http://blog.gmane.org/gmane.network.dns.c-ares hourly 1 1901-01-01T00:00+00:00 http://gmane.org/img/gmane-25t.png http://gmane.org http://comments.gmane.org/gmane.network.dns.c-ares/912 <pre>Hi, After using c-ares' library in an application I see that relative names (names without the domain part) are not handled. Relative names are converted into full names appending the value of domain (/etc/resolv.conf), also known as "DNS Suffix" (Windows). Is there a current way to enable handling relative names? To give an example, host vs. ahost, the former handles relative names and has option: "The -N option sets the number of dots that have to be in name for it to be considered absolute. The default value is that defined using the ndots statement in /etc/resolv.conf, or 1 if no ndots statement is present. Names with fewer dots are interpreted as relative names and will be searched for in the domains listed in the search or domain directive in /etc/resolv.conf." The later doesn't handle relative names. </pre> René Berber 2012-05-24T17:45:12 http://comments.gmane.org/gmane.network.dns.c-ares/909 <pre>Hi all The link on http://c-ares.haxx.se/download/ for c-ares 1.8.0 is incorrect. The link is pointing to http://c-ares.haxx.se/download/download/c-ares-1.8.0.tar.gz but the tarball is at http://c-ares.haxx.se/download/c-ares-1.8.0.tar.gz BTW, I'm maintaining c-ares for gentoo these days. </pre> Anthony G. Basile 2012-04-30T10:11:24 http://comments.gmane.org/gmane.network.dns.c-ares/908 <pre>Hi! I'm happy to announce that I've just packaged and uploaded c-ares 1.8.0 and it is available at http://c-ares.haxx.se/ as usual. c-ares version 1.8.0 Changed: o Added ares_parse_naptr_reply() Fixed: o handle CNAME-only in ares_parse_aaaa_reply() o support multiple DNS servers on Android o check for __ANDROID__ in addition to ANDROID macro o port numbers: convert them to network order o get_iphlpapi_dns_info: fix buffer overrun o configure: make CURL_CHECK_DEF ignore leading whitespace o segfault triggered in ares_init_options() o ares_getnameinfo's memcpy did not copy enough bytes o ares_destroy: fix segfault in ares_destroy_options() o CHANGES: generate from script o configure: fix symbol hiding usability check Thanks go to these friendly people for their efforts and contributions: Geert Uytterhoeven, Guenter Knauf, Yang Tse, Poul Thomas Lomholt, Peter Griess, Albert Chin, Denis Bilenko Have fun! </pre> Daniel Stenberg 2012-04-27T20:14:54 http://comments.gmane.org/gmane.network.dns.c-ares/903 <pre>From the commit log: * add . to include path so ares_build.h is picked up * make ar configurable to ease cross-compiling On a side note, can you guys start using GH pull requests? It's kind of annoying to have to subscribe to yet another mailing list, especially when it's a trivial patch like this. </pre> Ben Noordhuis 2012-04-25T12:14:28 http://comments.gmane.org/gmane.network.dns.c-ares/896 <pre>I'm using libev in my application and trying to figure out if c-ares is suitable for DNS. My general thinking is that I'll use a libev io watcher per socket created by c-ares. Then I'll just need to call ares_process_fd when the watcher wakes up. I'll also likely need a timer, which is probably as simple as reseting a libev timer to ares_timeout() every time an I/O callback wakes up. It looks like I can use the "ARES_OPT_SOCK_STATE_CB" to detect when I need to enable/disable my watchers (i.e. event listeners) for each socket. It also appears that ares_set_socket_callback can be used to detect when new sockets are created and in turn initialize my libev watchers. However, I'm not sure of the correct way to detect when sockets are closed so that I can safely destroy libev watchers. Is there a callback I'm missing that provides that information? If not, any thoughts on where to insert it? </pre> Brian McFarland 2012-04-23T20:31:26 http://comments.gmane.org/gmane.network.dns.c-ares/893 <pre>Hi, I just found an issue with 'make install' when cross-compiling: the libs are stored in $prefix/lib64 if the build system is 64-bit while they are stored in $prefix/lib on a 32-bit build system. It seems that the decision if the subfolder is named 'lib' or 'lib64' is made based on the build system and not on the host system as it should be. I've no idea if that can be fixed at all for cross-compilation since I doubt that configure (or where is it decided?) knows if the host compiler produces 64 or 32 bit libraries ... but its probably a good idea to always use 'lib' when cross-compiling ... anyone have some ideas? This happens with both c-ares and curl, and IIRC with libssh2 too ... Gün. </pre> Guenter 2012-04-22T12:35:10 http://comments.gmane.org/gmane.network.dns.c-ares/892 <pre>Hi friends, If you have anything particular pending you want included in the next release, please speak up! If nothing major happens, I aim to make a new release eary next week. </pre> Daniel Stenberg 2012-04-20T20:12:15 http://comments.gmane.org/gmane.network.dns.c-ares/889 <pre>Before, c-ares always used the first DNS server on Android, causing network problems if this DNS server was not available. Signed-off-by: Geert Uytterhoeven &lt;Geert.Uytterhoeven-osDt5Q4Chk1BDgjK7y7TUQ&lt; at &gt;public.gmane.org&gt; --- ares_init.c | 18 +++++++++++++++--- 1 files changed, 15 insertions(+), 3 deletions(-) diff --git a/ares_init.c b/ares_init.c index a0bfc83..fa9e1d7 100644 --- a/ares_init.c +++ b/ares_init.c &lt; at &gt;&lt; at &gt; -62,6 +62,7 &lt; at &gt;&lt; at &gt; #if defined(ANDROID) || defined(__ANDROID__) #include &lt;sys/system_properties.h&gt; +#define MAX_DNS_PROPERTIES8/* From the Bionic sources */ #endif #include "ares.h" &lt; at &gt;&lt; at &gt; -953,11 +954,22 &lt; at &gt;&lt; at &gt; DhcpNameServer status = ARES_EOF; #elif defined(ANDROID) || defined(__ANDROID__) + unsigned int i; + char name[PROP_NAME_MAX]; char value[PROP_VALUE_MAX]=""; - __system_property_get("net.dns1", value); - status = config_nameserver(&amp;servers, &amp;nservers, value); - if (status == ARES_SUCCESS) + int len; + for (i = 1; i &lt;= MAX_DNS_PROPERTIES; i++) { + snprintf(name, sizeof(n</pre> Geert Uytterhoeven 2012-04-19T13:48:00 http://comments.gmane.org/gmane.network.dns.c-ares/884 <pre>---  ares_parse_aaaa_reply.c |    4 +++-  1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/ares_parse_aaaa_reply.c b/ares_parse_aaaa_reply.c index 1fbe838..09d6af8 100644 --- a/ares_parse_aaaa_reply.c +++ b/ares_parse_aaaa_reply.c &lt; at &gt;&lt; at &gt; -204,7 +204,9 &lt; at &gt;&lt; at &gt; int ares_parse_aaaa_reply(const unsigned char *abuf, int alen,          }      } -  if (status == ARES_SUCCESS &amp;&amp; naddrs == 0) +  if (status == ARES_SUCCESS &amp;&amp; naddrs == 0 &amp;&amp; naliases == 0) +    /* the check for naliases to be zero is to make sure CNAME responses +       don't get caught here */      status = ARES_ENODATA;    if (status == ARES_SUCCESS)      { -- 1.7.5.4 </pre> Peter Griess 2012-04-17T00:53:06 http://comments.gmane.org/gmane.network.dns.c-ares/880 <pre>Hi Daniel, I just found that we do not yet have an INSTALL file as we have with curl - what do you think about copying it over to c-ares? I think that most if not all in it applies to c-ares as well, and since we have now c-ares living as standalone project I believe it makes sense to have it in the c-ares repo too ... we can just copy it over, and tweak a bit if needed once its in ... Gün. </pre> Guenter 2012-04-16T09:34:17 http://comments.gmane.org/gmane.network.dns.c-ares/874 <pre>Hi everyone, I have noticed a problem with building c-ares on the Android NDK. The problem is that c-ares currently only checks for #ifdef ANDROID but it really also needs to check for __ANDROID__ as well. I noticed recently that a patch was submitted by Cedric Deltheil to the CURL sources on 20 Dec 2011, which is designed to correctly test for building with the Android NDK. I have included a copy of his comments and patches for everyone to review at the bottom of this email. I think a similar patch should also be applied to c-ares as well: diff --git a/ares_init.c b/ares_init.c index ea2a978..3d29cfe 100644 --- a/ares_init.c +++ b/ares_init.c &lt; at &gt;&lt; at &gt; -60,7 +60,7 &lt; at &gt;&lt; at &gt; #include &lt;ctype.h&gt; #include &lt;time.h&gt; -#ifdef ANDROID +#if defined(ANDROID) || defined(__ANDROID__) #include &lt;sys/system_properties.h&gt; #endif &lt; at &gt;&lt; at &gt; -971,7 +971,7 &lt; at &gt;&lt; at &gt; DhcpNameServer } status = ARES_EOF; -#elif defined(ANDROID) +#elif defined(ANDROID) || defined(__ANDROID__) char value[PROP_VALUE_MAX]=""; __system_property_get("net.dns</pre> wayne-tecwZl0ki1bR7s880joybQ< at >public.gmane.org 2012-04-12T21:59:42 http://comments.gmane.org/gmane.network.dns.c-ares/872 <pre></pre> SF Markus Elfring 2012-03-15T15:20:45 http://comments.gmane.org/gmane.network.dns.c-ares/868 <pre>C-ares version: c-ares-1.7.5, c-ares-1.7.6-(daily-snapshot) Test system: debian GNU/Linux (lenny x86) Build options: ./configure --disable-shared &amp;&amp; make check Problem1: ========= the -U option argument is used in inverse byte order How to reproduce: $ strace -f ./adig -U 14666 -s 192.168.1.113 www.cnn.com 2&gt;&amp;1 | grep connect connect(3, {sa_family=AF_INET, sin_port=htons(19001), sin_addr=inet_addr("192.168.1.113")}, 16) = 0 connect(3, {sa_family=AF_INET, sin_port=htons(19001), sin_addr=inet_addr("192.168.1.113")}, 16) = 0 connect(3, {sa_family=AF_INET, sin_port=htons(19001), sin_addr=inet_addr("192.168.1.113")}, 16) = 0 connect(3, {sa_family=AF_INET, sin_port=htons(19001), sin_addr=inet_addr("192.168.1.113")}, 16) = 0 adig attempts to connect to port 19001 (0x4A39) instead of port 14666 (0x394A). Problem2: ========= the -T option does not force using TCP for queries How to reproduce: $ strace -f ./adig -T 14666 -s 192.168.1.113 www.cnn.com 2&gt;&amp;1|grep connect connect(3, {sa_family=AF_INET, sin_por</pre> cares2011.mlist.ps2-2pMamKoQTv4< at >public.gmane.org 2012-02-18T18:38:03 http://comments.gmane.org/gmane.network.dns.c-ares/865 <pre>Hi, I experienced a buffer overrun exception in c-ares on Windows and tracked it down to be an error in the calculation of the 'left' variable in get_iphlpapi_dns_info(). The following patch fixed the problem for me, feel free to incorporate it as you see fit I changed the variable type of 'left' to a _signed_ type because of the subtraction arithmetic; not sure if a long is the best choice Thanks Index: ares_init.c =================================================================== --- ares_init.c(revision 100) +++ ares_init.c(working copy) &lt; at &gt;&lt; at &gt; -612,7 +612,7 &lt; at &gt;&lt; at &gt; { const size_t ipv4_size = INET_ADDRSTRLEN + 1; /* +1 for ',' at end */ const size_t ipv6_size = INET6_ADDRSTRLEN + 12; /* +12 for "%0123456789," at end */ - size_t left = ret_size; + long left = ret_size; char *ret = ret_buf; int count = 0; &lt; at &gt;&lt; at &gt; -687,7 +687,7 &lt; at &gt;&lt; at &gt; ret[ stringlen ] = ','; ret[ stringlen + 1 ] = '\0'; ret += stringlen + 1; - left -= ret - </pre> Poul Thomas Lomholt 2012-02-04T06:30:31 http://comments.gmane.org/gmane.network.dns.c-ares/864 <pre>--- Makefile.am | 11 ++++--- c-ares.map | 72 +++++++++++++++++++++++++++++++++++++++++++++++ configure.ac | 9 +---- m4/ld-version-script.m4 | 53 ++++++++++++++++++++++++++++++++++ 4 files changed, 133 insertions(+), 12 deletions(-) create mode 100644 c-ares.map create mode 100644 m4/ld-version-script.m4 --- c-ares-1.7.5.orig/Makefile.am +++ c-ares-1.7.5/Makefile.am &lt; at &gt;&lt; at &gt; -93,12 +93,13 &lt; at &gt;&lt; at &gt; VER=-version-info 2:0:0 # set age to 0. (c:r:a=0) # -if NO_UNDEFINED -# The -no-undefined flag is crucial for this to build fine on some platforms -UNDEF = -no-undefined -endif +libcares_la_LDFLAGS = -no-undefined $(UNDEF) $(VER) -libcares_la_LDFLAGS = $(UNDEF) $(VER) +if HAVE_LD_VERSION_SCRIPT +libcares_la_LDFLAGS += -Wl,--version-script=$(srcdir)/c-ares.map +else +libcares_la_LDFLAGS += -export-symbols-regex '^ares_.*' +endif # Add -Werror if defined CFLAGS += &lt; at &gt;CARES_CFLAG_EXTRAS&lt; at &gt; --- /dev/null +++ c-ares-1.7.5/c-ares.map &lt; at &gt;&lt; at &gt; -0,0 +1,72 &lt; at &gt;&lt; at &gt; +CARES_1.2.0 { +global: +</pre> Cristian Rodríguez 2012-02-03T21:01:21 http://comments.gmane.org/gmane.network.dns.c-ares/857 <pre>I am working on trying to build a project that uses c-ares. However, I am currently getting this error: ld: library not found for -lcares Any insight as to what this means and/or how to fix it would be appreciated. Kristin </pre> Kristin Roher 2012-01-10T12:01:32 http://comments.gmane.org/gmane.network.dns.c-ares/849 <pre>Hi, I am trying to use C-ARES for an application that will do DNS lookups in parallel with other network I/O (it is a network test-tool, meant to check the availability of various network services). And either I do not understand how the timeout mechanism is supposed to work in C-ARES, or there is a bug somewhere in the timeout handling. To demonstrate this, I have written a small test application based mostly on the "adig" utility that comes with C-ARES. This application does the following: - use ares_library_init() to initialize c-ares library; - create a channel with the options including ARES_OPT_TIMEOUT with a value of 10; - use ares_set_servers() to point the channel to an IP that does not respond; - use ares_query() to send a single query to the non-responding DNS server. It then loops doing ares_fds(), ares_timeout(), select() and ares_process(). I pass a "maxtv" value to ares_timeout(), because I only want to wait 1 second between each select() call (to do other I/O between the ARES ch</pre> Henrik Størner 2012-01-02T21:01:22 http://comments.gmane.org/gmane.network.dns.c-ares/848 <pre>Hi guys, I'm using the asynchronous features of ares to resolve IP addresses back to domain names (ares_gethostbyaddr). Is there a way to get the reply TTL of an IP directly in the callback? On a related note, IIRC I read somewhere in the documentation that "the library respects TTL and does caching" (I hope that was not in that other library I was looking into ;-) ). If so, can I get the TTL from the ares-channel struct somehow? thanks, Clemens </pre> Clemens Kolbitsch 2011-11-30T02:06:58 http://comments.gmane.org/gmane.network.dns.c-ares/847 <pre> HI,&amp;nbsp; My application is using c-ares.1.6.0 for resolving FQDN . For resolving FQDN, my application first &amp;nbsp;initiates &amp;nbsp;SRV query for FQDN and when SRV is successful, it initiates &amp;nbsp;A-query for &amp;nbsp;fetched records to get IP address. However in high load , C-ares only handle SRV query. Upon receiving SRV response , my application initiates A-QUERY. For this it passes required data to c-ares but no A query is goes on the wire in network. &amp;nbsp; Please let me know if any packet trace is required for this .&amp;nbsp; Thanks in Advance&amp;nbsp;Brijesh</pre> brijesh gautam 2011-11-25T06:07:45 http://comments.gmane.org/gmane.network.dns.c-ares/844 <pre>Hi, I can't manage how to compile the c-ares library for Linux and Windows. I'm trying to do a multiplataform DNS client that mekes NAPTR queries. I also check that c-ares has support for NAPTR. Somewhere in the mailing list I check that there is a piece of code to decode de replies to these queries. However, in the code I downloaded it is not presented. Can anyone give me some tips about the c-ares? Best regards, Carlos Guimarães </pre> Carlos Guimarães 2011-11-03T14:04:40 http://comments.gmane.org/gmane.network.dns.c-ares/843 <pre>Hi list. A bit quite here. A problem with C-ares has been bugging me ever since I got a VPN connection (from StrongVPN.com). On Win-XP / Win-7 this requires the installation of an PPP-adapter that comes and goes into play as the VPN goes up/down (when back from sleep-mode I use MS Outlook Express to force it back up). Ok so, the routing table is automatically updated depending on the state of the VPN. Normally any Winsock program is unaware of this and any created socket() get to it's destination okay. Obviously I want everything to be encrypted; even DNS-lookups (they're a big give-away if someone wants to snoop). Therein is the problem with C-ares. It uses GetAdaptersAddresses() and loops over all the adapters to extract a server list. Since my ISP ignores DNS-request from non-ISP-customers, C-ares programs hangs for a long time (I'm virtually in New York when the VPN is up). Then according to "ipconfig" I should use 98.158.112.60 / 216.131.94.5. Otherwise (vpn down) I should use my ISP's serv</pre> Gisle Vanem 2011-11-02T22:58:35 Search the mailing list at Gmane query http://search.gmane.org/?group=$group=gmane.network.dns.c-ares