There's too high a chance that the openssl app and perl get different
messages for some error numbers.
[extended tests]
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7699)
(cherry picked from commit 0777de15ff)
This ensures we collected them properly and and as completely as can
be tested safely.
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7681)
(cherry picked from commit 4b801fdcf4)
When libssl and libcrypto are compiled on Linux with "-rpath", but
not "--enable-new-dtags", the RPATH takes precedence over
LD_LIBRARY_PATH, and we end up running with the wrong libraries.
This is resolved by using full (or at least relative, rather than
just the filename to be found on LD_LIBRARY_PATH) paths to the
shared objects.
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7626)
(cherry picked from commit 1828939974)
SSL_get_signature_nid() -- local signature algorithm
SSL_get_signature_type_nid() -- local signature algorithm key type
SSL_get_peer_tmp_key() -- Peer key-exchange public key
SSL_get_tmp_key -- local key exchange public key
Aliased pre-existing SSL_get_server_tmp_key(), which was formerly
just for clients, to SSL_get_peer_tmp_key(). Changed internal
calls to use the new name.
Reviewed-by: Matt Caswell <matt@openssl.org>
Setting the SipHash hash size and setting its key is done with two
independent functions... and yet, the internals depend on both.
Unfortunately, the function to change the size wasn't adapted for the
possibility that the key was set first, with a different hash size.
This changes the hash setting function to fix the internal values
(which is easy, fortunately) according to the hash size.
evpmac.txt value for digestsize:8 is also corrected.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7613)
(cherry picked from commit 425036130d)
pkey_test_ctrl() was designed for parsing values, not for using in
test runs. Relying on its returned value when it returned 1 even for
control errors made it particularly useless for mac_test_run().
Here, it gets replaced with a MAC specific control function, that
parses values the same way but is designed for use in a _run() rather
than a _parse() function.
This uncovers a SipHash test with an invalid control that wasn't
caught properly. After all, that stanza is supposed to test that
invalid control values do generate an error. Now we catch that.
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7500)
(cherry picked from commit ce5d64c79c)
Commit 5b4cb385c1 (#7382) introduced a bug which had the effect
that RAND_add()/RAND_seed() failed for buffer sizes less than
32 bytes. The reason was that now the added random data was used
exlusively as entropy source for reseeding. When the random input
was too short or contained not enough entropy, the DRBG failed
without querying the available entropy sources.
This commit makes drbg_add() act smarter: it checks the entropy
requirements explicitely. If the random input fails this check,
it won't be added as entropy input, but only as additional data.
More precisely, the behaviour depends on whether an os entropy
source was configured (which is the default on most os):
- If an os entropy source is avaible then we declare the buffer
content as additional data by setting randomness to zero and
trigger a regular reseeding.
- If no os entropy source is available, a reseeding will fail
inevitably. So drbg_add() uses a trick to mix the buffer contents
into the DRBG state without forcing a reseeding: it generates a
dummy random byte, using the buffer content as additional data.
Related-to: #7449
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7456)
(cherry picked from commit 8817215d5c)
In commit 8bf3665196 some renamings andd typo fixes were made
while adding back the DRBG-HMAC and DRBG-HASH implementation.
Since the commit could not be backported, a lot of unnecessary
differences between master and 1.1.1 were introduced.
These differences result in tiresome merge conflicts when
cherry-picking. To minimize these merge-conflicts, this patch
ports all 'non-feature' changes of commit 8bf3665196
(e.g., renamings of private variables, fixes of typographical
errors, comment changes) manually back to 1.1.1.
The commits a83dc59afa (#7399) and 8817215d5c (#7456)
failed to cherry-pick previously to 1.1.1, with this patch
they both cherry-pick without conflicts.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
(Merged from https://github.com/openssl/openssl/pull/7505)
When building shared libraries on Windows, we had a clash between
'libcrypto.lib' the static routine library and 'libcrypto.lib' the
import library.
We now change it so the static versions of our libraries get '_static'
appended to their names. These will never get installed, but can
still be used for our internal purposes, such as internal tests.
When building non-shared, the renaming mechanism doesn't come into
play. In that case, the static libraries 'libcrypto.lib' and
'libssl.lib' are installed, just as always.
Fixes#7492
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7496)
(cherry picked from commit b3023ced6b)
In pull request #4328 the seeding of the DRBG via RAND_add()/RAND_seed()
was implemented by buffering the data in a random pool where it is
picked up later by the rand_drbg_get_entropy() callback. This buffer
was limited to the size of 4096 bytes.
When a larger input was added via RAND_add() or RAND_seed() to the DRBG,
the reseeding failed, but the error returned by the DRBG was ignored
by the two calling functions, which both don't return an error code.
As a consequence, the data provided by the application was effectively
ignored.
This commit fixes the problem by a more efficient implementation which
does not copy the data in memory and by raising the buffer the size limit
to INT32_MAX (2 gigabytes). This is less than the NIST limit of 2^35 bits
but it was chosen intentionally to avoid platform dependent problems
like integer sizes and/or signed/unsigned conversion.
Additionally, the DRBG is now less permissive on errors: In addition to
pushing a message to the openssl error stack, it enters the error state,
which forces a reinstantiation on next call.
Thanks go to Dr. Falko Strenzke for reporting this issue to the
openssl-security mailing list. After internal discussion the issue
has been categorized as not being security relevant, because the DRBG
reseeds automatically and is fully functional even without additional
randomness provided by the application.
Fixes#7381
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7382)
(cherry picked from commit 3064b55134)
Hash can be longer than EC group degree and it will be truncated.
CLA: trivial
Reviewed-by: Paul Yang <yang.yang@baishancloud.com>
Reviewed-by: Paul Dale <paul.dale@oracle.com>
(Merged from https://github.com/openssl/openssl/pull/7329)
(cherry picked from commit b770a80f6d)
Check that different return values passed to the BIO callback are
correctly handled.
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7344)
(cherry picked from commit 52d78cc5eb)
This commit reuses a variable instead of reevaluating the expression
and updates an outdated comment in the EVP test.
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/7242)
(cherry picked from commit 523fcfb4c0)
It's actually not a real issue but caused by the absence of the default case
which does not occur in reality but which makes coverity see a code path where
pkey remains unassigned.
Reported by Coverity Scan (CID 1423323)
[extended tests]
Reviewed-by: Nicola Tuveri <nic.tuv@gmail.com>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7158)
This was originally part of SipHash_Init. However, there are cases
where there isn't any key material to initialize from when setting the
hash size, and we do allow doing so with a EVP_PKEY control. The
solution is to provide a separate hash_size setter and to use it in
the corresponding EVP_PKEY_METHOD.
Fixes#7143
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7145)
PR #3783 introduce coded to reset the server side SNI state in
SSL_do_handshake() to ensure any erroneous config time SNI changes are
cleared. Unfortunately SSL_do_handshake() can be called mid-handshake
multiple times so this is the wrong place to do this and can mean that
any SNI data is cleared later on in the handshake too.
Therefore move the code to a more appropriate place.
Fixes#7014
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/7149)
Even though we already sent close_notify the server may not have recieved
it yet and could issue a CertificateRequest to us. Since we've already
sent close_notify we can't send any reasonable response so we just ignore
it.
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7114)
This test case is originally submitted in #6757, by Jack Lloyd. The test
case has been modified to use the a different method to set the ID when
computing the Z hash of SM2 signature.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Paul Yang <yang.yang@baishancloud.com>
(Merged from https://github.com/openssl/openssl/pull/7113)
zero-length ID is allowed, but it's not allowed to skip the ID.
Fixes: #6534
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7113)
Check that we use an RSA certificate if an RSA key exchange ciphersuite
is being used and we have both RSA and RSA-PSS certificates configured.
Reviewed-by: Paul Dale <paul.dale@oracle.com>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7099)
Add a test to check that we create the correct number of tickets after a
TLSv1.3 PSK.
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7097)
Added NIST test cases for these two as well.
Additionally deprecate the public definiton of HMAC_MAX_MD_CBLOCK in 1.2.0.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6972)
Add one more unit test case
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Paul Yang <yang.yang@baishancloud.com>
(Merged from https://github.com/openssl/openssl/pull/6892)
The shared libraries are now stored as members of archives, as it is usual
on AIX. To correctly address this the custom dladdr()-implementation as
well as the dlfcn_load() routine need to be able to cope with such a
construct: libname.a(libname.so).
Signed-off-by: Matthias Kraft <Matthias.Kraft@softwareag.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6872)
We recently turned on the TLSv1.3 downgrade sentinels by default.
Unfortunately we are using a very old version of the BoringSSL test
runner which uses an old draft implementation of TLSv1.3 that also
uses the downgrade sentinels by default. The two implementations do
not play well together and were causing spurious test failures. Until
such time as we update the BoringSSL test runner we disable the failing
tests:
SendFallbackSCSV
In this test the client is OpenSSL and the server is the boring test runner.
The client and server fail to negotiate TLSv1.3 because the test runner is
using an old draft TLSv1.3 version. The server does however add the
TLSv1.3->TLSv1.2 downgrade sentinel in the ServerHello random. Since we
recently turned on checking of the downgrade sentinels on the client side
this causes the connection to fail.
VersionNegotiationExtension-TLS11
In this test the test runner is the client and OpenSSL is the server. The
test modifies the supported_versions extension sent by the client to only
include TLSv1.1 (and some other spurious versions), even though the client
does actually support TLSv1.2. The server successfully selects TLSv1.1, but
adds the TLSv1.3->TLSv1.1 downgrade sentinel. This behaviour was recently
switched on by default. The test runner then checks the downgrade sentinel
and aborts the connection because it knows that it really supports TLSv1.2.
VersionNegotiationExtension-TLS1
VersionNegotiationExtension-SSL3
The same as VersionNegotiationExtension-TLS11 but for TLSv1 and SSLv3.
ConflictingVersionNegotiation
In this test the client is the test runner, and OpenSSL is the server. The
client offers TLSv1.2 in ClientHello.version, but also adds a
supported_versions extension that only offers TLSv1.1. The
supported_versions extension takes precedence and the server (correctly)
selects TLSv1.1. However it also adds the TLSv1.3->TLSv1.1 downgrade
sentinel. On the client side it knows it actually offered TLSv1.2 and so the
downgrade sentinel check fails.
[extended tests]
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7013)
This commit destroys the free list pointers which would otherwise be
present in the returned memory blocks. This in turn helps prevent
information leakage from the secure memory area.
Note: CRYPTO_secure_malloc is not guaranteed to return zeroed memory:
before the secure memory system is initialised or if it isn't implemented.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/7011)
The EFD database does not state that the "ladd-2002-it-3" algorithm
assumes X1 != 0.
Consequently the current implementation, based on it, fails to compute
correctly if the affine x coordinate of the scalar multiplication input
point is 0.
We replace this implementation using the alternative algorithm based on
Eq. (9) and (10) from the same paper, which being derived from the
additive relation of (6) does not incur in this problem, but costs one
extra field multiplication.
The EFD entry for this algorithm is at
https://hyperelliptic.org/EFD/g1p/auto-shortw-xz.html#ladder-ladd-2002-it-4
and the code to implement it was generated with tooling.
Regression tests add one positive test for each named curve that has
such a point. The `SharedSecret` was generated independently from the
OpenSSL codebase with sage.
This bug was originally reported by Dmitry Belyavsky on the
openssl-users maling list:
https://mta.openssl.org/pipermail/openssl-users/2018-August/008540.html
Co-authored-by: Billy Brumley <bbrumley@gmail.com>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/7000)
We already have SSL_set_post_handshake_auth(). This just adds the SSL_CTX
equivalent.
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6938)
Having post handshake auth automatically switched on breaks some
applications written for TLSv1.2. This changes things so that an explicit
function call is required for a client to indicate support for
post-handshake auth.
Fixes#6933.
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6938)
spurious output when checking for error conditions.
Reviewed-by: Tim Hudson <tjh@openssl.org>
Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/6899)
The TLSv1.4 tolerance test wasn't testing what we thought it was.
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6741)
Test that a server can handle an unecrypted alert when normally the next
message is encrypted.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6887)
In 38eca7fed0 a new check for the pem_str member of the entries of the
ASN1 method table was introduced. Because the test condition was split
into two TEST_true(...) conditions, the test outputs error diagnostics
for all entries which have pem_str != NULL. This commit joins the two
test conditions into a single condition.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6888)
The CRYPTO_memcmp test isn't testing the test framework.
It would seem to better belong in the sanity tests.
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6878)
Ensure that the certificate required alert actually gets sent (and doesn't
get translated into handshake failure in TLSv1.3).
Ensure that proper reason codes are given for the new TLSv1.3 alerts.
Remove an out of date macro for TLS13_AD_END_OF_EARLY_DATA. This is a left
over from an earlier TLSv1.3 draft that is no longer used.
Fixes#6804
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6809)
To start with, actually set an SNI callback (copied from bssl_shim); we
weren't actually testing much otherwise (and just happened to have been
passing due to buggy libssl behavior prior to
commit 1c4aa31d79).
Also use proper C++ code for handling C strings -- when a C API
(SSL_get_servername()) returns NULL instead of a string, special-case
that instead of blindly trying to compare NULL against a std::string,
and perform the comparsion using the std::string operators instead of
falling back to pointer comparison.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6792)
The spec says that a client MUST set legacy_version to TLSv1.2, and
requires servers to verify that it isn't SSLv3.
Fixes#6600
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6747)
The GOST ciphers are dynamically loaded via the GOST engine, so we must
be able to support that. The engine also uses DSA and CMS symbols, so we
skip the test on no-dsa or no-cms.
Reviewed-by: Tim Hudson <tjh@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6730)
and catch corner cases better and earlier
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6690)
Fix the NULL check lack in a different way that is more compatible with
non-NULL branch. Refer #6632
Also mark and pop the error stack instead of clearing all errors when something
goes awry in CONF_get_number.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6643)
The problematic case falls back to a NULL conf which returns the result
of getenv(2). If this returns NULL, everything was good. If this returns
a string an attempt to convert it to a number is made using the function
pointers from conf.
This fix uses the strtol(3) function instead, we don't have the
configuration settings and this behaves as the default would.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6632)
This also adds the ability to control this through s_server
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6469)
1. For every named curve, two "golden" keypair positive tests.
2. Also two "golden" stock ECDH positive tests.
3. For named curves with non-trivial cofactors, additionally two "golden"
ECC CDH positive tests.
4. For named curves with non-trivial cofactors, additionally two negative
tests.
There is some overlap with existing EVP tests, especially for the NIST
curves (for example, positive testing ECC CDH KATs for NIST curves).
"Golden" here means all the values are independent from OpenSSL's ECC
code. I used sage to calculate them. What comes from OpenSSL is:
1. The OIDs (parsed by tooling)
2. The curve parameters (parsing ecparam output with tooling)
The values inside the PEMs (private keys, public keys) and shared keys
are from sage. The PEMs themselves are the output of asn1parse, with
input taken from sage.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6608)
Currently if you encounter application data while waiting for a
close_notify from the peer, and you have called SSL_shutdown() then
you will get a -1 return (fatal error) and SSL_ERROR_SYSCALL from
SSL_get_error(). This isn't accurate (it should be SSL_ERROR_SSL) and
isn't persistent (you can call SSL_shutdown() again and it might then work).
We change this into a proper fatal error that is persistent.
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/6340)
In the case where we are shutdown for writing and awaiting a close_notify
back from a subsequent SSL_shutdown() call we skip over handshake data
that is received. This should not be treated as an error - instead it
should be signalled with SSL_ERROR_WANT_READ.
Reviewed-by: Bernd Edlinger <bernd.edlinger@hotmail.de>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/6340)
Implement support for stateful TLSv1.3 tickets, and use them if
SSL_OP_NO_TICKET is set.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6563)
Fix prototype warnings triggered by -Wstrict-prototypes when configuring
with `enable-ec_nistp_64_gcc_128`
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
Reviewed-by: Matthias St. Pierre <Matthias.St.Pierre@ncp-e.com>
(Merged from https://github.com/openssl/openssl/pull/6556)
(introduced by commit 9186016582, which added -Wstrict-prototypes)
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Kurt Roeckx <kurt@roeckx.be>
(Merged from https://github.com/openssl/openssl/pull/6555)
Use EVP_PKEY_set_alias_type to access
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Matt Caswell <matt@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6443)
We generate the secrets based on the nonce immediately so there is no
need to keep the nonce.
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6415)
Under a mingw shell, the command line path conversion either mangles
file: URIs to something useless (file;C:\...) or not at all (which
can't be opened by the Windows C RTL unless we're really lucky), so we
simply skip testing them in that environment.
Fixes#6369
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6376)
TlsGetValue clears the last error even on success, so that callers may
distinguish it successfully returning NULL or failing. This error-mangling
behavior interferes with the caller's use of GetLastError. In particular
SSL_get_error queries the error queue to determine whether the caller should
look at the OS's errors. To avoid destroying state, save and restore the
Windows error.
Fixes#6299.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6316)
Per SEC 1, the curve coefficients must be padded up to size. See C.2's
definition of Curve, C.1's definition of FieldElement, and 2.3.5's definition
of how to encode the field elements in http://www.secg.org/sec1-v2.pdf.
This comes up for P-521, where b needs a leading zero.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6314)
Don't apply DNS name constraints to the subject CN when there's a
least one DNS-ID subjectAlternativeName.
Don't apply DNS name constraints to subject CN's that are sufficiently
unlike DNS names. Checked name must have at least two labels, with
all labels non-empty, no trailing '.' and all hyphens must be
internal in each label. In addition to the usual LDH characters,
we also allow "_", since some sites use these for hostnames despite
all the standards.
Reviewed-by: Matt Caswell <matt@openssl.org>
Reviewed-by: Tim Hudson <tjh@openssl.org>
Because TLS 1.3 sends more non-application data records some clients run
into problems because they don't expect SSL_read() to return and set
SSL_ERROR_WANT_READ after processing it.
This can cause problems for clients that use blocking I/O and use
select() to see if data is available. It can be cleared using
SSL_CTX_clear_mode().
Reviewed-by: Matt Caswell <matt@openssl.org>
GH: #6260
In TLSv1.2 and below we always cache new sessions by default on the server
side in the internal cache (even when we're using session tickets). This is
in order to support resumption from a session id.
In TLSv1.3 there is no session id. It is only possible to resume using the
ticket. Therefore, in the default case, there is no point in caching the
session in the internal store.
There is still a reason to call the external cache new session callback
because applications may be using the callbacks just to know about when
sessions are created (and not necessarily implementing a full cache). If
the application also implements the remove session callback then we are
forced to also store it in the internal cache so that we can create
timeout events. Otherwise the external cache could just fill up
indefinitely.
This mostly addresses the issue described in #5628. That issue also proposes
having an option to not create full stateless tickets when using the
internal cache. That aspect hasn't been addressed yet.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/6293)
Don't call the decrypt ticket callback if we've already encountered a
fatal error. Do call it if we have an empty ticket present.
Change the return code to have 5 distinct returns codes and separate it
from the input status value.
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6198)
The recent change in behaviour where you do not get a NewSessionTicket
message sent if you established the connection using a PSK caused a mem
leak to be triggered in sslapitest. It was actually a latent bug and we
were just lucky we never hit it before. The problem is due to complexity
with the way PSK sessions were set up in the early_data tests. PSK session
reference counting was handled differently to normal session reference
counting. This meant there were lots of special cases in the code where
we don't free a session if it is a PSK. It makes things easier if we just
handle PSK reference counts in the same way as other session reference
counts, and then we can remove all of the special case code.
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6198)
The return value from the ticket_key callback was not properly handled in
TLSv1.3, so that a ticket was *always* renewed even if the callback
requested that it should not be.
Also the ticket decrypt callback was not being called at all in TLSv1.3.
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6198)
If using an old style PSK callback and no certificate is configured for
the server, we should prefer ciphersuites based on SHA-256, because that
is the default hash for those callbacks as specified in the TLSv1.3 spec.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6215)
In certain circumstances in the DTLS code we have to drop a record (e.g. if
it is a stale retransmit). We then have to move on to try and read the next
record. Some applications using blocking sockets (e.g. s_server/s_client
will hang if there isn't actually any data to be read from the socket yet).
Others can tolerate this. Therefore SSL_read()/SSL_write() can sometimes
return SSL_ERROR_WANT_READ/SSL_ERROR_WANT_WRITE even when using blocking
sockets. Applications can use the mode SSL_MODE_AUTO_RETRY, to switch this
behaviour off so that we never return unless we have read the data we
wanted to.
Commit ad96225285 fixed a DTLS problem where we always retried even if
SSL_MODE_AUTO_RETRY was not set. However that fix caused the Boring
ossl_shim to fail in some tests because it was relying on the previous
(buggy) behaviour. The ossl_shim should be set into SSL_MODE_AUTO_RETRY if
it is not operating asynchronously to avoid this problem.
[extended tests]
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6216)
Drop a record from a handshake and check that we can still complete the
handshake. Repeat for all records in the handshake.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6170)
