aboutsummaryrefslogtreecommitdiff
path: root/tests/server/sws.c
AgeCommit message (Collapse)Author
2008-09-20fix compiler warning: external definition with no prior declarationYang Tse
2008-09-18fix compiler warning: external declaration in primary source fileYang Tse
2008-09-04fix print formatting string directivesYang Tse
2008-08-29- When libcurl was doing a HTTP POST and the server would respond withDaniel Stenberg
"Connection: close" and actually close the connection after the response-body, libcurl could still have outstanding data to send and it would not properly notice this and stop sending. This caused weirdness and sad faces. http://curl.haxx.se/bug/view.cgi?id=2080222 Note that there are still reasons to consider libcurl's behavior when getting a >= 400 response code while sending data, as Craig Perras' note "http upload: how to stop on error" specifies: http://curl.haxx.se/mail/archive-2008-08/0138.html
2008-07-30Added test case 1051 to test Location: following with PUT, as reportedDan Fandrich
by Ben Sutcliffe. The test when run manually shows a problem in curl, but the test harness web server doesn't run the test correctly so it's disabled for now.
2008-04-23improve synchronization between test harness runtests.pl scriptYang Tse
and test harness servers to minimize risk of false test failures. http://curl.haxx.se/mail/lib-2008-04/0392.html
2008-04-22fix minor memory leak triggered upon test failureYang Tse
2008-04-22Remove fflush() + fsync() previously introduced accelerated writing ofYang Tse
server input and response request files of the test harness sws server. Reintroduce, for test # 1001, the <postcheck> small delay. The delay is needed even with the accelerated writing of server input and response request files in test harness sws server. http://curl.haxx.se/mail/lib-2008-04/0385.html
2008-04-20accelerate the writing of server input and response request filesYang Tse
to disk, trying to defeat file and disk write-behind algorithms
2008-02-28header inclusion cleanupYang Tse
2008-02-26refactor some code out to write_pidfile() in util.cYang Tse
2008-02-06Use a long int data type to handle getpid() resultYang Tse
2008-02-05Fix buffer size specification.Yang Tse
Improve handling of boundary conditions for huge requests.
2008-02-05Minor variable type cleanups.Yang Tse
Disable "swsbounce" mode when the received request isn't for the same test and part number.
2008-02-05proper initialization of httprequest, no longer zeroing out twiceYang Tse
the whole 150000+ bytes struct, and also removing an equally big additional buffer for pipelining treatment.
2008-01-25improve request initialization for test harness HTTP serverYang Tse
2008-01-25Dmitry Kurochkin's test harness HTTP server pipelining fix fot test 530Yang Tse
2007-10-26Detect, log and avoid storing a request with a negative size.Yang Tse
2007-10-25Allow test server to handle binary POSTs.Patrick Monnerat
Tests 35, 544 545 added: binary data POSTs.
2007-10-09also log error message stringYang Tse
2007-10-04On error, close "log/server.response"Yang Tse
2007-10-03If TCP_NODELAY is not defined we can't disable the Nagle algorithmYang Tse
2007-10-02Disable the Nagle algorithm and send back responses in small chunks in anDaniel Stenberg
attempt to force smaller bits to get read by clients.
2007-02-19Oops missing varYang Tse
2007-02-19add debug messages for initialization failuresYang Tse
2007-02-16use macros ERRNO, SET_ERRNO(), SOCKERRNO and SET_SOCKERRNO() for errno handlingYang Tse
2007-02-14compiler warning fixYang Tse
2007-02-14enhance HTTP server request input writing,Yang Tse
retrying upon EINTR errors.
2006-11-25Venkat Akella found out that libcurl did not like HTTP responses that simplyDaniel Stenberg
responded with a single status line and no headers nor body. Starting now, a HTTP response on a persistent connection (i.e not set to be closed after the response has been taken care of) must have Content-Length or chunked encoding set, or libcurl will simply assume that there is no body. To my horror I learned that we had no less than 57(!) test cases that did bad HTTP responses like this, and even the test http server (sws) responded badly when queried by the test system if it is the test system. So although the actual fix for the problem was tiny, going through all the newly failing test cases got really painful and boring.
2006-11-13Tor Arntsen spotted this mistakeDaniel Stenberg
2006-10-23Avoid trying to compare more than strlen bytes.Yang Tse
2006-10-23Replace is*() macros with our own IS*() ones.Yang Tse
2006-10-11Remove redundant __CYGWIN__ symbol checkYang Tse
2006-09-12Cygwin preprocessor adjustmentsYang Tse
2006-09-08test 530 is the first ever HTTP pipelining test for libcurlDaniel Stenberg
2006-08-14Replace exit() with return() in main()Yang Tse
2006-07-19Abort if unable to write pid file.Yang Tse
2006-07-17-Use curl_socket_t instead of int.Yang Tse
-Log errno in message if setsockopt() fails. -Close listener socket on major errors.
2006-07-12Fix compiler warning: comparison between signed and unsignedYang Tse
2006-07-12Place parenthesis surrounding macro parameters so that the use of sread and ā†µYang Tse
swrite is more intuitive.
2006-07-12Pay attention when typecasting an operationYang Tse
2006-07-12sread now returns ssize_tYang Tse
2006-07-12Use platform's native types for recv() and send() arguments.Yang Tse
2006-05-10removed variable declarations shadowing previously declared variablesDaniel Stenberg
2006-04-10if configure found a fork(), sws supports --fork which is *NOT* used by theDaniel Stenberg
ordinary test suite. Also removed the perror() calls and instead made the logging output the errno code to ease error tracking using logs.
2006-02-25Fix typo.Gisle Vanem
2006-02-18Ulf Härnhammar fixed a format string (printf style) problem in the NegotiateDaniel Stenberg
code. It should however not be the cause of any troubles. He also fixed a few similar problems in the HTTP test server code.
2006-01-09Made the copyright year match the latest modification's year.Daniel Stenberg
2006-01-03modified to hush compiler warningsDaniel Stenberg
2006-01-021. sws now supports two new "commands" and 2. if built withDaniel Stenberg
CURL_SWS_FORK_ENABLED defined it forks for each new connection and thus can support any amount of connection clients (used for hiper tests and not for the standard plain curl test suite)