]>
granicus.if.org Git - curl/log
Daniel Stenberg [Mon, 17 May 2004 10:54:00 +0000 (10:54 +0000)]
'make test-torture' in the root now runs a full torture test
Daniel Stenberg [Mon, 17 May 2004 10:53:34 +0000 (10:53 +0000)]
removed debug output and trailing whitespace
Daniel Stenberg [Mon, 17 May 2004 10:51:51 +0000 (10:51 +0000)]
grrr, fix the check again if no ftp server at all is running
Daniel Stenberg [Mon, 17 May 2004 10:39:44 +0000 (10:39 +0000)]
torture testing and a moved CVS repo
Daniel Stenberg [Mon, 17 May 2004 08:07:07 +0000 (08:07 +0000)]
new Curl_done() proto
Daniel Stenberg [Mon, 17 May 2004 08:05:46 +0000 (08:05 +0000)]
I made Curl_done() take a pointer-pointer in the first argument instead, and
if the connection is killed it blanks the pointer it points to, to make it
easier to detect usage problems whereever Curl_done() is used.
Daniel Stenberg [Mon, 17 May 2004 08:04:42 +0000 (08:04 +0000)]
better bailing out on memory failure
Daniel Stenberg [Mon, 17 May 2004 08:02:23 +0000 (08:02 +0000)]
When waiting for the second connect, we now use alarm to timeout the waiting.
This is necessary in case the client never connects or somehow fails to do
it timely. The timeout used now is only 2 seconds, which might cause problems
on really slow hosts but longer times are painful when doing torture testing
on FTP test cases.
I'm not sure how this 'alarm' functionality works on Windows or other systems
that don't actually have the alarm() function.
Daniel Stenberg [Mon, 17 May 2004 07:59:10 +0000 (07:59 +0000)]
improved the check for our own ftp server
Daniel Stenberg [Mon, 17 May 2004 07:45:20 +0000 (07:45 +0000)]
modified to not leak memory if a libcurl function returns failure, for better
memory leak detection
Daniel Stenberg [Mon, 17 May 2004 07:12:32 +0000 (07:12 +0000)]
Seshubabu Pasam's format fixes and added notes about DER not working for
some (SSL-)options.
Daniel Stenberg [Mon, 17 May 2004 06:55:04 +0000 (06:55 +0000)]
make it not leak memory when it returns prematurely
Daniel Stenberg [Mon, 17 May 2004 06:54:20 +0000 (06:54 +0000)]
automatically delete trailing white space on save in curl-mode
Daniel Stenberg [Mon, 17 May 2004 06:53:41 +0000 (06:53 +0000)]
bail out nicely if strdup() returns NULL, removed trailing whitespace
Daniel Stenberg [Mon, 17 May 2004 06:50:32 +0000 (06:50 +0000)]
deleted trailing whitespace
Daniel Stenberg [Mon, 17 May 2004 06:50:08 +0000 (06:50 +0000)]
fixed a warning on IRIX, deleted trailing whitespace
Daniel Stenberg [Fri, 14 May 2004 11:46:35 +0000 (11:46 +0000)]
added string for the new share error code
Daniel Stenberg [Fri, 14 May 2004 09:30:31 +0000 (09:30 +0000)]
Added CURLSHE_NOMEM
Daniel Stenberg [Fri, 14 May 2004 09:22:12 +0000 (09:22 +0000)]
clean up properly on failure to enable easier libcurl leak detection
Daniel Stenberg [Fri, 14 May 2004 09:21:42 +0000 (09:21 +0000)]
new cvs instructions
Daniel Stenberg [Fri, 14 May 2004 08:40:33 +0000 (08:40 +0000)]
enable memory debugging the same way the curl command line tool already does
Daniel Stenberg [Thu, 13 May 2004 15:19:02 +0000 (15:19 +0000)]
memory cleanup and check fix
Daniel Stenberg [Thu, 13 May 2004 15:18:29 +0000 (15:18 +0000)]
check that memory allocation functions truly return good data or bail out
Daniel Stenberg [Thu, 13 May 2004 15:17:49 +0000 (15:17 +0000)]
return on memory alloc fail
Daniel Stenberg [Thu, 13 May 2004 15:17:07 +0000 (15:17 +0000)]
better bailing out in case of no memory
Daniel Stenberg [Thu, 13 May 2004 15:16:36 +0000 (15:16 +0000)]
curl_free() doesn't free(NULL) but just returns
Daniel Stenberg [Thu, 13 May 2004 15:16:10 +0000 (15:16 +0000)]
deal with input arguments as NULL
Daniel Stenberg [Thu, 13 May 2004 14:14:03 +0000 (14:14 +0000)]
return CURLDIGEST_NOMEM when a memory function fails to deliver
Daniel Stenberg [Thu, 13 May 2004 14:13:12 +0000 (14:13 +0000)]
mark a value as alloced when strdup()ed to prevent memory leaks
Daniel Stenberg [Thu, 13 May 2004 14:12:49 +0000 (14:12 +0000)]
bail out when no memory occurs
Daniel Stenberg [Thu, 13 May 2004 10:40:17 +0000 (10:40 +0000)]
Gisle Vamem reintroduced the verifyconnect() call on windows as well, and
we now use it to provide more info back on connect failures.
Daniel Stenberg [Thu, 13 May 2004 10:38:57 +0000 (10:38 +0000)]
Gisle: minor fix
Daniel Stenberg [Thu, 13 May 2004 10:38:37 +0000 (10:38 +0000)]
no more Curl_ldap_done
Daniel Stenberg [Thu, 13 May 2004 09:01:15 +0000 (09:01 +0000)]
Added two two missing header files I missed when I removed the noinst_HEADERS
Daniel Stenberg [Thu, 13 May 2004 08:23:09 +0000 (08:23 +0000)]
added https.c
Daniel Stenberg [Thu, 13 May 2004 08:22:40 +0000 (08:22 +0000)]
basic https fetching script
Daniel Stenberg [Thu, 13 May 2004 07:52:33 +0000 (07:52 +0000)]
made 'runtests.pl -t' run over all the tests just like other command lines
Also made -t imply -n to disable valgrind, it runs sloooow otherwise.
This now manages to run all tests OK up to test case 100 (the first FTP one)
for me.
Daniel Stenberg [Thu, 13 May 2004 06:53:29 +0000 (06:53 +0000)]
James Bursa's patch to avoid free(NULL) (mainly because the libcurl memdebug
system thinks free(NULL) is badness)
Daniel Stenberg [Wed, 12 May 2004 13:24:40 +0000 (13:24 +0000)]
Check that memory functions return non-NULL or return error.
Daniel Stenberg [Wed, 12 May 2004 13:23:17 +0000 (13:23 +0000)]
make sure the returned pointer is NULL when encoding fails
Daniel Stenberg [Wed, 12 May 2004 13:05:01 +0000 (13:05 +0000)]
clean up and return better on out of memory
Daniel Stenberg [Wed, 12 May 2004 13:04:30 +0000 (13:04 +0000)]
return NULL on out of memory
Daniel Stenberg [Wed, 12 May 2004 12:06:39 +0000 (12:06 +0000)]
Curl_done() and the protocol-specific conn->curl_done() functions now all
take a CURLcode as a second argument, that is non-zero when Curl_done()
is called after an error was returned from Curl_do() (or similar).
Daniel Stenberg [Wed, 12 May 2004 12:05:13 +0000 (12:05 +0000)]
return faster when we "hit a wall" while printfing
Daniel Stenberg [Wed, 12 May 2004 12:04:38 +0000 (12:04 +0000)]
general cleanup to bail out nice and clean when a memory function fails
to deliver
Daniel Stenberg [Wed, 12 May 2004 09:02:54 +0000 (09:02 +0000)]
even if Curl_do() fails, we must call Curl_done() to do proper cleaning up
Daniel Stenberg [Wed, 12 May 2004 09:02:23 +0000 (09:02 +0000)]
improved cleaning up in case of memory allocation failures
Daniel Stenberg [Wed, 12 May 2004 08:26:56 +0000 (08:26 +0000)]
bail out if we can't allocate the new range string, and make use of aprintf()
instead of using snprintf() + strdup().
Daniel Stenberg [Wed, 12 May 2004 08:22:04 +0000 (08:22 +0000)]
Disable memdebug for the allocs done by the app, unless CURLTOOLDEBUG is
defined (which it never is atm).
Now, we can focus on making 'runtests -t [num]' work on all test cases and
we should never leak nor crash.
Daniel Stenberg [Wed, 12 May 2004 08:10:25 +0000 (08:10 +0000)]
new man page
Daniel Stenberg [Wed, 12 May 2004 08:00:21 +0000 (08:00 +0000)]
use size_t better for buffer and alloc lengths
Daniel Stenberg [Wed, 12 May 2004 07:56:01 +0000 (07:56 +0000)]
removed another jhrg-reference in a comment
Daniel Stenberg [Wed, 12 May 2004 07:55:05 +0000 (07:55 +0000)]
Edited comments only.
Daniel Stenberg [Wed, 12 May 2004 07:54:44 +0000 (07:54 +0000)]
The Curl_unencode_XXX_write() function take a ssize_t as third argument, so
we typecast on invoke.
Daniel Stenberg [Wed, 12 May 2004 06:27:40 +0000 (06:27 +0000)]
Left-over from before the return-code fix. This is probably the code that
causes xlc and gcc act differently on AIX.
Daniel Stenberg [Tue, 11 May 2004 21:17:03 +0000 (21:17 +0000)]
fixed Curl_open() to not leak anything if one malloc() fails, fix by
James Bursa only modified by me.
Daniel Stenberg [Tue, 11 May 2004 21:12:10 +0000 (21:12 +0000)]
- Nico Stappenbelt reported that when processing domain and search lines in
the resolv.conf file, the first entry encountered is processed and used as
the search list. According to the manual pages for both Linux, Solaris and
Tru64, the last entry of either a domain or a search field is used.
Daniel Stenberg [Tue, 11 May 2004 18:57:03 +0000 (18:57 +0000)]
revert the accidentally added use of strace
Daniel Stenberg [Tue, 11 May 2004 14:53:24 +0000 (14:53 +0000)]
minor leak in case of error, thanks to "./runtests.pl -n -t 25"
Daniel Stenberg [Tue, 11 May 2004 14:48:53 +0000 (14:48 +0000)]
clear up memory on failure a little better
Daniel Stenberg [Tue, 11 May 2004 14:22:51 +0000 (14:22 +0000)]
make the libidn pointer in the version struct a const
Daniel Stenberg [Tue, 11 May 2004 14:21:34 +0000 (14:21 +0000)]
they're at least 36 functions now
Daniel Stenberg [Tue, 11 May 2004 14:15:49 +0000 (14:15 +0000)]
updated to reflect reality!
Daniel Stenberg [Tue, 11 May 2004 12:22:50 +0000 (12:22 +0000)]
cut out the changelog, it is far from accurate anyway
Daniel Stenberg [Tue, 11 May 2004 11:48:09 +0000 (11:48 +0000)]
added curl_global_init_mem
Daniel Stenberg [Tue, 11 May 2004 11:30:23 +0000 (11:30 +0000)]
curl_global_init_mem() allows the memory functions to be replaced.
memory.h is included everywhere for this.
Daniel Stenberg [Tue, 11 May 2004 11:29:31 +0000 (11:29 +0000)]
*** empty log message ***
Daniel Stenberg [Tue, 11 May 2004 11:29:02 +0000 (11:29 +0000)]
Make this source code use our internal *printf().
Also some minor edits.
Daniel Stenberg [Tue, 11 May 2004 08:10:32 +0000 (08:10 +0000)]
Added recent events to the log
Daniel Stenberg [Tue, 11 May 2004 08:10:15 +0000 (08:10 +0000)]
more changes, more news, more people
Daniel Stenberg [Tue, 11 May 2004 08:09:56 +0000 (08:09 +0000)]
another official download mirror
Daniel Stenberg [Tue, 11 May 2004 07:54:55 +0000 (07:54 +0000)]
new attempt at an improved DSP-file generation
Daniel Stenberg [Mon, 10 May 2004 14:45:11 +0000 (14:45 +0000)]
slightly better dealing of bad mem situations
Daniel Stenberg [Mon, 10 May 2004 14:22:20 +0000 (14:22 +0000)]
Moved the fetching of the list of matching cookies to make it easier to free
that list in case something goes wrong in the function and we must bail out.
Courtesy of the torture testing.
Daniel Stenberg [Mon, 10 May 2004 14:21:19 +0000 (14:21 +0000)]
curl_slist_append() fixed to clear up properly if a memory function fails
Daniel Stenberg [Mon, 10 May 2004 14:04:35 +0000 (14:04 +0000)]
better detect if/when curl_slist_append() returns failure, and bail out
accordingly
Daniel Stenberg [Mon, 10 May 2004 14:04:06 +0000 (14:04 +0000)]
if a malloc fails, clear up the memory and return failure
Daniel Stenberg [Mon, 10 May 2004 10:52:29 +0000 (10:52 +0000)]
typo
Daniel Stenberg [Mon, 10 May 2004 10:50:43 +0000 (10:50 +0000)]
the aprintf() versions now return NULL if _any_ alloc along the way failed,
previously they could return a piece of the string, making it impossible
for the caller to detect errors.
Daniel Stenberg [Mon, 10 May 2004 10:49:35 +0000 (10:49 +0000)]
better detection for when add_buffer() returns failure, and return when that
happens
Daniel Stenberg [Mon, 10 May 2004 09:17:50 +0000 (09:17 +0000)]
Curl_hash_add() was modified to clear up better in case of internal failure.
When failing, it should not tamper at all with the data it was supposed to
add to the cache.
Daniel Stenberg [Mon, 10 May 2004 09:16:21 +0000 (09:16 +0000)]
checkpasswd() prevents segfault by checking that input argument is non-NULL
Daniel Stenberg [Mon, 10 May 2004 09:01:45 +0000 (09:01 +0000)]
James Bursa's adjustments to make the -t option work for any test case.
The -t is the "torture" test that first runs the test and counts the number
of allocations performed during it, then it runs the test repeatedly over and
over again and makes alloc number N fail to verify that we detect and return
properly from error cases everywhere.
Daniel Stenberg [Mon, 10 May 2004 08:57:37 +0000 (08:57 +0000)]
make Curl_llist_insert_next() fail properly if malloc() fails
Daniel Stenberg [Mon, 10 May 2004 08:57:18 +0000 (08:57 +0000)]
better checking that strdup() works
Daniel Stenberg [Mon, 10 May 2004 08:09:15 +0000 (08:09 +0000)]
Luca fixed the nc= in the digest line since it apparantly should not have
quotes...
Daniel Stenberg [Mon, 10 May 2004 07:11:52 +0000 (07:11 +0000)]
James Bursa added better error checking for failer memory calls when
building formposts
Daniel Stenberg [Mon, 10 May 2004 07:03:46 +0000 (07:03 +0000)]
build curllib.dsp from these
Daniel Stenberg [Mon, 10 May 2004 06:29:52 +0000 (06:29 +0000)]
don't use -i when checking for our own server
Daniel Stenberg [Fri, 7 May 2004 20:08:38 +0000 (20:08 +0000)]
James' and Gisle' reports/fixes
Daniel Stenberg [Fri, 7 May 2004 18:56:33 +0000 (18:56 +0000)]
James Bursa changed two error message to use the display-name instead of the
internally-used name.
Daniel Stenberg [Fri, 7 May 2004 18:54:09 +0000 (18:54 +0000)]
James Bursa's fix to make this deal with malloc(0) as OK to free()
Daniel Stenberg [Fri, 7 May 2004 18:46:28 +0000 (18:46 +0000)]
We don't support any long protocol names so we can use a smaller buffer.
Also, make sure we have room for the trailing zero, only scan to size-1.
Gisle Vanem reported.
Daniel Stenberg [Fri, 7 May 2004 09:50:49 +0000 (09:50 +0000)]
count the formdata size using a 64bit size if avaialble
Daniel Stenberg [Fri, 7 May 2004 09:45:33 +0000 (09:45 +0000)]
Made the lib/curllib.dsp file get generated automaticly
Daniel Stenberg [Fri, 7 May 2004 09:42:23 +0000 (09:42 +0000)]
don't use a magic define name
Daniel Stenberg [Fri, 7 May 2004 09:41:28 +0000 (09:41 +0000)]
This file is now generated at dist-time.
Daniel Stenberg [Fri, 7 May 2004 09:41:05 +0000 (09:41 +0000)]
Generate curllib.dsp on dist.time from msvcproj.head msvcproj.foot and the
known source files. Not actually verified to work yet.
Daniel Stenberg [Fri, 7 May 2004 06:18:47 +0000 (06:18 +0000)]
get the display host name properly
Daniel Stenberg [Thu, 6 May 2004 15:17:10 +0000 (15:17 +0000)]
removed two odd comments