aboutsummaryrefslogtreecommitdiff
path: root/lib
AgeCommit message (Collapse)Author
2002-01-16- shrunk the BUFSIZE define from 50K to 20KDaniel Stenberg
- made a separate buffer for uploads (due to the non-blocking stuff) - added two connectdata struct fields for non-blocking uploads
2002-01-16modified to deal with the new non-blocking versions of Curl_read() andDaniel Stenberg
Curl_write().
2002-01-16Curl_write's 5th argument now is signedDaniel Stenberg
2002-01-16Curl_read() and Curl_write() are both now adjusted to return properly inDaniel Stenberg
cases where EWOULDBLOCK or equivalent is returned. We must not block.
2002-01-16Curl_write() now takes a different 5th argumentDaniel Stenberg
2002-01-16Somewhat ugly fix to deal with non-blocking sockets. We just loop and tryDaniel Stenberg
again. THIS IS NOT A NICE FIX.
2002-01-16Somewhat ugly fix to deal with non-blocking sockets. We just loop and tryDaniel Stenberg
again. THIS IS NOT A NICE FIX. We should/must make a select() then and only retry when we can write to the socket again.
2002-01-14Curl_read() now returns a negative return code if EWOULDBLOCK or similarDaniel Stenberg
2002-01-14a memory leak when name lookup failed is now removedDaniel Stenberg
2002-01-14Remove erreaneous include, setup.h is included one line aboveSterling Hughes
2002-01-09Cris Bailiff found out that when the SSL session cache was filled, libcurlDaniel Stenberg
would crash. This corrects the problem.
2002-01-08added a missing failf() before returning an error codeDaniel Stenberg
2002-01-08this seems to correct the SSL reading problem introduced when switchingDaniel Stenberg
over to non-blocking sockets, but this loops very nastily. We should return back to the select() and wait there until more data arrives, not just blindly attempt again and again...
2002-01-08added two typecasts to prevent compiler (gcc3) warningsDaniel Stenberg
2002-01-081) the dns_cache_timeout should be an integer, not a boolSterling Hughes
2) in the curl_dns_cache_entry structure, timestamp should be a time_t instead of an integer (although I doubt it matters).
2002-01-08Add support for DNS cache timeouts via the CURLOPT_DNS_CACHE_TIMEOUT option.Sterling Hughes
The default cache timeout for this is 60 seconds, which is arbitrary and completely subject to change :)
2002-01-07As identified in bug report #495290, the last "name=value" pair in aDaniel Stenberg
Set-Cookie: line was ignored if they didn't end with a trailing semicolon. This is indeed wrong syntax, but there are high-profile web sites out there sending cookies like that so we must make a best-effort to parse them.
2002-01-07the bool typedef is now made unsigned, to make sure it stays that on allDaniel Stenberg
platforms, unrelated to what they might prefer by default
2002-01-07removed silly check for >=0 of a supposedly unsigned value!Daniel Stenberg
2002-01-07Probably not necessary, but good practice.Sterling Hughes
2002-01-07Make cach'ing work with threads now, there are now three cases:Sterling Hughes
- Use a global dns cache (via setting the tentatively named, CURLOPT_DNS_USE_GLOBAL_CACHE option to true) - Use a per-handle dns cache, by default - Use a pooled dns cache when in the "multi" interface
2002-01-07Modified to use non-blocking sockets all the time.Daniel Stenberg
2002-01-07added proper breaks in the switch()Daniel Stenberg
2002-01-07HTTP response 204 should be treated similar to 304, that is we must notDaniel Stenberg
expect (nor read) any response-body
2002-01-07added precautions to not go insane when two matching cookies end up in theDaniel Stenberg
cookie list, even though they're not supposed to do that...
2002-01-04Add hash and llist to VC dsp filesm
2002-01-04Add hash and llist to VC makefilesm
2002-01-04replaced printf() => Curl_sendf()Daniel Stenberg
2002-01-04removed a commented lineDaniel Stenberg
2002-01-04fixed an inet_ntoa() occurance to use inet_ntoa_r() if it is available.Daniel Stenberg
I also replaced all printf() calls with calls to Curl_failf()
2002-01-04an unconditional occurance of inet_ntoa() now uses inet_ntoa_r() on allDaniel Stenberg
platforms that have such a function. This affects multi-thread running libcurls on IPv4 systems that have VERBOSE switched on. The previous version was risking that another thread overwrote the data before it was read out in this thread. There could possibly also be a slight risk that the data isn't zero terminated for a short while and thus could cause the thread to crash...
2002-01-04#include the local "inet_ntoa_r.h" file if no proto was found in the globalDaniel Stenberg
header directory but the function *is* present!
2002-01-04The buffer in ftp_pasv_verbose(), used for gethostbyaddr_r(), is now definedDaniel Stenberg
to become properly 8-byte aligned on 64-bit archs. Philip Gladstone reported.
2002-01-04The FTP response lines are now passed to the function callback registered forDaniel Stenberg
headers.
2002-01-03merged the multi-dev branch back into MAIN againDaniel Stenberg
2002-01-03Sterling Hughes' provided initial DNS cache source code.Daniel Stenberg
2002-01-03minor changesDaniel Stenberg
2002-01-03pass an 'int' as the third argument to bind()Daniel Stenberg
2002-01-02Philip Gladstone's 64-bit sparc native compiler compatibility issues fixed.Daniel Stenberg
2001-12-20*cool* fix by Björn Stenberg, makes proxy transfers work better...! :-)Daniel Stenberg
2001-12-20If nobody is set we won't download any FTP file. If include_header is set,Daniel Stenberg
we return a set of headers not more. This enables FTP operations that don't transfer any data, only perform FTP commands.
2001-12-17Götz Babin-Ebell's OpenSSL ENGINE patchDaniel Stenberg
2001-12-14Marcus Webster's newly added CURLFORM_CONTENTHEADERDaniel Stenberg
2001-12-13minor changesDaniel Stenberg
2001-12-11solaris 2.5.1 needs the sys/types.h file before the sys/socket.hDaniel Stenberg
2001-12-11failf() calls should not have newlines in the message string!Daniel Stenberg
2001-12-07corrected the comment above gmtime_rDaniel Stenberg
2001-12-07added gmtime_r checkDaniel Stenberg
2001-12-05Jon Travis suggested fix. when CURLOPT_HTTPGET is used we must assignDaniel Stenberg
set.upload to FALSE or else we might still get an upload if the previous operation was an upload!
2001-12-04Eric's #include fixes for better macos compilesDaniel Stenberg