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>

A few were missed in the previous commit.

Closes RT#4412

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

Use SSL_get_SSL_CTX instead of passing SSL_CTX to s_client.c:print_stuff

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

Pass entire CTLOG_STORE to SCT_print, rather than just the SCT's CTLOG

SCT_print now looks up the correct CT log for you.

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

Remove 'log' field from SCT and related accessors

In order to still have access to an SCT's CTLOG when calling SCT_print,
SSL_CTX_get0_ctlog_store has been added.

Improved documentation for some CT functions in openssl/ssl.h.

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

Closes RT#4406

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

Reviewed-by: Stephen Henson <steve@openssl.org>

Reviewed-by: Stephen Henson <steve@openssl.org>

The AFALG engine created a global EVP_CIPHER instance but was not freeing
it up when the engine was destroyed.

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

The help text for -d in 'config' was aged, and the option processing
in 'config.com' was just different.  This harmonizes 'config.com' with
the instructions in INSTALL and both current reality.

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

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

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

UEFI needs this too. Don't keep it only in the Windows/DOS ifdef block.

This is a fixed version of what was originally commit 963bb621 and
subsequently reverted in commit 37b1f8bd

. Somewhere along the way, the
Windows/DOS ifdef actually got removed, leaving it just broken. It should
have been turned into an #elif, not removed.

This one correctly changes the logic from

    # if WINDOWS|DOS
    #  if OPENSSL_NO_SOCK
        ... no-sock ...
    #  elif !DJGPP
        ... native windows ...

to

    # if OPENSSL_NO_SOCK
       ... no-sock ...
    # elif WINDOWS|DOS
    #  if !DJGPP
        ... native windows ...

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

A line from cryptlib.h was missed during the old Thread API removal. This
breaks no-deprecated builds.

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

The function SRP_VBASE_get_by_user() is declared as deprecated but the
implementation was not.

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