The commit
9dd85bc unintentionally changed the way we compute the time
spent waiting for 100-continue. In particular, when using a SSL client
certificate, the time spent by SSL handshake was included and could
cause the CURL_TIMEOUT_EXPECT_100 timeout to be mistakenly fired up.
Bug: https://bugzilla.redhat.com/767490
Reported by: Mamoru Tasaka
o timer: restore PRETRANSFER timing [27]
o libcurl.m4: Fix quoting arguments of AC_LANG_PROGRAM [28]
o appconnect time fixed for non-blocking connect ssl backends [29]
+ o do not include SSL handshake into time spent waiting for 100-continue [32]
This release includes the following known bugs:
[29] = http://curl.haxx.se/mail/lib-2011-12/0211.html
[30] = http://curl.haxx.se/libcurl/c/curl_easy_setopt.html#CURLOPTACCEPTTIMOUTMS
[31] = http://curl.haxx.se/mail/lib-2011-12/0133.html
+ [32] = https://bugzilla.redhat.com/767490
(data->state.proto.http->sending == HTTPSEND_BODY)) {
/* wait with write until we either got 100-continue or a timeout */
k->exp100 = EXP100_AWAITING_CONTINUE;
- k->start100 = k->start;
+ k->start100 = Curl_tvnow();
/* set a timeout for the multi interface */
Curl_expire(data, CURL_TIMEOUT_EXPECT_100);