1996-12-22 - Re: DES implementation in C

Header Data

From: Dale Thorn <dthorn@gte.net>
To: Eric Young <eay@mincom.com>
Message Hash: 2db1c2ce194322fabaf463aec2d64666acb0bbf36dc60b2bcdb3dbf8ab3bc8b4
Message ID: <32BD7DC4.745A@gte.net>
Reply To: <Pine.SOL.3.95.961223003733.19844A-100000@orb.mincom.oz.au>
UTC Datetime: 1996-12-22 18:29:47 UTC
Raw Date: Sun, 22 Dec 1996 10:29:47 -0800 (PST)

Raw message

From: Dale Thorn <dthorn@gte.net>
Date: Sun, 22 Dec 1996 10:29:47 -0800 (PST)
To: Eric Young <eay@mincom.com>
Subject: Re: DES implementation in C
In-Reply-To: <Pine.SOL.3.95.961223003733.19844A-100000@orb.mincom.oz.au>
Message-ID: <32BD7DC4.745A@gte.net>
MIME-Version: 1.0
Content-Type: text/plain


Eric Young wrote:
> On Fri, 20 Dec 1996, Peter Trei wrote:
> > > So yet again my dislike of doing assember has been justified.  One just
> > > needs a good compiler and be willing to put in the correct C code :-).
> > > While I know the asm is faster, the C compiler does a better job of
> > > the chaining between the 16 inlined inner loops.

[snippo]

If you could produce a 100% 'C' version, with in-line provisions for
replaceable 'C' or assembler functions, and insure that there is no
significant performance loss for having this portability, then you
could at least issue the 100% 'C' code for any machine or O/S.

This would make a lot of people happy if the application has wide
interest.  Is this possible, even if it's more work?






Thread