platform/upstream/curl.git
20 years agoRemove the attempt to detect if we already tested the same source setup.
Daniel Stenberg [Fri, 6 Feb 2004 07:15:27 +0000 (07:15 +0000)]
Remove the attempt to detect if we already tested the same source setup.
We really don't care, and so many other things could've changed to make the
new test interesting anyway.

20 years agonumerous things went in today
Daniel Stenberg [Thu, 5 Feb 2004 21:52:53 +0000 (21:52 +0000)]
numerous things went in today

20 years agoadded the missing stdin section
Daniel Stenberg [Thu, 5 Feb 2004 21:51:45 +0000 (21:51 +0000)]
added the missing stdin section

20 years agoAn attempt to only set both libz-related defines at the same time. We need
Daniel Stenberg [Thu, 5 Feb 2004 21:40:05 +0000 (21:40 +0000)]
An attempt to only set both libz-related defines at the same time. We need
both the lib and the header present for both defines to be set. If only one
of the files is found, we issue a warning and set no define.

20 years agoAndrés García's updates
Daniel Stenberg [Thu, 5 Feb 2004 21:03:53 +0000 (21:03 +0000)]
Andrés García's updates

20 years agoif an empty 'transfer-encoding:' header is provided, we switch off the
Daniel Stenberg [Thu, 5 Feb 2004 15:50:16 +0000 (15:50 +0000)]
if an empty 'transfer-encoding:' header is provided, we switch off the
chunky coding of uploads

20 years agoMade a test that sends data on stdin to PUT, with a given length and
Daniel Stenberg [Thu, 5 Feb 2004 15:21:46 +0000 (15:21 +0000)]
Made a test that sends data on stdin to PUT, with a given length and
chunked transfer-encoding disabled. Fixed to work after Len Krause's
bug report.

20 years agoGisle Vanem fixed a windows compiler warning
Daniel Stenberg [Thu, 5 Feb 2004 13:25:14 +0000 (13:25 +0000)]
Gisle Vanem fixed a windows compiler warning

20 years agochanged the test() function to return type int
Daniel Stenberg [Thu, 5 Feb 2004 12:34:17 +0000 (12:34 +0000)]
changed the test() function to return type int

20 years agoinclude process.h to get the _getpid() proto
Daniel Stenberg [Thu, 5 Feb 2004 12:19:13 +0000 (12:19 +0000)]
include process.h to get the _getpid() proto

20 years agowhen using --enable-debug and gcc, provide the -Wno-format-nonliteral option
Daniel Stenberg [Thu, 5 Feb 2004 10:38:32 +0000 (10:38 +0000)]
when using --enable-debug and gcc, provide the -Wno-format-nonliteral option
to prevent the warning in mprintf.c:

(currently line 930) "format not a string literal, argument types not checked"

20 years agooptions we get as longs need to be typecasted when assigned to prevent
Daniel Stenberg [Thu, 5 Feb 2004 09:38:56 +0000 (09:38 +0000)]
options we get as longs need to be typecasted when assigned to prevent
picky compiler warnings

20 years agofix return type to silence compiler warnings
Daniel Stenberg [Thu, 5 Feb 2004 09:38:10 +0000 (09:38 +0000)]
fix return type to silence compiler warnings

20 years agoprevent warning from that picky MIPSpro compiler
Daniel Stenberg [Thu, 5 Feb 2004 09:37:46 +0000 (09:37 +0000)]
prevent warning from that picky MIPSpro compiler

20 years agocompiler warning fix, compare struct pointers of the same type
Daniel Stenberg [Thu, 5 Feb 2004 09:37:04 +0000 (09:37 +0000)]
compiler warning fix, compare struct pointers of the same type

20 years agouse the timeout options when waiting for the server to connect when using
Daniel Stenberg [Thu, 5 Feb 2004 09:26:01 +0000 (09:26 +0000)]
use the timeout options when waiting for the server to connect when using
PORT
Provide better error messages to allow debugging if one if the ipv6-related
name functions fail in the ftp_use_port() function. This might help us
diagnose the problems on AIX. Also make sure getaddrinfo() uses NULL and
not "0" for the service argument.

20 years agohide the pack_hostent proto if ipv6 is enabled, as figured out by Tor Arntsen
Daniel Stenberg [Thu, 5 Feb 2004 08:34:31 +0000 (08:34 +0000)]
hide the pack_hostent proto if ipv6 is enabled, as figured out by Tor Arntsen

20 years agoFix the socklen_t type too. Needed only when built with memory debugging
Daniel Stenberg [Thu, 5 Feb 2004 08:27:47 +0000 (08:27 +0000)]
Fix the socklen_t type too. Needed only when built with memory debugging
as then we include the memdebug.h header from the lib dir, and it then
requires this type... (fails on IRIX 6.5 without this)

20 years agofixed "comparison between signed and unsigned" complaints
Daniel Stenberg [Wed, 4 Feb 2004 10:24:23 +0000 (10:24 +0000)]
fixed "comparison between signed and unsigned" complaints

20 years agodon't use 'sin' as variable name as the picky compiler warnings complain
Daniel Stenberg [Wed, 4 Feb 2004 10:23:15 +0000 (10:23 +0000)]
don't use 'sin' as variable name as the picky compiler warnings complain
about it shadowing the function sin()

20 years agofixing
Daniel Stenberg [Wed, 4 Feb 2004 09:16:05 +0000 (09:16 +0000)]
fixing

20 years agoprevent a compiler warning about a macro definition
Daniel Stenberg [Wed, 4 Feb 2004 08:04:28 +0000 (08:04 +0000)]
prevent a compiler warning about a macro definition

20 years agoprevent the windows version to use global symbol names
Daniel Stenberg [Wed, 4 Feb 2004 08:00:25 +0000 (08:00 +0000)]
prevent the windows version to use global symbol names
added prototypes for the strcasecmp() functions

20 years agowe use the more aggressive compiler warnings
Daniel Stenberg [Wed, 4 Feb 2004 07:54:42 +0000 (07:54 +0000)]
we use the more aggressive compiler warnings

20 years agomade more pointers unsigned, as they were mostly used passed in to functions
Daniel Stenberg [Wed, 4 Feb 2004 07:52:13 +0000 (07:52 +0000)]
made more pointers unsigned, as they were mostly used passed in to functions
that assume them to be unsigned. Stops compiler warnings.

20 years agotypecast comparision between signed and unsigned
Daniel Stenberg [Wed, 4 Feb 2004 07:51:30 +0000 (07:51 +0000)]
typecast comparision between signed and unsigned

20 years agotry_config() takes a second parameter as const, to prevent picky compiler
Daniel Stenberg [Wed, 4 Feb 2004 07:50:51 +0000 (07:50 +0000)]
try_config() takes a second parameter as const, to prevent picky compiler
warnings

20 years agotakes a void *, not a char * anymore
Daniel Stenberg [Wed, 4 Feb 2004 07:50:18 +0000 (07:50 +0000)]
takes a void *, not a char * anymore

20 years agoares_free_string() now takes a void * instead
Daniel Stenberg [Wed, 4 Feb 2004 07:49:33 +0000 (07:49 +0000)]
ares_free_string() now takes a void * instead

20 years agodon't compare signed and unsigned
Daniel Stenberg [Wed, 4 Feb 2004 07:48:46 +0000 (07:48 +0000)]
don't compare signed and unsigned

20 years agoadded a tags target
Daniel Stenberg [Wed, 4 Feb 2004 07:48:20 +0000 (07:48 +0000)]
added a tags target

20 years agonew configure file
Daniel Stenberg [Wed, 4 Feb 2004 07:48:00 +0000 (07:48 +0000)]
new configure file

20 years agouse configure.ac instead of configure.in
Daniel Stenberg [Wed, 4 Feb 2004 07:47:04 +0000 (07:47 +0000)]
use configure.ac instead of configure.in
support --enable-debug to switch on picky compiler options

20 years agostop a compiler warning
Daniel Stenberg [Wed, 4 Feb 2004 07:40:25 +0000 (07:40 +0000)]
stop a compiler warning

20 years agono one uses libcurl before 7.7 anyway...
Daniel Stenberg [Tue, 3 Feb 2004 15:59:06 +0000 (15:59 +0000)]
no one uses libcurl before 7.7 anyway...

20 years agolibcares.a is here
Daniel Stenberg [Tue, 3 Feb 2004 14:06:16 +0000 (14:06 +0000)]
libcares.a is here

20 years agolink with libcares instead of libares
Daniel Stenberg [Tue, 3 Feb 2004 13:58:38 +0000 (13:58 +0000)]
link with libcares instead of libares

20 years agoWe now produce 'libcares.a' instead, to make it possible to have both c-ares
Daniel Stenberg [Tue, 3 Feb 2004 13:58:08 +0000 (13:58 +0000)]
We now produce 'libcares.a' instead, to make it possible to have both c-ares
and the original ares installed in the same lib dir.

20 years agothe configure and memdebug fixes of yday
Daniel Stenberg [Tue, 3 Feb 2004 10:07:26 +0000 (10:07 +0000)]
the configure and memdebug fixes of yday

20 years agothe unused quit-function didn't pass a correct variable type to the
Daniel Stenberg [Tue, 3 Feb 2004 09:52:32 +0000 (09:52 +0000)]
the unused quit-function didn't pass a correct variable type to the
response reading function

20 years agoadd link to the new cares web site
Daniel Stenberg [Tue, 3 Feb 2004 09:40:19 +0000 (09:40 +0000)]
add link to the new cares web site

20 years agoadded notes about incompatible functions
Daniel Stenberg [Tue, 3 Feb 2004 09:16:40 +0000 (09:16 +0000)]
added notes about incompatible functions

20 years agoadjusted to the new single-parameter version of this function
Daniel Stenberg [Tue, 3 Feb 2004 08:58:10 +0000 (08:58 +0000)]
adjusted to the new single-parameter version of this function

20 years agoremoved odd newline
Daniel Stenberg [Tue, 3 Feb 2004 08:47:26 +0000 (08:47 +0000)]
removed odd newline

20 years agoadded the missing ares numerical version initialiser
Daniel Stenberg [Tue, 3 Feb 2004 06:39:37 +0000 (06:39 +0000)]
added the missing ares numerical version initialiser

20 years agomodified *strerror()
Daniel Stenberg [Tue, 3 Feb 2004 06:38:38 +0000 (06:38 +0000)]
modified *strerror()

20 years agoUse the nroff option figured out by the configure script. An attempt to
Daniel Stenberg [Mon, 2 Feb 2004 22:39:37 +0000 (22:39 +0000)]
Use the nroff option figured out by the configure script. An attempt to
make this better on more systems.

20 years agotry to figure out if -man or -mandoc works to get text with the NROFF
Daniel Stenberg [Mon, 2 Feb 2004 22:39:04 +0000 (22:39 +0000)]
try to figure out if -man or -mandoc works to get text with the NROFF
utility

20 years agoundef accept before defining it, since AIX 5.2 has it as a define!
Daniel Stenberg [Mon, 2 Feb 2004 21:34:01 +0000 (21:34 +0000)]
undef accept before defining it, since AIX 5.2 has it as a define!

20 years agomore test91 tweaks and some c-ares stuff
Daniel Stenberg [Mon, 2 Feb 2004 16:29:58 +0000 (16:29 +0000)]
more test91 tweaks and some c-ares stuff

20 years agonow provides c-ares version info in both version calls
Daniel Stenberg [Mon, 2 Feb 2004 16:24:41 +0000 (16:24 +0000)]
now provides c-ares version info in both version calls

20 years agoIntroducing the SECOND version of the version_info struct. This should
Daniel Stenberg [Mon, 2 Feb 2004 16:24:18 +0000 (16:24 +0000)]
Introducing the SECOND version of the version_info struct. This should
be backwards compatible with older libcurls just fine.

20 years agouse the new single-argument ares_strerror()
Daniel Stenberg [Mon, 2 Feb 2004 16:15:58 +0000 (16:15 +0000)]
use the new single-argument ares_strerror()

20 years agoadjusted to the modified ares_strerror() function
Daniel Stenberg [Mon, 2 Feb 2004 16:00:31 +0000 (16:00 +0000)]
adjusted to the modified ares_strerror() function

NOTE that this breaks ares-compatibility, we have now officially taken the
turn into the c-ares path. We will now officially depend on c-ares for asynch
name resolves.

20 years agoremove rcsid stuff from c files, it serves no useful purpose
Daniel Stenberg [Mon, 2 Feb 2004 15:59:35 +0000 (15:59 +0000)]
remove rcsid stuff from c files, it serves no useful purpose

20 years agoremoved the silly second argument to ares_strerror()
Daniel Stenberg [Mon, 2 Feb 2004 15:59:12 +0000 (15:59 +0000)]
removed the silly second argument to ares_strerror()

This breaks the API and ABI with the existing ares library. We hereby require
the upcoming c-ares 1.0 for asynch name resolves!

20 years agoDirk Manske fixed the ares usage even more. We could get a timeout from ares
Daniel Stenberg [Mon, 2 Feb 2004 15:53:50 +0000 (15:53 +0000)]
Dirk Manske fixed the ares usage even more. We could get a timeout from ares
as well, and when failing and not getting a timeout we now include the
error message ares can provide us with.

20 years agomention mpe/ix
Daniel Stenberg [Mon, 2 Feb 2004 15:34:09 +0000 (15:34 +0000)]
mention mpe/ix

20 years agoset the 'retry' bit to TRUE when the connection is about to be retried,
Daniel Stenberg [Mon, 2 Feb 2004 14:49:54 +0000 (14:49 +0000)]
set the 'retry' bit to TRUE when the connection is about to be retried,
this allows the HTTP code to *not* return a failure just because no data
has been received from the server

20 years agoclear the sockerror if no error was returned
Daniel Stenberg [Mon, 2 Feb 2004 14:49:10 +0000 (14:49 +0000)]
clear the sockerror if no error was returned

20 years agoares resolve timeout and ca bundle include fixes
Daniel Stenberg [Mon, 2 Feb 2004 12:46:58 +0000 (12:46 +0000)]
ares resolve timeout and ca bundle include fixes

20 years agowe include errno.h to truly know if we have ECONNRESET or not
Daniel Stenberg [Mon, 2 Feb 2004 11:59:42 +0000 (11:59 +0000)]
we include errno.h to truly know if we have ECONNRESET or not

20 years agoTimeout slow ares name lookups. This is based on the patch brought by
Daniel Stenberg [Mon, 2 Feb 2004 10:15:22 +0000 (10:15 +0000)]
Timeout slow ares name lookups. This is based on the patch brought by
Dirk Manske, but modified by me.

20 years agowhen including ca-bundle.h, don't look in the current dir first, simply use
Daniel Stenberg [Mon, 2 Feb 2004 10:13:54 +0000 (10:13 +0000)]
when including ca-bundle.h, don't look in the current dir first, simply use
the search path since we want the build-version rather than the one in the
source dir

20 years agofixed the win32 function to use the correct proto, as pointed out by Gisle
Daniel Stenberg [Mon, 2 Feb 2004 07:24:30 +0000 (07:24 +0000)]
fixed the win32 function to use the correct proto, as pointed out by Gisle
Vanem

20 years agotest case 97
Daniel Stenberg [Mon, 2 Feb 2004 07:21:21 +0000 (07:21 +0000)]
test case 97

20 years agoadded test 97, a simple test with -d post and a replacede content-type
Daniel Stenberg [Mon, 2 Feb 2004 07:13:29 +0000 (07:13 +0000)]
added test 97, a simple test with -d post and a replacede content-type
header using -H

20 years agoattempt to fix the notorious test 91 failures
Daniel Stenberg [Fri, 30 Jan 2004 12:43:49 +0000 (12:43 +0000)]
attempt to fix the notorious test 91 failures

20 years agomake sure the connection is closed when it was detected reset!
Daniel Stenberg [Fri, 30 Jan 2004 12:41:27 +0000 (12:41 +0000)]
make sure the connection is closed when it was detected reset!

20 years agoSomewhat crude attempt at fixing the test 91 failures. I commit this now
Daniel Stenberg [Fri, 30 Jan 2004 12:08:18 +0000 (12:08 +0000)]
Somewhat crude attempt at fixing the test 91 failures. I commit this now
so that the automatic testing hosts will test these changes over the weekend.

20 years agohide the grep result
Daniel Stenberg [Fri, 30 Jan 2004 09:48:55 +0000 (09:48 +0000)]
hide the grep result

20 years agocatching up with reality
Daniel Stenberg [Fri, 30 Jan 2004 09:31:44 +0000 (09:31 +0000)]
catching up with reality

20 years agotest case 510, mpeix fix
Daniel Stenberg [Fri, 30 Jan 2004 09:31:23 +0000 (09:31 +0000)]
test case 510, mpeix fix

20 years agowhen we receive a request overflow, we still dump the incoming request to
Daniel Stenberg [Fri, 30 Jan 2004 09:27:27 +0000 (09:27 +0000)]
when we receive a request overflow, we still dump the incoming request to
the dump file to make it easier to understand and debug the situation

20 years agoadded lib510.c for callback POST using chunked encoding
Daniel Stenberg [Fri, 30 Jan 2004 09:26:23 +0000 (09:26 +0000)]
added lib510.c for callback POST using chunked encoding

20 years agoadded test 510, callback-based POST using chunked encoding
Daniel Stenberg [Fri, 30 Jan 2004 09:25:45 +0000 (09:25 +0000)]
added test 510, callback-based POST using chunked encoding

20 years ago--socks
Daniel Stenberg [Fri, 30 Jan 2004 08:54:00 +0000 (08:54 +0000)]
--socks

20 years agoAdded --socks
Daniel Stenberg [Fri, 30 Jan 2004 08:51:24 +0000 (08:51 +0000)]
Added --socks

20 years agoonly do the verifyconnect() clear magic on mpeix
Daniel Stenberg [Fri, 30 Jan 2004 07:51:51 +0000 (07:51 +0000)]
only do the verifyconnect() clear magic on mpeix

20 years agothe thread and SSL issue is now fixed
Daniel Stenberg [Thu, 29 Jan 2004 16:17:51 +0000 (16:17 +0000)]
the thread and SSL issue is now fixed

20 years agonote about the need for extra functions set to OpenSSL if you use OpenSSL
Daniel Stenberg [Thu, 29 Jan 2004 16:17:25 +0000 (16:17 +0000)]
note about the need for extra functions set to OpenSSL if you use OpenSSL
multi-threaded

20 years agoupdate with recent info
Daniel Stenberg [Thu, 29 Jan 2004 16:00:10 +0000 (16:00 +0000)]
update with recent info

20 years agodon't include config.g in src/hugehelp.c unless HAVE_CONFIG_H is defined
Daniel Stenberg [Thu, 29 Jan 2004 15:48:27 +0000 (15:48 +0000)]
don't include config.g in src/hugehelp.c unless HAVE_CONFIG_H is defined

20 years agoares fix, warnings fixed, mpeix fixes
Daniel Stenberg [Thu, 29 Jan 2004 15:41:56 +0000 (15:41 +0000)]
ares fix, warnings fixed, mpeix fixes

20 years agocheck for sys/ioctl.h as well
Daniel Stenberg [Thu, 29 Jan 2004 15:38:27 +0000 (15:38 +0000)]
check for sys/ioctl.h as well
added commented more aggressive compiler options for gcc, subject to be used
instead of the current ones when --enable-debug is used

20 years agoadded verifyconnect proto and use it correctly in the waitconnect function
Daniel Stenberg [Thu, 29 Jan 2004 15:37:21 +0000 (15:37 +0000)]
added verifyconnect proto and use it correctly in the waitconnect function

20 years agomoved the definitions of IOCTL_3_ARGS to setup.h
Daniel Stenberg [Thu, 29 Jan 2004 15:35:42 +0000 (15:35 +0000)]
moved the definitions of IOCTL_3_ARGS to setup.h

20 years ago1. changed order of two include files to build fine on MPE/iX
Daniel Stenberg [Thu, 29 Jan 2004 15:29:47 +0000 (15:29 +0000)]
1. changed order of two include files to build fine on MPE/iX
2. now reads the socket error before check connect status, also to make us
   run fine on MPE/iX

20 years agoKen Hirsch says he basicly needs all 'mv' to be 'mv -f' for configure to
Daniel Stenberg [Thu, 29 Jan 2004 15:28:53 +0000 (15:28 +0000)]
Ken Hirsch says he basicly needs all 'mv' to be 'mv -f' for configure to
run in a sane manner on his MPE/iX operating system.

20 years agoDan Fandrich's cleanup patch to make pedantic compiler options cause less
Daniel Stenberg [Thu, 29 Jan 2004 13:56:45 +0000 (13:56 +0000)]
Dan Fandrich's cleanup patch to make pedantic compiler options cause less
warnings. Minor edits by me.

20 years agomake pedantic compiler options generate less warnings
Daniel Stenberg [Thu, 29 Jan 2004 13:54:08 +0000 (13:54 +0000)]
make pedantic compiler options generate less warnings

20 years agoadded hugehelp.h
Daniel Stenberg [Thu, 29 Jan 2004 13:53:34 +0000 (13:53 +0000)]
added hugehelp.h

20 years agononsence change to make -Wunreachable-code get happy
Daniel Stenberg [Thu, 29 Jan 2004 13:49:41 +0000 (13:49 +0000)]
nonsence change to make -Wunreachable-code get happy

20 years agouse hugehelp.h to silence picky compiler warnings
Daniel Stenberg [Thu, 29 Jan 2004 13:48:36 +0000 (13:48 +0000)]
use hugehelp.h to silence picky compiler warnings

20 years agodocumented ares_version()
Daniel Stenberg [Thu, 29 Jan 2004 12:07:48 +0000 (12:07 +0000)]
documented ares_version()

20 years agoreturn a const char *
Daniel Stenberg [Thu, 29 Jan 2004 12:07:34 +0000 (12:07 +0000)]
return a const char *

20 years agouse the name 'c-ares' for this package
Daniel Stenberg [Thu, 29 Jan 2004 11:33:31 +0000 (11:33 +0000)]
use the name 'c-ares' for this package

20 years agoinclude the new README.cares
Daniel Stenberg [Thu, 29 Jan 2004 11:33:16 +0000 (11:33 +0000)]
include the new README.cares

20 years agopoint out that this is a forked project, all c-ares specific stuff is now
Daniel Stenberg [Thu, 29 Jan 2004 11:32:45 +0000 (11:32 +0000)]
point out that this is a forked project, all c-ares specific stuff is now
in README.cares

20 years agoc-ares specific README, leaving most of the original README as-is
Daniel Stenberg [Thu, 29 Jan 2004 11:32:24 +0000 (11:32 +0000)]
c-ares specific README, leaving most of the original README as-is