On Wed, May 25, 2022 at 10:05 PM Gurjeet Singh <gurjeet@singh.im> wrote:
> I have added a comment to be_tls_init(), which I hope explains this
> difference in treatment of errors. I have also added comments to
> be_tls_init(), explaining why we don't destroy/free the global
> SSL_context variable in case of an error in re-initialization of SSL;
> it's not just an optimization, it's essential to normal server
> operation.
Please see attached patch that reverts the unintentional removal of a
comment in be_tls_init(). Forgot to put that comment back in after my
edits.
Best regards,
Gurjeet
http://Gurje.et