Hi,

On Wed, Jan 12, 2005 at 08:13:43AM +0100, Gert Doering wrote:
> Unfortunately, it doesn't.  So the problem is not the assembly code:
> 
> error calculating RIPEMD160 on ''
> got c12836ad0d061da6ccde02fb0b5be87f0c62a4a5 instead of 9c1185a5c5e9fc54612808977ee8f548b2258d31
[..]
> I think I'll try to take this "upstream" to the OpenSSL people, and 
> see whether they have any idea about it.

The OpenSSL people say "this is a known issue with gcc on (some) sparc64
platforms, it's a compiler optimization issue".

http://www.aet.tu-cottbus.de/rt2/Ticket/Display.html?id=1000
is the OpenSSL ticket I opened.

If one compiles the two .o files in crypto/ripemd/ without "-O3", the
ripemd tests pass.  Maybe this would be doable in an automated way for 
the pkgsrc version?

gert

-- 
USENET is *not* the non-clickable part of WWW!
                                                           //www.muc.de/~gert/
Gert Doering - Munich, Germany                             gert@greenie.muc.de
fax: +49-89-35655025                        gert@net.informatik.tu-muenchen.de