Fix a ** 0 mod 1 = 0 for real this time.

Commit 2b0180c37fa6ffc48ee40caa831ca398b828e680 attempted to do this but
only hit one of many BN_mod_exp codepaths. Fix remaining variants and add
a test for each method.

Thanks to Hanno Boeck for reporting this issue.

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Dr. Stephen Henson <steve@openssl.org>
(cherry picked from commit d911097d7c93e4cfeab624b34d73fe51da158b69)

Fix tarball production to keep test/bctest and util/pod2mantest

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

x86_64 assembly pack: tune clang version detection even further.

RT#4171

Reviewed-by: Kurt Roeckx <kurt@openssl.org>
(cherry picked from commit b9749432346f69b29d82070041e71b237d718ce7)

ec/ecp_nistz256_table.c: fix potential misalignment problem with Sun C.

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

Ensure |rwstate| is set correctly on BIO_flush

A BIO_flush call in the DTLS code was not correctly setting the |rwstate|
variable to SSL_WRITING. This means that SSL_get_error() will not return
SSL_ERROR_WANT_WRITE in the event of an IO retry.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 67f60be8c9ae5ff3129fcd6238baf124385a41d8)

Fix DTLS handshake fragment retries

If using DTLS and NBIO then if a second or subsequent handshake message
fragment hits a retry, then the retry attempt uses the wrong fragment
offset value. This commit restores the fragment offset from the last
attempt.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 2ad226e88bee97847496e542d63c67997d5beda6)

Add a return value check

If the call to OBJ_find_sigid_by_algs fails to find the relevant NID then
we should set the NID to NID_undef.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 330dcb09b2df7e1e6d1d3d14a5df7269aebd9a68)

Don't use applink for static builds.

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

Refer to website for acknowledgements.

Reviewed-by: Steve Marquess <marquess@openssl.com>
(cherry picked from commit ab29c82a55f3583a490733dd521ea6c486e8e2fb)

Not all 'find's know -xtype, use -type instead

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 3cd7aef34d0d414d27ab00abadb99265a2cffde9)

Adapt the OS X build to use the OS X tar

As part of this, move release creation to a script to be called from
.travis.yml.  That makes it much easier to test outside of travis.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 382af61f6213e975b4c2a50fd8b9fedd23d86ab5)

Make it possible to affect the way dists are made

Introducing DISTTARVARS to propagate changed variables down to the
tar-making target.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 4d3c30a1799bf7b4dc7223b84417c4de992a6b9c)

Cleanup the EVP_MD_CTX before exit rather than after

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

Change tar owner and group to just 0

It seems like some tar versions don't like the name:id form for
--owner and --group.  The closest known anonymous user being 0 (root),
that seems to be the most appropriate user/group to assign ownership
to.  It matters very little when unpacking either way.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit b91dd150d2b9b5ddca37722e7f52ea59ba7f80da)

Do not add symlinks in the source release

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 451a5bdf0386d7acf091c3e3b39107e5ed8be25d)

In travis, build from a "source release" rather than from the build tree

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 475fc3d8729190fd12b4ff23d6ec488439fb78f9)

Small changes to creating dists

Make TARFILE include ../ instead of having that hard coded all over the place.
When transforming file names in TAR_COMMAND, use $(NAME) instead of openssl-$(VERSION)

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 4a544810f08539f1549eea9be36bd878c67c8e26)

Prepare for 1.0.2f-dev

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

Prepare for 1.0.2e release

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

make update

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

Update CHANGES and NEWS

Update the CHANGES and NEWS files for the new release.

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

Add test for CVE-2015-3194

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

Add PSS parameter check.

Avoid seg fault by checking mgf1 parameter is not NULL. This can be
triggered during certificate verification so could be a DoS attack
against a client or a server enabling client authentication.

Thanks to Loïc Jonas Etienne (Qnective AG) for discovering this bug.

CVE-2015-3194

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

bn/asm/x86_64-mont5.pl: fix carry propagating bug (CVE-2015-3193).

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit e7c078db57908cbf16074c68034977565ffaf107)

Fix leak with ASN.1 combine.

When parsing a combined structure pass a flag to the decode routine
so on error a pointer to the parent structure is not zeroed as
this will leak any additional components in the parent.

This can leak memory in any application parsing PKCS#7 or CMS structures.

CVE-2015-3195.

Thanks to Adam Langley (Google/BoringSSL) for discovering this bug using
libFuzzer.

PR#4131

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

Add cleanup of *.s

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

perlasm/ppc-xlate.pl: comply with ABIs that specify vrsave as reserved.

RT#4162

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit b5516cfbd65de9331d827012fc5bdace1953613e)

_BSD_SOURCE is deprecated, use _DEFAULT_SOURCE instead

The feature_test_macros(7) manual tells us that _BSD_SOURCE is
deprecated since glibc 2.20 and that the compiler will warn about it
being used, unless _DEFAULT_SOURCE is defined as well.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit f9fd35248c9a3b1125d9ab82ffb19d62e86533ac)

crypto/sparcv9cap.c: add SIGILL-free feature detection for Solaris.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 2238e0e45d03ec9ba15e9547f7739cc6987a8f57)

Remove cookie validation return value trick

In the DTLS ClientHello processing the return value is stored in |ret| which
by default is -1. |ret| is only updated to a positive value once we are past
all points where we could hit an error. We wish to return 1 on success or 2
on success *and* we have validated the DTLS cookie. Previously on successful
validation of the cookie we were setting |ret| to -2, and then once we were
past all error points we set |ret = -ret|. This is non-obvious behaviour and
could be error prone. This commit tries to make this a bit more intuitive.

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

x86_64 assembly pack: tune clang version detection.

RT#4142

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 76eba0d94bb418325be6409b272eac5e2bd4a0a9)

Resolved conflicts:
crypto/bn/asm/x86_64-mont.pl
crypto/bn/asm/x86_64-mont5.pl

Reviewed-by: Kurt Roeckx <kurt@openssl.org>

Use both sun and __sun

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

mark openssl configuration as loaded at end of OPENSSL_config

Signed-off-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@akamai.com>
GH: #466
(cherry picked from commit 434b58457c8cca4d09f77a1774d6d2c446604bae)

ssl3_free(): Return if it wasn't created

If somewhere in SSL_new() there is a memory allocation failure, ssl3_free() can
get called with s->s3 still being NULL.

Patch also provided by Willy Tarreau <wtarreau@haproxy.com>

Signed-off-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Viktor Dukhovni <openssl-users@dukhovni.org>
(cherry picked from commit 3e7bd2ce0b16f8611298175d6dc7cb35ee06ea6d)

Set reference count earlier

Backport of 0e04674e964b905e67e3d215bcf888932c92765f

Reviewed-by: Steve Henson <steve@openssl.org>
RT #4047, #4110, MR #1356

Fix uninitialised p error.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 63eb10a07ee29a312e50a227f5b3a290b1ad22b4)

Limit depth of ASN1 parse printing.

Thanks to Guido Vranken <guidovranken@gmail.com> for reporting this issue.

Reviewed-by: Tim Hudson <tjh@openssl.org>
(cherry picked from commit 158e5207a794603f5d64ffa95e0247c7808ab445)

Conflicts:
crypto/asn1/asn1_par.c

Makefile.org: add LC_ALL=C to unify error [and other] messages.

RT#4138

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 193ed2d74b1a643b5f32bd46bf50a858fcdd2da4)

Use defined(__sun) instead of defined(sun)

Strict ISO confirming C compilers only define __sun

Reviewed-by: Viktor Dukhovni <openssl-users@dukhovni.org>
RT #4144, MR #1353

(cherry picked from commit 3d32218812e87221344f2985512e42e4aaa88745)

Add initial AppVeyor configuration

Original patch by Frank Morgner.

Signed-off-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@akamai.com>
GH: #456
(cherry picked from commit 68db80e2d1accdd4c4a6b4763559c6cfe9663820)

Good hygiene with size_t output argument.

Though the callers check the function return value and ignore the
size_t output argument on failure, it is still often not ideal to
store -1 in a size_t on error.  That might signal an unduly large
buffer.  Instead set the size_t to 0, to indicate no space.

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

Ensure all EVP calls have their returns checked where appropriate

There are lots of calls to EVP functions from within libssl There were
various places where we should probably check the return value but don't.
This adds these checks.

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

Fix README version typo

Reviewed-by: Steve Marquess <marquess@openssl.com>

Fix release in README

Reviewed-by: Steve Marquess <marquess@openssl.com>

aes/asm/vpaes-ppc.pl: eliminate overhung stores in misaligned cases.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit ce24d2ed232a095349886aecc8cb959d5dda8342)

bn/asm/ppc64-mont.pl: adapt for little-endian.

The problem remained unnoticed so far, because it's never called by default.
You have to craft OPENSSL_ppccap environment variable to trigger the problem.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit e4693b4e2a0c3f6241d4d3e61460c34c7e0013f6)

bn/asm/s390x.S: improve performance on z196 and z13 by up to 26%. [even z10 is couple percent faster]. Triggered by RT#4128, but solves the problem by real modulo-scheduling.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit 9d0e4dc6351df7d0c08400c4b4cf17c017022e50)

aesni-sha256-x86_64.pl: fix crash on AMD Jaguar.

It was also found that stich performs suboptimally on AMD Jaguar, hence
execution is limited to XOP-capable and Intel processors.

Reviewed-by: Kurt Roeckx <kurt@openssl.org>
(cherry picked from commit a5fd24d19bbb586b1c6d235c2021e9bead22c9f5)

RT32671: wrong multiple errs TS_check_status_info

Manual cherry-pick of 0704343f138a38d3882d5af2a4ebe8821e9a8f3a.

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

modes/asm/ghash-armv4.pl: make it compile by clang. RT#3989

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

Stop DTLS servers asking for unsafe legacy renegotiation

If a DTLS client that does not support secure renegotiation connects to an
OpenSSL DTLS server then, by default, renegotiation is disabled. If a
server application attempts to initiate a renegotiation then OpenSSL is
supposed to prevent this. However due to a discrepancy between the TLS and
DTLS code, the server sends a HelloRequest anyway in DTLS.

This is not a security concern because the handshake will still fail later
in the process when the client responds with a ClientHello.

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

Only call ssl3_init_finished_mac once for DTLS

In DTLS if an IO retry occurs during writing of a fragmented ClientHello
then we can end up reseting the finish mac variables on the retry, which
causes a handshake failure. We should only reset on the first attempt not
on retries.

Thanks to BoringSSL for reporting this issue.

RT#4119

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

Remove redundant check from tls1_get_curvelist

The function tls1_get_curvelist() has an explicit check to see if s->cert
is NULL or not. However the check appears *after* calling the tls1_suiteb
macro which derefs s->cert. In reality s->cert can never be NULL because
it is created in SSL_new(). If the malloc fails then the SSL_new call fails
and no SSL object is created.

Reviewed-by: Tim Hudson <tjh@openssl.org>
(cherry picked from commit 6329b6092b28b656be8a1e4a8363d2e3bcc32053)

Conflicts:
ssl/t1_lib.c

Fix missing malloc return value checks

During work on a larger change in master a number of locations were
identified where return value checks were missing. This backports the
relevant fixes.

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

Make the match for previous cflags a bit more strict

./Configure [target] --strict-warnings -Wno-pedantic-ms-format
would not add '-pedantic' because it matches '-Wno-pedantic-ms-format',
which was added first.

Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 6703c4ea87b30554283deaa5df1f8d68725d3ee4)

Reject TLS 1.2 ciphersuites if not allowed.

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

Ensure the dtls1_get_*_methods work with DTLS_ANY_VERSION

The various dtls1_get*_methods did not handle the DTLS_ANY_VERSION case,
so this needed to be added.

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

Minor EVP_SignInit_ex doc fix

EVP_SignInit_ex was missing from the NAME section of its man page so
typing "man EVP_SignInit_ex" failed to load the page.

Reviewed-by: Stephen Henson <steve@openssl.org>
(cherry picked from commit 3d866ea67e8b19777e88ac2a78ee4188e0983168)

Clarify the preferred way of creating patch files

Clarify that git format-patch output is preferred for creating patch files.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit f89ee71bc81017e04ac50f570d8aed87f495bcf2)

Various README and CONTRIBUTING updates

Close GH Issue 69
Close GH PR 457/RT4113
Some other updates

By Rich Salz, Alessandro Ghedini, Steve Marquess, Collin Anderson

(manual cherry-pick of a2aaf8be7e3c22a61ef89f273aa85f482b955336 and
b06935f439af7150d3ae566922353c3f210e63ed)

Reviewed-by: Kurt Roeckx <kurt@openssl.org>

BN_GF2m_mod_inv(): check bn_wexpand return value

Signed-off-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@akamai.com>
MR #1276, RT #4107

(cherry picked from commit 94b3664a528258df5ebcaae213d19bf6568cc47d)

Remove useless code

RT#4081

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 070c23325af4526c9a8532a60d63522c58d5554b)

Check memory allocation

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 8cf9d71a3a43d9b98a8a278d47dc08088a954a7b)

Fix references to various RFCs

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 3240e7cf5f651d9d94814b4d494fbe294e463b72)

Fix typos

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit d900a015b5a89ea2018e6122dd2738925e4d68ab)

Set salt length after the malloc has succeeded

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit c2319cf9fce87a2e82efb6e58ced11a85190dc3d)

Fix memory leaks and other mistakes on errors

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 3f6c7691870d1cd2ad0e0c83638cef3f35a0b548)

Replace malloc+strlcpy with strdup

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 8acaabec429b39f9436f6a88006384d72d292539)

Do not treat 0 return value from BIO_get_fd() as error

0 is a valid file descriptor.

RT#4068

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 4428c7dba8f6f407d915c1226f4e0f673e8be241)

Clarify return values for EVP_DigestVerifyFinal.

Previous language was unclear.  New language isn't pretty but I believe
it is more accurate.

Signed-off-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Ben Laurie <ben@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit 8cbb048c3ea416f2bd8a3706d027f3aa26ef08d9)

Fix self signed handling.

Don't mark a certificate as self signed if keyUsage is present and
certificate signing not asserted.

PR#3979

Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit f51e5ed6b4b91d12228da873db72aa28109d1797)

make depend: prefer clang over makedepend

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

RFC5753 compliance.

RFC5753 requires that we omit parameters for AES key wrap and set them
to NULL for 3DES wrap. OpenSSL decrypt uses the received algorithm
parameters so can transparently handle either form.

Reviewed-by: Andy Polyakov <appro@openssl.org>
(cherry picked from commit 4ec36aff2a3c221b640dafa56ac838312e6724f4)

RT3346: Fix test_bn regexp for Windows using MSYS.

(cherry picked from commit 028bac0670c167f154438742eb4d0fbed73df209)

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

Typo.

PR#4079

Reviewed-by: Tim Hudson <tjh@openssl.org>
(cherry picked from commit c69ce9351336f5b4a8b33890756b3fd185528210)

Don't treat a bare OCTETSTRING as DigestInfo in int_rsa_verify

The function int_rsa_verify is an internal function used for verifying an
RSA signature. It takes an argument |dtype| which indicates the digest type
that was used. Dependant on that digest type the processing of the
signature data will vary. In particular if |dtype == NID_mdc2| and the
signature data is a bare OCTETSTRING then it is treated differently to the
default case where the signature data is treated as a DigestInfo (X509_SIG).

Due to a missing "else" keyword the logic actually correctly processes the
OCTETSTRING format signature first, and then attempts to continue and
process it as DigestInfo. This will invariably fail because we already know
that it is a bare OCTETSTRING.

This failure doesn't actualy make a real difference because it ends up at
the |err| label regardless and still returns a "success" result. This patch
just cleans things up to make it look a bit more sane.

RT#4076

Reviewed-by: Richard Levitte <levitte@openssl.org>
(cherry picked from commit dffe51091f412dcbc18f6641132f0b4f0def6bce)

When ENGINE_add finds that id or name is missing, actually return

Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 5850cc75ea0c1581a9034390f1ca77cadc596238)

Fix travis 1.0.2 builds

Disable -Wshadow error when building with GCC
Add support for linu-x86_64-clang debug; this is needed for Travis CI.
Disable linux-clang and mingw debug builds on Travis CI; not supported.
Fix Travis email notifications config

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

Move BN_CTX_start() call so the error case can always call BN_CTX_end().

Signed-off-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
MR #1231

(cherry picked from commit 99c203337574d967c86ffbfa13f40ace51048485)

Properly check return type of DH_compute_key()

It returns -1 on error, not 0.

Signed-off-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
MR #1231

(cherry picked from commit d6e92c0bd6c36fc68291e79ef5753fd7f0420695)

Set flags to 0 before calling BN_with_flags()

BN_with_flags() will read the dest->flags to keep the BN_FLG_MALLOCED but
overwrites everything else.

Signed-off-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
MR #1231

(cherry picked from commit f92768e6f5259069bd21dbed2b98b3423c1dfca4)

Don't try and parse boolean type.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit e58c4d3cdde7a0a01df2884bfeec31a2b07be22d)

Change functions to pass in a limit rather than calculate it

Some extension handling functions were passing in a pointer to the start
of the data, plus the length in order to calculate the end, rather than
just passing in the end to start with. This change makes things a little
more readable.

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

Validate ClientHello extension field length

RT#4069

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

Fix more d2i cases to properly update the input pointer

Thanks to David Benjamin <davidben@google.com> for pointing them out.

Reviewed-by: Steve Henson <steve@openssl.org>
MR #1198

(cherry picked from commit 605236f6a8fe0743af2f63d93239a74c69dae137)

Link in applink with fips_premain_dso

PR#4042

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

GH367: use random data if seed too short.

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

RT2772: accept empty SessionTicket

RFC 5077 section 3.3 says: If the server determines that it does not
want to include a ticket after it has included the SessionTicket
extension in the ServerHello, then it sends a zero-length ticket in the
NewSessionTicket handshake message.

Previously the client would fail upon attempting to allocate a
zero-length buffer. Now, we have the client ignore the empty ticket and
keep the existing session.

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

SRP memory leak fix

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

Fix typo in previous merge.

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

Change --debug to -d for compat with old releases.

Reviewed-by: Tim Hudson <tjh@openssl.org>
(cherry picked from commit 1d4ddb4e1a088f1333c4bb155c52c7f94e572bca)

Document BUF_strnlen

Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 21b0fa91186ff1c1c3d956c0593ef4ac02521695)

BUF_strdup and friends: update docs

Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 58e3457a82e8940ff36b36949f9c7a60e7614b2c)

BUF_strndup: tidy

Fix comment, add another overflow check, tidy style

Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit de8883e11befde31d9b6cfbbd1fc017c365e0bbf)

Make BUF_strndup() read-safe on arbitrary inputs

BUF_strndup was calling strlen through BUF_strlcpy, and ended up reading
past the input if the input was not a C string.

Make it explicitly part of BUF_strndup's contract to never read more
than |siz| input bytes. This augments the standard strndup contract to
be safer.

The commit also adds a check for siz overflow and some brief documentation
for BUF_strndup().

Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 110f7b37de9feecfb64950601cc7cec77cf6130b)

GH398: Add mingw cross-compile, etc.

For all release branches.  It adds travis build support. If you don't
have a config file it uses the default (because we enabled travis for the
project), which uses ruby/rake/rakefiles, and you get confusing "build
still failing" messages.

Reviewed-by: Andy Polyakov <appro@openssl.org>
(cherry picked from commit db9defdfe306e1adf0af7188b187d535eb0268da)

RT3479: Add UTF8 support to BIO_read_filename()

If we use BIO_new_file(), on Windows it'll jump through hoops to work
around their unusual charset/Unicode handling. it'll convert a UTF-8
filename to UCS-16LE and attempt to use _wfopen().

If you use BIO_read_filename(), it doesn't do this. Shouldn't it be
consistent?

It would certainly be nice if SSL_use_certificate_chain_file() worked.

Also made BIO_C_SET_FILENAME work (rsalz)

Signed-off-by: Rich Salz <rsalz@akamai.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
(cherry picked from commit ff03599a2f518dbdf13bca0bb0208e431b892fe9)

RT3823: Improve the robustness of event logging

There are a couple of minor fixes here:

1) Handle the case when RegisterEventSource() fails (which it may for
various reasons) and do the work of logging the event only if it succeeds.

2) Handle the case when ReportEvent() fails and do our best in debug builds
to at least attempt somehow indicate that something has gone wrong. The
typical situation would be someone running tools like DbMon, DBWin32,
DebugView or just having the debugger attached. The intent is to make sure
that at least some data will be captured so that we can save hours and days
of debugging time.

3) Minor fix to change the MessageBox() flag to MB_ICONERROR. Though the
value of MB_ICONERROR is the same value as MB_ICONSTOP, the intent is
better conveyed by using MB_ICONERROR.

Testing performed:

1) Clean compilation for debug-VC-WIN32 and VC-WIN32.

2) Good test results (nmake -f ms\ntdll.mak test) for debug-VC-WIN32 and
VC-WIN32.

3) Stepped through relevant changes using WinDBG and exercised the impacted
code paths.

Signed-off-by: Rich Salz <rsalz@akamai.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(cherry picked from commit 4cd94416a452c3a3e0df24c297f7d2f0e6d5bb5f)

Fix SRP memory leaks

There were some memory leaks in the creation of an SRP verifier (both on
successful completion and also on some error paths).

Reviewed-by: Emilia Käsper <emilia@openssl.org>
(cherry picked from commit bf95cde28712cfcad90cb3975cdcb8e5c0f20fde)

Handle SSL_ERROR_WANT_X509_LOOKUP

Reviewed-by: Rich Salz <rsalz@openssl.org>
(cherry picked from commit f1c412c9e63f7c9cac2c723bff09cce563dda1b0)