Commit Graph

13771 Commits

Author SHA1 Message Date
Yang Tse 61623b74e3 build: BCC - require Borlands's MAKE for Makefile.b32 processing. 2011-01-12 20:53:28 +01:00
Yang Tse 5f7d34811a build: BCC - recover lost functionality from commit 3d81320426
Borland's $(MAKEDIR) expands to the path where make.exe is located,
use this feature to define BCCDIR when user has not defined BCCDIR.
2011-01-12 18:07:04 +01:00
Yang Tse 42f5e8a0f1 build: use external preprocessor cpp32 when building with Borland C 2011-01-12 02:35:14 +01:00
Yang Tse c5d9cd5c1b build: allow usage of Borland 5.5.1 external preprocessor cpp32 2011-01-12 02:33:39 +01:00
Guenter Knauf b5cc77bd25 Disable LDAP support since BCC headers are insufficient. 2011-01-11 18:32:38 +01:00
Guenter Knauf 3d81320426 Instead of exiting with error lets set BCCDIR self. 2011-01-11 18:22:42 +01:00
Guenter Knauf 308db9d780 Some Borland C++ makefile tweaks. 2011-01-11 17:51:14 +01:00
Guenter Knauf 03ea06b8e3 Borland C++ doesnt have struct sockaddr_storage. 2011-01-11 17:40:13 +01:00
Daniel Stenberg 08a65b10fb connect: use UDP correctly
The idea that the protocol and socktype is part of name resolving in the
libc functions is nuts. We keep the name resolver functions assume
TCP/STREAM and we make sure that when we want to connect to a UDP
service we use the correct UDP/DGRAM set instead. This bug was because
the ->protocol field was not always set correctly.

This bug was only affecting ipv6-disabled non-cares non-threaded builds.

Bug: http://curl.haxx.se/bug/view.cgi?id=3154436
Reported by: "dperham"
2011-01-11 00:14:36 +01:00
Quinn Slack 1238edaeaf SSL: fix memory leak
In OOM situation. Follow-up fix to commit a9cd4f4ed4.
2011-01-10 13:53:57 +01:00
Daniel Stenberg a9cd4f4ed4 gtls: fix memory leak
Bug: http://curl.haxx.se/mail/lib-2011-01/0079.html
Reported by: Quinn Slack
2011-01-08 19:14:28 +01:00
Marcel Roelofs 1d28efb9d1 HTTP: HTTP Negotiate authentication using SSPI
Only under Windows
2011-01-07 23:48:40 +01:00
Daniel Stenberg 9ba42a023d cmake: removed two files
CMake/CheckTypeSize.c.in and CMake/CheckTypeSize.cmake were removed in
the previous cmake commit
2011-01-06 23:41:45 +01:00
Dan Fandrich ebb9c7ae04 Made unit_setup() return an error code to abort the test early
This makes it possible to skip the call to unit_stop() in such
cases.  Also use Curl_safefree() in unit test 1302 so it will
pass the memory torture test.
2011-01-05 23:53:24 -08:00
Daniel Stenberg 53014175e8 SSH: speedcheck clobbered existing error
The just added speedcheck must not ruin the error code if already set
due to a problem.
2011-01-06 00:19:17 +01:00
Brad King 651c0bcdf2 CMake: Use upstream CheckTypeSize module
The CheckTypeSize module that comes with CMake 2.6.2 and above does
everything we need and also supports cross-compiling.  Avoid duplicating
an older version of it here.  This also fixes a cross-compiling error
because the old line

  include ("${CMAKE_MODULE_PATH}/CheckTypeSize.cmake")

failed because CMAKE_MODULE_PATH is a search path and not a directory.

Signed-off-by: Brad King <brad.king@kitware.com>
2011-01-06 00:06:29 +01:00
Daniel Stenberg d45ed0ecf5 SSH: acknowledge speedcheck
Check for speedcheck limits during the state machine traversals
2011-01-05 14:09:53 +01:00
Daniel Stenberg e21b103c3e tests: <info> tag corrections 2011-01-04 23:13:10 +01:00
Daniel Stenberg 0029b2f042 unittest: 1303 tests Curl_timeleft
I came up with 33 different ways to call it and verify that it returns the
correct return code.
2011-01-04 23:13:10 +01:00
Daniel Stenberg 7c5d888ea6 curlcheck.h: add fail()
fail is a new function/macro that a test case can use to indicate a test
failure for cases when the standard macros are not sufficient.
2011-01-04 23:13:10 +01:00
Daniel Stenberg adb49ad8bb Curl_timeleft: s/conn/data in first argument
As the function doesn't really use the connectdata struct but only the
SessionHanadle struct I modified what argument it wants.
2011-01-04 23:13:10 +01:00
Dan Fandrich a2c8966d50 Fixed path to allow out-of-tree builds 2011-01-04 12:59:44 -08:00
Kamil Dudka d8f6d1c334 nss: avoid CURLE_OUT_OF_MEMORY given a file name without any slash
Bug: https://bugzilla.redhat.com/623663
2011-01-04 17:20:43 +01:00
Daniel Stenberg 1e52ea92eb ignore: all executable unit test cases 2011-01-04 16:51:41 +01:00
Daniel Stenberg 28888a0b41 tests: add 1302 to the package 2011-01-04 16:50:59 +01:00
Daniel Stenberg 1022e754f4 unittest: test base64 encode/decode 2011-01-04 16:42:31 +01:00
Daniel Stenberg 703573c72b curlcheck.h: avoid using NULL pointers 2011-01-04 16:41:42 +01:00
Daniel Stenberg 7af54ef9a5 curlcheck.h: add verify_memory
This check verifies that a pointer contains the correct data.
2011-01-04 16:31:54 +01:00
Daniel Stenberg 1602ed6ba1 curlcheck.h: add newlines in error messages 2011-01-04 16:14:23 +01:00
Daniel Stenberg 9e46318a03 unittest: verify curl_strequal 2011-01-04 16:13:58 +01:00
Daniel Stenberg c0c89cd44e get_cert_chain: support larger data sets
512 bytes turned out too short for some data, so now we allocate a
larger buffer instead

Bug: http://curl.haxx.se/mail/archive-2011-01/0002.html
2011-01-04 10:20:28 +01:00
Daniel Stenberg 6cf35852ad RELEASE-NOTES: synced with 83e9fb21aa 2011-01-04 00:07:30 +01:00
Daniel Stenberg 83e9fb21aa curlcheck.h: add fail_if() fix code
The UNITTEST_START and UNITTEST_STOP defines needed to do a new brace
level so that test cases can declare variables fine and still remain
fine C89 code.
2011-01-03 23:47:34 +01:00
Daniel Stenberg 5c42b2ceae unittests: basic docs 2011-01-03 23:47:13 +01:00
Daniel Stenberg 82aa386732 ignore: unit test files 2011-01-03 22:52:50 +01:00
Daniel Stenberg c43ad0f972 unittests: a dedicated feature in tests
The test runner script now knows if unittests can run and the unit test
setup file says it is one. I also made runtests.pl deal with no
<command> tag set, so that the description file can get even simpler.
2011-01-03 22:42:46 +01:00
Daniel Stenberg 53640a3ce0 unittesting: build a separate static lib
When configure --enable-debug has been used, all files in lib/ are now
built twice and a separate static library crafted for unit-testing will
be linked. The unit tests in the tests/unit subdir will use that
library.
2011-01-03 22:16:16 +01:00
Daniel Stenberg 35e1d6538a unittest: framework for unit-testing
This is the first approach at doing fairly clean and easy to write and
debug unit tests.
2011-01-03 19:38:10 +01:00
Daniel Stenberg 45cea71968 SSH: avoid PATH_MAX with alloc
We cannot assume that PATH_MAX will be enough for the remote path name
so allocating room for it is the only sensible approach.
2011-01-03 13:36:45 +01:00
Daniel Stenberg 569202c9a4 TODO: get rid of PATH_MAX 2011-01-03 09:40:40 +01:00
Daniel Stenberg 2b3fbc8cdb Curl_nss_connect: avoid PATH_MAX
Since some systems don't have PATH_MAX and it isn't that clever to
assume a fixed maximum path length, the code now allocates buffer space
instead of using stack.

Reported by: Samuel Thibault
Bug: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=608521
2011-01-02 23:43:03 +01:00
Daniel Stenberg 1ad5764feb SFTP: make pwd output result to header callback
Sending "pwd" as a QUOTE command only sent the reply to the
DEBUGFUNCTION. Now it also sends an FTP-like header to the header
callback to allow similar operations as with FTP, and apps can re-use
the same parser.
2011-01-01 15:35:53 +01:00
Daniel Stenberg ae29142198 pubkey_show: allocate buffer to fit any-size result
The loop condition was wrong so keys larger than 340 bits would overflow
the local stack-based buffer.
2011-01-01 15:33:57 +01:00
Daniel Stenberg cd045e24a0 CURLINFO_FTP_ENTRY_PATH: sftp support 2010-12-30 23:49:03 +01:00
Luke Amery 0dc8479b89 ssh: honour the CURLINFO_FTP_ENTRY_PATH curl_getinfo option 2010-12-30 23:42:44 +01:00
Pierre Joye 420eac5542 IDN: use win32 API if told to
The functionality is provided in a new source file: lib/idn_win32.c
2010-12-28 20:01:03 +01:00
Yang Tse 15aeb94f3f test harness: take in account that Windows does not support LD_PRELOAD
configure.ac: Test harness libhostname library will not be built for Windows.

runtests.pl: LD_PRELOAD mechanism will not be used to load libhostname
library on operating systems which lack LD_PRELOAD support.
2010-12-28 16:52:20 +01:00
Daniel Stenberg b6a3e2be8e c-ares: fix cancelled resolves
When built IPv6-enabled, we could do Curl_done() with one of the two
resolves having returned already, so when ares_cancel() is called the
resolve callback ends up doing funny things (sometimes resulting in a
segfault) since it would try to actually store the previous resolve even
though we're shutting down the resolve.

This bug was introduced in commit 8ab137b2bc so it hasn't been
included in any public release.

Bug: http://curl.haxx.se/bug/view.cgi?id=3145445
Reported by: Pedro Larroy
2010-12-27 10:01:44 +01:00
Brad Hards 9f64bbd6d8 Typo / spelling fixes. 2010-12-25 23:07:16 +01:00
Brad Hards 879914def3 Use angle address, as for the rest of the example.
Also spelling fix for RECIPIENT #define.
2010-12-25 23:06:53 +01:00