Richard Levitte [Fri, 11 Nov 2016 09:23:26 +0000 (10:23 +0100)]
Fix the effect of no-dso in crypto/init.c
When configured no-dso, there are no DSO_{whatever} macros defined.
Therefore, before checking those, you have to check if OPENSSL_NO_DSO
is defined.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1902)
Richard Levitte [Thu, 10 Nov 2016 21:03:28 +0000 (22:03 +0100)]
Small fixups of util/process_docs.pl
- the pod path hadn't been changed with the directory layout change
- apparently, pod2html doesn't add ".html" at the end of links, making
them useless, so we need to fix that
With thanks for the report to Michel <michel.sales@free.fr>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1896)
Matt Caswell [Thu, 10 Nov 2016 15:35:42 +0000 (15:35 +0000)]
Fix test_sslcorrupt when using TLSv1.3
The test loops through all the ciphers, attempting to test each one in turn.
However version negotiation happens before cipher selection, so with TLSv1.3
switched on if we use a non-TLSv1.3 compatible cipher suite we get "no
share cipher".
Reviewed-by: Rich Salz <rsalz@openssl.org>
Richard Levitte [Thu, 10 Nov 2016 15:21:54 +0000 (16:21 +0100)]
Revert "Move algorithm specific ppccap code from crypto/ppccap.c"
Now that we can link specifically with static libraries, the immediate
need to split ppccap.c (and eventually other *cap.c files) is no more.
This reverts commit
e3fb4d3d52e188b83ccb8506aa2f16cb686f4d6c.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Wed, 9 Nov 2016 14:51:06 +0000 (14:51 +0000)]
Rename the Elliptic Curves extension to supported_groups
This is a skin deep change, which simply renames most places where we talk
about curves in a TLS context to groups. This is because TLS1.3 has renamed
the extension, and it can now include DH groups too. We still only support
curves, but this rename should pave the way for a future extension for DH
groups.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Richard Levitte [Wed, 9 Nov 2016 19:09:03 +0000 (20:09 +0100)]
Link internal tests with static OpenSSL libraries when needed
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1889)
Richard Levitte [Thu, 10 Nov 2016 00:00:17 +0000 (01:00 +0100)]
Document how to force linking with static libraries
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1889)
Richard Levitte [Wed, 9 Nov 2016 19:01:51 +0000 (20:01 +0100)]
Building: make it possible to force linking with static OpenSSL libs
Very simply, support having the .a extension to denote depending on
static libraries. Note that this is not supported on native Windows
when building shared libraries, as there is not static library then,
just an import library with the same name.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1889)
Richard Levitte [Thu, 10 Nov 2016 00:33:54 +0000 (01:33 +0100)]
Fix no-ct in test/ct_test.c
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1890)
Richard Levitte [Wed, 9 Nov 2016 23:41:53 +0000 (00:41 +0100)]
descrip.mms.tmpl: Simplify fiddling of logical names record and statem
Instead of enumerating exactly those files in test/ that include
../ssl/ssl_locl.h, assume they all do.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1891)
Matt Caswell [Thu, 10 Nov 2016 11:49:06 +0000 (11:49 +0000)]
Update CHANGES and NEWS
Reviewed-by: Richard Levitte <levitte@openssl.org>
Matt Caswell [Thu, 10 Nov 2016 11:27:07 +0000 (11:27 +0000)]
Fix the no-tls option
Reviewed-by: Richard Levitte <levitte@openssl.org>
Richard Levitte [Thu, 10 Nov 2016 00:49:47 +0000 (01:49 +0100)]
Fix no-cms (CVE-2016-7053)
Reviewed-by: Matt Caswell <matt@openssl.org>
Andy Polyakov [Tue, 1 Nov 2016 21:06:42 +0000 (22:06 +0100)]
test/evptests.txt: add negative tests for AEAD ciphers.
This is done by taking one vector, "corrupting" last bit of the
tag value and verifying that decrypt fails.
Reviewed-by: Emilia Käsper <emilia@openssl.org>
Andy Polyakov [Mon, 31 Oct 2016 20:50:26 +0000 (21:50 +0100)]
test: add TLS application data corruption test.
Reviewed-by: Emilia Käsper <emilia@openssl.org>
Dr. Stephen Henson [Fri, 14 Oct 2016 11:02:12 +0000 (12:02 +0100)]
add test for CVE-2016-7053
Reviewed-by: Richard Levitte <levitte@openssl.org>
Dr. Stephen Henson [Fri, 14 Oct 2016 10:51:43 +0000 (11:51 +0100)]
Don't set choice selector on parse failure.
Don't set choice selector on parse failure: this can pass unexpected
values to the choice callback. Instead free up partial structure
directly.
CVE-2016-7053
Thanks to Tyler Nighswander of ForAllSecure for reporting this issue.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Richard Levitte [Fri, 4 Nov 2016 13:21:46 +0000 (14:21 +0100)]
chacha20/poly1305: make sure to clear the buffer at correct position
The offset to the memory to clear was incorrect, causing a heap buffer
overflow.
CVE-2016-7054
Thanks to Robert Święcki for reporting this
Reviewed-by: Rich Salz <rsalz@openssl.org>
Andy Polyakov [Tue, 8 Nov 2016 19:25:09 +0000 (20:25 +0100)]
aes/asm/aesp8-ppc.pl: improve [backward] portability.
Some of stone-age assembler can't cope with r0 in address. It's actually
sensible thing to do, because r0 is shunted to 0 in address arithmetic
and by refusing r0 assembler effectively makes you understand that.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Andy Polyakov [Sun, 6 Nov 2016 17:33:17 +0000 (18:33 +0100)]
bn/asm/x86_64-mont.pl: fix for CVE-2016-7055 (Low severity).
Reviewed-by: Rich Salz <rsalz@openssl.org>
Andy Polyakov [Sun, 6 Nov 2016 17:31:14 +0000 (18:31 +0100)]
test/bntest.c: regression test for CVE-2016-7055.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Richard Levitte [Thu, 10 Nov 2016 09:03:37 +0000 (10:03 +0100)]
Fix the evp_test Ctrl keyword processing
Skip the test if the value after ":" is a disabled algorithm, rather
than failing it
Reviewed-by: Matt Caswell <matt@openssl.org>
Richard Levitte [Thu, 10 Nov 2016 01:08:22 +0000 (02:08 +0100)]
Fix no-dso (shlibloadtest)
Reviewed-by: Matt Caswell <matt@openssl.org>
Matt Caswell [Wed, 9 Nov 2016 14:43:05 +0000 (14:43 +0000)]
Address some supported_versions review comments
Added some TODOs, refactored a couple of things and added a SSL_IS_TLS13()
macro.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Wed, 9 Nov 2016 09:52:37 +0000 (09:52 +0000)]
Remove some redundant trace code
No need to have a supported versions table and a versions table. They
should be the same.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 13:48:07 +0000 (13:48 +0000)]
Fix some missing checks for TLS1_3_VERSION_DRAFT
There were a few places where we weren't checking to see if we were using
the draft TLS1.3 version or not.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 15:36:00 +0000 (15:36 +0000)]
Give the test with only TLS1.1 and TLS1.0 a better name
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Wed, 2 Nov 2016 11:33:20 +0000 (11:33 +0000)]
Add a TODO(TLS1.3) about renegotation
Renegotiation does not exist in TLS1.3, so we need to disable it at some
point.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Wed, 2 Nov 2016 09:09:02 +0000 (09:09 +0000)]
Add some more version tests
Send a TLS1.4 ClientHello with supported_versions and get TLS1.3
Send a TLS1.3 ClientHello without supported_versions and get TLS1.2
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Tue, 1 Nov 2016 00:37:23 +0000 (00:37 +0000)]
A style tweak based on feedback received
Replace a bare ";" with "continue;" for the body of a for loop.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 31 Oct 2016 18:12:33 +0000 (18:12 +0000)]
Look at the supported_versions extension even if the server <TLS1.3
If supported_versions is present it takes precedence.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Wed, 2 Nov 2016 11:15:08 +0000 (11:15 +0000)]
Ensure that vent->smeth != NULL before we call vent->smeth()
We can end up with a NULL SSL_METHOD function if a method has been
disabled. If that happens then we shouldn't call vent->smeth().
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 31 Oct 2016 17:05:20 +0000 (17:05 +0000)]
Address some review feedback comments for supported_versions
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Thu, 27 Oct 2016 17:32:19 +0000 (18:32 +0100)]
Add a test for the supported_versions extension
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Thu, 27 Oct 2016 15:30:36 +0000 (16:30 +0100)]
Update TLS1.3 draft version numbers for latest draft
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Sat, 22 Oct 2016 23:41:11 +0000 (00:41 +0100)]
Add server side support for supported_versions extension
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Fri, 21 Oct 2016 22:41:18 +0000 (23:41 +0100)]
Ensure that the -trace option can interpret the supported_versions extension
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Fri, 21 Oct 2016 22:40:52 +0000 (23:40 +0100)]
Add the ability to send the supported_versions extension
Reviewed-by: Rich Salz <rsalz@openssl.org>
EasySec [Sat, 5 Nov 2016 21:56:13 +0000 (22:56 +0100)]
When no SRP identity is found, no error was reported server side
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1859)
Emilia Kasper [Mon, 7 Nov 2016 15:53:15 +0000 (16:53 +0100)]
Add main() test methods to reduce test boilerplate.
Simple tests only need to implement register_tests().
Tests that need a custom main() should implement test_main(). This will
be wrapped in a main() that performs common setup/teardown (currently
crypto-mdebug).
Note that for normal development, enable-asan is usually
sufficient for detecting leaks, and more versatile.
enable-crypto-mdebug is stricter as it will also
insist that all static variables be freed. This is useful for debugging
library init/deinit; however, it also means that test_main() must free
everything it allocates.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
EasySec [Thu, 3 Nov 2016 14:42:11 +0000 (15:42 +0100)]
dtl_mtu_test doesn't follow BIO_* conventions and make Windows build fail
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Richard Levitte <levitte@openssl.org>
Matt Caswell [Tue, 8 Nov 2016 23:22:11 +0000 (23:22 +0000)]
Ensure the key and iv labels are declared as static
Fixes a travis failure
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 10:18:41 +0000 (10:18 +0000)]
Add a test for the TLS1.3 secret generation
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Tue, 8 Nov 2016 10:34:28 +0000 (10:34 +0000)]
Add support for TLS1.3 secret generation
Nothing is using this yet, it just adds the underlying functions necesary
for generating the TLS1.3 secrets.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Tue, 8 Nov 2016 10:33:35 +0000 (10:33 +0000)]
Add support for initialising WPACKETs from a static buffer
Normally WPACKETs will use a BUF_MEM which can grow as required. Sometimes
though that may be overkill for what is needed - a static buffer may be
sufficient. This adds that capability.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Tue, 8 Nov 2016 10:25:21 +0000 (10:25 +0000)]
Add some documentation for the new HKDF modes
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Tue, 8 Nov 2016 11:48:33 +0000 (11:48 +0000)]
Following the changes to HKDF to accept a mode, add some tests for this
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 10:16:57 +0000 (10:16 +0000)]
Update HKDF to support separte Extract and Expand steps
At the moment you can only do an HKDF Extract and Expand in one go. For
TLS1.3 we need to be able to do an Extract first, and the subsequently do
a number of Expand steps on the same PRK.
Reviewed-by: Rich Salz <rsalz@openssl.org>
FdaSilvaYY [Thu, 15 Sep 2016 19:42:53 +0000 (21:42 +0200)]
Simplify and clean X509_VERIFY_PARAM new/free code.
Split x509_verify_param_zero code to the right place
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Matt Caswell [Tue, 8 Nov 2016 13:43:12 +0000 (13:43 +0000)]
Make some CLIENTHELLO_MSG function arguments const
There were a few places where they could be declared const so this commit
does that.
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 15:15:06 +0000 (15:15 +0000)]
Update a comment
The name and type of the argument to ssl_check_for_safari() has changed.
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 15:13:04 +0000 (15:13 +0000)]
Swap back to using SSL3_RANDOM_SIZE instead of sizeof(clienthello.random)
The size if fixed by the protocol and won't change even if
sizeof(clienthello.random) does.
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 15:09:19 +0000 (15:09 +0000)]
Move setting the session_id_len until after we filled the session_id
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 15:07:56 +0000 (15:07 +0000)]
Load the sessionid directly in SSLv2 compat ClientHello
Don't use a sub-packet, just load it.
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 14:59:12 +0000 (14:59 +0000)]
Rename clienthello.version to clienthello.legacy_version
For consistency with the TLSv1.3 spec.
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Wed, 2 Nov 2016 14:14:23 +0000 (14:14 +0000)]
Fix some minor style issues
Add a blank line, take one away - due to feedback received during review.
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 31 Oct 2016 16:36:30 +0000 (16:36 +0000)]
Add some function documentation and update some existing comments
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 31 Oct 2016 14:52:22 +0000 (14:52 +0000)]
Fix make update following extensions refactor
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 31 Oct 2016 13:20:03 +0000 (13:20 +0000)]
Improve some comment documentation following the extensions refactor
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 31 Oct 2016 13:11:17 +0000 (13:11 +0000)]
Fix various style issues in the extension parsing refactor
Based on review feedback received.
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 31 Oct 2016 12:50:05 +0000 (12:50 +0000)]
Use an explicit name for the struct for definition of RAW_EXTENSION
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 31 Oct 2016 12:48:37 +0000 (12:48 +0000)]
Fix a memory leak in the ClientHello extension parsing
We should be freeing up the raw extension data after we've finished with it.
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 31 Oct 2016 12:47:20 +0000 (12:47 +0000)]
Handle compression methods correctly with SSLv2 compat ClientHello
In the case of an SSLv2 compat ClientHello we weren't setting up the
compression methods correctly, which could lead to uninit reads or crashes.
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Sat, 22 Oct 2016 16:24:37 +0000 (17:24 +0100)]
Refactor ClientHello processing so that extensions get parsed earlier
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
Richard Levitte [Tue, 8 Nov 2016 22:55:51 +0000 (23:55 +0100)]
Move algorithm specific ppccap code from crypto/ppccap.c
Having that code in one central object file turned out to cause
trouble when building test/modes_internal_test.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1883)
Richard Levitte [Tue, 8 Nov 2016 23:14:56 +0000 (00:14 +0100)]
Unix Makefile: Make sure to use $(PERL) when running ./Configure
For consistency, it's better to use the perl that was specified to
Configure last time it was called.
Use case:
perl v5.8.8 was first along $PATH, perl v5.22.2 was available and
specified as: PERL=/opt/local/bin/perl ./config. When make wanted to
reconfigure and called './Configure reconf', configuration broke down,
complaining about a perl that's too old.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1884)
FdaSilvaYY [Tue, 8 Nov 2016 18:22:09 +0000 (19:22 +0100)]
Missing BN_RECP_CTX field init.
BN_RECP_CTX_new direclty use bn_init to avoid twice memset calls
Reviewed-by: Kurt Roeckx <kurt@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1879)
Rich Salz [Tue, 8 Nov 2016 20:56:04 +0000 (15:56 -0500)]
Zero stack variable with DSA nonce
Thanks to Falko Strenzke for bringing this to our attention.
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1882)
Matt Caswell [Tue, 8 Nov 2016 13:52:30 +0000 (13:52 +0000)]
Fix zlib BIO_METHOD for latest BIO_METHOD structure changes
Reviewed-by: Richard Levitte <levitte@openssl.org>
Richard Levitte [Tue, 8 Nov 2016 09:17:20 +0000 (10:17 +0100)]
INSTALL: small typo
libssl, not libddl.
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1871)
FdaSilvaYY [Tue, 27 Sep 2016 21:36:37 +0000 (23:36 +0200)]
Allow null in X509_CRL_METHOD_free
and fix documentation.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1634)
Andrea Grandi [Thu, 3 Nov 2016 04:42:07 +0000 (04:42 +0000)]
Improve PRF documentation
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
GH: #1834
David Benjamin [Mon, 7 Nov 2016 00:12:47 +0000 (19:12 -0500)]
Improve RSA test coverage.
MD5/SHA1 and MDC-2 have special-case logic beyond the generic DigestInfo
wrapping. Test that each of these works, including hash and length
mismatches (both input and signature). Also add VerifyRecover tests. It
appears
5824cc298174d462c827cd090675e30fc03f0caf added support for
VerifyRecover, but forgot to add the test data.
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
GH: #1474
David Benjamin [Sat, 20 Aug 2016 19:48:56 +0000 (15:48 -0400)]
Make RSA_sign.pod less confusing.
PKCS #1 v2.0 is the name of a document which specifies an algorithm
RSASSA-PKCS1-v1_5, often referred to as "PKCS #1 v1.5" after an earlier
document which specified it. This gets further confusing because the
document PKCS #1 v2.1 specifies two signature algorithms,
RSASSA-PKCS1-v1_5 and RSASSA-PSS. RSA_sign implements RSASSA-PKCS1-v1_5.
Refer to the document using the RFC number which is easier to find
anyway, and refer to the algorithm by its name.
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
GH: #1474
David Benjamin [Sat, 20 Aug 2016 17:35:17 +0000 (13:35 -0400)]
Implement RSASSA-PKCS1-v1_5 as specified.
RFC 3447, section 8.2.2, steps 3 and 4 states that verifiers must encode
the DigestInfo struct and then compare the result against the public key
operation result. This implies that one and only one encoding is legal.
OpenSSL instead parses with crypto/asn1, then checks that the encoding
round-trips, and allows some variations for the parameter. Sufficient
laxness in this area can allow signature forgeries, as described in
https://www.imperialviolet.org/2014/09/26/pkcs1.html
Although there aren't known attacks against OpenSSL's current scheme,
this change makes OpenSSL implement the algorithm as specified. This
avoids the uncertainty and, more importantly, helps grow a healthy
ecosystem. Laxness beyond the spec, particularly in implementations
which enjoy wide use, risks harm to the ecosystem for all. A signature
producer which only tests against OpenSSL may not notice bugs and
accidentally become widely deployed. Thus implementations have a
responsibility to honor the specification as tightly as is practical.
In some cases, the damage is permanent and the spec deviation and
security risk becomes a tax all implementors must forever pay, but not
here. Both BoringSSL and Go successfully implemented and deployed
RSASSA-PKCS1-v1_5 as specified since their respective beginnings, so
this change should be compatible enough to pin down in future OpenSSL
releases.
See also https://tools.ietf.org/html/draft-thomson-postel-was-wrong-00
As a bonus, by not having to deal with sign/verify differences, this
version is also somewhat clearer. It also more consistently enforces
digest lengths in the verify_recover codepath. The NID_md5_sha1 codepath
wasn't quite doing this right.
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Rich Salz <rsalz@openssl.org>
GH: #1474
Matt Caswell [Thu, 27 Oct 2016 09:46:25 +0000 (10:46 +0100)]
Partial revert of "Fix client verify mode to check SSL_VERIFY_PEER"
This partially reverts commit
c636c1c47. It also tweaks the documentation
and comments in this area. On the client side the documented interface for
SSL_CTX_set_verify()/SSL_set_verify() is that setting the flag
SSL_VERIFY_PEER causes verfication of the server certificate to take place.
Previously what was implemented was that if *any* flag was set then
verification would take place. The above commit improved the semantics to
be as per the documented interface.
However, we have had a report of at least one application where an
application was incorrectly using the interface and used *only*
SSL_VERIFY_FAIL_IF_NO_PEER_CERT on the client side. In OpenSSL prior to
the above commit this still caused verification of the server certificate
to take place. After this commit the application silently failed to verify
the server certificate.
Ideally SSL_CTX_set_verify()/SSL_set_verify() could be modified to indicate
if invalid flags were being used. However these are void functions!
The simplest short term solution is to revert to the previous behaviour
which at least means we "fail closed" rather than "fail open".
Thanks to Cory Benfield for reporting this issue.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Emilia Kasper [Fri, 4 Nov 2016 15:06:12 +0000 (16:06 +0100)]
Simplify tests part 2
1) Remove some unnecessary fixtures
2) Add EXECUTE_TEST_NO_TEARDOWN shorthand when a fixture exists but has
no teardown.
3) Fix return values in ct_test.c (introduced by an earlier refactoring,
oops)
Note that for parameterized tests, the index (test vector) usually holds all the
customization, and there should be no need for a separate test
fixture. The CTS test is an exception: it demonstrates how to combine
customization with parameterization.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 14:26:41 +0000 (14:26 +0000)]
Add a test for the wrong version number in a record
Prior to TLS1.3 we check that the received record version number is correct.
In TLS1.3 we need to ignore the record version number. This adds a test to
make sure we do it correctly.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 13:49:18 +0000 (13:49 +0000)]
Ignore the record version in TLS1.3
The record layer version field must be ignored in TLSv1.3, so we remove the
check when using that version.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 14:44:38 +0000 (14:44 +0000)]
test_sslcbcpadding only makes sense <TLS1.3
We may get failures if we run it in TLS1.3, and it makes no sense anyway
so force TLS1.2
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Mon, 7 Nov 2016 13:44:56 +0000 (13:44 +0000)]
Correct the Id for the TLS1.3 ciphersuite
We have one TLS1.3 ciphersuite, but there is a typo in the id that should
be corrected.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Thu, 3 Nov 2016 13:21:28 +0000 (13:21 +0000)]
Always ensure that init_msg is initialised for a CCS
We read it later in grow_init_buf(). If CCS is the first thing received in
a flight, then it will use the init_msg from the last flight we received. If
the init_buf has been grown in the meantime then it will point to some
arbitrary other memory location. This is likely to result in grow_init_buf()
attempting to grow to some excessively large amount which is likely to
fail. In practice this should never happen because the only time we receive
a CCS as the first thing in a flight is in an abbreviated handshake. None
of the preceding messages from the server flight would be large enough to
trigger this.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Richard Levitte [Mon, 24 Oct 2016 13:11:29 +0000 (15:11 +0200)]
Windows: use default ZLIB1 unless --with-zlib-lib is set
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1772)
Richard Levitte [Mon, 24 Oct 2016 13:03:57 +0000 (15:03 +0200)]
Fix the LIBZ macro on VC config targets
If zlib-dynamic was given but not --with-zlib-lib, LIBZ was defined to
the empty string. Instead, give it the default "ZLIB1".
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1772)
Richard Levitte [Sun, 6 Nov 2016 17:35:01 +0000 (18:35 +0100)]
VMS: pretend to use -znodelete
VMS only unloads shared libraries at process rundown, so tell the
OpenSSL code so by pretending we linked with -znodelete.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1862)
Richard Levitte [Sat, 5 Nov 2016 10:38:29 +0000 (11:38 +0100)]
constant time test: include our internal/numbers.h rather than limits.h
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1856)
Richard Levitte [Fri, 4 Nov 2016 18:11:11 +0000 (19:11 +0100)]
VMS build file template: assign 'arch' to local symbol table
Since the local symbol table is looked up before the global symbol
table, 'arch' assigned in the local symbol table of the DCL where MMS
is called would be seen before the 'arch' defined in descrip.mms.
Assigning it to the local symbol table in descrip.mms removes that
issue.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1853)
Rich Salz [Fri, 4 Nov 2016 14:27:47 +0000 (10:27 -0400)]
Missed a mention of RT
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1849)
Richard Levitte [Fri, 4 Nov 2016 14:26:57 +0000 (15:26 +0100)]
Correct internal tests sources
The sources for internal tests were sometimes badly formed, assuming
perl variables such as $target{cpuid_asm_src} contains only one file
name. This change correctly massages all file names in such a
variable.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1850)
Richard Levitte [Fri, 4 Nov 2016 13:08:55 +0000 (14:08 +0100)]
VMS: update the list of files that need some extra treatment
This is related to a lack in path merging involding includes of includes
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1846)
Richard Levitte [Fri, 4 Nov 2016 13:08:25 +0000 (14:08 +0100)]
VMS: correct the logic around linking executables
The logic around avoiding MULDEF warnings was flawed. Simplifying it
makes it better.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1846)
Emilia Kasper [Thu, 3 Nov 2016 16:15:41 +0000 (17:15 +0100)]
Don't create fixtures for simple tests
The test fixtures are (meant to be) useful for sharing common
setup. Don't bother when we don't have any setup/teardown.
This only addresses simple tests. Parameterized tests (ADD_ALL_TESTS)
will be made more user-friendly in a follow-up.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Emilia Kasper [Thu, 3 Nov 2016 13:27:05 +0000 (14:27 +0100)]
testutil: always print errors on failure
Reviewed-by: Richard Levitte <levitte@openssl.org>
Richard Levitte [Thu, 3 Nov 2016 15:46:14 +0000 (16:46 +0100)]
Travis: add a strict build
Clang on Linux seems to catch things that we might miss otherwise.
Also, throw in 'no-deprecated' to make sure we test that as well.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1839)
Matt Caswell [Fri, 4 Nov 2016 10:26:57 +0000 (10:26 +0000)]
Fix a missed size_t variable declaration
pqueue_size() now returns a size_t, but the variable that gets returned
was still declared as an int.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Fri, 4 Nov 2016 10:25:03 +0000 (10:25 +0000)]
Fix some style issues from libssl size_tify review
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Wed, 26 Oct 2016 19:59:49 +0000 (20:59 +0100)]
Tweak the SSL_read()/SSL_write() text based on feedback received.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Wed, 26 Oct 2016 09:43:34 +0000 (10:43 +0100)]
Rename all "read" variables with "readbytes"
Travis is reporting one file at a time shadowed variable warnings where
"read" has been used. This attempts to go through all of libssl and replace
"read" with "readbytes" to fix all the problems in one go.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Tue, 25 Oct 2016 22:46:27 +0000 (23:46 +0100)]
Clarify the return values for SSL_read_ex()/SSL_write_ex()
Give more detail on what constitutes success/failure.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Matt Caswell [Tue, 25 Oct 2016 22:27:16 +0000 (23:27 +0100)]
Fix a shadowed variable declaration warning picked up by Travis
Rename "read" to "readbytes"
Reviewed-by: Rich Salz <rsalz@openssl.org>