apparently impossible to compose blended code with would perform
satisfactory on all x86 and x86_64 cores, an extra RC4_CHAR
code-path is introduced and P4 core is detected at run-time. This
way we keep original performance on non-P4 implementations and
turbo-charge P4 performance by factor of 2.8x (on 32-bit core).

results in even higher performance gain of 3.3x:-) At least on
Opteron...

if $(EXHEADER) is empty.

Notified by many, solution suggested by Carson Gaspar <carson@taltos.org>

"Proper" means "compiles and passes test." Versioning is broken (I think).

COFF and a.out targets [similar to ELF targets]. You might notice some
rudementary support for shared mingw builds under cygwin. It works (it
produces cryptoeay32.dll and ssleay32.dll with everything exported by
name), but it's primarily for testing/debugging purposes, at least for
now...

linux-x86_64 target.

This is part of a large change submitted by Markus Friedl <markus@openbsd.org>

PR: 780
Submitted by: Verdon Walker <VWalker@novell.com>
Reviewed by: Richard Levitte

Submitted by: Tim Rice <tim@multitalents.net>

PR: #735, #765

FreeBSD. Goal is to extend support even to SCO5, UnixWare/OpenUnix...

always give the expected result on some platforms.

I've covered all the memset()s I felt safe modifying, but may have missed some.

give it.
For 0.9.7 and up, that means util/domd needs to remove those double
dashes from the argument list when gcc is used to find the
dependencies.

handled properly.
Part of PR 75

Change all calls to low level digest routines in the library and
applications to use EVP. Add missing calls to HMAC_cleanup() and
don't assume HMAC_CTX can be copied using memcpy().

Note: this is almost identical to the patch submitted to openssl-dev
by Verdon Walker <VWalker@novell.com> except some redundant
EVP_add_digest_()/EVP_cleanup() calls were removed and some changes
made to avoid compiler warnings.

errors can be tolerated, hide the error from 'make'.
This gives shorter output both if ranlib fails and if
it works.

Submitted by Brad Dominy <jdominy@darwinuser.org>

missed any.

This compiles and runs on Linux, and external applications have no
problems with it.  The definite test will be to build this on VMS.

sure they are available in opensslconf.h, by giving them names starting
with "OPENSSL_" to avoid conflicts with other packages and by making
sure e_os2.h will cover all platform-specific cases together with
opensslconf.h.

I've checked fairly well that nothing breaks with this (apart from
external software that will adapt if they have used something like
NO_KRB5), but I can't guarantee it completely, so a review of this
change would be a good thing.

acceptable, since all that happens if it fails is a library with
an index, which makes linking slower, but still working correctly.

"Jan Mikkelsen" <janm@transactionsite.com> correctly states that the
OpenSSL header files have #include's and extern "C"'s in an incorrect
order.  Thusly fixed.

This was meant for building individual ciphers separately;
but nothing of this is maintained, it does not work
because we rely on central configuration by the Configure
utility with <openssl/opensslconf.h> etc., so the files
are only wasting space and time.

choice of parameters) when they are needed.

proposed by Holger Reif

Submitted by: Lennart Bång, Bodo Möller