I have some code that does a sort of read buffering scheme. I need to be able to tell if n characters are available to read, and if so to read them. I tried to use SSL_pending() to do this, but found that upon entering for the first time, SSL_pending() _always_ returned 0. I am also not convinced that SSL_pending() will ever equal n if n is large. So now I am doing an SSL_read(_,&toss,0) before the SSL_pending() as a way to "kick start" things and keeping a buffer of data read so far. If the buffer isn't large enough and if SSL_pending() says there's something there (after being kick-started), then the buffer is extended until either it's long enough or until kick-starting fails to yield a non-zero SSL_pending(). I am not convinced this is the best way to achieve the desired result. 1. Is SSL_read() really non-blocking? It will really return 0 if there's nothing to read? Will it return short reads (SSL_read(_,_,65536)<65536)? Always or only if it runs out of data to read? 2. Can I presume that SSL_read() will return non-zero if either SSL_pending() returns non-zero or if select() says that the underlying file descriptor has data waiting? How else would I? SSL_peek(_,toss,1)? 3. Will a 0 length read assure that SSL_pending() returns non-zero if there's anything available? ______________________________________________________________________ OpenSSL Project http://www.openssl.org User Support Mailing List [EMAIL PROTECTED] Automated List Manager [EMAIL PROTECTED]
