CURLcode curl_easy_setopt(CURL *handle, CURLOPT_TIMEOUT_MS, long timeout);
If libcurl is built to use the standard system name resolver, that portion of the transfer will still use full-second resolution for timeouts with a minimum timeout allowed of one second.
In unix-like systems, this might cause signals to be used unless CURLOPT_NOSIGNAL(3) is set.
If both CURLOPT_TIMEOUT(3) and CURLOPT_TIMEOUT_MS(3) are set, the value set last will be used.
Since this puts a hard limit for how long time a request is allowed to take, it has limited use in dynamic use cases with varying transfer times. You are then advised to explore CURLOPT_LOW_SPEED_LIMIT(3), CURLOPT_LOW_SPEED_TIME(3) or using CURLOPT_PROGRESSFUNCTION(3) to implement your own timeout logic.
CURL *curl = curl_easy_init(); if(curl) { curl_easy_setopt(curl, CURLOPT_URL, "https://example.com"); /* complete within 20000 milliseconds */ curl_easy_setopt(curl, CURLOPT_TIMEOUT_MS, 20000L); curl_easy_perform(curl); }