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)
When a server call SSL_write_early_data() to write to an unauthenticated
client the buffering BIO is still in place, so we should ensure we flush
the write.
Reviewed-by: Viktor Dukhovni <viktor@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6198)
This was preventing DTLS connections from being made from the command line.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6159)
The TLS code marks records as read when its finished using a record. The DTLS code did
not do that. However SSL_has_pending() relies on it. So we should make DTLS consistent.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6159)
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)
If we have no certificate and we are using "old style" PSKs then we will
always default to using SHA-256 for that PSK. However we may have selected
a ciphersuite that is not based on SHA-256. Therefore if we see that there
are no certificates and we have been configured for "old style" PSKs then
we should prefer SHA-256 based ciphersuites during the selection process.
Fixes#6197
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)
One is clang --strict-warnings and one gcc sanitizer extended test.
Sanitizer build is quite expensive, can take >30 mins and is commented
for occasions when there is reason to believe that PPC-specific problem
can be diagnosed with sanitizer.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6192)
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6070)
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6070)
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6070)
* EC_POINT_mul is now responsible for constant time point multiplication
(for single fixed or variable point multiplication, when the scalar is
in the range [0,group_order), so we need to strip the nonce padding
from ECDSA.
* Entry added to CHANGES
* Updated EC_POINT_mul documentation
- Integrate existing EC_POINT_mul and EC_POINTs_mul entries in the
manpage to reflect the shift in constant-time expectations when
performing a single fixed or variable point multiplication;
- Add documentation to ec_method_st to reflect the updated "contract"
between callers and implementations of ec_method_st.mul.
Reviewed-by: Richard Levitte <levitte@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6070)
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)
During a full handshake the server is the last one to "speak". The timer
should continue to run until we know that the client has received our last
flight (e.g. because we receive some application data).
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6170)
Previous tests only invoked CMS via the command line app. This test uses
the CMS API directly to do and encrypt and decrypt operation. This test
would have caught the memory leak fixed by the previous commit (when
building with enable-crypto-mdebug).
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6142)
The function CMS_RecipientInfo_set0_pkey() is a "set0" and therefore
memory management passes to OpenSSL. If the same function is called again
then we should ensure that any previous value that was set is freed first
before we set it again.
Fixes#5052
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6142)
deletion of *.exp files in krb5 sub-module.
Reviewed-by: Andy Polyakov <appro@openssl.org>
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6186)
Linking fails with "unrecognized option '--push-state--no-as-needed'",
which is beyond our control.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6185)
Apparently trusty image has newer clang, there is no need to pull
clang-3.9 packages. It's clang-5.0.0, installation is a bit quirky,
as it fails to compile for example strcmp(s,"-") without warning,
and complains about unreferred -I flags. But it's argued that benefits
of exercising newer sanitizer outweights the inconvenience of
additional -D__NO_STRING_INLINE and -Wno-unused-command-line-argument.
Also pull golang when actually needed.
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6185)
Earlier Apple Xcode compilers, e.g. one targeting Mac OS X 10.7, don't
support dependency generation and one still has to use makedepend. It's
unclear when it was fixed, but all clang-based Apple compilers seem to
support -M options.
Reviewed-by: Rich Salz <rsalz@openssl.org>
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6169)
config probe doesn't work in cross-compile scenarios or with clang.
In addition consolidate -Qunused-arguments handling.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6174)
When the input length is zero, just return zero early. Otherwise,
there's a small chance that memory allocation is engaged, fails and
returns -1, which is a bit confusing when nothing should be written.
Fixes#4782#4827
Reviewed-by: Ben Kaduk <kaduk@mit.edu>
(Merged from https://github.com/openssl/openssl/pull/6175)
Stop redefining structures that are already defined in system
headers. This also means we can stop setting the pointer size
globally, because the system structures will have the correct pointer
sizes either way. The only exception is passing the right pointer
size to a function.
Stop trying to twist things around with rand(), that's the job of the
DRBG that we feed.
Stop assuming the location of the JPI$_FINALEXC item, look it up
instead.
Signal an exception if the sys$getjpiw call fails (it means the item
list isn't set up right, so works as an assertion, but using VMS
methodology).
Reviewed-by: Andy Polyakov <appro@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6151)
Calculating BN_mod_inverse where n is 1 (or -1) doesn't make sense. We
should return an error in that case. Instead we were returning a valid
result with value 0.
Fixes#6004
Reviewed-by: Rich Salz <rsalz@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6119)
The max protocol version was only being set on the server side. It should
have been done on both the client and the server.
Reviewed-by: Richard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/6113)
