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.
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.
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)
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)
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.
Matt Caswell [Wed, 19 Oct 2016 13:09:02 +0000 (14:09 +0100)]
Convert SSL BIO to use SSL_write_ex().
We also modify the SSL_get_error() function to handle the fact that with
SSL_write_ex() the error return is 0 not -1, and fix some bugs in the
SSL BIO reading.
Richard Levitte [Fri, 14 Oct 2016 14:52:50 +0000 (15:52 +0100)]
Convert 90-test_external.t to using "executable" rather than "system"
Use the newly added "executable" function rather than "system". Also filter
the output to add a prefix to every line so that the "ok" doesn't confuse
Test::More
Reviewed-by: Richard Levitte <levitte@openssl.org> Reviewed-by: Matt Caswell <matt@openssl.org>
Matt Caswell [Wed, 12 Oct 2016 14:43:46 +0000 (15:43 +0100)]
Add a shim config file
This just disables all tests that fail at the moment. Over time we will
want to go over these and figure out why they are failing (and fix them if
appropriate)
Reviewed-by: Richard Levitte <levitte@openssl.org>
Matt Caswell [Mon, 10 Oct 2016 11:24:00 +0000 (12:24 +0100)]
Add a test to call the BoringSSL test suite
This adds a test to the OpenSSL test suite to invoke the BoringSSL test
suite.
It assumes you have already compiled the ossl_shim (see previous commit).
It also assumes that you have an environment variable BORING_RUNNER_DIR
set up to point to the ssl/test/runner directory of a checkout of BoringSSL.
This has only been tested with a very old version of BoringSSL (from commit f277add6c) - since that was the last known checkout where the shim compiles
successfully. Even with that version of BoringSSL this test will fail. There
are lots of Boring tests that are failing for various reasons. Some might
be due to bugs in OpenSSL, some might be due to features that BoringSSL has
that OpenSSL doesn't, some are due to assumptions about the way BoringSSL
behaves that are not true for OpenSSL etc.
To get the verbose BoringSSL test output, run like this:
VERBOSE=1 BORING_RUNNER_DIR=/path/to/boring/ssl/test/runner make \
TESTS="test_external" test
Reviewed-by: Richard Levitte <levitte@openssl.org>
Matt Caswell [Mon, 10 Oct 2016 11:17:21 +0000 (12:17 +0100)]
Integrate BoringSSL shim
The BoringSSL test suite contains numerous tests which OpenSSL does not.
The BoringSSL test runner uses a shim to launch the library and execute the
tests. This is a version of the BoringSSL shim converted to compile against
OpenSSL instead.
This is primarily based on the work of David Benjamin from the BoringSSL
project who did most of the necessary conversion. It also includes a few
other tweaks for opacity changes etc.
This is based on a *very* old version of BoringSSL from commit f277add6c.
That was the last commit known to work with this patched shim. Later
versions may also work but lots of merge conflicts occur when trying to
bring it up to date.
At the moment this has not been integrated into the build system. There is
a very simple standalone makefile in the ossl_shim directory which should
be executed directly before tyring to use the shim.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Richard Levitte [Thu, 3 Nov 2016 17:48:23 +0000 (18:48 +0100)]
Don't assume to know the shared library extension
test/shlibloadtest.c assumes all Unix style platforms use .so as
shared library extension. This is not the case for Mac OS X, which
uses .dylib. Instead of this, have the test recipe find out the
extension from configuration data.
Reviewed-by: Matt Caswell <matt@openssl.org> Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1844)
Richard Levitte [Thu, 3 Nov 2016 16:08:10 +0000 (17:08 +0100)]
Enable memory debugging while testing
Pre 1.1.0, 'make test' would set the environment variable
OPENSSL_DEBUG_MEMORY to "on". This got lost when translating the old
build files to the new templates. This changes reintroduces that
variable.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/1840)