A couple of '$(PERLASM_SCHEM' had sneaked in.

Reviewed-by: Tim Hudson <tjh@openssl.org>

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

The reason to do so is that some of the generators detect PIC flags
like -fPIC and -KPIC, and those are normally delivered in LD_CFLAGS.

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

When passing down values to Makefile.shared, do so with single quotes
as much as possible to avoid having the shell create a mess of quotes.

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

The variable SHARED_CFLAGS and SHARD_LDFLAGS were used in the Unix
template because they normally contain options used when building
"shared".  The Windows template, on the other hand, uses LIB_CFLAGS,
to express the intended use of those flags rather than their content.
The Windows template still used SHARED_LDFLAGS, which seems
inconsistent.

To harmonize the two, any SHARED_CFLAGS gets renamed to LIB_CFLAGS and
SHARED_LDFLAGS to LIB_LDFLAGS.  That makes the intent consistent along
with BIN_{C,LD}FLAGS and DSO_{C,LD}FLAGS.

Finally, make sure to pass down $(LIB_CFLAGS) or $(DSO_CFLAGS) along
with $(CFLAGS) when using Makefile.shared.

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

ENGINE_cleanup calls CRYPTO_free_ex_data and therefore,
CRYPTO_cleanup_all_ex_data - which cleans up the method pointers - must
run after ENGINE_cleanup.

Additionally, don't needlessly initialize the EX_CALLBACKS stack during
e.g. CRYPTO_free_ex_data. The only time this is actually needed is when
reserving the first ex data index. Specifically, since sk_num returns -1
on NULL input, the rest of the code already handles a NULL method stack
correctly.

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

With help from Viktor.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>

ccache + clang produces a false strcmp warning, see
https://llvm.org/bugs/show_bug.cgi?id=20144



Since this only happens with ccache and --strict-warnings, and
only with certain versions of glibc / clang, disabling
ccache is a reasonable short-term workaround.

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

Reviewed-by: Matt Caswell <matt@openssl.org>

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

While insignificant on Unix like systems, this is significant on
systems like VMS.

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>

It turns out that different sed implementations treat -i differently
to cause issues.  make it simpler by avoiding it entirely and give
perl the trust to be consistent enough.

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

That doesn't change even to make a dummy to hide its unavailability.

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

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

On Windows we call WSAGetLastError() to find out the last error that
happened on a socket operation. We use this to find out whether we can
retry the operation or not. You are supposed to call this immediately
however in a couple of places we logged an error first. This can end up
making other Windows system calls to get the thread local error state.
Sometimes that can clobber the error code, so if you call WSAGetLastError()
later on you get a spurious response and the socket operation looks like
a fatal error.

Really we shouldn't be logging an error anyway if its a retryable issue.
Otherwise we could end up with stale errors on the error queue.

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

verified that build succeeds without the extra define
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>

Fix no-blake2 for Windows classic build
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>

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

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

The issue is demonstrated as follows:

  On Linux:

    $ echo ': foo.h /usr/include/stddef.h bar.h' | sed -e 's/ \/\(\\.\|[^ ]\)*//g'
    : foo.h bar.h

  On MacOS X:

    $ echo ': foo.h /usr/include/stddef.h bar.h' | sed -e 's/ \/\(\\.\|[^ ]\)*//g'
    : foo.husr/include/stddef.h bar.h

Perl is more consistent:

  On Linux:

    $ echo ': foo.h /usr/include/stddef.h bar.h' | perl -pe 's/ \/(\\.|[^ ])*//g;'
    : foo.h bar.h

  On MacOS X:

    $ echo ': foo.h /usr/include/stddef.h bar.h' | perl -pe 's/ \/(\\.|[^ ])*//g;'
    : foo.h bar.h

Reviewed-by: Andy Polyakov <appro@openssl.org>

Reviewed-by: Emilia Käsper <emilia@openssl.org>

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>

They are not numbers in the machine byte order.

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>

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

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

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

If pre-processor failed, an empty .s file could be left behind,
which could get successfully compiled if one simply re-ran make
and cause linking failures. Not anymore. Remove even intermediate .S
in case of pre-processor failure.

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

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

Reviewed-by: Emilia Käsper <emilia@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>

Reviewed-by: Emilia Käsper <emilia@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>

Running test_ssl with HARNESS_VERBOSE results in lots of spurious warnings
about an inability to load the CT config file. This fixes it.

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

Issue identified on Solaris by Erik Forsberg <erik@efca.com>

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

During auto de-init we were calling ENGINE_cleanup(), and then later
CONF_modules_free(). However the latter function can end up calling
engine code, which can lead to a use of the global_engine_lock after it
has already been freed. Therefore we should swap the calling order of
these two functions.

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

no-ripemd was unified a while ago but the change was not done in mk1mf.pl. This commit changes the no-ripemd string to no-rmd160 which fixes the no-rmd160 classic build.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>

Experience has shown that dynamic engines with their own copy of
libcrypto is problematic, so we disable that possibility.

Reviewed-by: Matt Caswell <matt@openssl.org>

One of them didn't clean away .d.tmp files properly.
The other would overwrite the .d files unconditionally, thereby
causing a possibly unnecessary dependency rebuild, which touches the
date of Makefile, which causes a possibly unnecessary rebuild of
buildinf.h and everything that depends on that.

Reviewed-by: Andy Polyakov <appro@openssl.org>

A few things were out of date and a few things were missing. Also some
formatting updates.

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

Some of these scripts would recognise an output parameter if it looks
like a file path.  That works both in both the classic and new build
schemes.  Some fo these scripts would only recognise it if it's a
basename (i.e. no directory component).  Those need to be corrected,
as the output parameter in the new build scheme is more likely to
contain a directory component than not.

Reviewed-by: Andy Polyakov <appro@openssl.org>