Commit Graph
730 Commits
Author SHA1 Message Date
Daniel Stenberg 2e31cc0be4 and we start on 1.5.2! 2007-11-21 10:16:44 +00:00
Daniel Stenberg 73b3390b83 change cares-1_5_1 2007-11-21 10:12:20 +00:00
Daniel Stenberg f4be2e0cbb oops 2007-11-21 10:12:06 +00:00
Daniel Stenberg 72a245668f start working on 1.5.1 now 2007-11-21 09:31:56 +00:00
Daniel Stenberg 178f8b401f this is what 1.5.0 is cares-1_5_0 2007-11-21 09:24:03 +00:00
Daniel Stenberg 346e989760 fill in missing copyrights 2007-11-19 15:47:01 +00:00
Gunter Knauf d1015f560d removed now obsolete defines;
updated external library versions to latest.
2007-11-18 22:48:39 +00:00
Steinar H. Gunderson c69f48b24f Fix a double free. 2007-11-16 09:36:00 +00:00
Yang Tse 8627b23715 Needed now that in6_addr is referenced in ares.h 2007-11-15 19:44:01 +00:00
Steinar H. Gunderson d9dd239771 When looking up in DNS and then in the hosts file, return the error code from DNS if both fail, instead of returning the error code from the hosts file, as today. Patch from the Google tree. 2007-11-15 09:16:48 +00:00
Steinar H. Gunderson b4bdb6d4d7 Return TTL data from ares_parse_{a,aaaa}_reply, if the user is so inclined. Patch from the Google tree. 2007-11-15 08:36:25 +00:00
Yang Tse 7cd35ce698 Define WIN32 when build target is Win32 API.
This also defines it for WinCE even though it is a subset of WIN32.
2007-11-08 18:13:54 +00:00
Yang Tse aa74d21c4e The only libraries actually needed for sample programs adig and
ahost are ws2_32.lib and advapi32.lib
2007-11-08 01:33:28 +00:00
Yang Tse 70f98a29cb MSVC versions prior to VS2005 do not complain about portable C functions 2007-11-07 18:18:42 +00:00
Yang Tse 269e082f77 Windows build targets have socklen_t definition in ws2tcpip.h but some
versions of ws2tcpip.h do not have the definition. It seems that when
the socklen_t definition is missing from ws2tcpip.h the definition for
INET_ADDRSTRLEN is also missing, and that when one definition is present
the other one also is available.
curl-7_17_1
2007-10-24 14:39:07 +00:00
Gunter Knauf b68a4326d2 removed dependency on gettimeofday() since we use only 1 sec resolution here. 2007-10-22 23:31:40 +00:00
Yang Tse 1fe4d44ed9 Fix compiler warning: conversion from "int" to "unsigned short" may lose significant bits 2007-10-20 15:11:51 +00:00
Yang Tse 746b3d7acc Fix message shown when detecting icc version 2007-10-19 12:15:00 +00:00
Yang Tse e2d80922fe Avoid shadowing a global declaration 2007-10-19 10:52:28 +00:00
Yang Tse 0e74c9c8cf Renamed a variable to avoid shadowing a global declaration 2007-10-18 17:31:19 +00:00
Yang Tse 3bac381e3f Renamed internal function to avoid a variable shadowing it 2007-10-18 17:17:18 +00:00
Yang Tse b8a1ea0e83 Fix compiler warning: feupdateenv is not implemented and will always fail.
Specifically for linux x86-64 with Intel's icc.
2007-10-18 16:24:14 +00:00
Yang Tse 788acb901c Sync PLATFORM_AIX_V3 detection and CURL_CC_DEBUG_OPTS()
icc warning level with libcurl's
2007-10-18 15:11:51 +00:00
Yang Tse 57abe9d22d Fix compiler warning: conversion from "int" to "unsigned char"
may lose significant bits
2007-10-18 01:01:20 +00:00
Yang Tse 2c8db1aec8 actually sync with lib/setup_once.h 2007-10-17 18:18:27 +00:00
Yang Tse 0a7eb8e30e sync with lib/setup_once.h 2007-10-17 16:59:24 +00:00
Steinar H. Gunderson 6bd2b895fb Fix a bug where fallback from AF_INET6 to AF_INET would not work properly together with relative search; if you had a search path of .a.com and .b.com, and foo.a.com would return ARES_ENODATA and foo.b.com would return ARES_ENOTFOUND, the lookup would not properly retry with AF_INET as it forgot the first ARES_ENODATA. 2007-10-16 21:27:51 +00:00
Dan Fandrich 92b8afea6a Mention first version with CURLOPT_COPYPOSTFIELDS.
Don't confuse NUL with NULL.
2007-10-15 21:19:40 +00:00
Gisle Vanem 8b280df27d Added needed 'HAVE_*' defines. 2007-10-08 14:39:52 +00:00
Gisle Vanem 6ff970c067 'FD_CLOXEC' is meaningless on MSDOS/Watt-32. 2007-10-08 14:38:51 +00:00
Steinar H. Gunderson f75d1172f8 Removed a piece of redundant code (process_answer already takes care of it). 2007-10-04 08:12:12 +00:00
Steinar H. Gunderson b7dec81498 Another timeout fix in ares_getnameinfo(). 2007-10-04 08:09:52 +00:00
Steinar H. Gunderson d5919bbe17 Send the timeout count in ares_getnameinfo(). 2007-10-04 08:09:04 +00:00
Steinar H. Gunderson 4f438605ca Moved the NULL check for channel upwards in ares_destroy(). 2007-10-04 08:07:47 +00:00
Steinar H. Gunderson 47e86f021f Clarified the comment over ares_cancel. 2007-10-04 08:06:25 +00:00
Yang Tse a15a88a4e6 Avoid a segfault when generating a DNS "Transaction ID" in internal
function init_id_key() under low memory conditions.
2007-10-02 18:26:48 +00:00
Yang Tse 962f9c220a Add ares_llist.c and ares_llist.h to MSCV project file. 2007-10-02 11:13:58 +00:00
Daniel Stenberg 338d59bd0b Fixed the problem where next_lookup would use 'status' uninitialized. Now
it gets passed the initial value as an argument.
2007-10-02 08:12:30 +00:00
Yang Tse 343d6f221f Avoid inline C99ism, and move c-ares routines for managing doubly-linked lists. 2007-10-02 02:18:01 +00:00
Daniel Stenberg e4eb2d0e1d ares_strerror() segfaulted if the input error number was out of the currently
supported range.
2007-10-01 22:52:31 +00:00
Daniel Stenberg 250e7c4020 Prevent ares_strerror() from segfaulting if an invalid error code is passed
in as argument!
2007-10-01 22:51:38 +00:00
Yang Tse 10df4bdba2 Fix compiler warning 2007-09-30 19:43:23 +00:00
Yang Tse 68aa62d76b check availability of <netinet/tcp.h> 2007-09-30 02:12:11 +00:00
Yang Tse a0b8ac2bdc improve portability, defining MAXDNAME and MAXCDNAME 2007-09-30 00:37:47 +00:00
Steinar H. Gunderson 879f89a989 Fix a memory leak that I recently inadvertedly introduced. 2007-09-30 00:08:01 +00:00
Steinar H. Gunderson 60519063ea Use ISDIGIT instead of isdigit; fixes a gcc warning. 2007-09-29 21:57:05 +00:00
Steinar H. Gunderson ff7137ad74 Port the TCP socket fix made in ares_fds() to ares_getsock() as well. 2007-09-29 19:26:59 +00:00
Steinar H. Gunderson 04e49e09dc Previously, processing a large batch of timeouts was O(n^2) in the number of
outstanding queries, and processing a DNS response packet was O(n) in the
number of outstanding queries. To speed things up in Google, we added a few circular,
doubly-linked lists of queries that are hash-bucketed based on
the attributes we care about, so most important operations are now O(1).

It might be that the number of buckets are higher than most people would need,
but on a quick calculation it should only be 100kB or so even on a 64-bit
system, so I've let it stay as-is.
2007-09-29 18:18:47 +00:00
Gisle Vanem edf1901077 We should standarise on C comments. 2007-09-29 14:37:47 +00:00
Gisle Vanem 053ffe405d Fix compiler warning in setsockopt(). 2007-09-29 14:34:59 +00:00