oweals/openssl.git
5 years agoAdapt SSL_DEBUG to the new generic trace API
Richard Levitte [Tue, 11 Dec 2018 23:04:44 +0000 (00:04 +0100)]
Adapt SSL_DEBUG to the new generic trace API

Co-authored-by: Dr. Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8198)

5 years agoMake it possible to disable the TRACE API
Richard Levitte [Wed, 12 Dec 2018 20:31:36 +0000 (21:31 +0100)]
Make it possible to disable the TRACE API

This disabled the tracing functionality by making functions do
nothing, and making convenience macros produce dead code.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8198)

5 years agoapps/openssl.c: Adapt to enable tracing output
Richard Levitte [Fri, 14 Dec 2018 14:48:53 +0000 (15:48 +0100)]
apps/openssl.c: Adapt to enable tracing output

Use the environment variables OPENSSL_TRACE to determine what's going
to be enabled.  The value of this variables is a comma separated list
of trace and debugging names, which correspond to the trace category
macros defined in include/openssl/trace.h.

For example, setting OPENSSL_DEBUG=TRACE,SSL will enable debugging output
for the types OSSL_TRACE_CATEGORY_TRACE and OSSL_TRACE_CATEGORY_SSL.

This also slightly changes the handling of the prefix method in
apps/apps.c.  This is for the better, as the prefix method pointer was
unneccessarily stored in two places.

Co-authored-by: Dr. Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8198)

5 years agoAdd generic trace API
Richard Levitte [Tue, 11 Dec 2018 22:58:29 +0000 (23:58 +0100)]
Add generic trace API

The idea is that the application shall be able to register output
channels or callbacks to print tracing output as it sees fit.

OpenSSL internals, on the other hand, want to print thoses texts using
normal printing routines, such as BIO_printf() or BIO_dump() through
well defined BIOs.

When the application registers callbacks, the tracing functionality
sets up an internal BIO that simply forwards received text to the
appropriate application provided callback.

Co-authored-by: Dr. Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8198)

5 years agoMake the sparse array code use ossl_uintmax_t as its index rather than size_t.
Pauli [Wed, 6 Mar 2019 03:50:54 +0000 (13:50 +1000)]
Make the sparse array code use ossl_uintmax_t as its index rather than size_t.

This should never reduce the range covered and might increase it on some
platforms.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8415)

5 years agopadlock: generate assembler source for static libraries too
Richard Levitte [Tue, 5 Mar 2019 17:38:16 +0000 (18:38 +0100)]
padlock: generate assembler source for static libraries too

The GENERATE lines for generating the padlock assembler source were
wrongly placed in such a way that they only applied to the shared
library build.

[extended tests]

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/8412)

5 years agoDon't write the tick_identity to the session
Matt Caswell [Fri, 1 Mar 2019 15:40:20 +0000 (15:40 +0000)]
Don't write the tick_identity to the session

Sessions must be immutable once they can be shared with multiple threads.
We were breaking that rule by writing the ticket index into it during the
handshake. This can lead to incorrect behaviour, including failed
connections in multi-threaded environments.

Reported by David Benjamin.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8383)

5 years agoAdd advice on setting CXX at the same time as CC
Richard Levitte [Fri, 1 Mar 2019 10:54:07 +0000 (11:54 +0100)]
Add advice on setting CXX at the same time as CC

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8370)

5 years agoTravis: use enable-buildtest-c++
Richard Levitte [Thu, 28 Feb 2019 12:35:32 +0000 (13:35 +0100)]
Travis: use enable-buildtest-c++

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8370)

5 years agoConfigure: make C++ build tests optional and configurable
Richard Levitte [Thu, 28 Feb 2019 12:28:43 +0000 (13:28 +0100)]
Configure: make C++ build tests optional and configurable

Disabled by default

Fixes #8360

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8370)

5 years agoVMS: fix util/[un]local.com.in to handle attributes
Richard Levitte [Fri, 1 Mar 2019 13:15:13 +0000 (14:15 +0100)]
VMS: fix util/[un]local.com.in to handle attributes

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8382)

5 years agotestutil: ensure good treatment of argv on non-Unix platforms
Richard Levitte [Sun, 3 Mar 2019 09:27:10 +0000 (10:27 +0100)]
testutil: ensure good treatment of argv on non-Unix platforms

From a Unix point of view, some other platform families have certain
quirks.  Windows command prompt doesn't expand globs into actual file
names, so we must do this.  VMS has some oddity with argv pointer size
that can cause crashes if you're not careful (by copying it to a less
surprising pointer size array).

The fixups already exist and are used in the apps/ code.  However, the
testutil code started using the opt routines from apps/ without
including the non-Unix fixups.  This change fixes that.

For VMS' sake, libtestutil gets an app_malloc() shim, to avoid sucking
in all of apps/apps.c.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8381)

5 years agoVMS: move copy_argc to its own module and make it an aux source
Richard Levitte [Sun, 3 Mar 2019 09:20:37 +0000 (10:20 +0100)]
VMS: move copy_argc to its own module and make it an aux source

copy_argv was never initialization code.

Make it self-cleaning too.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8381)

5 years agoDeprecate the "hw" configuration options, make "padlockeng" disablable
Richard Levitte [Fri, 1 Mar 2019 12:18:11 +0000 (13:18 +0100)]
Deprecate the "hw" configuration options, make "padlockeng" disablable

The "hw" and "hw-.*" style options are historical artifacts, sprung
from the time when ENGINE was first designed, with hardware crypto
accelerators and HSMs in mind.

Today, these options have largely lost their value, replaced by
options such as "no-{foo}eng" and "no-engine".

This completes the transition by making "hw" and "hw-.*" deprecated,
but automatically translated into more modern variants of the same.

In the process, we get rid of the last regular expression in
Configure's @disablables, a feature that was ill supported anyway.
Also, padlock now gets treated just as every other engine.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8380)

5 years agoopenssl_strerror_r: Fix handling of GNU strerror_r
Vitezslav Cizek [Thu, 28 Feb 2019 12:47:18 +0000 (13:47 +0100)]
openssl_strerror_r: Fix handling of GNU strerror_r

GNU strerror_r may return either a pointer to a string that the function
stores in buf, or a pointer to some (immutable) static string in which case
buf is unused.

In such a case we need to set buf manually.

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8371)

5 years agoCheck for negative return for signature size.Addresses Coverity 1442933
Pauli [Mon, 4 Mar 2019 03:37:39 +0000 (13:37 +1000)]
Check for negative return for signature size.Addresses Coverity 1442933

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8392)

5 years agoFix seeding from random device w/o getrandom syscall
Bernd Edlinger [Thu, 14 Feb 2019 23:03:50 +0000 (00:03 +0100)]
Fix seeding from random device w/o getrandom syscall

Use select to wait for /dev/random in readable state,
but do not actually read anything from /dev/random,
use /dev/urandom first.

Use linux define __NR_getrandom instead of the
glibc define SYS_getrandom, in case the kernel headers
are more current than the glibc headers.

Fixes #8215

Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/8251)

5 years agoFix no-ec, no-sm2 and no-sm3
Matt Caswell [Thu, 28 Feb 2019 13:47:26 +0000 (13:47 +0000)]
Fix no-ec, no-sm2 and no-sm3

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8372)

5 years agoAdd section order check in util/find-doc-nits
Paul Yang [Tue, 26 Feb 2019 05:51:02 +0000 (13:51 +0800)]
Add section order check in util/find-doc-nits

This patch checks if the EXAMPLES section in a pod file is placed
before the RETURN VALUES section.

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8338)

5 years agoPlace return values after examples in doc
Paul Yang [Tue, 26 Feb 2019 05:11:10 +0000 (13:11 +0800)]
Place return values after examples in doc

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8338)

5 years agoapps/ca.c: only output DER with SPKAC input and when -out is chosen
Richard Levitte [Thu, 28 Feb 2019 10:30:19 +0000 (11:30 +0100)]
apps/ca.c: only output DER with SPKAC input and when -out is chosen

So say the docs

Fixes #8055

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8368)

5 years agoThe use of the likes of UINT32_MAX requires internal/numbers.h
Richard Levitte [Thu, 28 Feb 2019 09:31:20 +0000 (10:31 +0100)]
The use of the likes of UINT32_MAX requires internal/numbers.h

Found a few more cases.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8366)

5 years agoConfigure: support a few more "make variables" defaulting from env
Richard Levitte [Tue, 26 Feb 2019 22:18:02 +0000 (23:18 +0100)]
Configure: support a few more "make variables" defaulting from env

CFLAGS, CXXFLAGS, CPPFLAGS, LDFLAGS, and LDLIBS

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8325)

5 years ago.travis.yml: change -std=c89 to -ansi
Richard Levitte [Tue, 26 Feb 2019 08:34:00 +0000 (09:34 +0100)]
.travis.yml: change -std=c89 to -ansi

For C, -ansi is equivalent to -std=c90
For C++, -ansi is equivalent to -std=c++98

We also place -ansi in CPPFLAGS instead of the usual command line config,
to avoid getting it when linking (clang complains)

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8325)

5 years agoConfiguration: divide devteam flags into language specific sets
Richard Levitte [Tue, 26 Feb 2019 08:32:52 +0000 (09:32 +0100)]
Configuration: divide devteam flags into language specific sets

Some of the devteam flags are not for C++

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8325)

5 years agoDo buildtests on our public header files with C++ as well
Richard Levitte [Mon, 25 Feb 2019 10:43:00 +0000 (11:43 +0100)]
Do buildtests on our public header files with C++ as well

This ensures that we don't mistakenly use C++ keywords anywhere public.

Related to #8313

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8325)

5 years agoOpenSSL::Util::Pod: allow slashes in names
Richard Levitte [Tue, 19 Feb 2019 11:10:14 +0000 (12:10 +0100)]
OpenSSL::Util::Pod: allow slashes in names

The names in the NAME section may describe headers, which contain a slash
for OpenSSL headers.  We deal with that by converting slashes to dashes
for the file names.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8286)

5 years agoReplumbing: Add include/openssl/core.h, initially with core types
Richard Levitte [Sun, 20 Jan 2019 12:23:30 +0000 (13:23 +0100)]
Replumbing: Add include/openssl/core.h, initially with core types

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8286)

5 years agoImplement SSH KDF
Simo Sorce [Wed, 19 Sep 2018 20:23:45 +0000 (16:23 -0400)]
Implement SSH KDF

SSH's KDF is defined in RFC 4253 in Section 7.2

Signed-off-by: Simo Sorce <simo@redhat.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7290)

5 years agoMake the padlock engine build correctly
Richard Levitte [Wed, 13 Feb 2019 03:23:14 +0000 (04:23 +0100)]
Make the padlock engine build correctly

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8220)

5 years agoEnsure configured module specific and application specific defines are used
Richard Levitte [Wed, 13 Feb 2019 03:21:59 +0000 (04:21 +0100)]
Ensure configured module specific and application specific defines are used

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8220)

5 years agoAdd PADLOCK_ASM to dso_defines rather than lib_defines
Richard Levitte [Wed, 13 Feb 2019 02:58:52 +0000 (03:58 +0100)]
Add PADLOCK_ASM to dso_defines rather than lib_defines

Since the padlock code is an engine, the assembler is for a module,
not a library link to when building a program...  there's a
distinction.

Fixes #2311

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8220)

5 years agoThe use of the likes of UINT32_MAX requires internal/numbers.h
Richard Levitte [Wed, 27 Feb 2019 08:53:17 +0000 (09:53 +0100)]
The use of the likes of UINT32_MAX requires internal/numbers.h

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8354)

5 years agoFix the default digest algorthm of SM2
Paul Yang [Mon, 25 Feb 2019 09:34:03 +0000 (17:34 +0800)]
Fix the default digest algorthm of SM2

Currently SM2 shares the ameth with EC, so the current default digest
algorithm returned is SHA256. This fixes the default digest algorithm of
SM2 to SM3, which is the only valid digest algorithm for SM2 signature.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8186)

5 years agoSupport raw input data in apps/pkeyutl
Paul Yang [Wed, 16 Jan 2019 08:16:28 +0000 (16:16 +0800)]
Support raw input data in apps/pkeyutl

Some signature algorithms require special treatment for digesting, such
as SM2. This patch adds the ability of handling raw input data in
apps/pkeyutl other than accepting only pre-hashed input data.

Beside, SM2 requries an ID string when signing or verifying a piece of data,
this patch also adds the ability for apps/pkeyutil to specify that ID
string.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8186)

5 years agoRevert "Configure: stop forcing use of DEFINE macros in headers"
Richard Levitte [Fri, 15 Feb 2019 10:48:24 +0000 (11:48 +0100)]
Revert "Configure: stop forcing use of DEFINE macros in headers"

Github PR #8246 provides a better solution to the problem.

This reverts commit f11ffa505f8a9345145a26a05bf77b012b6941bd.

[extended tests]

Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/8247)

5 years agoAdd BN_native2bn and BN_bn2nativepad, for native BIGNUM import/export
Richard Levitte [Tue, 26 Feb 2019 12:08:31 +0000 (13:08 +0100)]
Add BN_native2bn and BN_bn2nativepad, for native BIGNUM import/export

These are a couple of utility functions, to make import and export of
BIGNUMs to byte strings in platform native for (little-endian or
big-endian) easier.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8346)

5 years ago[test] modernize ecdsatest and extend ECDSA sign KATs
Billy Brumley [Sat, 23 Feb 2019 08:12:35 +0000 (10:12 +0200)]
[test] modernize ecdsatest and extend ECDSA sign KATs

Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8314)

5 years agoFix trivial typo in EVP_DigestVerifyInit doc
Nicola Tuveri [Sun, 24 Feb 2019 19:29:32 +0000 (21:29 +0200)]
Fix trivial typo in EVP_DigestVerifyInit doc

Reviewed-by: Paul Yang <yang.yang@baishancloud.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8319)

5 years agoClarify that SSL_shutdown() must not be called after a fatal error
Matt Caswell [Wed, 20 Feb 2019 14:21:36 +0000 (14:21 +0000)]
Clarify that SSL_shutdown() must not be called after a fatal error

Follow on from CVE-2019-1559

Reviewed-by: Richard Levitte <levitte@openssl.org>
5 years agoDisable 02-test_errstr.t on msys/mingw as well as MSWin32
Richard Levitte [Tue, 26 Feb 2019 10:22:16 +0000 (11:22 +0100)]
Disable 02-test_errstr.t on msys/mingw as well as MSWin32

There is too high a risk that perl and OpenSSL are linked with
different C RTLs, and thereby get different messages for even the most
mundane error numbers.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8343)

5 years agoRearrange the inclusion of curve448/curve448_lcl.h
Richard Levitte [Mon, 25 Feb 2019 18:27:42 +0000 (19:27 +0100)]
Rearrange the inclusion of curve448/curve448_lcl.h

The real cause for this change is that test/ec_internal_test.c
includes ec_lcl.h, and including curve448/curve448_lcl.h from there
doesn't work so well with compilers who always do inclusions relative
to the C file being compiled.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8334)

(cherry picked from commit f408e2a352b59f2e7aa2160bfb6285725fe88ea7)

5 years agoEnsure bn_cmp_words can handle the case where n == 0
Matt Caswell [Mon, 25 Feb 2019 11:28:32 +0000 (11:28 +0000)]
Ensure bn_cmp_words can handle the case where n == 0

Thanks to David Benjamin who reported this, performed the analysis and
suggested the patch. I have incorporated some of his analysis in the
comments below.

This issue can cause an out-of-bounds read. It is believed that this was
not reachable until the recent "fixed top" changes. Analysis has so far
only identified one code path that can encounter this - although it is
possible that others may be found. The one code path only impacts 1.0.2 in
certain builds. The fuzzer found a path in RSA where iqmp is too large. If
the input is all zeros, the RSA CRT logic will multiply a padded zero by
iqmp. Two mitigating factors:

- Private keys which trip this are invalid (iqmp is not reduced mod p).
Only systems which take untrusted private keys care.
- In OpenSSL 1.1.x, there is a check which rejects the oversize iqmp,
so the bug is only reproducible in 1.0.2 so far.

Fortunately, the bug appears to be relatively harmless. The consequences of
bn_cmp_word's misbehavior are:

- OpenSSL may crash if the buffers are page-aligned and the previous page is
non-existent.
- OpenSSL will incorrectly treat two BN_ULONG buffers as not equal when they
are equal.
- Side channel concerns.

The first is indeed a concern and is a DoS bug. The second is fine in this
context. bn_cmp_word and bn_cmp_part_words are used to compute abs(a0 - a1)
in Karatsuba. If a0 = a1, it does not matter whether we use a0 - a1 or
a1 - a0. The third would be worth thinking about, but it is overshadowed
by the entire Karatsuba implementation not being constant time.

Due to the difficulty of tripping this and the low impact no CVE is felt
necessary for this issue.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8326)

5 years agofix x509 -force_pubkey option to take effect with cert input or self-signing; improve...
David von Oheimb [Sun, 3 Feb 2019 06:57:59 +0000 (07:57 +0100)]
fix x509 -force_pubkey option to take effect with cert input or self-signing; improve its doc

Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8165)

5 years agoCID 1442838: API usage errors
Pauli [Sun, 24 Feb 2019 11:27:02 +0000 (21:27 +1000)]
CID 1442838:  API usage errors

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/8318)

5 years agoCID 1442835: Integer Overflow
Pauli [Sun, 24 Feb 2019 11:26:39 +0000 (21:26 +1000)]
CID 1442835:  Integer Overflow

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/8318)

5 years agoCID 1442836: Resource leaks
Pauli [Sun, 24 Feb 2019 11:26:20 +0000 (21:26 +1000)]
CID 1442836:  Resource leaks

Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/8318)

5 years agoWindows: Call TerminateProcess, not ExitProcess
Richard Levitte [Thu, 21 Feb 2019 17:25:50 +0000 (18:25 +0100)]
Windows: Call TerminateProcess, not ExitProcess

Ty Baen-Price explains:

> Problem and Resolution:
> The following lines of code make use of the Microsoft API ExitProcess:
>
> ```
> Apps\Speed.c line 335: ExitProcess(ret);
> Ms\uplink.c line 22: ExitProcess(1);
> ```
>
> These function calls are made after fatal errors are detected and
> program termination is desired. ExitProcess(), however causes
> _orderly_ shutdown of a process and all its threads, i.e. it unloads
> all dlls and runs all destructors. See MSDN for details of exactly
> what happens
> (https://msdn.microsoft.com/en-us/library/windows/desktop/ms682658(v=vs.85).aspx).
> The MSDN page states that ExitProcess should never be called unless
> it is _known to be safe_ to call it. These calls should simply be
> replaced with calls to TerminateProcess(), which is what should be
> called for _disorderly_ shutdown.
>
> An example of usage:
>
> ```
> TerminateProcess(GetCurrentProcess(), exitcode);
> ```
>
> Effect of Problem:
> Because of a compilation error (wrong c++ runtime), my program
> executed the uplink.c ExitProcess() call. This caused the single
> OpenSSL thread to start executing the destructors of all my dlls,
> and their objects. Unfortunately, about 30 other threads were
> happily using those objects at that time, eventually causing a
> 0xC0000005 ACCESS_VIOLATION. Obviously an ACCESS_VIOLATION is the
> best case scenario, as I'm sure you can imagine at the consequences
> of undiscovered memory corruption, even in a terminating process.

And on the subject of `TerminateProcess()` being asynchronous:

> That is technically true, but I think it's probably synchronous
> "enough" for your purposes, since a call to TerminateProcess
> suspends execution of all threads in the target process. This means
> it's really only asynchronous if you're calling TerminateProcess one
> some _other_ process. If you're calling TerminateProcess on your own
> process, you'll never return from the TerminateProcess call.

Fixes #2489
Was originally RT-4526

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8301)

5 years agoDon't restrict the number of KeyUpdate messages we can process
Matt Caswell [Thu, 21 Feb 2019 16:02:24 +0000 (16:02 +0000)]
Don't restrict the number of KeyUpdate messages we can process

Prior to this commit we were keeping a count of how many KeyUpdates we
have processed and failing if we had had too many. This simplistic approach
is not sufficient for long running connections. Since many KeyUpdates
would not be a particular good DoS route anyway, the simplest solution is
to simply remove the key update count.

Fixes #8068

Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/8299)

5 years agoengines/dasync: add explaining comments about AES-128-CBC-HMAC-SHA1
Dr. Matthias St. Pierre [Fri, 22 Feb 2019 12:08:54 +0000 (13:08 +0100)]
engines/dasync: add explaining comments about AES-128-CBC-HMAC-SHA1

Fixes #7950

It was reported that there might be a null pointer dereference in the
implementation of the dasync_aes_128_cbc_hmac_sha1() cipher, because
EVP_aes_128_cbc_hmac_sha1() can return a null pointer if AES-NI is
not available. It took some analysis to find out that this is not
an issue in practice, and these comments explain the reason to comfort
further NPD hunters.

Detected by GitHub user @wurongxin1987 using the Sourcebrella Pinpoint
static analyzer.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8305)

5 years agoengines/e_devcrypto: set digest input_blocksize
Eneas U de Queiroz [Thu, 21 Feb 2019 17:16:12 +0000 (14:16 -0300)]
engines/e_devcrypto: set digest input_blocksize

This restores the behavior of previous versions of the /dev/crypto
engine, in alignment with the default implementation.

Reported-by: Gerard Looije <lglooije@hotmail.com>
Signed-off-by: Eneas U de Queiroz <cote2004-github@yahoo.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8213)

5 years agoengines/e_devcrypto: fixes logic in close_devcrypto
Eneas U de Queiroz [Fri, 15 Feb 2019 17:51:59 +0000 (15:51 -0200)]
engines/e_devcrypto: fixes logic in close_devcrypto

Call close(cfd) before setting cfd = -1.

Signed-off-by: Eneas U de Queiroz <cote2004-github@yahoo.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8213)

5 years agoengines/e_devcrypto.c: fix cipher_ctrl function
Eneas U de Queiroz [Tue, 12 Feb 2019 18:53:15 +0000 (16:53 -0200)]
engines/e_devcrypto.c: fix cipher_ctrl function

This fixes commit c703a80, which had a mistake in cipher_ctrl function.

Move the /dev/crypto session cleanup code to its own function.

Signed-off-by: Eneas U de Queiroz <cote2004-github@yahoo.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8213)

5 years agoengines/build.info: fix devcrypto MODULES entry
Eneas U de Queiroz [Tue, 12 Feb 2019 20:02:51 +0000 (18:02 -0200)]
engines/build.info: fix devcrypto MODULES entry

The devcrypto MODULES line was missing the "engine" attribute.

Signed-off-by: Eneas U de Queiroz <cote2004-github@yahoo.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8213)

5 years agoFix a grammar nit in CRYPTO_get_ex_new_index.pod
Paul Yang [Fri, 22 Feb 2019 06:27:39 +0000 (14:27 +0800)]
Fix a grammar nit in CRYPTO_get_ex_new_index.pod

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8303)

5 years agoMake sure that generated POD files are actually created before we run doc-nits
Kurt Roeckx [Tue, 19 Feb 2019 19:29:53 +0000 (20:29 +0100)]
Make sure that generated POD files are actually created before we run doc-nits

Reviewed-by: Richard Levitte <levitte@openssl.org>
GH: #8285

5 years agoIndent with 4
Kurt Roeckx [Tue, 19 Feb 2019 19:24:08 +0000 (20:24 +0100)]
Indent with 4

doc-nits says that over needs a parameter

Reviewed-by: Richard Levitte <levitte@openssl.org>
GH: #8285

5 years agoFix dasync engine
Matt Caswell [Wed, 20 Feb 2019 11:11:04 +0000 (11:11 +0000)]
Fix dasync engine

The aes128_cbc_hmac_sha1 cipher in the dasync engine is broken. Probably
by commit e38c2e8535 which removed use of the "enc" variable...but not
completely.

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/8291)

5 years agoSSL_CONF_cmd: fix doc for NoRenegotiation
Hubert Kario [Wed, 20 Feb 2019 15:21:18 +0000 (16:21 +0100)]
SSL_CONF_cmd: fix doc for NoRenegotiation

The option is a flag for Options, not a standalone setting.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8292)

5 years agoMIPS32R3 provides the EXT instruction to extract bits from
Markus Stockhausen [Mon, 11 Feb 2019 17:38:46 +0000 (18:38 +0100)]
MIPS32R3 provides the EXT instruction to extract bits from
registers. As the AES table is already 1K aligned we can
use it everywhere and speedup table address calculation by
10%. Performance numbers:

decryption         16B       64B      256B     1024B     8192B
-------------------------------------------------------------------
aes-256-cbc   5636.84k  6443.26k  6689.02k  6752.94k  6766.59k bef.
aes-256-cbc   6200.31k  7195.71k  7504.30k  7585.11k  7599.45k aft.
-------------------------------------------------------------------
aes-128-cbc   7313.85k  8653.67k  9079.55k  9188.35k  9205.08k bef.
aes-128-cbc   7925.38k  9557.99k 10092.37k 10232.15k 10272.77k aft.

encryption         16B       64B      256B     1024B     8192B
-------------------------------------------------------------------
aes-256 cbc   6009.65k  6592.70k  6766.59k  6806.87k  6815.74k bef.
aes-256 cbc   6643.93k  7388.69k  7605.33k  7657.81k  7675.90k aft.
-------------------------------------------------------------------
aes-128 cbc   7862.09k  8892.48k  9214.04k  9291.78k  9311.57k bef.
aes-128 cbc   8639.29k  9881.17k 10265.86k 10363.56k 10392.92k aft.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8206)

5 years agocfi build fixes in x86-64 ghash assembly
Shane Lontis [Tue, 19 Feb 2019 03:56:33 +0000 (13:56 +1000)]
cfi build fixes in x86-64 ghash assembly

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8281)

5 years agoClear BN_FLG_CONSTTIME on BN_CTX_get()
Nicola Tuveri [Fri, 8 Feb 2019 10:42:25 +0000 (12:42 +0200)]
Clear BN_FLG_CONSTTIME on BN_CTX_get()

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8253)

5 years agoTest for constant-time flag leakage in BN_CTX
Nicola Tuveri [Mon, 11 Feb 2019 22:37:25 +0000 (00:37 +0200)]
Test for constant-time flag leakage in BN_CTX

This commit adds a simple unit test to make sure that the constant-time
flag does not "leak" among BN_CTX frames:

- test_ctx_consttime_flag() initializes (and later frees before
  returning) a BN_CTX object, then it calls in sequence
  test_ctx_set_ct_flag() and test_ctx_check_ct_flag() using the same
  BN_CTX object. The process is run twice, once with a "normal"
  BN_CTX_new() object, then with a BN_CTX_secure_new() one.
- test_ctx_set_ct_flag() starts a frame in the given BN_CTX and sets the
  BN_FLG_CONSTTIME flag on some of the BIGNUMs obtained from the frame
  before ending it.
- test_ctx_check_ct_flag() then starts a new frame and gets a number of
  BIGNUMs from it. In absence of leaks, none of the BIGNUMs in the new
  frame should have BN_FLG_CONSTTIME set.

In actual BN_CTX usage inside libcrypto the leak could happen at any
depth level in the BN_CTX stack, with varying results depending on the
patterns of sibling trees of nested function calls sharing the same
BN_CTX object, and the effect of unintended BN_FLG_CONSTTIME on the
called BN_* functions.

This simple unit test abstracts away this complexity and verifies that
the leak does not happen between two sibling functions sharing the same
BN_CTX object at the same level of nesting.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8253)

5 years agotest/context_internal_test.c: don't initialize as a separate test
Richard Levitte [Wed, 20 Feb 2019 07:21:02 +0000 (08:21 +0100)]
test/context_internal_test.c: don't initialize as a separate test

Because test order can be randomized, running foo_init() as a separate
test is unsafe practice.  Instead, we make it possible to call it
multiple times, and call it at the start of each separate test.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8288)

5 years agosha/keccak1600.c: subscribe more platforms for "complementing" optimization.
Andy Polyakov [Sun, 17 Feb 2019 17:10:12 +0000 (18:10 +0100)]
sha/keccak1600.c: subscribe more platforms for "complementing" optimization.

E.g. on MIPS64 it gives >20% improvement...

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8261)

5 years agoDon't set SNI by default if hostname is not dNS name
Ionut Mihalcea [Wed, 6 Feb 2019 21:09:15 +0000 (21:09 +0000)]
Don't set SNI by default if hostname is not dNS name

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8175)

5 years agoFix reference to symbol 'main'.
Matthias Kraft [Tue, 19 Feb 2019 12:22:35 +0000 (13:22 +0100)]
Fix reference to symbol 'main'.

The AIX binder needs to be instructed that the output will have no entry
point (see AIX' ld manual: -e in the Flags section; autoexp and noentry
in the Binder section).

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8282)

5 years agoAdd a test for interleaving app data with handshake data in TLSv1.3
Matt Caswell [Fri, 8 Feb 2019 17:25:58 +0000 (17:25 +0000)]
Add a test for interleaving app data with handshake data in TLSv1.3

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/8191)

5 years agoDon't interleave handshake and other record types in TLSv1.3
Matt Caswell [Fri, 8 Feb 2019 16:36:32 +0000 (16:36 +0000)]
Don't interleave handshake and other record types in TLSv1.3

In TLSv1.3 it is illegal to interleave handshake records with non handshake
records.

Fixes #8189

Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/8191)

5 years agoFixup internal documentation
Richard Levitte [Mon, 18 Feb 2019 15:28:00 +0000 (16:28 +0100)]
Fixup internal documentation

There were some faults that got caught by the updated doc-nits

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8270)

5 years agoFix a test ordering issue.
Pauli [Tue, 19 Feb 2019 01:48:51 +0000 (11:48 +1000)]
Fix a test ordering issue.

A randomised order causes failure due to unintentional dependencies between
two of the test cases.

[extended tests]

Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/8279)

5 years agoutil/find-docs-nits: Recognise SPARSE_ARRAY_OF
Richard Levitte [Mon, 18 Feb 2019 15:25:47 +0000 (16:25 +0100)]
util/find-docs-nits: Recognise SPARSE_ARRAY_OF

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8269)

5 years agoutil/find-docs-nits: Extend to handle internal documentation
Richard Levitte [Mon, 18 Feb 2019 15:00:06 +0000 (16:00 +0100)]
util/find-docs-nits: Extend to handle internal documentation

While we're at it, we also check for names that contain white-space,
as they are invalid.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8269)

5 years agocygwin: drop explicit O_TEXT
Corinna Vinschen [Fri, 15 Feb 2019 11:24:47 +0000 (12:24 +0100)]
cygwin: drop explicit O_TEXT

Cygwin binaries should not enforce text mode these days, just
use text mode if the underlying mount point requests it

CLA: trivial

Signed-off-by: Corinna Vinschen <vinschen@redhat.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8248)

5 years agoConform to proper NAME section format
Richard Levitte [Mon, 18 Feb 2019 13:32:58 +0000 (14:32 +0100)]
Conform to proper NAME section format

The NAME section format is comma separated names to the left of the
left of the dash, free form on the right.  If we don't follow that
form, programs like apropos(1) and whatis(1) can't do their job
properly.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8267)

5 years agoProperty: naming and manual clarifiations
Richard Levitte [Mon, 18 Feb 2019 08:40:07 +0000 (09:40 +0100)]
Property: naming and manual clarifiations

- Add a bit more text about that is expected of the user or
  OSSL_METHOD_STOREs.
- Clarify what a method and what a numeric identity are.
- Change all mentions of 'implementation' and 'result' to 'method'.

To clarify further: OpenSSL has used the term 'method' for structures
that mainly contains function pointers.  Those are the methods that
are expected to be stored away in OSSL_METHOD_STOREs.  In the end,
however, it's the caller's responsibility to define exactly what they
want to store, as long as its 'methods' are associated with a numeric
identity and properties.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8265)

5 years agoDon't leak EVP_KDF_CTX on error
Matt Caswell [Sun, 17 Feb 2019 16:13:08 +0000 (16:13 +0000)]
Don't leak EVP_KDF_CTX on error

Found by Coverity

Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/8260)

5 years agogenerated files
Pauli [Sat, 16 Feb 2019 08:33:37 +0000 (18:33 +1000)]
generated files

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8224)

5 years agoProperties for implementation selection.
Pauli [Fri, 16 Nov 2018 01:44:30 +0000 (11:44 +1000)]
Properties for implementation selection.

Properties are a sequence of comma separated name=value pairs.  A name
without a corresponding value is assumed to be a Boolean and have the
true value 'yes'.  Values are either strings or numbers.  Strings can be
quoted either _"_ or _'_ or unquoted (with restrictions).  There are no
escape characters inside strings.  Number are either decimal digits or
'0x' followed by hexidecimal digits.  Numbers are represented internally
as signed sixty four bit values.

Queries on properties are a sequence comma separated conditional tests.
These take the form of name=value (equality test), name!=value (inequality
test) or name (Boolean test for truth).  Queries can be parsed, compared
against a definition or merged pairwise.

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8224)

5 years agoAdd missing dots in dgst man page
Vedran Miletić [Fri, 1 Feb 2019 14:03:09 +0000 (15:03 +0100)]
Add missing dots in dgst man page

CLA: trivial

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
GH: #8142

5 years agoFixed typo
Jan Macku [Wed, 30 Jan 2019 15:09:50 +0000 (16:09 +0100)]
Fixed typo

CLA: trivial

Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
GH: #8121

5 years agoCheck for unpaired .cfi_remember_state
David Benjamin [Tue, 29 Jan 2019 23:41:39 +0000 (17:41 -0600)]
Check for unpaired .cfi_remember_state

Reviewed-by: Richard Levitte <levitte@openssl.org>
GH: #8109

5 years agoFix some CFI issues in x86_64 assembly
David Benjamin [Tue, 29 Jan 2019 05:12:15 +0000 (05:12 +0000)]
Fix some CFI issues in x86_64 assembly

The add/double shortcut in ecp_nistz256-x86_64.pl left one instruction
point that did not unwind, and the "slow" path in AES_cbc_encrypt was
not annotated correctly. For the latter, add
.cfi_{remember,restore}_state support to perlasm.

Next, fill in a bunch of functions that are missing no-op .cfi_startproc
and .cfi_endproc blocks. libunwind cannot unwind those stack frames
otherwise.

Finally, work around a bug in libunwind by not encoding rflags. (rflags
isn't a callee-saved register, so there's not much need to annotate it
anyway.)

These were found as part of ABI testing work in BoringSSL.

Reviewed-by: Richard Levitte <levitte@openssl.org>
GH: #8109

5 years ago[test] unit test for field_inv function pointer in EC_METHOD
Billy Brumley [Tue, 12 Feb 2019 14:00:20 +0000 (16:00 +0200)]
[test] unit test for field_inv function pointer in EC_METHOD

Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/8254)

5 years agoSCA hardening for mod. field inversion in EC_GROUP
Billy Brumley [Sat, 2 Feb 2019 08:53:29 +0000 (10:53 +0200)]
SCA hardening for mod. field inversion in EC_GROUP

This commit adds a dedicated function in `EC_METHOD` to access a modular
field inversion implementation suitable for the specifics of the
implemented curve, featuring SCA countermeasures.

The new pointer is defined as:
`int (*field_inv)(const EC_GROUP*, BIGNUM *r, const BIGNUM *a, BN_CTX*)`
and computes the multiplicative inverse of `a` in the underlying field,
storing the result in `r`.

Three implementations are included, each including specific SCA
countermeasures:
  - `ec_GFp_simple_field_inv()`, featuring SCA hardening through
    blinding.
  - `ec_GFp_mont_field_inv()`, featuring SCA hardening through Fermat's
    Little Theorem (FLT) inversion.
  - `ec_GF2m_simple_field_inv()`, that uses `BN_GF2m_mod_inv()` which
    already features SCA hardening through blinding.

From a security point of view, this also helps addressing a leakage
previously affecting conversions from projective to affine coordinates.

This commit also adds a new error reason code (i.e.,
`EC_R_CANNOT_INVERT`) to improve consistency between the three
implementations as all of them could fail for the same reason but
through different code paths resulting in inconsistent error stack
states.

Co-authored-by: Nicola Tuveri <nic.tuv@gmail.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/8254)

5 years agoARM64 assembly pack: make it Windows-friendly.
Andy Polyakov [Fri, 15 Feb 2019 21:16:41 +0000 (22:16 +0100)]
ARM64 assembly pack: make it Windows-friendly.

"Windows friendliness" means a) unified PIC-ification, unified across
all platforms; b) unified commantary delimiter; c) explicit ldur/stur,
as Visual Studio assembler can't automatically encode ldr/str as
ldur/stur when needed.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8256)

5 years agoARM assembly pack: make it Windows-friendly.
Andy Polyakov [Fri, 15 Feb 2019 08:44:39 +0000 (09:44 +0100)]
ARM assembly pack: make it Windows-friendly.

"Windows friendliness" means a) flipping .thumb and .text directives,
b) always generate Thumb-2 code when asked(*); c) Windows-specific
references to external OPENSSL_armcap_P.

(*) so far *some* modules were compiled as .code 32 even if Thumb-2
was targeted. It works at hardware level because processor can alternate
between the modes with no overhead. But clang --target=arm-windows's
builtin assembler just refuses to compile .code 32...

Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8252)

5 years agos390x assembly pack: fix formal interface bug in chacha module
Patrick Steuer [Fri, 15 Feb 2019 21:59:09 +0000 (22:59 +0100)]
s390x assembly pack: fix formal interface bug in chacha module

Signed-off-by: Patrick Steuer <patrick.steuer@de.ibm.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8257)

5 years agoAdd an OpenSSL library context
Richard Levitte [Wed, 6 Feb 2019 16:42:50 +0000 (17:42 +0100)]
Add an OpenSSL library context

The context builds on CRYPTO_EX_DATA, allowing it to be dynamically
extended with new data from the different parts of libcrypto.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8225)

5 years agoAdd CRYPTO_alloc_ex_data()
Richard Levitte [Fri, 8 Feb 2019 15:46:28 +0000 (16:46 +0100)]
Add CRYPTO_alloc_ex_data()

This allows allocation of items at indexes that were created after the
CRYPTO_EX_DATA variable was initialized, using the exact same method
that was used then.

Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8225)

5 years agoAdd EC_GROUP_get0_field
David Asraf [Thu, 7 Feb 2019 09:51:39 +0000 (11:51 +0200)]
Add  EC_GROUP_get0_field

New function to return internal pointer for field.

Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8195)

5 years agoMark generated functions unused (applies to safestack, lhash, sparse_array)
Richard Levitte [Fri, 15 Feb 2019 07:06:36 +0000 (08:06 +0100)]
Mark generated functions unused (applies to safestack, lhash, sparse_array)

safestack.h, lhash.h and sparse_array.h all define macros to generate
a full API for the containers as static inline functions.  This
potentially generates unused code, which some compilers may complain
about.

We therefore need to mark those generated functions as unused, so the
compiler knows that we know, and stops complaining about it.

Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/8246)

5 years agoAdd option to disable Extended Master Secret
Todd Short [Mon, 10 Jul 2017 17:28:35 +0000 (13:28 -0400)]
Add option to disable Extended Master Secret

Add SSL_OP64_NO_EXTENDED_MASTER_SECRET, that can be set on either
an SSL or an SSL_CTX. When processing a ClientHello, if this flag
is set, do not indicate that the EMS TLS extension was received in
either the ssl3 object or the SSL_SESSION.  Retain most of the
sanity checks between the previous and current session during
session resumption, but weaken the check when the current SSL
object is configured to not use EMS.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/3910)

5 years agoUse order not degree to calculate a buffer size in ecdsatest
Matt Caswell [Thu, 14 Feb 2019 12:21:20 +0000 (12:21 +0000)]
Use order not degree to calculate a buffer size in ecdsatest

Otherwise this can result in an incorrect calculation of the maximum
encoded integer length, meaning an insufficient buffer size is allocated.

Thanks to Billy Brumley for helping to track this down.

Fixes #8209

Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8237)

5 years agoFix no-stdio
Matt Caswell [Thu, 14 Feb 2019 15:22:59 +0000 (15:22 +0000)]
Fix no-stdio

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8238)

5 years agoConfigure: make --strict-warnings a regular user provided compiler option
Richard Levitte [Thu, 14 Feb 2019 15:26:40 +0000 (16:26 +0100)]
Configure: make --strict-warnings a regular user provided compiler option

This makes `--strict-warnings` into a compiler pseudo-option, i.e. it
gets treated the same way as any other compiler option given on the
configuration command line, but is retroactively replaced by actual
compiler warning options, depending on what compiler is used.

This makes it easier to see in what order options are given to the
compiler from the configuration command line, i.e. this:

    ./config -Wall --strict-warnings

would give the compiler flags in the same order as they're given,
i.e.:

    -Wall -Werror -Wno-whatever ...

instead of what we got previously:

    -Werror -Wno-whatever ... -Wall

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8239)

5 years agoFix -verify_return_error in s_client
Matt Caswell [Thu, 24 Jan 2019 12:21:39 +0000 (12:21 +0000)]
Fix -verify_return_error in s_client

The "verify_return_error" option in s_client is documented as:

 Return verification errors instead of continuing. This will typically
 abort the handshake with a fatal error.

In practice this option was ignored unless also accompanied with the
"-verify" option. It's unclear what the original intention was. One fix
could have been to change the documentation to match the actual behaviour.
However it seems unecessarily complex and unexpected that you should need
to have both options. Instead the fix implemented here is make the option
match the documentation so that "-verify" is not also required.

Note that s_server has a similar option where "-verify" (or "-Verify") is
still required. This makes more sense because those options additionally
request a certificate from the client. Without a certificate there is no
possibility of a verification failing, and so "-verify_return_error" doing
nothing seems ok.

Fixes #8079

Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
(Merged from https://github.com/openssl/openssl/pull/8080)

5 years agoDon't signal SSL_CB_HANDSHAKE_START for TLSv1.3 post-handshake messages
Matt Caswell [Sun, 27 Jan 2019 11:00:16 +0000 (11:00 +0000)]
Don't signal SSL_CB_HANDSHAKE_START for TLSv1.3 post-handshake messages

The original 1.1.1 design was to use SSL_CB_HANDSHAKE_START and
SSL_CB_HANDSHAKE_DONE to signal start/end of a post-handshake message
exchange in TLSv1.3. Unfortunately experience has shown that this confuses
some applications who mistake it for a TLSv1.2 renegotiation. This means
that KeyUpdate messages are not handled properly.

This commit removes the use of SSL_CB_HANDSHAKE_START and
SSL_CB_HANDSHAKE_DONE to signal the start/end of a post-handshake
message exchange. Individual post-handshake messages are still signalled in
the normal way.

This is a potentially breaking change if there are any applications already
written that expect to see these TLSv1.3 events. However, without it,
KeyUpdate is not currently usable for many applications.

Fixes #8069

Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/8096)

5 years agoIgnore cipher suites when setting cipher list
Sam Roberts [Mon, 26 Nov 2018 21:58:52 +0000 (13:58 -0800)]
Ignore cipher suites when setting cipher list

set_cipher_list() sets TLSv1.2 (and below) ciphers, and its success or
failure should not depend on whether set_ciphersuites() has been used to
setup TLSv1.3 ciphers.

Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7759)

5 years agoConfigure: stop forcing use of DEFINE macros in headers
Richard Levitte [Thu, 14 Feb 2019 08:25:40 +0000 (09:25 +0100)]
Configure: stop forcing use of DEFINE macros in headers

There are times when one might want to use something like
DEFINE_STACK_OF in a .c file, because it defines a stack for a type
defined in that .c file.  Unfortunately, when configuring with
`--strict-warnings`, clang aggressively warn about unused functions in
such cases, which forces the use of such DEFINE macros to header
files.

We therefore disable this warning from the `--strict-warnings`
definition for clang.

(note for the curious: `-Wunused-function` is enabled via `-Wall`)

Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/8234)