On Mon, Apr 09, 2001 at 04:23:41PM +0200, Richard Levitte - VMS Whacker wrote:
> From: Bodo Moeller <[EMAIL PROTECTED]>
>
> moeller> I don't think it's good to have active assert()s in default
> moeller> compilations of libraries -- in particular, libraries for
> moeller> multi-threaded programs. Everywhere else, we define NDEBUG unless an
> moeller> appropriate ..._DEBUG preprocessor symbol is defined (otherwise we
> moeller> should use -DNDEBUG in all non-debug configurations). With that
> moeller> SSLerr(), the problem will still be noticed when testing, and in
> moeller> addition to this the test is active even if NDEBUG is defined.
>
> Then, the right thing would also have been to add a small section at
> the beginning of s3_enc.c that defines NDEBUG if SSL_DEBUG (or
> something like that) is undefined, wouldn't it? I'm not saying that
> an error shouldn't be returned, however...
Sure, we can use both kinds of bug checking -- assert() if SSL_DEBUG
is defined, and SSLerr() in case assert() is disabled. However there
is no SSL_DEBUG symbol yet, so we'd have to change various Configure
entries, so in the end it would probably be more convenient to simply
add a loop to s3_enc.c so that it works even if buf[] is smaller than
it would ideally be.
--
Bodo Möller <[EMAIL PROTECTED]>
PGP http://www.informatik.tu-darmstadt.de/TI/Mitarbeiter/moeller/0x36d2c658.html
* TU Darmstadt, Theoretische Informatik, Alexanderstr. 10, D-64283 Darmstadt
* Tel. +49-6151-16-6628, Fax +49-6151-16-6036
______________________________________________________________________
OpenSSL Project http://www.openssl.org
Development Mailing List [EMAIL PROTECTED]
Automated List Manager [EMAIL PROTECTED]
