This should fix the recent AppVeyor failures.

[extended tests]
Reviewed-by: NAndy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4171)

Reviewed-by: NAndy Polyakov <appro@openssl.org>
Reviewed-by: NRichard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4167)

$(SHLIB_MAJOR).$(SHLIB_MINOR) is really a synonym for
$(SHLIB_VERSION_NUMBER), and is therefore an added complexity,
so better to use $(SHLIB_VERSION_NUMBER) directly.  SHLIB_MAJOR and
SHLIB_MINOR are now unused, but are kept around purely as information
in case someone relies on their existence.

At the same time, add support for custom shared library extensions
with the three new Makefile variables SHLIB_EXT, SHLIB_EXT_SIMPLE and
SHLIB_EXT_IMPORT.  By default, they hold the variants of shared
library extensions we support.  On mingw and cygwin, SHLIB_EXT_IMPORT
is defined; on all other Unix platforms, it's empty.

An example to get shared libraries with a slightly different SOVER name:

    $ make SHLIB_EXT='.$(SHLIB_VERSION_NUMBER).so'

Fixes #3902
Reviewed-by: NAndy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3964)

Makefile.shared was designed to figure out static library names,
shared library names, library version compatibility, import library
names and the like on its own.  This was a design for pre-1.1.0
OpenSSL because the main Makefile didn't have all that knowledge.

With 1.1.0, the situation isn't the same, a lot more knowledge is
included in the main Makefile, and while Makefile.shared did things
right most of the time (there are some corner cases, such as the
choice of .sl or .so as DSO extension on some HPUX versions), there's
still an inherent fragility when one has to keep an eye on
Makefile.shared to make sure it produces what the main Makefile
produces.

This change simplifies Makefile.shared by removing all its
"intelligence" and have it depend entirely on the input from the main
Makefile instead.  That way, all the naming is driven from
configuration data.
Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3983)

Reviewed-by: NTim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3880)

Remove some incorrect copyright references.
Move copyright to standard place
Add OpenSSL copyright where missing.
Remove copyrighted file that we don't use any more
Remove Itanium assembler for RC4 and MD5 (assembler versions of old and
weak algorithms for an old chip)
Standardize apps/rehash copyright comment; approved by Timo
Put dual-copyright notice on mkcert
Reviewed-by: NRichard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3691)

Reviewed-by: NRichard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3801)

Based on cryptodev-linux
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3744)

Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3745)

Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3650)

This new target is used to build all generated files and only that.
This can be used to prepare everything that requires things like perl
for a system that lacks perl and then move everything to that system
and do the rest of the build there.
Reviewed-by: NTim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3695)

Because apps/progs.h isn't configuration agnostic, it's not at all
suited for 'make update' or being versioned, so change it to be
dynamically generated.
Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3688)

Fix test for "documenting private functions"
And add -p flag to doc-nits recipe
Mark when things were deprecated, if doc'd as such
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3624)

The list of programs hit nmake's maximum line length, so we split up the
line in smaller chunks.

Fixes #3634
Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3636)

Run perltidy on util/mkerr
Change some mkerr flags, write some doc comments
Make generated tables "const" when genearting lib-internal ones.
Add "state" file for mkerr
Renerate error tables and headers
Rationalize declaration of ERR_load_XXX_strings
Fix out-of-tree build
Add -static; sort flags/vars for options.
Also tweak code output
Moved engines/afalg to engines (from master)
Use -static flag
Standard engine #include's of errors
Don't linewrap err string tables unless necessary
Reviewed-by: NRichard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3392)

Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3243)

The trick is to use the .a extension explicitely in the build.info files.
Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3243)

... on the theme "I could have sworn I saved that fix!"
Reviewed-by: NMatt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3285)

As far as I know, there is no MMS / MMK with parallellism today.
However, it might be added in the future (perhaps in MMK at least), so
we may as well prepare for it now.
Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3282)

jom is an nmake clone that does parallell building, via the same -j
argument as GNU make.  To make it work, we need to apply the same
dependeency build up as done in 27c40a93

Fixes #3272
Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3277)

When building object files for libraries, information whether the
library would be installed or not wasn't passed down to the object
file building rules.

Also, make it so settings like |no_inst_lib_cflags| can be the empty
string.
Reviewed-by: NTim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3247)

Fixes GH#3116.
Reviewed-by: NRichard Levitte <levitte@openssl.org>

Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2962)

It turns out that /DSF didn't do any good for our purposes.  Instead,
remove the CALL_DEBUG flag from any image we link.  This ensures that
we can have debugging information in the image files, but don't
automatically end up in a debugging session upon image activation.

Unfortunately, this means the CALL_DEBUG must be turned on when there
is a need to run with the debugger activated, and to turn it off when
done.  This has been documented in NOTES.VMS.
Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2957)

That makes it possible to run images without automagically ending up
in a debug session, while still being able to debug when required.

All .DSF files must reside in the same directory to be useful.
Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2947)

Reviewed-by: NRichard Levitte <levitte@openssl.org>

Add 2017 copyright year
Add missing typedef to NAME
Remove ec(7) and bn(7) doc links
Remove .pod link errors, bogus links, make a few typo corrections
Fix some typo's in links and some missing items.
Don't link to C runtime functions (See OPENSSL_malloc for example/precedent)
Document ASN1_tag2str(), add a few typedef's that were missing from NAME
Update doc-nits target; addresses
    https://github.com/openssl/openssl/pull/1900#issuecomment-259943891,
Merge check-doc-links into find-doc-nits; if run regularly, would have found
    https://github.com/openssl/openssl/pull/2825Reviewed-by: NRichard Levitte <levitte@openssl.org>
    (Merged from https://github.com/openssl/openssl/pull/2862)

Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2863)

Reviewed-by: NAndy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2851)

For each platform, we may need to perform some basic checks to see
that available tools perform as we expect them.

For the moment, the added checkers test that Perl gives the expected
path format.  This should help MingW users to see if they run an
appropriate Perl implementation, for example.
Reviewed-by: NAndy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2851)

GH#2816
Reviewed-by: NRich Salz <rsalz@openssl.org>
Reviewed-by: NRichard Levitte <levitte@openssl.org>

Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2810)

One of the reasons for why masm/ml64 is not [fully] supported is that
it's problematic to support multiple versions. But latest one usually
works and/or it's lesser problem to make it work. So idea here is to
have a "whistle" when it breaks, so that problems can be evaluated as
they emerge. It's kind of "best effort" thing, as opposite to "full
support".
Reviewed-by: NRichard Levitte <levitte@openssl.org>

Reviewed-by: NRichard Levitte <levitte@openssl.org>
Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2344)

Reviewed-by: NRichard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2202)

Reviewed-by: NRichard Levitte <levitte@openssl.org>
Reviewed-by: NRich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/2202)

b6d5ba1a forgot to update some linux arches.
Reviewed-by: NRichard Levitte <levitte@openssl.org>

GH: #1977

'linux-x86' is similar to 'linux-x86_64' but uses -m32 rather than -m64.
Reviewed-by: NAndy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1924)

Reviewed-by: NRich Salz <rsalz@openssl.org>

Reviewed-by: NRich Salz <rsalz@openssl.org>