aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2004-05-12Check that memory functions return non-NULL or return error.Daniel Stenberg
2004-05-12make sure the returned pointer is NULL when encoding failsDaniel Stenberg
2004-05-12clean up and return better on out of memoryDaniel Stenberg
2004-05-12return NULL on out of memoryDaniel Stenberg
2004-05-12Curl_done() and the protocol-specific conn->curl_done() functions now allDaniel Stenberg
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).
2004-05-12return faster when we "hit a wall" while printfingDaniel Stenberg
2004-05-12general cleanup to bail out nice and clean when a memory function failsDaniel Stenberg
to deliver
2004-05-12even if Curl_do() fails, we must call Curl_done() to do proper cleaning upDaniel Stenberg
2004-05-12improved cleaning up in case of memory allocation failuresDaniel Stenberg
2004-05-12bail out if we can't allocate the new range string, and make use of aprintf()Daniel Stenberg
instead of using snprintf() + strdup().
2004-05-12Disable memdebug for the allocs done by the app, unless CURLTOOLDEBUG isDaniel Stenberg
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.
2004-05-12new man pageDaniel Stenberg
2004-05-12use size_t better for buffer and alloc lengthsDaniel Stenberg
2004-05-12removed another jhrg-reference in a commentDaniel Stenberg
2004-05-12Edited comments only.Daniel Stenberg
2004-05-12The Curl_unencode_XXX_write() function take a ssize_t as third argument, soDaniel Stenberg
we typecast on invoke.
2004-05-12Left-over from before the return-code fix. This is probably the code thatDaniel Stenberg
causes xlc and gcc act differently on AIX.
2004-05-11fixed Curl_open() to not leak anything if one malloc() fails, fix byDaniel Stenberg
James Bursa only modified by me.
2004-05-11- Nico Stappenbelt reported that when processing domain and search lines inDaniel Stenberg
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.
2004-05-11revert the accidentally added use of straceDaniel Stenberg
2004-05-11minor leak in case of error, thanks to "./runtests.pl -n -t 25"Daniel Stenberg
2004-05-11clear up memory on failure a little betterDaniel Stenberg
2004-05-11make the libidn pointer in the version struct a constDaniel Stenberg
2004-05-11they're at least 36 functions nowDaniel Stenberg
2004-05-11updated to reflect reality!Daniel Stenberg
2004-05-11cut out the changelog, it is far from accurate anywayDaniel Stenberg
2004-05-11added curl_global_init_memDaniel Stenberg
2004-05-11curl_global_init_mem() allows the memory functions to be replaced.Daniel Stenberg
memory.h is included everywhere for this.
2004-05-11*** empty log message ***Daniel Stenberg
2004-05-11Make this source code use our internal *printf().Daniel Stenberg
Also some minor edits.
2004-05-11Added recent events to the logDaniel Stenberg
2004-05-11more changes, more news, more peopleDaniel Stenberg
2004-05-11another official download mirrorDaniel Stenberg
2004-05-11new attempt at an improved DSP-file generationDaniel Stenberg
2004-05-10slightly better dealing of bad mem situationsDaniel Stenberg
2004-05-10Moved the fetching of the list of matching cookies to make it easier to freeDaniel Stenberg
that list in case something goes wrong in the function and we must bail out. Courtesy of the torture testing.
2004-05-10curl_slist_append() fixed to clear up properly if a memory function failsDaniel Stenberg
2004-05-10better detect if/when curl_slist_append() returns failure, and bail outDaniel Stenberg
accordingly
2004-05-10if a malloc fails, clear up the memory and return failureDaniel Stenberg
2004-05-10typoDaniel Stenberg
2004-05-10the aprintf() versions now return NULL if _any_ alloc along the way failed,Daniel Stenberg
previously they could return a piece of the string, making it impossible for the caller to detect errors.
2004-05-10better detection for when add_buffer() returns failure, and return when thatDaniel Stenberg
happens
2004-05-10Curl_hash_add() was modified to clear up better in case of internal failure.Daniel Stenberg
When failing, it should not tamper at all with the data it was supposed to add to the cache.
2004-05-10checkpasswd() prevents segfault by checking that input argument is non-NULLDaniel Stenberg
2004-05-10James Bursa's adjustments to make the -t option work for any test case.Daniel Stenberg
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.
2004-05-10make Curl_llist_insert_next() fail properly if malloc() failsDaniel Stenberg
2004-05-10better checking that strdup() worksDaniel Stenberg
2004-05-10Luca fixed the nc= in the digest line since it apparantly should not haveDaniel Stenberg
quotes...
2004-05-10James Bursa added better error checking for failer memory calls whenDaniel Stenberg
building formposts
2004-05-10build curllib.dsp from theseDaniel Stenberg