platform/upstream/curl.git
14 years agofix compiler warning
Yang Tse [Tue, 23 Feb 2010 16:07:54 +0000 (16:07 +0000)]
fix compiler warning

14 years agofix socket data type
Yang Tse [Tue, 23 Feb 2010 13:35:01 +0000 (13:35 +0000)]
fix socket data type

14 years agofix socket data type and logging format in debug tracking socket functions
Yang Tse [Mon, 22 Feb 2010 23:28:56 +0000 (23:28 +0000)]
fix socket data type and logging format in debug tracking socket functions

14 years agoconvert Curl_ultous() and Curl_ultouc() functions to curlx_ultous() and
Yang Tse [Mon, 22 Feb 2010 18:56:29 +0000 (18:56 +0000)]
convert Curl_ultous() and Curl_ultouc() functions to curlx_ultous() and
curlx_ultouc(), exposing them through curlx.h to allow proper code reuse
later in our test harness.

14 years agoupdated sources
Yang Tse [Mon, 22 Feb 2010 16:16:35 +0000 (16:16 +0000)]
updated sources

14 years ago_ Adjusted RFC821 HELO fallback and enabled test804
Patrick Monnerat [Mon, 22 Feb 2010 13:57:57 +0000 (13:57 +0000)]
_ Adjusted RFC821 HELO fallback and enabled test804

14 years ago- Proper handling of STARTTLS on SMTP, taking CURLUSESSL_TRY into account.
Patrick Monnerat [Mon, 22 Feb 2010 12:41:02 +0000 (12:41 +0000)]
- Proper handling of STARTTLS on SMTP, taking CURLUSESSL_TRY into account.
- SMTP falls back to RFC821 HELO when EHLO fails (and SSL is not required).
- Use of true local host name (i.e.: via gethostname()) when available, as default argument to SMTP HELO/EHLO.
- Test case 804 for HELO fallback.

14 years agoadd header inclusion
Yang Tse [Mon, 22 Feb 2010 03:41:02 +0000 (03:41 +0000)]
add header inclusion

14 years agofix compiler warning
Yang Tse [Mon, 22 Feb 2010 02:37:13 +0000 (02:37 +0000)]
fix compiler warning

14 years agofix compiler warning
Yang Tse [Sun, 21 Feb 2010 19:59:09 +0000 (19:59 +0000)]
fix compiler warning

14 years agoclarify more details on section "2.1 More non-blocking"
Daniel Stenberg [Sun, 21 Feb 2010 14:42:59 +0000 (14:42 +0000)]
clarify more details on section "2.1 More non-blocking"

14 years agoTFTP transfers are not blocking since 7.20.0
Daniel Stenberg [Sun, 21 Feb 2010 14:40:22 +0000 (14:40 +0000)]
TFTP transfers are not blocking since 7.20.0

14 years ago- Fixed the SMTP compliance by making sure RCPT TO addresses are specified
Daniel Stenberg [Sat, 20 Feb 2010 22:29:59 +0000 (22:29 +0000)]
- Fixed the SMTP compliance by making sure RCPT TO addresses are specified
  properly in angle brackets. Recipients provided with CURLOPT_MAIL_RCPT now
  get angle bracket wrapping automatically by libcurl unless the recipient
  starts with an angle bracket as then the app is assumed to deal with that
  properly on its own.

14 years ago- I made the SMTP code expect a 250 response back from the server after the
Daniel Stenberg [Sat, 20 Feb 2010 21:56:48 +0000 (21:56 +0000)]
- I made the SMTP code expect a 250 response back from the server after the
  full DATA has been sent, and I modified the test SMTP server to also send
  that response. As usual, the DONE operation that is made after a completed
  transfer is still not doable in a non-blocking way so this waiting for 250
  is unfortunately made blockingly.

14 years agocorected a comment and wrapped a few longish lines
Daniel Stenberg [Sat, 20 Feb 2010 21:41:25 +0000 (21:41 +0000)]
corected a comment and wrapped a few longish lines

14 years agofix compiler warning
Yang Tse [Sat, 20 Feb 2010 19:51:02 +0000 (19:51 +0000)]
fix compiler warning

14 years agofix compiler warning
Yang Tse [Sat, 20 Feb 2010 11:58:26 +0000 (11:58 +0000)]
fix compiler warning

14 years agofix compiler warning
Yang Tse [Sat, 20 Feb 2010 01:15:10 +0000 (01:15 +0000)]
fix compiler warning

14 years agofix compiler warning
Yang Tse [Fri, 19 Feb 2010 18:02:38 +0000 (18:02 +0000)]
fix compiler warning

14 years agofix compiler warning
Yang Tse [Thu, 18 Feb 2010 12:31:24 +0000 (12:31 +0000)]
fix compiler warning

14 years agoares_reinit()
Daniel Stenberg [Wed, 17 Feb 2010 22:15:23 +0000 (22:15 +0000)]
ares_reinit()

- To allow an app to force a re-read of /etc/resolv.conf etc, pretty much
  like the res_init() resolver function offers

14 years agouse curl standard indentation and line lengths
Daniel Stenberg [Wed, 17 Feb 2010 12:13:55 +0000 (12:13 +0000)]
use curl standard indentation and line lengths

14 years agoreplaced tabs with spaces
Yang Tse [Tue, 16 Feb 2010 13:32:45 +0000 (13:32 +0000)]
replaced tabs with spaces

14 years agofix Content-Length validation
Yang Tse [Tue, 16 Feb 2010 11:17:00 +0000 (11:17 +0000)]
fix Content-Length validation

14 years agouse (void) in front of fwrite() calls that ignore the return code
Daniel Stenberg [Mon, 15 Feb 2010 21:29:19 +0000 (21:29 +0000)]
use (void) in front of fwrite() calls that ignore the return code

14 years agofix compiler warning: conversion from "long" to "size_t" may lose sign
Yang Tse [Mon, 15 Feb 2010 17:40:35 +0000 (17:40 +0000)]
fix compiler warning: conversion from "long" to "size_t" may lose sign

14 years agofix compiler warning: conversion from "long" to "size_t" may lose sign
Yang Tse [Mon, 15 Feb 2010 16:18:52 +0000 (16:18 +0000)]
fix compiler warning: conversion from "long" to "size_t" may lose sign

14 years ago-w speed_download and speed_upload are measured in bytes per second
Daniel Stenberg [Mon, 15 Feb 2010 07:48:28 +0000 (07:48 +0000)]
-w speed_download and speed_upload are measured in bytes per second

14 years ago75. NTLM authentication involving unicode user name or password.
Daniel Stenberg [Sun, 14 Feb 2010 22:51:43 +0000 (22:51 +0000)]
75. NTLM authentication involving unicode user name or password.
  http://curl.haxx.se/mail/lib-2009-10/0024.html
  http://curl.haxx.se/bug/view.cgi?id=2944325

14 years agoremoved trailing whitespace
Yang Tse [Sun, 14 Feb 2010 19:40:18 +0000 (19:40 +0000)]
removed trailing whitespace

14 years agofix compiler warning
Yang Tse [Sun, 14 Feb 2010 16:37:22 +0000 (16:37 +0000)]
fix compiler warning

14 years agoOverhauled test suite getpart() function. Fixing potential out of bounds
Yang Tse [Sun, 14 Feb 2010 13:14:17 +0000 (13:14 +0000)]
Overhauled test suite getpart() function. Fixing potential out of bounds
stack and memory overwrites triggered with huge test case definitions.

14 years ago- Martin Hager reported and fixed a problem with a missing quote in libcurl.m4
Daniel Stenberg [Sat, 13 Feb 2010 21:30:16 +0000 (21:30 +0000)]
- Martin Hager reported and fixed a problem with a missing quote in libcurl.m4

  (http://curl.haxx.se/bug/view.cgi?id=2951319)

14 years agoused allways #ifdef / #ifndef;
Gunter Knauf [Sat, 13 Feb 2010 18:39:47 +0000 (18:39 +0000)]
used allways #ifdef / #ifndef;
moved gethostbyname_thread() to #else case to avoid 'not used' compiler warning.

14 years agoreplaced tabs by spaces, removed trailing tabs/spaces.
Gunter Knauf [Sat, 13 Feb 2010 18:29:59 +0000 (18:29 +0000)]
replaced tabs by spaces, removed trailing tabs/spaces.

14 years ago- Tom Donovan fixed the CURL_FORMAT_* defines when building with cmake.
Daniel Stenberg [Sat, 13 Feb 2010 17:51:30 +0000 (17:51 +0000)]
- Tom Donovan fixed the CURL_FORMAT_* defines when building with cmake.

14 years ago- Jack Zhang reported a problem with SMTP: we wrongly used multiple addresses
Daniel Stenberg [Fri, 12 Feb 2010 22:23:46 +0000 (22:23 +0000)]
- Jack Zhang reported a problem with SMTP: we wrongly used multiple addresses
  in the same RCPT TO line, when they should be sent in separate single
  commands. I updated test case 802 to verify this.

- I also fixed a bad use of my_setopt_str() of CURLOPT_MAIL_RCPT in the curl
  tool which made it try to output it as string for the --libcurl feature
  which could lead to crashes.

14 years agoCURLOPT_MAIL_RCPT is *not* a string so we must not try to printf() it
Daniel Stenberg [Fri, 12 Feb 2010 22:19:48 +0000 (22:19 +0000)]
CURLOPT_MAIL_RCPT is *not* a string so we must not try to printf() it

14 years agofree --mail-from strings properly
Daniel Stenberg [Fri, 12 Feb 2010 22:16:21 +0000 (22:16 +0000)]
free --mail-from strings properly

14 years ago_ Make it compilable again on OS400.
Patrick Monnerat [Thu, 11 Feb 2010 18:49:06 +0000 (18:49 +0000)]
_ Make it compilable again on OS400.
_ Upgrade OS400 EBCDIC wrappers for new options.
_ Upgrade ILE/RPG bindings to current state.

14 years agomention last changes
Yang Tse [Thu, 11 Feb 2010 07:36:34 +0000 (07:36 +0000)]
mention last changes

14 years agoSteven M. Schweda updated VMS readme file
Yang Tse [Thu, 11 Feb 2010 07:29:38 +0000 (07:29 +0000)]
Steven M. Schweda updated VMS readme file

14 years agoSteven M. Schweda removed batch_compile.com and defines.com
Yang Tse [Thu, 11 Feb 2010 07:27:42 +0000 (07:27 +0000)]
Steven M. Schweda removed batch_compile.com and defines.com

14 years agoSteven M. Schweda fixed:
Yang Tse [Thu, 11 Feb 2010 07:24:29 +0000 (07:24 +0000)]
Steven M. Schweda fixed:

VMS builder bad behavior when used in a batch job.

Various ".LIS" and ".MAP" files created without being requested
by a "LIST" command-line option, and in the wrong place, too.

Some minor typographical changes.

14 years agoMention the minimum size of CURL_MAX_WRITE_SIZE
Dan Fandrich [Wed, 10 Feb 2010 21:17:53 +0000 (21:17 +0000)]
Mention the minimum size of CURL_MAX_WRITE_SIZE

14 years ago- remove extra "\r\n" from doc404_RTSP
Yang Tse [Wed, 10 Feb 2010 13:03:32 +0000 (13:03 +0000)]
- remove extra "\r\n" from doc404_RTSP

- avoid memory alignment issue when setting RTSP packet length

14 years agoRemoved some erroneous "compressed" key words
Dan Fandrich [Tue, 9 Feb 2010 22:50:52 +0000 (22:50 +0000)]
Removed some erroneous "compressed" key words

14 years agostart working on 7.20.1
Daniel Stenberg [Tue, 9 Feb 2010 11:37:10 +0000 (11:37 +0000)]
start working on 7.20.1

14 years agospell and 7.20.0
Daniel Stenberg [Tue, 9 Feb 2010 10:06:48 +0000 (10:06 +0000)]
spell and 7.20.0

14 years ago- When downloading compressed content over HTTP and the app as asked libcurl
Daniel Stenberg [Tue, 9 Feb 2010 09:35:48 +0000 (09:35 +0000)]
- When downloading compressed content over HTTP and the app as asked libcurl
  to automatically uncompress it with the CURLOPT_ENCODING option, libcurl
  could wrongly provide the callback with more data than what the maximum
  documented amount. An application could thus get tricked into badness if the
  maximum limit was trusted to be enforced by libcurl itself (as it is
  documented).

  This is further detailed and explained in the libcurl security advisory
  20100209 at

    http://curl.haxx.se/docs/adv_20100209.html

14 years agoset VERSIONINFO accordingly for the 7.20.0 release
Daniel Stenberg [Tue, 9 Feb 2010 09:35:07 +0000 (09:35 +0000)]
set VERSIONINFO accordingly for the 7.20.0 release

14 years agowarning fix
Yang Tse [Sat, 6 Feb 2010 17:31:09 +0000 (17:31 +0000)]
warning fix

14 years agoOOM handling fix
Yang Tse [Sat, 6 Feb 2010 17:30:06 +0000 (17:30 +0000)]
OOM handling fix

14 years agoOOM handling fix
Yang Tse [Sat, 6 Feb 2010 13:21:45 +0000 (13:21 +0000)]
OOM handling fix

14 years agofix compiler warning
Yang Tse [Fri, 5 Feb 2010 20:52:00 +0000 (20:52 +0000)]
fix compiler warning

14 years agofix compiler warning
Yang Tse [Fri, 5 Feb 2010 19:24:22 +0000 (19:24 +0000)]
fix compiler warning

14 years agofix compiler warning
Yang Tse [Fri, 5 Feb 2010 19:19:34 +0000 (19:19 +0000)]
fix compiler warning

14 years agoAddes OOM handling for curl_easy_setopt() calls in test
Yang Tse [Fri, 5 Feb 2010 18:07:19 +0000 (18:07 +0000)]
Addes OOM handling for curl_easy_setopt() calls in test

14 years ago- avoid OpenSSL 0.9.8 ENGINE_by_id memory leak
Yang Tse [Fri, 5 Feb 2010 09:33:36 +0000 (09:33 +0000)]
- avoid OpenSSL 0.9.8 ENGINE_by_id memory leak

- cleanup parenthesis usage in return statements

14 years ago- attempt to workaround icc 9.1 optimizer issue
Yang Tse [Fri, 5 Feb 2010 09:32:11 +0000 (09:32 +0000)]
- attempt to workaround icc 9.1 optimizer issue

14 years agofix printf-style format strings
Yang Tse [Thu, 4 Feb 2010 19:44:31 +0000 (19:44 +0000)]
fix printf-style format strings

14 years agoValidate server port argument
Yang Tse [Thu, 4 Feb 2010 17:17:19 +0000 (17:17 +0000)]
Validate server port argument

14 years agoFix variable initialization
Yang Tse [Thu, 4 Feb 2010 14:50:58 +0000 (14:50 +0000)]
Fix variable initialization

14 years agoModified test case 557 to additionally verify libcurl's internal curl_m*printf()
Yang Tse [Thu, 4 Feb 2010 14:41:01 +0000 (14:41 +0000)]
Modified test case 557 to additionally verify libcurl's internal curl_m*printf()
functions formatting functionality when handling signed and unsigned shorts.

14 years agoAdded size check for 'short' data type
Yang Tse [Thu, 4 Feb 2010 14:40:11 +0000 (14:40 +0000)]
Added size check for 'short' data type

14 years agoFix compiler warning: unused variable
Yang Tse [Thu, 4 Feb 2010 10:08:39 +0000 (10:08 +0000)]
Fix compiler warning: unused variable

14 years agoadded an additional second to allow test to pass on heavily loaded servers
Yang Tse [Thu, 4 Feb 2010 09:55:45 +0000 (09:55 +0000)]
added an additional second to allow test to pass on heavily loaded servers

14 years agoChanged the Watcom makefiles to make them easier to keep in sync with
Dan Fandrich [Wed, 3 Feb 2010 23:07:11 +0000 (23:07 +0000)]
Changed the Watcom makefiles to make them easier to keep in sync with
Makefile.inc since that can't be included directly.

14 years agomore symbols added in 7.20.0
Daniel Stenberg [Wed, 3 Feb 2010 18:18:13 +0000 (18:18 +0000)]
more symbols added in 7.20.0

14 years agoFix OOM handling
Yang Tse [Wed, 3 Feb 2010 13:28:47 +0000 (13:28 +0000)]
Fix OOM handling

14 years agoFix progressmode Configurable struct member data type. Changed to
Yang Tse [Wed, 3 Feb 2010 10:57:42 +0000 (10:57 +0000)]
Fix progressmode Configurable struct member data type. Changed to
'int' which fits better with existing CURL_PROGRESS_* definitions.

14 years agoFix portability issue related with unaligned memory access
Yang Tse [Wed, 3 Feb 2010 06:49:27 +0000 (06:49 +0000)]
Fix portability issue related with unaligned memory access

14 years agoFix compiler warnings: conversion from 'const int ' to 'unsigned char ', possible...
Yang Tse [Wed, 3 Feb 2010 06:44:18 +0000 (06:44 +0000)]
Fix compiler warnings: conversion from 'const int ' to 'unsigned char ', possible loss of data

14 years agocookies with same path length might get sorted in different order when
Yang Tse [Wed, 3 Feb 2010 01:53:47 +0000 (01:53 +0000)]
cookies with same path length might get sorted in different order when
using different qsort implementations. In order to make this test give
same results on different systems, paths now have different lengths.

14 years agoadded an additional second to allow test to pass on heavily loaded servers
Yang Tse [Wed, 3 Feb 2010 01:41:07 +0000 (01:41 +0000)]
added an additional second to allow test to pass on heavily loaded servers

14 years agoFix compiler warning: conditional expression is constant
Yang Tse [Tue, 2 Feb 2010 16:25:07 +0000 (16:25 +0000)]
Fix compiler warning: conditional expression is constant

14 years agoFix compiler warning: local variable may be used without having been initialized.
Yang Tse [Tue, 2 Feb 2010 16:23:42 +0000 (16:23 +0000)]
Fix compiler warning: local variable may be used without having been initialized.

14 years agoFix compiler warnings: conversion from 'const int ' to 'unsigned char ', possible...
Yang Tse [Tue, 2 Feb 2010 16:23:01 +0000 (16:23 +0000)]
Fix compiler warnings: conversion from 'const int ' to 'unsigned char ', possible loss of data

14 years agoFix compiler warnings:
Yang Tse [Tue, 2 Feb 2010 16:21:37 +0000 (16:21 +0000)]
Fix compiler warnings:

(1) conversion from 'const int ' to 'unsigned char ', possible loss of data
(2) conditional expression is constant

14 years agomention a couple of changes back from November
Yang Tse [Tue, 2 Feb 2010 15:23:13 +0000 (15:23 +0000)]
mention a couple of changes back from November

14 years agomention run time statistics options
Yang Tse [Tue, 2 Feb 2010 14:46:55 +0000 (14:46 +0000)]
mention run time statistics options

14 years ago- Symbol CURL_FORMAT_OFF_T now obsoleted, will be removed in a future release,
Yang Tse [Tue, 2 Feb 2010 12:58:49 +0000 (12:58 +0000)]
- Symbol CURL_FORMAT_OFF_T now obsoleted, will be removed in a future release,
  symbol will not be available when building with CURL_NO_OLDIES defined. Use
  of CURL_FORMAT_CURL_OFF_T is preferred since 7.19.0

14 years agoavoid possibility of using obsoleted stuff
Yang Tse [Tue, 2 Feb 2010 12:39:10 +0000 (12:39 +0000)]
avoid possibility of using obsoleted stuff

14 years agoremove setup.h inclusion which is already done from test.h inclusion
Yang Tse [Tue, 2 Feb 2010 12:36:16 +0000 (12:36 +0000)]
remove setup.h inclusion which is already done from test.h inclusion

14 years agomention SOCKS related problems in the curl(1) man page
Kamil Dudka [Tue, 2 Feb 2010 12:27:12 +0000 (12:27 +0000)]
mention SOCKS related problems in the curl(1) man page

14 years agoFix compiler warning: variable was set but never used
Yang Tse [Tue, 2 Feb 2010 09:15:52 +0000 (09:15 +0000)]
Fix compiler warning: variable was set but never used

Simplify preprocessor symbol checking

14 years agoJulien Chaffraix pointed out a comment mistake, and I re-indented the code
Daniel Stenberg [Tue, 2 Feb 2010 08:48:58 +0000 (08:48 +0000)]
Julien Chaffraix pointed out a comment mistake, and I re-indented the code
slightly while editing

14 years agoinclude headers
Yang Tse [Tue, 2 Feb 2010 07:01:21 +0000 (07:01 +0000)]
include headers

14 years agoConroy added a check to the coded message size since the docs stipulate
Yang Tse [Mon, 1 Feb 2010 22:14:22 +0000 (22:14 +0000)]
Conroy added a check to the coded message size since the docs stipulate
that each call will contain a full protocol packet.

14 years agoConroy's fix to make the code match with the RTP documentation regarding
Yang Tse [Mon, 1 Feb 2010 22:13:55 +0000 (22:13 +0000)]
Conroy's fix to make the code match with the RTP documentation regarding
writing out the whole header. The docs say it writes the whole header,
but the code (before this patch) did not write out the leading $.

14 years agoWe introduce a loop in lib/multi.c around all calls to multi_runsingle() and
Daniel Stenberg [Mon, 1 Feb 2010 21:42:44 +0000 (21:42 +0000)]
We introduce a loop in lib/multi.c around all calls to multi_runsingle() and
simply check for CURLM_CALL_MULTI_PERFORM internally. This has the added
benefit that this goes in line with my long-term wishes to get rid of the
CURLM_CALL_MULTI_PERFORM all together from the public API.

14 years agoupdate rtsp server header field from SWS to RTSPD
Yang Tse [Mon, 1 Feb 2010 19:57:11 +0000 (19:57 +0000)]
update rtsp server header field from SWS to RTSPD

14 years agoTest suite support for RTSP
Yang Tse [Mon, 1 Feb 2010 12:05:08 +0000 (12:05 +0000)]
Test suite support for RTSP

14 years agoMake Curl_pop3_write() additionally truncate trailing POP3_EOB from received
Yang Tse [Sat, 30 Jan 2010 05:22:30 +0000 (05:22 +0000)]
Make Curl_pop3_write() additionally truncate trailing POP3_EOB from received
string buffer, otherwise Curl_client_write() call with zero size would write
to the end of string buffer including matched POP3_EOB.

14 years agoWIN32 fix, _beginthreadex() may return either 0 or -1L upon failure
Yang Tse [Fri, 29 Jan 2010 17:47:54 +0000 (17:47 +0000)]
WIN32 fix, _beginthreadex() may return either 0 or -1L upon failure

14 years agofix errno usage for WIN32 builds
Yang Tse [Fri, 29 Jan 2010 16:38:43 +0000 (16:38 +0000)]
fix errno usage for WIN32 builds

14 years agofix printf-style format strings
Yang Tse [Fri, 29 Jan 2010 12:29:26 +0000 (12:29 +0000)]
fix printf-style format strings

14 years agoChris Conroy fixed test #568 issues with carriage returns
Yang Tse [Fri, 29 Jan 2010 02:58:39 +0000 (02:58 +0000)]
Chris Conroy fixed test #568 issues with carriage returns

14 years agoRTSP followup fix. Both the pipelined and non-pipelined case need to
Yang Tse [Fri, 29 Jan 2010 01:16:23 +0000 (01:16 +0000)]
RTSP followup fix. Both the pipelined and non-pipelined case need to
check for (excess > 0 && !k->ignorebody).

14 years ago- Tommie Gannert pointed out a silly bug in ares_process_fd() since it didn't
Daniel Stenberg [Thu, 28 Jan 2010 22:03:48 +0000 (22:03 +0000)]
- Tommie Gannert pointed out a silly bug in ares_process_fd() since it didn't
  check for broken connections like ares_process() did. Based on that, I
  merged the two functions into a single generic one with two front-ends.