A C library for asynchronous DNS requests (grpc依赖)
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

869 lines
28 KiB

/* MIT License
*
* Copyright (c) The c-ares project and its contributors
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice (including the next
* paragraph) shall be included in all copies or substantial portions of the
* Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*
* SPDX-License-Identifier: MIT
*/
// -*- mode: c++ -*-
#ifndef ARES_TEST_H
#define ARES_TEST_H
#include "ares_setup.h"
#include "dns-proto.h"
// Include ares internal file for DNS protocol constants
#include "ares_nameser.h"
#include "gtest/gtest.h"
#include "gmock/gmock.h"
#if defined(HAVE_USER_NAMESPACE) && defined(HAVE_UTS_NAMESPACE)
# define HAVE_CONTAINER
#endif
#include <functional>
#include <list>
#include <map>
#include <memory>
#include <set>
#include <string>
#include <mutex>
#include <thread>
#include <utility>
#include <vector>
#include <chrono>
namespace ares {
typedef unsigned char byte;
namespace test {
extern bool verbose;
extern unsigned short mock_port;
extern const std::vector<int> both_families;
extern const std::vector<int> ipv4_family;
extern const std::vector<int> ipv6_family;
extern const std::vector<std::pair<int, bool>> both_families_both_modes;
extern const std::vector<std::pair<int, bool>> ipv4_family_both_modes;
extern const std::vector<std::pair<int, bool>> ipv6_family_both_modes;
9 months ago
extern const std::vector<std::tuple<ares_evsys_t, int, bool>>
all_evsys_ipv4_family_both_modes;
extern const std::vector<std::tuple<ares_evsys_t, int, bool>>
all_evsys_ipv6_family_both_modes;
extern const std::vector<std::tuple<ares_evsys_t, int, bool>>
all_evsys_both_families_both_modes;
extern const std::vector<std::tuple<ares_evsys_t, int>> all_evsys_ipv4_family;
extern const std::vector<std::tuple<ares_evsys_t, int>> all_evsys_ipv6_family;
extern const std::vector<std::tuple<ares_evsys_t, int>> all_evsys_both_families;
// Which parameters to use in tests
extern std::vector<int> families;
extern std::vector<std::tuple<ares_evsys_t, int>> evsys_families;
extern std::vector<std::pair<int, bool>> families_modes;
extern std::vector<std::tuple<ares_evsys_t, int, bool>> evsys_families_modes;
// Hopefully a more accurate sleep than sleep_for()
void ares_sleep_time(unsigned int ms);
// Process all pending work on ares-owned file descriptors, plus
// optionally the given set-of-FDs + work function.
1 year ago
void ProcessWork(ares_channel_t *channel,
std::function<std::set<ares_socket_t>()> get_extrafds,
std::function<void(ares_socket_t)> process_extra,
unsigned int cancel_ms = 0);
std::set<ares_socket_t> NoExtraFDs();
9 months ago
const char *af_tostr(int af);
const char *mode_tostr(bool mode);
std::string
PrintFamilyMode(const testing::TestParamInfo<std::pair<int, bool>> &info);
std::string PrintFamily(const testing::TestParamInfo<int> &info);
// Test fixture that ensures library initialization, and allows
// memory allocations to be failed.
class LibraryTest : public ::testing::Test {
public:
LibraryTest()
{
EXPECT_EQ(ARES_SUCCESS, ares_library_init_mem(
ARES_LIB_INIT_ALL, &LibraryTest::amalloc,
&LibraryTest::afree, &LibraryTest::arealloc));
}
~LibraryTest()
{
ares_library_cleanup();
ClearFails();
}
// Set the n-th malloc call (of any size) from the library to fail.
// (nth == 1 means the next call)
static void SetAllocFail(int nth);
// Set the next malloc call for the given size to fail.
static void SetAllocSizeFail(size_t size);
// Remove any pending alloc failures.
static void ClearFails();
static void *amalloc(size_t size);
static void *arealloc(void *ptr, size_t size);
static void afree(void *ptr);
private:
static bool ShouldAllocFail(size_t size);
static unsigned long long fails_;
static std::map<size_t, int> size_fails_;
static std::mutex lock_;
};
// Test fixture that uses a default channel.
class DefaultChannelTest : public LibraryTest {
public:
DefaultChannelTest() : channel_(nullptr)
{
/* Enable query cache for live tests */
struct ares_options opts;
memset(&opts, 0, sizeof(opts));
1 year ago
opts.qcache_max_ttl = 300;
int optmask = ARES_OPT_QUERY_CACHE;
EXPECT_EQ(ARES_SUCCESS, ares_init_options(&channel_, &opts, optmask));
EXPECT_NE(nullptr, channel_);
}
~DefaultChannelTest()
{
ares_destroy(channel_);
channel_ = nullptr;
}
// Process all pending work on ares-owned file descriptors.
void Process(unsigned int cancel_ms = 0);
protected:
`ares_channel` -> `ares_channel_t *`: don't bury the pointer (#595) `ares_channel` is defined as `typedef struct ares_channeldata *ares_channel;`. The problem with this, is it embeds the pointer into the typedef, which means an `ares_channel` can never be declared as `const` as if you write `const ares_channel channel`, that expands to `struct ares_channeldata * const ares_channel` and not `const struct ares_channeldata *channel`. We will now typedef `ares_channel_t` as `typedef struct ares_channeldata ares_channel_t;`, so if you write `const ares_channel_t *channel`, it properly expands to `const struct ares_channeldata *channel`. We are maintaining the old typedef for API compatibility with existing integrations, and due to typedef expansion this should not even cause any compiler warnings for existing code. There are no ABI implications with this change. I could be convinced to keep existing public functions as `ares_channel` if a sufficient argument exists, but internally we really need make this change for modern best practices. This change will allow us to internally use `const ares_channel_t *` where appropriate. Whether or not we decide to change any public interfaces to use `const` may require further discussion on if there might be ABI implications (I don't think so, but I'm also not 100% sure what a compiler internally does with `const` when emitting machine code ... I think more likely ABI implications would occur going the opposite direction). FYI, This PR was done via a combination of sed and clang-format, the only manual code change was the addition of the new typedef, and a couple doc fixes :) Fix By: Brad House (@bradh352)
1 year ago
ares_channel_t *channel_;
};
Replace hosts parser, add caching capabilities (#591) HOSTS FILE PROCESSING OVERVIEW ============================== The hosts file on the system contains static entries to be processed locally rather than querying the nameserver. Each row is an IP address followed by a list of space delimited hostnames that match the ip address. This is used for both forward and reverse lookups. We are caching the entire parsed hosts file for performance reasons. Some files may be quite sizable and as per Issue #458 can approach 1/2MB in size, and the parse overhead on a rapid succession of queries can be quite large. The entries are stored in forwards and backwards hashtables so we can get O(1) performance on lookup. The file is cached until the file modification timestamp changes (or 60s if there is no implemented stat() capability). The hosts file processing is quite unique. It has to merge all related hosts and ips into a single entry due to file formatting requirements. For instance take the below: ``` 127.0.0.1 localhost.localdomain localhost ::1 localhost.localdomain localhost 192.168.1.1 host.example.com host 192.168.1.5 host.example.com host 2620:1234::1 host.example.com host6.example.com host6 host ``` This will yield 2 entries. 1) ips: `127.0.0.1,::1` hosts: `localhost.localdomain,localhost` 2) ips: `192.168.1.1,192.168.1.5,2620:1234::1` hosts: `host.example.com,host,host6.example.com,host6` It could be argued that if searching for `192.168.1.1` that the `host6` hostnames should not be returned, but this implementation will return them since they are related (both ips have the fqdn of host.example.com). It is unlikely this will matter in the real world. Fix By: Brad House (@bradh352)
1 year ago
// Test fixture that uses a file-only channel.
class FileChannelTest : public LibraryTest {
public:
FileChannelTest() : channel_(nullptr)
{
struct ares_options opts;
memset(&opts, 0, sizeof(opts));
1 year ago
opts.lookups = strdup("f");
int optmask = ARES_OPT_LOOKUPS;
Replace hosts parser, add caching capabilities (#591) HOSTS FILE PROCESSING OVERVIEW ============================== The hosts file on the system contains static entries to be processed locally rather than querying the nameserver. Each row is an IP address followed by a list of space delimited hostnames that match the ip address. This is used for both forward and reverse lookups. We are caching the entire parsed hosts file for performance reasons. Some files may be quite sizable and as per Issue #458 can approach 1/2MB in size, and the parse overhead on a rapid succession of queries can be quite large. The entries are stored in forwards and backwards hashtables so we can get O(1) performance on lookup. The file is cached until the file modification timestamp changes (or 60s if there is no implemented stat() capability). The hosts file processing is quite unique. It has to merge all related hosts and ips into a single entry due to file formatting requirements. For instance take the below: ``` 127.0.0.1 localhost.localdomain localhost ::1 localhost.localdomain localhost 192.168.1.1 host.example.com host 192.168.1.5 host.example.com host 2620:1234::1 host.example.com host6.example.com host6 host ``` This will yield 2 entries. 1) ips: `127.0.0.1,::1` hosts: `localhost.localdomain,localhost` 2) ips: `192.168.1.1,192.168.1.5,2620:1234::1` hosts: `host.example.com,host,host6.example.com,host6` It could be argued that if searching for `192.168.1.1` that the `host6` hostnames should not be returned, but this implementation will return them since they are related (both ips have the fqdn of host.example.com). It is unlikely this will matter in the real world. Fix By: Brad House (@bradh352)
1 year ago
EXPECT_EQ(ARES_SUCCESS, ares_init_options(&channel_, &opts, optmask));
EXPECT_NE(nullptr, channel_);
free(opts.lookups);
}
~FileChannelTest()
{
ares_destroy(channel_);
channel_ = nullptr;
}
// Process all pending work on ares-owned file descriptors.
void Process(unsigned int cancel_ms = 0);
Replace hosts parser, add caching capabilities (#591) HOSTS FILE PROCESSING OVERVIEW ============================== The hosts file on the system contains static entries to be processed locally rather than querying the nameserver. Each row is an IP address followed by a list of space delimited hostnames that match the ip address. This is used for both forward and reverse lookups. We are caching the entire parsed hosts file for performance reasons. Some files may be quite sizable and as per Issue #458 can approach 1/2MB in size, and the parse overhead on a rapid succession of queries can be quite large. The entries are stored in forwards and backwards hashtables so we can get O(1) performance on lookup. The file is cached until the file modification timestamp changes (or 60s if there is no implemented stat() capability). The hosts file processing is quite unique. It has to merge all related hosts and ips into a single entry due to file formatting requirements. For instance take the below: ``` 127.0.0.1 localhost.localdomain localhost ::1 localhost.localdomain localhost 192.168.1.1 host.example.com host 192.168.1.5 host.example.com host 2620:1234::1 host.example.com host6.example.com host6 host ``` This will yield 2 entries. 1) ips: `127.0.0.1,::1` hosts: `localhost.localdomain,localhost` 2) ips: `192.168.1.1,192.168.1.5,2620:1234::1` hosts: `host.example.com,host,host6.example.com,host6` It could be argued that if searching for `192.168.1.1` that the `host6` hostnames should not be returned, but this implementation will return them since they are related (both ips have the fqdn of host.example.com). It is unlikely this will matter in the real world. Fix By: Brad House (@bradh352)
1 year ago
protected:
`ares_channel` -> `ares_channel_t *`: don't bury the pointer (#595) `ares_channel` is defined as `typedef struct ares_channeldata *ares_channel;`. The problem with this, is it embeds the pointer into the typedef, which means an `ares_channel` can never be declared as `const` as if you write `const ares_channel channel`, that expands to `struct ares_channeldata * const ares_channel` and not `const struct ares_channeldata *channel`. We will now typedef `ares_channel_t` as `typedef struct ares_channeldata ares_channel_t;`, so if you write `const ares_channel_t *channel`, it properly expands to `const struct ares_channeldata *channel`. We are maintaining the old typedef for API compatibility with existing integrations, and due to typedef expansion this should not even cause any compiler warnings for existing code. There are no ABI implications with this change. I could be convinced to keep existing public functions as `ares_channel` if a sufficient argument exists, but internally we really need make this change for modern best practices. This change will allow us to internally use `const ares_channel_t *` where appropriate. Whether or not we decide to change any public interfaces to use `const` may require further discussion on if there might be ABI implications (I don't think so, but I'm also not 100% sure what a compiler internally does with `const` when emitting machine code ... I think more likely ABI implications would occur going the opposite direction). FYI, This PR was done via a combination of sed and clang-format, the only manual code change was the addition of the new typedef, and a couple doc fixes :) Fix By: Brad House (@bradh352)
1 year ago
ares_channel_t *channel_;
Replace hosts parser, add caching capabilities (#591) HOSTS FILE PROCESSING OVERVIEW ============================== The hosts file on the system contains static entries to be processed locally rather than querying the nameserver. Each row is an IP address followed by a list of space delimited hostnames that match the ip address. This is used for both forward and reverse lookups. We are caching the entire parsed hosts file for performance reasons. Some files may be quite sizable and as per Issue #458 can approach 1/2MB in size, and the parse overhead on a rapid succession of queries can be quite large. The entries are stored in forwards and backwards hashtables so we can get O(1) performance on lookup. The file is cached until the file modification timestamp changes (or 60s if there is no implemented stat() capability). The hosts file processing is quite unique. It has to merge all related hosts and ips into a single entry due to file formatting requirements. For instance take the below: ``` 127.0.0.1 localhost.localdomain localhost ::1 localhost.localdomain localhost 192.168.1.1 host.example.com host 192.168.1.5 host.example.com host 2620:1234::1 host.example.com host6.example.com host6 host ``` This will yield 2 entries. 1) ips: `127.0.0.1,::1` hosts: `localhost.localdomain,localhost` 2) ips: `192.168.1.1,192.168.1.5,2620:1234::1` hosts: `host.example.com,host,host6.example.com,host6` It could be argued that if searching for `192.168.1.1` that the `host6` hostnames should not be returned, but this implementation will return them since they are related (both ips have the fqdn of host.example.com). It is unlikely this will matter in the real world. Fix By: Brad House (@bradh352)
1 year ago
};
// Test fixture that uses a default channel with the specified lookup mode.
class DefaultChannelModeTest
: public LibraryTest,
public ::testing::WithParamInterface<std::string> {
public:
DefaultChannelModeTest() : channel_(nullptr)
{
struct ares_options opts;
memset(&opts, 0, sizeof(opts));
1 year ago
opts.lookups = strdup(GetParam().c_str());
int optmask = ARES_OPT_LOOKUPS;
EXPECT_EQ(ARES_SUCCESS, ares_init_options(&channel_, &opts, optmask));
EXPECT_NE(nullptr, channel_);
free(opts.lookups);
}
~DefaultChannelModeTest()
{
ares_destroy(channel_);
channel_ = nullptr;
}
// Process all pending work on ares-owned file descriptors.
void Process(unsigned int cancel_ms = 0);
protected:
`ares_channel` -> `ares_channel_t *`: don't bury the pointer (#595) `ares_channel` is defined as `typedef struct ares_channeldata *ares_channel;`. The problem with this, is it embeds the pointer into the typedef, which means an `ares_channel` can never be declared as `const` as if you write `const ares_channel channel`, that expands to `struct ares_channeldata * const ares_channel` and not `const struct ares_channeldata *channel`. We will now typedef `ares_channel_t` as `typedef struct ares_channeldata ares_channel_t;`, so if you write `const ares_channel_t *channel`, it properly expands to `const struct ares_channeldata *channel`. We are maintaining the old typedef for API compatibility with existing integrations, and due to typedef expansion this should not even cause any compiler warnings for existing code. There are no ABI implications with this change. I could be convinced to keep existing public functions as `ares_channel` if a sufficient argument exists, but internally we really need make this change for modern best practices. This change will allow us to internally use `const ares_channel_t *` where appropriate. Whether or not we decide to change any public interfaces to use `const` may require further discussion on if there might be ABI implications (I don't think so, but I'm also not 100% sure what a compiler internally does with `const` when emitting machine code ... I think more likely ABI implications would occur going the opposite direction). FYI, This PR was done via a combination of sed and clang-format, the only manual code change was the addition of the new typedef, and a couple doc fixes :) Fix By: Brad House (@bradh352)
1 year ago
ares_channel_t *channel_;
};
DNS 0x20 implementation (#800) This PR enables DNS 0x20 as per https://datatracker.ietf.org/doc/html/draft-vixie-dnsext-dns0x20-00 . DNS 0x20 adds additional entropy to the request by randomly altering the case of the DNS question to help prevent cache poisoning attacks. Google DNS has implemented this support as of 2023, even though this is a proposed and expired standard from 2008: https://groups.google.com/g/public-dns-discuss/c/KxIDPOydA5M There have been documented cases of name server and caching server non-conformance, though it is expected to become more rare, especially since Google has started using this. This can be enabled via the `ARES_FLAG_DNS0x20` flag, which is currently disabled by default. The test cases do however enable this flag to validate this feature. Implementors using this flag will notice that responses will retain the mixed case, but since DNS names are case-insensitive, any proper implementation should not be impacted. There is currently no fallback mechanism implemented as it isn't immediately clear how this may affect a stub resolver like c-ares where we aren't querying the authoritative name server, but instead an intermediate recursive resolver where some domains may return invalid results while others return valid results, all while querying the same nameserver. Likely using DNS cookies as suggested by #620 is a better mechanism to fight cache poisoning attacks for stub resolvers. TCP queries do not use this feature even if the `ARES_FLAG_DNS0x20` flag is specified since they are not subject to cache poisoning attacks. Fixes Issue: #795 Fix By: Brad House (@bradh352)
5 months ago
// Mock DNS server to allow responses to be scripted by tests.
class MockServer {
public:
MockServer(int family, unsigned short port);
~MockServer();
// Mock method indicating the processing of a particular <name, RRtype>
// request.
MOCK_METHOD2(OnRequest, void(const std::string &name, int rrtype));
// Set the reply to be sent next; the query ID field will be overwritten
// with the value from the request.
void SetReplyData(const std::vector<byte> &reply)
{
DNS 0x20 implementation (#800) This PR enables DNS 0x20 as per https://datatracker.ietf.org/doc/html/draft-vixie-dnsext-dns0x20-00 . DNS 0x20 adds additional entropy to the request by randomly altering the case of the DNS question to help prevent cache poisoning attacks. Google DNS has implemented this support as of 2023, even though this is a proposed and expired standard from 2008: https://groups.google.com/g/public-dns-discuss/c/KxIDPOydA5M There have been documented cases of name server and caching server non-conformance, though it is expected to become more rare, especially since Google has started using this. This can be enabled via the `ARES_FLAG_DNS0x20` flag, which is currently disabled by default. The test cases do however enable this flag to validate this feature. Implementors using this flag will notice that responses will retain the mixed case, but since DNS names are case-insensitive, any proper implementation should not be impacted. There is currently no fallback mechanism implemented as it isn't immediately clear how this may affect a stub resolver like c-ares where we aren't querying the authoritative name server, but instead an intermediate recursive resolver where some domains may return invalid results while others return valid results, all while querying the same nameserver. Likely using DNS cookies as suggested by #620 is a better mechanism to fight cache poisoning attacks for stub resolvers. TCP queries do not use this feature even if the `ARES_FLAG_DNS0x20` flag is specified since they are not subject to cache poisoning attacks. Fixes Issue: #795 Fix By: Brad House (@bradh352)
5 months ago
exact_reply_ = reply;
reply_ = nullptr;
}
void SetReply(const DNSPacket *reply)
{
DNS 0x20 implementation (#800) This PR enables DNS 0x20 as per https://datatracker.ietf.org/doc/html/draft-vixie-dnsext-dns0x20-00 . DNS 0x20 adds additional entropy to the request by randomly altering the case of the DNS question to help prevent cache poisoning attacks. Google DNS has implemented this support as of 2023, even though this is a proposed and expired standard from 2008: https://groups.google.com/g/public-dns-discuss/c/KxIDPOydA5M There have been documented cases of name server and caching server non-conformance, though it is expected to become more rare, especially since Google has started using this. This can be enabled via the `ARES_FLAG_DNS0x20` flag, which is currently disabled by default. The test cases do however enable this flag to validate this feature. Implementors using this flag will notice that responses will retain the mixed case, but since DNS names are case-insensitive, any proper implementation should not be impacted. There is currently no fallback mechanism implemented as it isn't immediately clear how this may affect a stub resolver like c-ares where we aren't querying the authoritative name server, but instead an intermediate recursive resolver where some domains may return invalid results while others return valid results, all while querying the same nameserver. Likely using DNS cookies as suggested by #620 is a better mechanism to fight cache poisoning attacks for stub resolvers. TCP queries do not use this feature even if the `ARES_FLAG_DNS0x20` flag is specified since they are not subject to cache poisoning attacks. Fixes Issue: #795 Fix By: Brad House (@bradh352)
5 months ago
reply_ = reply;
exact_reply_.clear();
}
Add function ares_search_dnrec() to search for records using the new DNS record parser (#719) This PR adds a new function `ares_search_dnsrec()` to search for records using the new DNS record parser. The function takes an arbitrary DNS record object to search (that must represent a query for a single name). The function takes a new callback type, `ares_callback_dnsrec`, that is invoked with a parsed DNS record object rather than the raw buffer(+length). The original motivation for this change is to provide support for [draft-kaplan-enum-sip-routing-04](https://datatracker.ietf.org/doc/html/draft-kaplan-enum-sip-routing-04); when routing phone calls using an ENUM server, it can be useful to include identifying source information in an OPT RR options value, to help select the appropriate route for the call. The new function allows for more customisable searches like this. **Summary of code changes** A new function `ares_search_dnsrec()` has been added and exposed. Moreover, the entire `ares_search_int()` internal code flow has been refactored to use parsed DNS record objects and the new DNS record parser. The DNS record object is passed through the `search_query` structure by encoding/decoding to/from a buffer (if multiple search domains are used). A helper function `ares_dns_write_query_altname()` is used to re-write the DNS record object with a new query name (used to append search domains). `ares_search()` is now a wrapper around the new internal code, where the DNS record object is created based on the name, class and type parameters. The new function uses a new callback type, `ares_callback_dnsrec`. This is invoked with a parsed DNS record object. For now, we convert from `ares_callback` to this new type using `ares__dnsrec_convert_cb()`. Some functions that are common to both `ares_query()` and `ares_search()` have been refactored using the new DNS record parser. See `ares_dns_record_create_query()` and `ares_dns_query_reply_tostatus()`. **Testing** A new FV has been added to test the new function, which searches for a DNS record containing an OPT RR with custom options value. As part of this, I needed to enhance the mock DNS server to expect request text (and assert that it matches actual request text). This is because the FV needs to check that the request contains the correct OPT RR. **Documentation** The man page docs have been updated to describe the new feature. **Futures** In the future, a new variant of `ares_send()` could be introduced in the same vein (`ares_send_dnsrec()`). This could be used by `ares_search_dnsrec()`. Moreover, we could migrate internal code to use `ares_callback_dnsrec` as the default callback. This will help to make the new DNS record parser the norm in C-Ares. --------- Co-authored-by: Oliver Welsh (@oliverwelsh)
8 months ago
// Set the reply to be sent next as well as the request (in string form) that
// the server should expect to receive; the query ID field in the reply will
// be overwritten with the value from the request.
void SetReplyExpRequest(const DNSPacket *reply, const std::string &request)
{
expected_request_ = request;
DNS 0x20 implementation (#800) This PR enables DNS 0x20 as per https://datatracker.ietf.org/doc/html/draft-vixie-dnsext-dns0x20-00 . DNS 0x20 adds additional entropy to the request by randomly altering the case of the DNS question to help prevent cache poisoning attacks. Google DNS has implemented this support as of 2023, even though this is a proposed and expired standard from 2008: https://groups.google.com/g/public-dns-discuss/c/KxIDPOydA5M There have been documented cases of name server and caching server non-conformance, though it is expected to become more rare, especially since Google has started using this. This can be enabled via the `ARES_FLAG_DNS0x20` flag, which is currently disabled by default. The test cases do however enable this flag to validate this feature. Implementors using this flag will notice that responses will retain the mixed case, but since DNS names are case-insensitive, any proper implementation should not be impacted. There is currently no fallback mechanism implemented as it isn't immediately clear how this may affect a stub resolver like c-ares where we aren't querying the authoritative name server, but instead an intermediate recursive resolver where some domains may return invalid results while others return valid results, all while querying the same nameserver. Likely using DNS cookies as suggested by #620 is a better mechanism to fight cache poisoning attacks for stub resolvers. TCP queries do not use this feature even if the `ARES_FLAG_DNS0x20` flag is specified since they are not subject to cache poisoning attacks. Fixes Issue: #795 Fix By: Brad House (@bradh352)
5 months ago
reply_ = reply;
Add function ares_search_dnrec() to search for records using the new DNS record parser (#719) This PR adds a new function `ares_search_dnsrec()` to search for records using the new DNS record parser. The function takes an arbitrary DNS record object to search (that must represent a query for a single name). The function takes a new callback type, `ares_callback_dnsrec`, that is invoked with a parsed DNS record object rather than the raw buffer(+length). The original motivation for this change is to provide support for [draft-kaplan-enum-sip-routing-04](https://datatracker.ietf.org/doc/html/draft-kaplan-enum-sip-routing-04); when routing phone calls using an ENUM server, it can be useful to include identifying source information in an OPT RR options value, to help select the appropriate route for the call. The new function allows for more customisable searches like this. **Summary of code changes** A new function `ares_search_dnsrec()` has been added and exposed. Moreover, the entire `ares_search_int()` internal code flow has been refactored to use parsed DNS record objects and the new DNS record parser. The DNS record object is passed through the `search_query` structure by encoding/decoding to/from a buffer (if multiple search domains are used). A helper function `ares_dns_write_query_altname()` is used to re-write the DNS record object with a new query name (used to append search domains). `ares_search()` is now a wrapper around the new internal code, where the DNS record object is created based on the name, class and type parameters. The new function uses a new callback type, `ares_callback_dnsrec`. This is invoked with a parsed DNS record object. For now, we convert from `ares_callback` to this new type using `ares__dnsrec_convert_cb()`. Some functions that are common to both `ares_query()` and `ares_search()` have been refactored using the new DNS record parser. See `ares_dns_record_create_query()` and `ares_dns_query_reply_tostatus()`. **Testing** A new FV has been added to test the new function, which searches for a DNS record containing an OPT RR with custom options value. As part of this, I needed to enhance the mock DNS server to expect request text (and assert that it matches actual request text). This is because the FV needs to check that the request contains the correct OPT RR. **Documentation** The man page docs have been updated to describe the new feature. **Futures** In the future, a new variant of `ares_send()` could be introduced in the same vein (`ares_send_dnsrec()`). This could be used by `ares_search_dnsrec()`. Moreover, we could migrate internal code to use `ares_callback_dnsrec` as the default callback. This will help to make the new DNS record parser the norm in C-Ares. --------- Co-authored-by: Oliver Welsh (@oliverwelsh)
8 months ago
}
void SetReplyQID(int qid)
{
qid_ = qid;
}
void Disconnect()
{
DNS 0x20 implementation (#800) This PR enables DNS 0x20 as per https://datatracker.ietf.org/doc/html/draft-vixie-dnsext-dns0x20-00 . DNS 0x20 adds additional entropy to the request by randomly altering the case of the DNS question to help prevent cache poisoning attacks. Google DNS has implemented this support as of 2023, even though this is a proposed and expired standard from 2008: https://groups.google.com/g/public-dns-discuss/c/KxIDPOydA5M There have been documented cases of name server and caching server non-conformance, though it is expected to become more rare, especially since Google has started using this. This can be enabled via the `ARES_FLAG_DNS0x20` flag, which is currently disabled by default. The test cases do however enable this flag to validate this feature. Implementors using this flag will notice that responses will retain the mixed case, but since DNS names are case-insensitive, any proper implementation should not be impacted. There is currently no fallback mechanism implemented as it isn't immediately clear how this may affect a stub resolver like c-ares where we aren't querying the authoritative name server, but instead an intermediate recursive resolver where some domains may return invalid results while others return valid results, all while querying the same nameserver. Likely using DNS cookies as suggested by #620 is a better mechanism to fight cache poisoning attacks for stub resolvers. TCP queries do not use this feature even if the `ARES_FLAG_DNS0x20` flag is specified since they are not subject to cache poisoning attacks. Fixes Issue: #795 Fix By: Brad House (@bradh352)
5 months ago
reply_ = nullptr;
exact_reply_.clear();
for (ares_socket_t fd : connfds_) {
sclose(fd);
}
connfds_.clear();
free(tcp_data_);
tcp_data_ = NULL;
tcp_data_len_ = 0;
}
// The set of file descriptors that the server handles.
std::set<ares_socket_t> fds() const;
// Process activity on a file descriptor.
1 year ago
void ProcessFD(ares_socket_t fd);
// Ports the server is responding to
1 year ago
unsigned short udpport() const
{
return udpport_;
}
unsigned short tcpport() const
{
return tcpport_;
}
private:
1 year ago
void ProcessRequest(ares_socket_t fd, struct sockaddr_storage *addr,
Add function ares_search_dnrec() to search for records using the new DNS record parser (#719) This PR adds a new function `ares_search_dnsrec()` to search for records using the new DNS record parser. The function takes an arbitrary DNS record object to search (that must represent a query for a single name). The function takes a new callback type, `ares_callback_dnsrec`, that is invoked with a parsed DNS record object rather than the raw buffer(+length). The original motivation for this change is to provide support for [draft-kaplan-enum-sip-routing-04](https://datatracker.ietf.org/doc/html/draft-kaplan-enum-sip-routing-04); when routing phone calls using an ENUM server, it can be useful to include identifying source information in an OPT RR options value, to help select the appropriate route for the call. The new function allows for more customisable searches like this. **Summary of code changes** A new function `ares_search_dnsrec()` has been added and exposed. Moreover, the entire `ares_search_int()` internal code flow has been refactored to use parsed DNS record objects and the new DNS record parser. The DNS record object is passed through the `search_query` structure by encoding/decoding to/from a buffer (if multiple search domains are used). A helper function `ares_dns_write_query_altname()` is used to re-write the DNS record object with a new query name (used to append search domains). `ares_search()` is now a wrapper around the new internal code, where the DNS record object is created based on the name, class and type parameters. The new function uses a new callback type, `ares_callback_dnsrec`. This is invoked with a parsed DNS record object. For now, we convert from `ares_callback` to this new type using `ares__dnsrec_convert_cb()`. Some functions that are common to both `ares_query()` and `ares_search()` have been refactored using the new DNS record parser. See `ares_dns_record_create_query()` and `ares_dns_query_reply_tostatus()`. **Testing** A new FV has been added to test the new function, which searches for a DNS record containing an OPT RR with custom options value. As part of this, I needed to enhance the mock DNS server to expect request text (and assert that it matches actual request text). This is because the FV needs to check that the request contains the correct OPT RR. **Documentation** The man page docs have been updated to describe the new feature. **Futures** In the future, a new variant of `ares_send()` could be introduced in the same vein (`ares_send_dnsrec()`). This could be used by `ares_search_dnsrec()`. Moreover, we could migrate internal code to use `ares_callback_dnsrec` as the default callback. This will help to make the new DNS record parser the norm in C-Ares. --------- Co-authored-by: Oliver Welsh (@oliverwelsh)
8 months ago
ares_socklen_t addrlen, const std::string &reqstr,
DNS 0x20 implementation (#800) This PR enables DNS 0x20 as per https://datatracker.ietf.org/doc/html/draft-vixie-dnsext-dns0x20-00 . DNS 0x20 adds additional entropy to the request by randomly altering the case of the DNS question to help prevent cache poisoning attacks. Google DNS has implemented this support as of 2023, even though this is a proposed and expired standard from 2008: https://groups.google.com/g/public-dns-discuss/c/KxIDPOydA5M There have been documented cases of name server and caching server non-conformance, though it is expected to become more rare, especially since Google has started using this. This can be enabled via the `ARES_FLAG_DNS0x20` flag, which is currently disabled by default. The test cases do however enable this flag to validate this feature. Implementors using this flag will notice that responses will retain the mixed case, but since DNS names are case-insensitive, any proper implementation should not be impacted. There is currently no fallback mechanism implemented as it isn't immediately clear how this may affect a stub resolver like c-ares where we aren't querying the authoritative name server, but instead an intermediate recursive resolver where some domains may return invalid results while others return valid results, all while querying the same nameserver. Likely using DNS cookies as suggested by #620 is a better mechanism to fight cache poisoning attacks for stub resolvers. TCP queries do not use this feature even if the `ARES_FLAG_DNS0x20` flag is specified since they are not subject to cache poisoning attacks. Fixes Issue: #795 Fix By: Brad House (@bradh352)
5 months ago
int qid, const char *name, int rrtype);
1 year ago
void ProcessPacket(ares_socket_t fd, struct sockaddr_storage *addr,
ares_socklen_t addrlen, byte *data, int len);
unsigned short udpport_;
unsigned short tcpport_;
ares_socket_t udpfd_;
ares_socket_t tcpfd_;
std::set<ares_socket_t> connfds_;
DNS 0x20 implementation (#800) This PR enables DNS 0x20 as per https://datatracker.ietf.org/doc/html/draft-vixie-dnsext-dns0x20-00 . DNS 0x20 adds additional entropy to the request by randomly altering the case of the DNS question to help prevent cache poisoning attacks. Google DNS has implemented this support as of 2023, even though this is a proposed and expired standard from 2008: https://groups.google.com/g/public-dns-discuss/c/KxIDPOydA5M There have been documented cases of name server and caching server non-conformance, though it is expected to become more rare, especially since Google has started using this. This can be enabled via the `ARES_FLAG_DNS0x20` flag, which is currently disabled by default. The test cases do however enable this flag to validate this feature. Implementors using this flag will notice that responses will retain the mixed case, but since DNS names are case-insensitive, any proper implementation should not be impacted. There is currently no fallback mechanism implemented as it isn't immediately clear how this may affect a stub resolver like c-ares where we aren't querying the authoritative name server, but instead an intermediate recursive resolver where some domains may return invalid results while others return valid results, all while querying the same nameserver. Likely using DNS cookies as suggested by #620 is a better mechanism to fight cache poisoning attacks for stub resolvers. TCP queries do not use this feature even if the `ARES_FLAG_DNS0x20` flag is specified since they are not subject to cache poisoning attacks. Fixes Issue: #795 Fix By: Brad House (@bradh352)
5 months ago
std::vector<byte> exact_reply_;
const DNSPacket *reply_;
Add function ares_search_dnrec() to search for records using the new DNS record parser (#719) This PR adds a new function `ares_search_dnsrec()` to search for records using the new DNS record parser. The function takes an arbitrary DNS record object to search (that must represent a query for a single name). The function takes a new callback type, `ares_callback_dnsrec`, that is invoked with a parsed DNS record object rather than the raw buffer(+length). The original motivation for this change is to provide support for [draft-kaplan-enum-sip-routing-04](https://datatracker.ietf.org/doc/html/draft-kaplan-enum-sip-routing-04); when routing phone calls using an ENUM server, it can be useful to include identifying source information in an OPT RR options value, to help select the appropriate route for the call. The new function allows for more customisable searches like this. **Summary of code changes** A new function `ares_search_dnsrec()` has been added and exposed. Moreover, the entire `ares_search_int()` internal code flow has been refactored to use parsed DNS record objects and the new DNS record parser. The DNS record object is passed through the `search_query` structure by encoding/decoding to/from a buffer (if multiple search domains are used). A helper function `ares_dns_write_query_altname()` is used to re-write the DNS record object with a new query name (used to append search domains). `ares_search()` is now a wrapper around the new internal code, where the DNS record object is created based on the name, class and type parameters. The new function uses a new callback type, `ares_callback_dnsrec`. This is invoked with a parsed DNS record object. For now, we convert from `ares_callback` to this new type using `ares__dnsrec_convert_cb()`. Some functions that are common to both `ares_query()` and `ares_search()` have been refactored using the new DNS record parser. See `ares_dns_record_create_query()` and `ares_dns_query_reply_tostatus()`. **Testing** A new FV has been added to test the new function, which searches for a DNS record containing an OPT RR with custom options value. As part of this, I needed to enhance the mock DNS server to expect request text (and assert that it matches actual request text). This is because the FV needs to check that the request contains the correct OPT RR. **Documentation** The man page docs have been updated to describe the new feature. **Futures** In the future, a new variant of `ares_send()` could be introduced in the same vein (`ares_send_dnsrec()`). This could be used by `ares_search_dnsrec()`. Moreover, we could migrate internal code to use `ares_callback_dnsrec` as the default callback. This will help to make the new DNS record parser the norm in C-Ares. --------- Co-authored-by: Oliver Welsh (@oliverwelsh)
8 months ago
std::string expected_request_;
1 year ago
int qid_;
unsigned char *tcp_data_;
size_t tcp_data_len_;
};
// Test fixture that uses a mock DNS server.
class MockChannelOptsTest : public LibraryTest {
public:
MockChannelOptsTest(int count, int family, bool force_tcp, bool honor_sysconfig,
struct ares_options *givenopts, int optmask);
~MockChannelOptsTest();
// Process all pending work on ares-owned and mock-server-owned file
// descriptors.
void Process(unsigned int cancel_ms = 0);
protected:
// NiceMockServer doesn't complain about uninteresting calls.
typedef testing::NiceMock<MockServer> NiceMockServer;
typedef std::vector<std::unique_ptr<NiceMockServer>> NiceMockServers;
std::set<ares_socket_t> fds() const;
1 year ago
void ProcessFD(ares_socket_t fd);
1 year ago
static NiceMockServers BuildServers(int count, int family,
unsigned short base_port);
NiceMockServers servers_;
// Convenience reference to first server.
NiceMockServer &server_;
1 year ago
ares_channel_t *channel_;
};
class MockChannelTest
: public MockChannelOptsTest,
public ::testing::WithParamInterface<std::pair<int, bool>> {
public:
MockChannelTest()
: MockChannelOptsTest(1, GetParam().first, GetParam().second, false, nullptr, 0)
{
}
};
class MockUDPChannelTest : public MockChannelOptsTest,
public ::testing::WithParamInterface<int> {
public:
MockUDPChannelTest() : MockChannelOptsTest(1, GetParam(), false, false, nullptr, 0)
{
}
};
class MockTCPChannelTest : public MockChannelOptsTest,
public ::testing::WithParamInterface<int> {
public:
MockTCPChannelTest() : MockChannelOptsTest(1, GetParam(), true, false, nullptr, 0)
{
}
};
class MockEventThreadOptsTest : public MockChannelOptsTest {
public:
9 months ago
MockEventThreadOptsTest(int count, ares_evsys_t evsys, int family,
bool force_tcp, struct ares_options *givenopts,
int optmask)
: MockChannelOptsTest(count, family, force_tcp, false,
9 months ago
FillOptionsET(&evopts_, givenopts, evsys),
optmask | ARES_OPT_EVENT_THREAD)
{
}
9 months ago
~MockEventThreadOptsTest()
{
}
9 months ago
static struct ares_options *FillOptionsET(struct ares_options *opts,
struct ares_options *givenopts,
ares_evsys_t evsys)
{
if (givenopts) {
memcpy(opts, givenopts, sizeof(*opts));
} else {
memset(opts, 0, sizeof(*opts));
}
opts->evsys = evsys;
return opts;
}
void Process(unsigned int cancel_ms = 0);
private:
struct ares_options evopts_;
};
class MockEventThreadTest
: public MockEventThreadOptsTest,
public ::testing::WithParamInterface<std::tuple<ares_evsys_t, int, bool>> {
public:
MockEventThreadTest()
9 months ago
: MockEventThreadOptsTest(1, std::get<0>(GetParam()),
std::get<1>(GetParam()), std::get<2>(GetParam()),
nullptr, 0)
{
}
};
9 months ago
class MockUDPEventThreadTest
: public MockEventThreadOptsTest,
public ::testing::WithParamInterface<std::tuple<ares_evsys_t, int>> {
public:
9 months ago
MockUDPEventThreadTest()
: MockEventThreadOptsTest(1, std::get<0>(GetParam()),
std::get<1>(GetParam()), false, nullptr, 0)
{
}
};
9 months ago
class MockTCPEventThreadTest
: public MockEventThreadOptsTest,
public ::testing::WithParamInterface<std::tuple<ares_evsys_t, int>> {
public:
9 months ago
MockTCPEventThreadTest()
: MockEventThreadOptsTest(1, std::get<0>(GetParam()),
std::get<1>(GetParam()), true, nullptr, 0)
{
}
};
// gMock action to set the reply for a mock server.
ACTION_P2(SetReplyData, mockserver, data)
{
mockserver->SetReplyData(data);
}
ACTION_P2(SetReply, mockserver, reply)
{
mockserver->SetReply(reply);
}
Add function ares_search_dnrec() to search for records using the new DNS record parser (#719) This PR adds a new function `ares_search_dnsrec()` to search for records using the new DNS record parser. The function takes an arbitrary DNS record object to search (that must represent a query for a single name). The function takes a new callback type, `ares_callback_dnsrec`, that is invoked with a parsed DNS record object rather than the raw buffer(+length). The original motivation for this change is to provide support for [draft-kaplan-enum-sip-routing-04](https://datatracker.ietf.org/doc/html/draft-kaplan-enum-sip-routing-04); when routing phone calls using an ENUM server, it can be useful to include identifying source information in an OPT RR options value, to help select the appropriate route for the call. The new function allows for more customisable searches like this. **Summary of code changes** A new function `ares_search_dnsrec()` has been added and exposed. Moreover, the entire `ares_search_int()` internal code flow has been refactored to use parsed DNS record objects and the new DNS record parser. The DNS record object is passed through the `search_query` structure by encoding/decoding to/from a buffer (if multiple search domains are used). A helper function `ares_dns_write_query_altname()` is used to re-write the DNS record object with a new query name (used to append search domains). `ares_search()` is now a wrapper around the new internal code, where the DNS record object is created based on the name, class and type parameters. The new function uses a new callback type, `ares_callback_dnsrec`. This is invoked with a parsed DNS record object. For now, we convert from `ares_callback` to this new type using `ares__dnsrec_convert_cb()`. Some functions that are common to both `ares_query()` and `ares_search()` have been refactored using the new DNS record parser. See `ares_dns_record_create_query()` and `ares_dns_query_reply_tostatus()`. **Testing** A new FV has been added to test the new function, which searches for a DNS record containing an OPT RR with custom options value. As part of this, I needed to enhance the mock DNS server to expect request text (and assert that it matches actual request text). This is because the FV needs to check that the request contains the correct OPT RR. **Documentation** The man page docs have been updated to describe the new feature. **Futures** In the future, a new variant of `ares_send()` could be introduced in the same vein (`ares_send_dnsrec()`). This could be used by `ares_search_dnsrec()`. Moreover, we could migrate internal code to use `ares_callback_dnsrec` as the default callback. This will help to make the new DNS record parser the norm in C-Ares. --------- Co-authored-by: Oliver Welsh (@oliverwelsh)
8 months ago
// gMock action to set the reply for a mock server, as well as the request (in
// string form) that the server should expect to receive.
ACTION_P3(SetReplyExpRequest, mockserver, reply, request)
{
mockserver->SetReplyExpRequest(reply, request);
}
ACTION_P2(SetReplyQID, mockserver, qid)
{
mockserver->SetReplyQID(qid);
}
// gMock action to cancel a channel.
ACTION_P2(CancelChannel, mockserver, channel)
{
ares_cancel(channel);
}
// gMock action to disconnect all connections.
ACTION_P(Disconnect, mockserver)
{
mockserver->Disconnect();
}
// C++ wrapper for struct hostent.
struct HostEnt {
HostEnt() : addrtype_(-1)
{
}
HostEnt(const struct hostent *hostent);
std::string name_;
std::vector<std::string> aliases_;
int addrtype_; // AF_INET or AF_INET6
std::vector<std::string> addrs_;
};
std::ostream &operator<<(std::ostream &os, const HostEnt &result);
// Structure that describes the result of an ares_host_callback invocation.
struct HostResult {
HostResult() : done_(false), status_(0), timeouts_(0)
{
}
// Whether the callback has been invoked.
bool done_;
// Explicitly provided result information.
int status_;
int timeouts_;
// Contents of the hostent structure, if provided.
HostEnt host_;
};
std::ostream &operator<<(std::ostream &os, const HostResult &result);
// C++ wrapper for ares_dns_record_t.
struct AresDnsRecord {
~AresDnsRecord()
{
ares_dns_record_destroy(dnsrec_);
dnsrec_ = NULL;
}
AresDnsRecord() : dnsrec_(NULL)
{
}
void SetDnsRecord(const ares_dns_record_t *dnsrec)
{
if (dnsrec_ != NULL) {
ares_dns_record_destroy(dnsrec_);
}
if (dnsrec == NULL) {
return;
}
dnsrec_ = ares_dns_record_duplicate(dnsrec);
}
ares_dns_record_t *dnsrec_ = NULL;
};
std::ostream &operator<<(std::ostream &os, const AresDnsRecord &result);
// Structure that describes the result of an ares_host_callback invocation.
struct QueryResult {
QueryResult() : done_(false), status_(ARES_SUCCESS), timeouts_(0)
{
}
// Whether the callback has been invoked.
bool done_;
// Explicitly provided result information.
ares_status_t status_;
size_t timeouts_;
// Contents of the ares_dns_record_t structure if provided
AresDnsRecord dnsrec_;
};
std::ostream &operator<<(std::ostream &os, const QueryResult &result);
// Structure that describes the result of an ares_callback invocation.
struct SearchResult {
// Whether the callback has been invoked.
bool done_;
// Explicitly provided result information.
int status_;
int timeouts_;
std::vector<byte> data_;
};
std::ostream &operator<<(std::ostream &os, const SearchResult &result);
// Structure that describes the result of an ares_nameinfo_callback invocation.
struct NameInfoResult {
// Whether the callback has been invoked.
bool done_;
// Explicitly provided result information.
int status_;
int timeouts_;
std::string node_;
std::string service_;
};
std::ostream &operator<<(std::ostream &os, const NameInfoResult &result);
struct AddrInfoDeleter {
void operator()(ares_addrinfo *ptr)
{
if (ptr) {
ares_freeaddrinfo(ptr);
}
}
};
// C++ wrapper for struct ares_addrinfo.
using AddrInfo = std::unique_ptr<ares_addrinfo, AddrInfoDeleter>;
std::ostream &operator<<(std::ostream &os, const AddrInfo &result);
// Structure that describes the result of an ares_addrinfo_callback invocation.
struct AddrInfoResult {
AddrInfoResult() : done_(false), status_(-1), timeouts_(0)
{
}
// Whether the callback has been invoked.
bool done_;
// Explicitly provided result information.
int status_;
int timeouts_;
// Contents of the ares_addrinfo structure, if provided.
AddrInfo ai_;
};
std::ostream &operator<<(std::ostream &os, const AddrInfoResult &result);
// Standard implementation of ares callbacks that fill out the corresponding
// structures.
void HostCallback(void *data, int status, int timeouts,
struct hostent *hostent);
void QueryCallback(void *data, ares_status_t status, size_t timeouts,
const ares_dns_record_t *dnsrec);
void SearchCallback(void *data, int status, int timeouts, unsigned char *abuf,
int alen);
Add function ares_search_dnrec() to search for records using the new DNS record parser (#719) This PR adds a new function `ares_search_dnsrec()` to search for records using the new DNS record parser. The function takes an arbitrary DNS record object to search (that must represent a query for a single name). The function takes a new callback type, `ares_callback_dnsrec`, that is invoked with a parsed DNS record object rather than the raw buffer(+length). The original motivation for this change is to provide support for [draft-kaplan-enum-sip-routing-04](https://datatracker.ietf.org/doc/html/draft-kaplan-enum-sip-routing-04); when routing phone calls using an ENUM server, it can be useful to include identifying source information in an OPT RR options value, to help select the appropriate route for the call. The new function allows for more customisable searches like this. **Summary of code changes** A new function `ares_search_dnsrec()` has been added and exposed. Moreover, the entire `ares_search_int()` internal code flow has been refactored to use parsed DNS record objects and the new DNS record parser. The DNS record object is passed through the `search_query` structure by encoding/decoding to/from a buffer (if multiple search domains are used). A helper function `ares_dns_write_query_altname()` is used to re-write the DNS record object with a new query name (used to append search domains). `ares_search()` is now a wrapper around the new internal code, where the DNS record object is created based on the name, class and type parameters. The new function uses a new callback type, `ares_callback_dnsrec`. This is invoked with a parsed DNS record object. For now, we convert from `ares_callback` to this new type using `ares__dnsrec_convert_cb()`. Some functions that are common to both `ares_query()` and `ares_search()` have been refactored using the new DNS record parser. See `ares_dns_record_create_query()` and `ares_dns_query_reply_tostatus()`. **Testing** A new FV has been added to test the new function, which searches for a DNS record containing an OPT RR with custom options value. As part of this, I needed to enhance the mock DNS server to expect request text (and assert that it matches actual request text). This is because the FV needs to check that the request contains the correct OPT RR. **Documentation** The man page docs have been updated to describe the new feature. **Futures** In the future, a new variant of `ares_send()` could be introduced in the same vein (`ares_send_dnsrec()`). This could be used by `ares_search_dnsrec()`. Moreover, we could migrate internal code to use `ares_callback_dnsrec` as the default callback. This will help to make the new DNS record parser the norm in C-Ares. --------- Co-authored-by: Oliver Welsh (@oliverwelsh)
8 months ago
void SearchCallbackDnsRec(void *data, ares_status_t status, size_t timeouts,
const ares_dns_record_t *dnsrec);
void NameInfoCallback(void *data, int status, int timeouts, char *node,
char *service);
void AddrInfoCallback(void *data, int status, int timeouts,
struct ares_addrinfo *res);
// Retrieve the name servers used by a channel.
std::string GetNameServers(ares_channel_t *channel);
// RAII class to temporarily create a directory of a given name.
class TransientDir {
public:
TransientDir(const std::string &dirname);
~TransientDir();
private:
std::string dirname_;
};
// C++ wrapper around tempnam()
std::string TempNam(const char *dir, const char *prefix);
// RAII class to temporarily create file of a given name and contents.
class TransientFile {
public:
TransientFile(const std::string &filename, const std::string &contents);
~TransientFile();
protected:
std::string filename_;
};
// RAII class for a temporary file with the given contents.
class TempFile : public TransientFile {
public:
TempFile(const std::string &contents);
const char *filename() const
{
return filename_.c_str();
}
};
#ifdef _WIN32
extern "C" {
static int setenv(const char *name, const char *value, int overwrite)
{
char *buffer;
size_t buf_size;
if (name == NULL) {
return -1;
}
if (value == NULL) {
value = ""; /* For unset */
}
if (!overwrite && getenv(name) != NULL) {
return -1;
}
buf_size = strlen(name) + strlen(value) + 1 /* = */ + 1 /* NULL */;
buffer = (char *)malloc(buf_size);
_snprintf(buffer, buf_size, "%s=%s", name, value);
_putenv(buffer);
free(buffer);
return 0;
}
static int unsetenv(const char *name)
{
return setenv(name, NULL, 1);
}
} /* extern "C" */
#endif
// RAII class for a temporary environment variable value.
class EnvValue {
public:
EnvValue(const char *name, const char *value) : name_(name), restore_(false)
{
char *original = getenv(name);
if (original) {
restore_ = true;
original_ = original;
}
setenv(name_.c_str(), value, 1);
}
~EnvValue()
{
if (restore_) {
setenv(name_.c_str(), original_.c_str(), 1);
} else {
unsetenv(name_.c_str());
}
}
private:
std::string name_;
bool restore_;
std::string original_;
};
#ifdef HAVE_CONTAINER
// Linux-specific functionality for running code in a container, implemented
// in ares-test-ns.cc
typedef std::function<int(void)> VoidToIntFn;
typedef std::vector<std::pair<std::string, std::string>> NameContentList;
class ContainerFilesystem {
public:
ContainerFilesystem(NameContentList files, const std::string &mountpt);
~ContainerFilesystem();
std::string root() const
{
return rootdir_;
}
std::string mountpt() const
{
return mountpt_;
}
private:
void EnsureDirExists(const std::string &dir);
std::string rootdir_;
std::string mountpt_;
std::list<std::string> dirs_;
std::vector<std::unique_ptr<TransientFile>> files_;
};
int RunInContainer(ContainerFilesystem *fs, const std::string &hostname,
const std::string &domainname, VoidToIntFn fn);
# define ICLASS_NAME(casename, testname) Contained##casename##_##testname
# define CONTAINED_TEST_F(casename, testname, hostname, domainname, files) \
class ICLASS_NAME(casename, testname) : public casename { \
public: \
ICLASS_NAME(casename, testname)() \
{ \
} \
static int InnerTestBody(); \
}; \
TEST_F(ICLASS_NAME(casename, testname), _) \
{ \
ContainerFilesystem chroot(files, ".."); \
VoidToIntFn fn(ICLASS_NAME(casename, testname)::InnerTestBody); \
EXPECT_EQ(0, RunInContainer(&chroot, hostname, domainname, fn)); \
} \
int ICLASS_NAME(casename, testname)::InnerTestBody()
/* Derived from googletest/include/gtest/gtest-param-test.h, specifically the
* TEST_P() macro, and some fixes to try to be compatible with different
* versions. */
#ifndef GTEST_ATTRIBUTE_UNUSED_
# define GTEST_ATTRIBUTE_UNUSED_
#endif
#ifndef GTEST_INTERNAL_ATTRIBUTE_MAYBE_UNUSED
# define GTEST_INTERNAL_ATTRIBUTE_MAYBE_UNUSED
#endif
# define CONTAINED_TEST_P(test_suite_name, test_name, hostname, domainname, \
files) \
class GTEST_TEST_CLASS_NAME_(test_suite_name, test_name) \
: public test_suite_name { \
public: \
GTEST_TEST_CLASS_NAME_(test_suite_name, test_name)() \
{ \
} \
int InnerTestBody(); \
void TestBody() \
{ \
ContainerFilesystem chroot(files, ".."); \
VoidToIntFn fn = [this](void) -> int { \
ares_reinit(this->channel_); \
ares_sleep_time(100); \
return this->InnerTestBody(); \
}; \
EXPECT_EQ(0, RunInContainer(&chroot, hostname, domainname, fn)); \
} \
\
private: \
static int AddToRegistry() \
{ \
::testing::UnitTest::GetInstance() \
->parameterized_test_registry() \
.GetTestSuitePatternHolder<test_suite_name>( \
GTEST_STRINGIFY_(test_suite_name), \
::testing::internal::CodeLocation(__FILE__, __LINE__)) \
->AddTestPattern( \
GTEST_STRINGIFY_(test_suite_name), GTEST_STRINGIFY_(test_name), \
new ::testing::internal::TestMetaFactory<GTEST_TEST_CLASS_NAME_( \
test_suite_name, test_name)>(), \
::testing::internal::CodeLocation(__FILE__, __LINE__)); \
return 0; \
} \
GTEST_INTERNAL_ATTRIBUTE_MAYBE_UNUSED static int \
gtest_registering_dummy_ GTEST_ATTRIBUTE_UNUSED_; \
}; \
int GTEST_TEST_CLASS_NAME_(test_suite_name, \
test_name)::gtest_registering_dummy_ = \
GTEST_TEST_CLASS_NAME_(test_suite_name, test_name)::AddToRegistry(); \
int GTEST_TEST_CLASS_NAME_(test_suite_name, test_name)::InnerTestBody()
#endif
/* Assigns virtual IO functions to a channel. These functions simply call
* the actual system functions.
*/
class VirtualizeIO {
public:
VirtualizeIO(ares_channel);
~VirtualizeIO();
static const ares_socket_functions default_functions;
private:
`ares_channel` -> `ares_channel_t *`: don't bury the pointer (#595) `ares_channel` is defined as `typedef struct ares_channeldata *ares_channel;`. The problem with this, is it embeds the pointer into the typedef, which means an `ares_channel` can never be declared as `const` as if you write `const ares_channel channel`, that expands to `struct ares_channeldata * const ares_channel` and not `const struct ares_channeldata *channel`. We will now typedef `ares_channel_t` as `typedef struct ares_channeldata ares_channel_t;`, so if you write `const ares_channel_t *channel`, it properly expands to `const struct ares_channeldata *channel`. We are maintaining the old typedef for API compatibility with existing integrations, and due to typedef expansion this should not even cause any compiler warnings for existing code. There are no ABI implications with this change. I could be convinced to keep existing public functions as `ares_channel` if a sufficient argument exists, but internally we really need make this change for modern best practices. This change will allow us to internally use `const ares_channel_t *` where appropriate. Whether or not we decide to change any public interfaces to use `const` may require further discussion on if there might be ABI implications (I don't think so, but I'm also not 100% sure what a compiler internally does with `const` when emitting machine code ... I think more likely ABI implications would occur going the opposite direction). FYI, This PR was done via a combination of sed and clang-format, the only manual code change was the addition of the new typedef, and a couple doc fixes :) Fix By: Brad House (@bradh352)
1 year ago
ares_channel_t *channel_;
};
/*
* Slightly white-box macro to generate two runs for a given test case:
* One with no modifications, and one with all IO functions set to use
* the virtual io structure.
* Since no magic socket setup or anything is done in the latter case
* this should probably only be used for test with very vanilla IO
* requirements.
*/
#define VCLASS_NAME(casename, testname) Virt##casename##_##testname
#define VIRT_NONVIRT_TEST_F(casename, testname) \
class VCLASS_NAME(casename, testname) : public casename { \
public: \
VCLASS_NAME(casename, testname)() \
{ \
} \
void InnerTestBody(); \
}; \
GTEST_TEST_(casename, testname, VCLASS_NAME(casename, testname), \
::testing::internal::GetTypeId<casename>()) \
{ \
InnerTestBody(); \
} \
GTEST_TEST_(casename, testname##_virtualized, \
VCLASS_NAME(casename, testname), \
::testing::internal::GetTypeId<casename>()) \
{ \
VirtualizeIO vio(channel_); \
InnerTestBody(); \
} \
void VCLASS_NAME(casename, testname)::InnerTestBody()
} // namespace test
} // namespace ares
#endif