openssl.git
18 months agoMake the s_server command listen on IPv6 only when requested
Bernd Edlinger [Tue, 23 Jan 2018 16:43:45 +0000 (17:43 +0100)]
Make the s_server command listen on IPv6 only when requested

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5152)

18 months agoFix setting of IPV6_V6ONLY on Windows
Bernd Edlinger [Thu, 25 Jan 2018 14:16:18 +0000 (15:16 +0100)]
Fix setting of IPV6_V6ONLY on Windows

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5139)

18 months agoFix signature of min/max proto getter
Christian Heimes [Sun, 21 Jan 2018 09:37:59 +0000 (10:37 +0100)]
Fix signature of min/max proto getter

The getters for min and max proto version wrongly passed NULL instead of
0 as third argument to SSL_ctrl() and SSL_CTX_ctrl(). The third argument
is not used, but the error results in a compiler warning:

warning: passing argument 3 of ‘SSL_CTX_ctrl’ makes integer from pointer without a cast [-Wint-conversion]
int v = SSL_CTX_get_max_proto_version(self->ctx);

See https://github.com/openssl/openssl/pull/4364

Signed-off-by: Christian Heimes <christian@python.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/5128)

18 months agoFix SSL_CTX_get_{min,max}_proto_version integer conversion warning
Steffan Karger [Wed, 17 Jan 2018 16:06:41 +0000 (17:06 +0100)]
Fix SSL_CTX_get_{min,max}_proto_version integer conversion warning

When using the SSL_CTX_get_min_min_version macro while compiling with
-Wall, my compiler rightfully complains about this construction:

warning: passing argument 3 of ‘SSL_CTX_ctrl’ makes integer from
pointer without a cast [-Wint-conversion]

These macro's should use 0, instead of NULL, for the third argument,
like most other SSL_CTX_ctrl 'get' wrappers do.

CLA: trivial
Signed-off-by: Steffan Karger <steffan.karger@fox-it.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/5099)

18 months agoAdded "B" tag to all variables.
David Cooper [Wed, 24 Jan 2018 17:27:19 +0000 (12:27 -0500)]
Added "B" tag to all variables.

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4190)

18 months agoAdd documentation for the OCSP_basic_sign() and OCSP_basic_sign_ctx() functions.
David Cooper [Wed, 24 Jan 2018 16:47:23 +0000 (11:47 -0500)]
Add documentation for the OCSP_basic_sign() and OCSP_basic_sign_ctx() functions.

Correct error return value in OCSP_basic_sign().

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4190)

18 months agoMake editorial changes suggested by Matt Caswell and fixed Travis failures.
David Cooper [Tue, 23 Jan 2018 19:22:17 +0000 (14:22 -0500)]
Make editorial changes suggested by Matt Caswell and fixed Travis failures.

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4190)

18 months agoMake editorial changes suggested by Rich Salz and add the -rsigopt option to the...
David Cooper [Tue, 12 Dec 2017 21:01:22 +0000 (16:01 -0500)]
Make editorial changes suggested by Rich Salz and add the -rsigopt option to the man page for the ocsp command.

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4190)

18 months agoAdd -rsigopt option to ocsp command
David Cooper [Fri, 18 Aug 2017 13:27:19 +0000 (09:27 -0400)]
Add -rsigopt option to ocsp command

Add a -rsigopt option to the ocsp command that allows signature parameters to be provided for the signing of OCSP responses. The parameters that may be provided to -rsigopt are the same as may be provided to -sigopt in the ca, req, and x509 commands.

This PR also defines a OCSP_basic_sign_ctx() function, which functions in the same way as OCSP_basic_sign(), except that it accepts a EVP_MD_CTX rather than a key and digest. The OCSP_basic_sign_ctx() function is used to implement the -rsigopt option in the ocsp command.

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4190)

18 months agoFix error-path memory leak in asn_mime.c
Todd Short [Mon, 22 Jan 2018 19:30:24 +0000 (14:30 -0500)]
Fix error-path memory leak in asn_mime.c

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5142)

18 months agoUpdates following review of SSL_stateless() code
Matt Caswell [Tue, 23 Jan 2018 12:23:23 +0000 (12:23 +0000)]
Updates following review of SSL_stateless() code

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoAdd a timestamp to the cookie
Matt Caswell [Wed, 17 Jan 2018 14:29:22 +0000 (14:29 +0000)]
Add a timestamp to the cookie

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoAdd the ability for s_server to operate statelessly
Matt Caswell [Fri, 29 Dec 2017 17:37:04 +0000 (17:37 +0000)]
Add the ability for s_server to operate statelessly

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoDon't send unexpected_message if we receive CCS while stateless
Matt Caswell [Fri, 29 Dec 2017 17:36:28 +0000 (17:36 +0000)]
Don't send unexpected_message if we receive CCS while stateless

Probably this is the CCS between the first and second ClientHellos. It
should be ignored.

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoAdd documentation for SSL_stateless()
Matt Caswell [Thu, 28 Sep 2017 14:05:58 +0000 (15:05 +0100)]
Add documentation for SSL_stateless()

Fixes #4283

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoFix the cookie/key_share extensions for use with SSL_stateless()
Matt Caswell [Thu, 28 Sep 2017 12:25:23 +0000 (13:25 +0100)]
Fix the cookie/key_share extensions for use with SSL_stateless()

Fixes some bugs identified during testing.

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoFix the SSL_stateless() return code
Matt Caswell [Thu, 28 Sep 2017 12:24:58 +0000 (13:24 +0100)]
Fix the SSL_stateless() return code

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoFix interaction between SSL_stateless() and SSL_clear()
Matt Caswell [Thu, 28 Sep 2017 12:23:49 +0000 (13:23 +0100)]
Fix interaction between SSL_stateless() and SSL_clear()

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoAdd some tests for the SSL_stateless() capability
Matt Caswell [Wed, 27 Sep 2017 08:46:38 +0000 (09:46 +0100)]
Add some tests for the SSL_stateless() capability

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoFix logic around when to send an HRR based on cookies
Matt Caswell [Wed, 13 Sep 2017 13:50:49 +0000 (14:50 +0100)]
Fix logic around when to send an HRR based on cookies

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoAdd the SSL_stateless() function
Matt Caswell [Wed, 13 Sep 2017 12:48:48 +0000 (13:48 +0100)]
Add the SSL_stateless() function

This enables sending and receiving of the TLSv1.3 cookie on the server side
as appropriate.

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoEnable the cookie callbacks to work even in TLS in the apps
Matt Caswell [Tue, 12 Sep 2017 15:19:09 +0000 (16:19 +0100)]
Enable the cookie callbacks to work even in TLS in the apps

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoAdd support for sending TLSv1.3 cookies
Matt Caswell [Mon, 11 Sep 2017 14:43:56 +0000 (15:43 +0100)]
Add support for sending TLSv1.3 cookies

This just adds the various extension functions. More changes will be
required to actually use them.

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/4435)

18 months agoSmall cleanup of some build.info files
Richard Levitte [Tue, 23 Jan 2018 18:16:29 +0000 (19:16 +0100)]
Small cleanup of some build.info files

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5153)

18 months agoConfigure: ensure that a DEPEND generates the correct inclusion directory
Richard Levitte [Tue, 23 Jan 2018 18:13:48 +0000 (19:13 +0100)]
Configure: ensure that a DEPEND generates the correct inclusion directory

We incorrectly assumed that explicit dependencies meant that the
source directory would be added for inclusion.  However, if the
dependent file is generated, it's stored in the build directory, and
that should be used for inclusion rather than the source directory.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5153)

18 months agoConfigure: let INCLUDEs set on binaries "trickle down" to the objects
Richard Levitte [Tue, 23 Jan 2018 18:07:14 +0000 (19:07 +0100)]
Configure: let INCLUDEs set on binaries "trickle down" to the objects

This ensures that only one set of includes is associated with each
object file, reagardless of where it's used.

For example, if apps/build.info has this:

    SOURCE[openssl]=foo.c
    INCLUDE[openssl]=.. ../include

and test/build.info has this:

    SOURCE[footest]=../apps/foo.c
    INCLUDE[footest]=../include

The inclusion directories used for apps/foo.o would differ depending
on which program's dependencies get generated first in the build file.

With this change, all those INCLUDEs get combined into one set of
inclusion directories tied to the object file.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5153)

18 months agoSHA512/224 and SHA512/256
Pauli [Wed, 17 Jan 2018 03:20:22 +0000 (13:20 +1000)]
SHA512/224 and SHA512/256

Support added for these two digests, available only via the EVP interface.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5093)

18 months agoHave EVP_PKEY_asn1_find_str() work more like EVP_PKEY_asn1_find()
Richard Levitte [Mon, 22 Jan 2018 18:03:37 +0000 (19:03 +0100)]
Have EVP_PKEY_asn1_find_str() work more like EVP_PKEY_asn1_find()

EVP_PKEY_asn1_find_str() would search through standard asn1 methods
first, then those added by the application, which EVP_PKEY_asn1_find()
worked the other way around.  Also, EVP_PKEY_asn1_find_str() didn't
handle aliases.

This change brings EVP_PKEY_asn1_find_str() closer to EVP_PKEY_asn1_find().

Fixes #5086

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5137)

18 months agoRevert "EVP_PKEY_asn1_add0(): Check that this method isn't already registered"
Richard Levitte [Mon, 22 Jan 2018 17:24:55 +0000 (18:24 +0100)]
Revert "EVP_PKEY_asn1_add0(): Check that this method isn't already registered"

This reverts commit d85722d31ac9ff0dc54c06cdc8d125acf56ca27a.

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5137)

18 months agoFix no-ec --strict-warnings builds
Benjamin Kaduk [Tue, 23 Jan 2018 13:31:36 +0000 (07:31 -0600)]
Fix no-ec --strict-warnings builds

The 'loop' variable is only used if EC is available.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5149)

18 months agoImprove some BN documentation.
Rich Salz [Mon, 22 Jan 2018 19:41:09 +0000 (14:41 -0500)]
Improve some BN documentation.

Thanks to Nicolas Schodet for pointing this out.

Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5141)

18 months agoDon't add $(EX_LIBS) to libssl.pc's Libs.private
Richard Levitte [Mon, 22 Jan 2018 15:53:23 +0000 (16:53 +0100)]
Don't add $(EX_LIBS) to libssl.pc's Libs.private

Since libssl requires libcrypto and libcrypto.pc already has
Libs.private set exactly the same, there's no reason to repeat it in
libssl.pc.

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5134)

18 months agoAdd anything specifying a threads library to ex_libs
Richard Levitte [Mon, 22 Jan 2018 15:50:54 +0000 (16:50 +0100)]
Add anything specifying a threads library to ex_libs

Even -pthread gets treated that way.  The reason to do this is so it
ends up in 'Libs.private' in libcrypto.pc.

Fixes #3884

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5134)

18 months agoWhen building shared libraries, only ln -s when simple and full name differ
Richard Levitte [Mon, 22 Jan 2018 21:02:36 +0000 (22:02 +0100)]
When building shared libraries, only ln -s when simple and full name differ

Fixes #5143

Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5144)

18 months agoMove fprintf after assignment to avoid crash.
Rich Salz [Mon, 22 Jan 2018 19:33:22 +0000 (14:33 -0500)]
Move fprintf after assignment to avoid crash.

Thanks to David Vernet for reporting this.

Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5140)

18 months agotest/ossl_shim/packeted_bio.h: don't include e_os.h
Richard Levitte [Mon, 22 Jan 2018 16:51:51 +0000 (17:51 +0100)]
test/ossl_shim/packeted_bio.h: don't include e_os.h

That inclusion turned out to be completely unnecessary

[extended tests]

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5136)

18 months agoAdd accessors for AdmissionSyntax
Rich Salz [Thu, 5 Oct 2017 23:52:12 +0000 (00:52 +0100)]
Add accessors for AdmissionSyntax

Based on code from Matthias Ballreich, Steve Henson, and Wolf Tobias.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4724)

18 months agoReduce the use of e_os.h in test programs
Richard Levitte [Mon, 22 Jan 2018 10:00:59 +0000 (11:00 +0100)]
Reduce the use of e_os.h in test programs

This includes unnecessary use of the top as inclusion directory

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5132)

18 months agoDon't attempt to use X25519 for ECDSA in speed
Matt Caswell [Wed, 17 Jan 2018 10:33:18 +0000 (10:33 +0000)]
Don't attempt to use X25519 for ECDSA in speed

Fixes #5090

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/5097)

18 months agoDon't crash on a missing Subject in index.txt
Matt Caswell [Fri, 19 Jan 2018 14:48:45 +0000 (14:48 +0000)]
Don't crash on a missing Subject in index.txt

An index.txt entry which has an empty Subject name field will cause ca
to crash. Therefore check it when we load it to make sure its not empty.

Fixes #5109

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5114)

18 months agoDon't allow an empty Subject when creating a Certificate
Matt Caswell [Fri, 19 Jan 2018 14:34:56 +0000 (14:34 +0000)]
Don't allow an empty Subject when creating a Certificate

Misconfiguration (e.g. an empty policy section in the config file) can
lead to an empty Subject. Since certificates should have unique Subjects
this should not be allowed.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5114)

18 months agoAdd a configure option to opt-out secure memory
Bernd Edlinger [Fri, 19 Jan 2018 13:00:14 +0000 (14:00 +0100)]
Add a configure option to opt-out secure memory

./config -DOPENSSL_NO_SECURE_MEMORY

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5113)

19 months agoUpdate the license end year
Richard Levitte [Sat, 20 Jan 2018 09:02:23 +0000 (10:02 +0100)]
Update the license end year

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5121)

19 months agoEnable TLSProxy tests on Windows
Richard Levitte [Wed, 17 Jan 2018 09:39:28 +0000 (10:39 +0100)]
Enable TLSProxy tests on Windows

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5094)

19 months agoenc(1): document that AEAD is not and will not be supported
Benjamin Kaduk [Tue, 9 Jan 2018 21:26:37 +0000 (15:26 -0600)]
enc(1): document that AEAD is not and will not be supported

Note the reasons, including streaming output issues and key/iv/nonce
management issues.

Recommend the use of cms(1) instead.

Fixes #471.

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/5048)

19 months agoCopyright update of more files that have changed this year
Richard Levitte [Fri, 19 Jan 2018 09:49:22 +0000 (10:49 +0100)]
Copyright update of more files that have changed this year

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/5110)

19 months agoSet OPENSSL_ENGINES for Windows
Bernd Edlinger [Fri, 19 Jan 2018 07:24:29 +0000 (08:24 +0100)]
Set OPENSSL_ENGINES for Windows

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5108)

19 months agoCreate one permanent proxy socket per TLSProxy::Proxy instance
Richard Levitte [Wed, 17 Jan 2018 10:22:47 +0000 (11:22 +0100)]
Create one permanent proxy socket per TLSProxy::Proxy instance

On Windows, we sometimes see a behavior with SO_REUSEADDR where there
remains lingering listening sockets on the same address and port as a
newly created one.

To avoid this scenario, we don't create a new proxy port for each new
client run.  Instead, we create one proxy socket when the proxy object
is created, and close it when destroying that object.

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5095)

19 months agoOnly implement secure malloc if _POSIX_VERSION allows
Richard Levitte [Thu, 18 Jan 2018 13:05:33 +0000 (14:05 +0100)]
Only implement secure malloc if _POSIX_VERSION allows

Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5060)

19 months agoThe Cygwin gcc doesn't define _WIN32, don't pretend it does
Richard Levitte [Mon, 15 Jan 2018 18:05:01 +0000 (19:05 +0100)]
The Cygwin gcc doesn't define _WIN32, don't pretend it does

Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5060)

19 months agoSimplify Cygwin checks, part 1
Richard Levitte [Mon, 15 Jan 2018 18:04:17 +0000 (19:04 +0100)]
Simplify Cygwin checks, part 1

Because OPENSSL_SYS_CYGWIN will keep OPENSSL_SYS_UNIX defined, there's
no point having checks of this form:

    #if (defined(OPENSSL_SYS_UNIX) || defined(OPENSSL_SYS_CYGWIN))

Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5060)

19 months agoCygwin is POSIX, don't say it isn't
Richard Levitte [Thu, 11 Jan 2018 21:01:44 +0000 (22:01 +0100)]
Cygwin is POSIX, don't say it isn't

More to the point, Cygwin is a POSIX API.  In our library, the use of
a POSIX API is marked by defining the macro OPENSSL_SYS_UNIX.
Therefore, that macro shouldn't be undefined when building for Cygwin.

Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5060)

19 months agoTLSProxy::Proxy: Don't use ReuseAddr on Windows
Richard Levitte [Thu, 18 Jan 2018 09:54:48 +0000 (10:54 +0100)]
TLSProxy::Proxy: Don't use ReuseAddr on Windows

On Windows, we sometimes see a behavior with SO_REUSEADDR where there
remains lingering listening sockets on the same address and port as a
newly created one.

An easy solution is not to use ReuseAddr on Windows.

Thanks Bernd Edlinger for the suggestion.

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5103)

19 months agodoc: Bad prototypes of EVP_PKEY_CTX_new()
Jakub Jelen [Thu, 18 Jan 2018 00:23:37 +0000 (19:23 -0500)]
doc: Bad prototypes of EVP_PKEY_CTX_new()

CLA: trivial

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4861)

19 months agoTLSProxy::Proxy: don't waste time redirecting STDOUT and STDERR
Richard Levitte [Wed, 17 Jan 2018 20:27:33 +0000 (21:27 +0100)]
TLSProxy::Proxy: don't waste time redirecting STDOUT and STDERR

On Windows, it seems that doing so in a forked (pseudo-)process
sometimes affects the parent, and thereby hides all the results that
are supposed to be seen by the running test framework (the "ok" and
"not ok" lines).

It turns out that our redirection isn't necessary, as the test
framework seems to swallow it all in non-verbose mode anyway.
It's possible that we did need this at some point, but the framework
has undergone some refinement since then...

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5100)

19 months agoTLSProxy::Proxy: If we don't support IPv6, force IPv4
Richard Levitte [Tue, 16 Jan 2018 18:50:10 +0000 (19:50 +0100)]
TLSProxy::Proxy:  If we don't support IPv6, force IPv4

We use the first we can of the following IO::Socket modules to create
sockets:

- IO::Socket::INET6
- IO::Socket::IP
- IO::Socket::INET

The last of them doesn't support IPv6, so if that's the one available,
we must force the s_client and s_server processes to use IPv4.

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5092)

19 months agoUpdate all affected files' copyright year to 2018
Paul Yang [Mon, 15 Jan 2018 17:01:46 +0000 (01:01 +0800)]
Update all affected files' copyright year to 2018

Because the related PR/commits are merged in 2018...

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4976)

19 months agoEnforce return values section check
Paul Yang [Tue, 2 Jan 2018 17:07:57 +0000 (01:07 +0800)]
Enforce return values section check

To avoid check failure, make dummy RETURN VALUES sections in the docs
which have no real functions decribed inside...

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4976)

19 months agoAdd missing 'RETURN VALUES' sections in doc
Paul Yang [Mon, 25 Dec 2017 09:50:39 +0000 (17:50 +0800)]
Add missing 'RETURN VALUES' sections in doc

All missing sections are added.

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4976)

19 months agoExtend timeout for TLSProxy
Matt Caswell [Tue, 16 Jan 2018 15:04:51 +0000 (15:04 +0000)]
Extend timeout for TLSProxy

I received this error from a TLSProxy test:

Failed to start up server (localhost,4443): Transport endpoint is not
connected

So, extend the timeout before we give up trying to connect to the server.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5088)

19 months agoRevert BN_copy() flag copy semantics change
Matt Caswell [Mon, 15 Jan 2018 11:23:07 +0000 (11:23 +0000)]
Revert BN_copy() flag copy semantics change

Commit 9f9442918a changed the semantics of BN_copy() to additionally
copy the BN_FLG_CONSTTIME flag if it is set. This turns out to be
ill advised as it has unintended consequences. For example calling
BN_mod_inverse_no_branch() can sometimes return a result with the flag
set and sometimes not as a result. This can lead to later failures if we
go down code branches that do not support constant time, but check for
the presence of the flag.

The original commit was made due to an issue in BN_MOD_CTX_set(). The
original PR fixed the problem in that function, but it was changed in
review to fix it in BN_copy() instead. The solution seems to be to revert
the BN_copy() change and go back to the originally proposed way.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/5080)

19 months agoFix memory leak in do_rand_drbg_init()
Dr. Matthias St. Pierre [Tue, 16 Jan 2018 07:03:37 +0000 (08:03 +0100)]
Fix memory leak in do_rand_drbg_init()

Fixes #5076

Since do_rand_drbg_init() allocates three locks, it needs to ensure
that OPENSSL_init_crypto() is called, otherwise these resources are
not cleaned up properly.

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/5083)

19 months agoFix intermittent Windows and Cygwin failures in s_server
Richard Levitte [Mon, 15 Jan 2018 09:40:24 +0000 (10:40 +0100)]
Fix intermittent Windows and Cygwin failures in s_server

The same kind of failure that has already been observed on the
s_client can sometimes also be observed on s_server, so we need to add
the same kind of 50ms delay as was previously added on s_client.

Ref: git commit cb2e10f257a464c6b475b321dd9e4769df84dbf6:

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5079)

19 months agoFix Windows build file template to recognise .res files
Richard Levitte [Sun, 14 Jan 2018 21:39:20 +0000 (22:39 +0100)]
Fix Windows build file template to recognise .res files

Only when building the main shared libraries

Fixes #5075

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5077)

19 months agoFix intermittent Cygwin failures in s_client
Richard Levitte [Sun, 14 Jan 2018 16:15:32 +0000 (17:15 +0100)]
Fix intermittent Cygwin failures in s_client

This was identified for Windows almost two years ago for VC and
msys/mingw.  It seems that Cygwin suffers from the same issue, and
since Cygwin doesn't define OPENSSL_SYS_WINDOWS, we need to make a
special case to have a 50ms pause before closing the TLS connection.

Ref: git commit cb2e10f257a464c6b475b321dd9e4769df84dbf6

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/5074)

19 months agoUse constant value 1 instead of SHUT_WR in do_server
Bernd Edlinger [Sun, 14 Jan 2018 17:17:59 +0000 (18:17 +0100)]
Use constant value 1 instead of SHUT_WR in do_server

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5072)

19 months agoExplicitly shut the socket down in s_client
Bernd Edlinger [Sat, 13 Jan 2018 17:41:08 +0000 (18:41 +0100)]
Explicitly shut the socket down in s_client

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5072)

19 months agoCatch SIGPIPE in TLSProxy::Proxy::clientstart
Bernd Edlinger [Sat, 13 Jan 2018 09:24:45 +0000 (10:24 +0100)]
Catch SIGPIPE in TLSProxy::Proxy::clientstart

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5071)

19 months agoChange the name of the SEC typedef to openssl_speed_sec_t.
Pauli [Fri, 12 Jan 2018 03:37:39 +0000 (13:37 +1000)]
Change the name of the SEC typedef to openssl_speed_sec_t.

This to avoid a clash with a #defined symbol on Solaris from time.h.

Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5064)

19 months agoMake data argument const in SSL_dane_tlsa_add
Patrick Schlangen [Wed, 10 Jan 2018 17:06:35 +0000 (12:06 -0500)]
Make data argument const in SSL_dane_tlsa_add

The data argument of SSL_dane_tlsa_add is used read-only, so it
should be const.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5056)

19 months agoEnhance util/openssl-update-copyright shell script
Dr. Matthias St. Pierre [Tue, 9 Jan 2018 15:09:14 +0000 (16:09 +0100)]
Enhance util/openssl-update-copyright shell script

- Avoid creating repeated year ranges <current_year>-<current_year> for
  documents which were added in the current year.

- Use extended regular expressions for better readability (less quoting)

- Use a shebang line which is more portable

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5044)

19 months agoTolerate DTLS alerts with an incorrect version number
Matt Caswell [Fri, 5 Jan 2018 10:12:29 +0000 (10:12 +0000)]
Tolerate DTLS alerts with an incorrect version number

In the case of a protocol version alert being sent by a peer the record
version number may not be what we are expecting. In DTLS records with an
unexpected version number are silently discarded. This probably isn't
appropriate for alerts, so we tolerate a mismatch in the minor version
number.

This resolves an issue reported on openssl-users where an OpenSSL server
chose DTLS1.0 but the client was DTLS1.2 only and sent a protocol_version
alert with a 1.2 record number. This was silently ignored by the server.

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5018)

19 months agoConfigure: try to make sure every config target name is unique
Richard Levitte [Tue, 9 Jan 2018 18:51:19 +0000 (19:51 +0100)]
Configure: try to make sure every config target name is unique

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5047)

19 months agoFix --strict-warnings with C90
Todd Short [Tue, 9 Jan 2018 15:46:44 +0000 (10:46 -0500)]
Fix --strict-warnings with C90

Found with gcc 4.8.4

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5045)

19 months agoUpdate copyright years on all files merged since Jan 1st 2018
Richard Levitte [Tue, 9 Jan 2018 04:49:01 +0000 (05:49 +0100)]
Update copyright years on all files merged since Jan 1st 2018

Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5038)

19 months agoVarious small build improvements on mkdef.pl, progs.pl, crypto/init.c, crypto/mem.c
David von Oheimb [Mon, 8 Jan 2018 13:13:51 +0000 (14:13 +0100)]
Various small build improvements on mkdef.pl, progs.pl, crypto/init.c, crypto/mem.c

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4994)

19 months agoDocument SSL_OP_NO_RENEGOTIATION as new in 1.1.1
Christian Heimes [Mon, 11 Dec 2017 06:00:29 +0000 (07:00 +0100)]
Document SSL_OP_NO_RENEGOTIATION as new in 1.1.1

Closes: https://github.com/openssl/openssl/issues/4897
Signed-off-by: Christian Heimes <christian@python.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4898)

19 months agoAvoid only exact duplicates when creating the accepted CA names list
Tomas Mraz [Mon, 11 Dec 2017 12:09:13 +0000 (13:09 +0100)]
Avoid only exact duplicates when creating the accepted CA names list

This avoids situations where third party client is unable to recognize
that the client certificate was issued by the same CA with name differring
only by case or insignificant characters.

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4731)

19 months agoDocument OPENSSL_ENGINES environment variable
Patrick Steuer [Mon, 8 Jan 2018 18:42:32 +0000 (13:42 -0500)]
Document OPENSSL_ENGINES environment variable

In man1/engine.pod and man3/ENGINE_add.pod

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4956)

19 months agocrypto/engine/eng_list.c: compare getenv rv to NULL instead of 0
Patrick Steuer [Mon, 18 Dec 2017 21:47:01 +0000 (22:47 +0100)]
crypto/engine/eng_list.c: compare getenv rv to NULL instead of 0

Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4958)

19 months agoClarify error for unrecognized arguments.
Jacob Hoffman-Andrews [Sat, 23 Dec 2017 00:18:22 +0000 (16:18 -0800)]
Clarify error for unrecognized arguments.

Many of the sub-commands under apps/ accept cipher or digest arguments like
"-sha256". These are implemented using a catchall flag that runs the result
through opt_md() or opt_cipher(). That means any unrecognized flag, including
typos, gets sent to those two functions, producing confusing error messages like
below:

    $ ./apps/openssl req -x590
    req: Unrecognized digest x590
    req: Use -help for summary.

This change switches these two functions to say "Unrecognized flag X" instead.
The new message deliberately leaves off the "-" from the flag name, because
there are some cases where opt_md() and opt_cipher() are passed a flag value
instead (for instance, openssl ca -md). I think the new message is generic
enough that it can serve both cases with improved clarity.

CLA: trivial

Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4972)

19 months agofix compile error 'intrinsic function not declared'
EasySec [Sat, 30 Dec 2017 17:48:23 +0000 (18:48 +0100)]
fix compile error 'intrinsic function not declared'

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5000)

19 months agoFix spelling: adroideabi -> androideabi
pass86 [Sun, 7 Jan 2018 13:57:25 +0000 (21:57 +0800)]
Fix spelling: adroideabi -> androideabi

CLA: trivial

Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/5029)

19 months agoUse the index that matches the key type (either SSL_PKEY_RSA_PSS_SIGN or SSL_PKEY_RSA).
Noah Robbin [Wed, 29 Nov 2017 21:58:25 +0000 (16:58 -0500)]
Use the index that matches the key type (either SSL_PKEY_RSA_PSS_SIGN or SSL_PKEY_RSA).

Extract the RSA key using EVP_PKEY_get0.  Type is checked externally to be either EVP_PKEY_RSA_PSS or EVP_PKEY_RSA.

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4389)

19 months agoUse size of server key when selecting signature algorithm.
Noah Robbin [Tue, 19 Sep 2017 16:15:42 +0000 (12:15 -0400)]
Use size of server key when selecting signature algorithm.

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4389)

19 months agoNUMERICSTRING support
Dmitry Belyavskiy [Mon, 8 Jan 2018 12:32:47 +0000 (15:32 +0300)]
NUMERICSTRING support

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5036)

19 months agoAdd util/openssl-update-copyright shell script
Dr. Matthias St. Pierre [Sun, 7 Jan 2018 01:29:01 +0000 (02:29 +0100)]
Add util/openssl-update-copyright shell script

usage: openssl-update-copyright [-h|--help] [file|directory] ...

Updates the year ranges of all OpenSSL copyright statements in the given
files or directories. (Directories are traversed recursively.)

Only copyright statements containing the string 'The OpenSSL Project' are
affected. The copyright time range is adjusted to include the current year.
If only a single year was specified, it is replaced by a time range starting
at that year and ending at the current year. All '(c)' and '(C)' signs are
preserved.

Signed-off-by: Dr. Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5027)

19 months agoSeparate general linking flags from extra libraries
Richard Levitte [Mon, 8 Jan 2018 11:28:08 +0000 (12:28 +0100)]
Separate general linking flags from extra libraries

So far, we've placed all extra library related flags together, ending
up in the make variable EX_LIBS.  This turns out to be problematic, as
for example, some compilers don't quite agree with something like
this:

    cc -o foo foo.o -L/whatever -lsomething

They prefer this:

    cc -L/whatever -o foo foo.o -lsomething

IBM's compiler on OS/390 is such a compiler that we know of, and we
have previously handled that as a previous case.

The answer here is to make a more general solution, where linking
options are divided in two parts, where one ends up in LDFLAGS and
the other in EX_LIBS (they corresponds to what is called LDFLAGS and
LDLIBS in the GNU world)

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5033)

19 months agoClean up uClinux targets
Richard Levitte [Mon, 8 Jan 2018 11:40:06 +0000 (12:40 +0100)]
Clean up uClinux targets

The uClinux targets included some attributes that would result in
circular references of CFLAGS and LDCLAGS.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5034)

19 months agoFix IPv6 define
Rich Salz [Sun, 7 Jan 2018 20:58:52 +0000 (15:58 -0500)]
Fix IPv6 define

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5030)

19 months agoCHANGES: Document the removal of OS390-Unix
Richard Levitte [Sun, 7 Jan 2018 21:36:12 +0000 (22:36 +0100)]
CHANGES: Document the removal of OS390-Unix

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5031)

19 months agos390x assembly pack: add KMA code path for aes-gcm.
Patrick Steuer [Mon, 2 Oct 2017 13:53:00 +0000 (15:53 +0200)]
s390x assembly pack: add KMA code path for aes-gcm.

Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4634)

19 months agocrypto/aes/asm/aes-s390x.pl: replace decrypt flag by macro.
Patrick Steuer [Tue, 24 Oct 2017 11:29:40 +0000 (13:29 +0200)]
crypto/aes/asm/aes-s390x.pl: replace decrypt flag by macro.

Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4634)

19 months agos390x assembly pack: add KMA code path for aes-ctr.
Patrick Steuer [Tue, 14 Feb 2017 01:07:37 +0000 (02:07 +0100)]
s390x assembly pack: add KMA code path for aes-ctr.

Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4634)

19 months agoec/curve25519.c: avoid 2^51 radix on SPARC.
Andy Polyakov [Sun, 31 Dec 2017 12:23:08 +0000 (13:23 +0100)]
ec/curve25519.c: avoid 2^51 radix on SPARC.

SPARC ISA doesn't have provisions to back up 128-bit multiplications
and additions. And so multiplications are done with library calls
and carries with comparisons and conditional moves. As result base
2^51 code is >40% slower...

Reviewed-by: Tim Hudson <tjh@openssl.org>
19 months agoec/ecp_nistz256.c: switch to faster addition chain in scalar inversion.
Andy Polyakov [Sat, 30 Dec 2017 19:15:44 +0000 (20:15 +0100)]
ec/ecp_nistz256.c: switch to faster addition chain in scalar inversion.

[and improve formatting]

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5001)

19 months agoec/asm/ecp_nistz256-armv8.pl: add optimized inversion.
Andy Polyakov [Sat, 30 Dec 2017 14:11:25 +0000 (15:11 +0100)]
ec/asm/ecp_nistz256-armv8.pl: add optimized inversion.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5001)

19 months agoec/asm/ecp_nistz256-x86_64.pl: add .cfi and SEH handlers to new functions.
Andy Polyakov [Sat, 30 Dec 2017 14:51:55 +0000 (15:51 +0100)]
ec/asm/ecp_nistz256-x86_64.pl: add .cfi and SEH handlers to new functions.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5001)

19 months agoec/ecp_nistz256.c: improve ECDSA sign by 30-40%.
Andy Polyakov [Sat, 30 Dec 2017 14:08:31 +0000 (15:08 +0100)]
ec/ecp_nistz256.c: improve ECDSA sign by 30-40%.

This is based on RT#3810, which added dedicated modular inversion.
ECDSA verify results improves as well, but not as much.

Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5001)

19 months agoRemove remaining NETWARE ifdef's
Rich Salz [Sat, 6 Jan 2018 16:49:53 +0000 (11:49 -0500)]
Remove remaining NETWARE ifdef's

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/5028)