For backward compatibility with Mbed TLS <=3.5.x, applications must be able
to make a TLS connection with a peer that supports both TLS 1.2 and TLS 1.3,
regardless of whether they call psa_crypto_init(). Since Mbed TLS 3.6.0,
we enable TLS 1.3 in the default configuration, so we must take care of
calling psa_crypto_init() if needed. This is a change from TLS 1.3 in
previous versions, where enabling MBEDTLS_SSL_PROTO_TLS1_3 was a user
choice and could have additional requirement.
This commit changes our test programs to validate that the library
does not have the compatibility-breaking requirement.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Previously the client was only printing them on handshake success, and
the server was printing them on success and some but not all failures.
This makes ssl-opt.sh more consistent as we can always check for the
presence of the expected message in the output, regardless of whether
the failure is hard or soft.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
If an IP address shares multiple domain names with different SSL
certificates and makes a GET request without the remote server name
(host), it will fail with a 421 Misdirect Request.
Signed-off-by: Javier Tia <javier.tia@linaro.org>
This is replaced with: mbedtls_pk_get_psa_attributes() +
mbedtls_pk_import_into_psa() + mbedtls_pk_setup_opaque().
Signed-off-by: Valerio Setti <valerio.setti@nordicsemi.no>
No need to define specific early data,
the idea is rather to just send the
usual request data as early data
instead of standard application data.
Signed-off-by: Ronald Cron <ronald.cron@arm.com>
We have two options for early data.
early_data to indicate early data enable or not.
early_data_file to provide path file to read early data from
Signed-off-by: Xiaokang Qian <xiaokang.qian@arm.com>
Rename mbedtls_ssl_tls13_conf_early_data as
mbedtls_ssl_conf_early_data since in the future this may not be
specific to TLS 1.3.
Signed-off-by: Yanray Wang <yanray.wang@arm.com>
mbedtls_ssl_session_save() always outputs the output length, even on error.
Here, we're only calling it to get the needed output length, so it's ok to
ignore the return value. Convey this to linters.
Signed-off-by: Gilles Peskine <Gilles.Peskine@arm.com>
Every now and then, I see of these programs failing with a super-long
usage message that gives no clue as to what went wrong. (Recently it
happened with a test case in ssl-opt.sh with a fairly long command line
that was entirely correct, except some options were not valid in this
config - the test should have been skipped but wasn't due to some other
bug. It took me longer to figure out than it should have, and could have
if the program had simply reported which param was not recognized.)
Also, have an explicit "help" command, separate "help_ciphersuites", and
have default usage message that's not multiple screens long.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
All options have reasonable default so the programs don't need arguments
to do something useful.
It is widely accepted for programs that can work without arguments need
not insist on the user passing arguments, see 'ls', 'wc', 'sort', 'more'
and any number of POSIX utilities that all work without arguments.
It is also the historical behaviour of those programs, and something
relied one by at least a few team members.
Signed-off-by: Manuel Pégourié-Gonnard <manuel.pegourie-gonnard@arm.com>
