internal symbols for calculating timeouts.
Disable those test features if symbol hiding is enabled.
Fixes Bug: #664
Fix By: Brad House (@bradh352)
When doing ares_gethostbyname() or ares_getaddrinfo() with AF_UNSPEC, if ares_cancel() was called after one address class was returned but before the other address class, it would return ARES_SUCCESS rather than ARES_ECANCELLED.
Test case has been added for this specific condition.
Fixes Bug: #662
Fix By: Brad House (@bradh352)
As of c-ares 1.22.0, server timeouts were erroneously not incrementing server failures meaning the server in use wouldn't rotate. There was apparently never a test case for this condition.
This PR fixes the bug and adds a test case to ensure it behaves properly.
Fixes Bug: #650
Fix By: Brad House (@bradh352)
Apparently nodejs is relying on the above behavior for legacy reasons. Add
sanity checks to the various optmask parameters where it makes sense.
See https://github.com/nodejs/node/pull/50800
Fix By: Brad House (@bradh352)
Regression from c-ares 1.19.1, ARES_OPT_UDP_PORT and ARES_OPT_TCP_PORT are
specified from the user in host-byte order, but there was a regression that
caused it to be read as if it was network byte order.
Fixes Bug: #640
Reported By: @Flow86
Fix By: Brad House (@bradh352)
c-ares does not have any concept of thread-safety. It has always been 100% up to the implementor to ensure they never call c-ares from more than one thread at a time. This patch adds basic thread-safety support, which can be disabled at compile time if not desired. It uses a single recursive mutex per channel, which should be extremely quick when uncontested so overhead should be minimal.
Fixes Bug: #610
Also sets the stage to implement #611
Fix By: Brad House (@bradh352)
c-ares init options defines a flag ARES_FLAG_NOSEARCH that is supposed to prevent search using configured domain suffixes, however when using ares_getaddrinfo() the flag was ignored and domain suffixes were used anyway.
Configuring zero domains to search also does not work (if ndomains == 0 default domain search list is loaded regardless of the flag ARES_OPT_DOMAINS being set).
This change adds a check for the ARES_FLAG_NOSEARCH in as_is_only() function that is used by ares_getaddrinfo() to decide if to try to query next possible name ( next_dns_lookup() )
Fix By: @petrvh
Brad House
Gregor Jasny
Vojtěch Vobr
petrvh