Bruce Momjian <pgman@candle.pha.pa.us> writes:
> Actually, the docs say SSL_read/write can error needing READ or WRITE:
> http://www.openssl.org/docs/ssl/SSL_read.html#
Only when using a non-blocking BIO, which the backend had better not be
doing. But a few lines of useless code aren't a big deal...
> The SSL_write docs are the same. I have applied the following patch to
> handle both new cases. Does this help the SSL test program you have?
Where is the "default:" case to preserve sanity when the SSL call returns
an error code other than the ones the switch knows about? ISTM the lack
of this defense *is* a big deal.
regards, tom lane