Original Patch From Brad Spencer:
https://c-ares.haxx.se/mail/c-ares-archive-2016-04/0000.shtml
My modifications include:
* Dynamically find GetBestRoute2 since it is a Windows Vista+ symbol, and will fall back to prior behavior when not available.
* Prefer get_DNS_AdaptersAddresses as the modifications should alleviate the concerns which caused us to prefer get_DNS_NetworkParams
* Update AppVeyor to use MinGW-w64 instead of the legacy MinGW
* Fix compile error in test suite for Windows.
Original message from patch below:
From: Brad Spencer <bspencer@blackberry.com>
Date: Fri, 29 Apr 2016 14:26:23 -0300
On Windows, the c-ares DNS resolver tries first to get a full list of
DNS server addresses by enumerating the system's IPv4/v6 interfaces and
then getting the per-interface DNS server lists from those interfaces
and joining them together. The OS, at least in the way the c-ares
prefers to query them (which also may be the only or best way in some
environments), does not provide a unified list of DNS servers ordered
according to "current network conditions". Currently, c-ares will then
try to use them in whatever order the nested enumeration produces, which
may result in DNS requests being sent to servers on one interface
(hosting the current default route, for example) that are only intended
to be used via another interface (intended to be used when the first
interface is not available, for example). This, in turn, can lead to
spurious failures and timeouts simply because of the server address
order that resulted because of the enumeration process.
This patch makes the (safe?) assumption that there is no other better
rule to chose which interface's DNS server list should be prioritized.
After all, a DNS lookup isn't something "per network"; applications
don't look up "these DNS names on this interface and those DNS names on
that interface". There is a single resource pool of DNS servers and the
application should presume that any server will give it the "right"
answer. However, even if all DNS servers are assumed to give equally
useful responses, it is reasonable to expect that some DNS servers will
not accept requests on all interfaces. This patch avoids the problem by
sorting the DNS server addresses using the Windows IPv4/v6 routing tables.
For example, a request to DNS server C on interface 2 that is actually
sent over interface 1 (which may happen to have the default route) may
be rejected by or not delivered to DNS server C. So, better to use DNS
servers A and B associated with interface 1, at least as a first try.
By using the metric of the route to the DNS server itself as a proxy for
priority of the DNS server in the list, this patch is able to adapt
dynamically to changes in the interface list, the DNS server lists per
interface, which interfaces are active, the routing table, and so on,
while always picking a good "best" DNS server first.
In cases where any DNS server on any interface will do, this patch still
seems useful because it will prioritize a lower-metric route's (and thus
interface's) servers.
* Added test case macro to automatically run tests twice, once "normal",
once with virtual IO.
* Changed most "live" query tests to run in dual mode to verify
at least simple socket IO via virtual functions
* Added test case for settings/duping socket functions & callback data
Different platforms will do different numbers of allocations
in the processing of a given API call; just check that the
return code is either success or ENOMEM, and free off any
returned state in the former case.
Also cope with ECONNREFUSED as well as ENOTFOUND.
- Initial nmake file based off library nmake file
- Cast socket call arguments to (char *)
- Use wrapper sclose() that maps to closesocket() or close()
- Build a config.h indicating presence of headers
- Conditionally include netdb.h
- Remove unnecessary include of sys/socket.h
- Force longer bitmask for allocation failure tracking
- Call WSAStartup() / WSACleanup() in main()
- Set TCP_NODELAY for mock server
- Turn on tests in AppVeyor build
Include tests of internal functions, based on the value of the
CARES_SYMBOL_HIDING macro; need to configure the library with
--disable-symbol-hiding to enable these tests.
The tests are written in C++11, using the GoogleTest and GoogleMock
frameworks. They have their own independent autoconf setup, so that
users of the library need not have a C++ compiler just to get c-ares
working (however, the test/configure.ac file does assume the use of
a shared top-level m4/ directory). However, this autoconf setup has
only been tested on Linux and OSX so far.
Run with "./arestest", or "./arestest -v" to see extra debug info.
The GoogleTest options for running specific tests are also
available (e.g. "./arestest --gtest_filter=*Live*").
The tests are nowhere near complete yet (currently hitting around
60% coverage as reported by gcov), but they do include examples
of a few different styles of testing:
- There are live tests (ares-test-live.cc), which assume that the
current machine has a valid DNS setup and connection to the
internet; these tests issue queries for real domains but don't
particularly check what gets returned. The tests will fail on
an offline machine.
- There a few mock tests (ares-test-mock.cc) that set up a fake DNS
server and inject its port into the c-ares library configuration.
These tests allow specific response messages to be crafted and
injected, and so are likely to be used for many more tests in
future.
- To make this generation/injection easier, the dns-proto.h file
includes C++ helper classes for building DNS packets.
- Other library entrypoints that don't require network activity
(e.g. ares_parse_*_reply) are tested directly.
- There are few tests of library-internal functions that are not
normally visible to API users (in ares-test-internal.cc).
- A couple of the tests use a helper method of the test fixture to
inject memory allocation failures, using the earlier change to the
library to allow override of malloc/realloc/free.
- There is also an entrypoint to allow Clang's libfuzzer to drive
the packet parsing code in ares_parse_*_reply, together with a
standalone wrapper for it (./aresfuzz) to allow use of afl-fuzz
for further fuzz testing.
Brad House
Calle Wilund
David Drysdale