always get included with the relevant files.

SGI' cc; enable workaround.
Submitted by: Martin Kraemer

in cryptlib.h (which is often included as "../cryptlib.h"), then the
question remains relative to which directory this is to be interpreted.
gcc went one further directory up, as intended; but makedepend thinks
differently, and so probably do some C compilers.  So the ../ must go away;
thus e_os.h goes back into include/openssl (but I now use
#include "openssl/e_os.h" instead of <openssl/e_os.h> to make the point) --
and we have another huge bunch of dependency changes.  Argh.

of type XXX ** not XXX *

There were problems with putting e_os.h just into the top directory,
because the test programs are compiled within test/ in the "standard"
case in in their original directories in the makefile.one case;
and in the latter symlinks may not be available.

include file.

Submitted by: Richard Levitte <levitte@stacken.kth.se>

Pointed out by Juergen Moellenhoff <jurgen@oic.de>

Submitted by: Martin Kraemer <Martin.Kraemer@MchP.Siemens.De>

Submitted by: Sylvain Robitaille <syl@alcor.concordia.ca>

Submitted by: Andy Polyakov <appro@fy.chalmers.se>

released version of mklink.sh -- the latter couldn't even handle
the ../../include/openssl case).

This also avoids the problems with SC4.2 and unpatched SC5.

Submitted by: Andy Polyakov <appro@fy.chalmers.se>

do sensible things with extensions.

better suited for finding what went wrong in case that some compiler
versions create an output we can't parse.

version), so we need
   echo $GCCVAR | sed ...
instead of
   echo "$GCCVAR" | sed ...
to process it as intended.