[openssl-dev] s_client uses SSL_write(..., 0) for renegotiation?

[Judson Wilson]

Should s_client really be using a zero-byte-length SSL_write(...) as a
followup to SSL_renegotiate(...)?  All versions of the man pages on the
OpenSSL web site state this will yield undefined behavior (which I find
odd, but whatever).

Is SSL_do_handshake(...) more appropriate?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mta.openssl.org/pipermail/openssl-dev/attachments/20151205/603be490/attachment.html>