Re: FE/BE protocol oddity

Tom Lane writes:

> Good point.  Probably, PQfinish should only send the X message if the
> connection has gotten past the authentication stage.  A separate but
> also useful change would be to do immediate socket close on detecting
> auth failure, before returning to the client application.

Further investigation shows that all of this used to work correctly until
the infamous async connection patch.  The comment now reads:
(fe-connect.c)
/* * We used to close the socket at this point, but that makes it * awkward for those above us if they wish to remove
thissocket from * their own records (an fd_set for example).  We'll just have this * socket closed when PQfinish is
called(which is compulsory even * after an error, since the connection structure must be freed). */
 

I guess there is sort of a point there.  So I'm leaning towards adding a
"startup complete" flag somewhere in PGconn and simply fix up
closePGconn().

> > Btw., is recv(sock, x, 1, MSG_PEEK) == 0 an appropriate way to check for a
> > closed connection without reading anything?
>
> Seems a little risky as far as portability goes; is MSG_PEEK supported
> on BeOS, Darwin, Cygwin, etc?

Darwin is FreeBSD, so yes.  Cygwin says it supports recv() and MSG_PEEK is
trivial from there.  BeOS is going backrupt before the next release
anyway. ;-)  Seriously, in the worst case we'll get EINVAL.

> Might be better to fix the backend libpq routines to understand
> whether a connection-close event is expected or not, and only emit a
> complaint to the log when it's not.  Not sure how far such a change
> would need to propagate though...

Deep...

-- 
Peter Eisentraut   peter_e@gmx.net   http://funkturm.homeip.net/~peter