openssl/doc/apps/s_client.pod

500 lines
15 KiB
Plaintext
Raw Normal View History

=pod
=head1 NAME
s_client - SSL/TLS client program
=head1 SYNOPSIS
B<openssl> B<s_client>
[B<-help>]
[B<-connect host:port>]
[B<-proxy host:port>]
[B<-servername name>]
2000-01-15 06:43:11 +08:00
[B<-verify depth>]
2014-04-07 20:02:10 +08:00
[B<-verify_return_error>]
[B<-cert filename>]
[B<-certform DER|PEM>]
[B<-key filename>]
[B<-keyform DER|PEM>]
[B<-pass arg>]
[B<-CApath directory>]
[B<-CAfile filename>]
[B<-no-CAfile>]
[B<-no-CApath>]
[B<-dane_tlsa_domain domain>]
[B<-dane_tlsa_rrdata rrdata>]
[B<-attime timestamp>]
[B<-check_ss_sig>]
[B<-crl_check>]
[B<-crl_check_all>]
[B<-explicit_policy>]
[B<-extended_crl>]
[B<-ignore_critical>]
[B<-inhibit_any>]
[B<-inhibit_map>]
[B<-issuer_checks>]
[B<-partial_chain>]
[B<-policy arg>]
[B<-policy_check>]
[B<-policy_print>]
[B<-purpose purpose>]
[B<-suiteB_128>]
[B<-suiteB_128_only>]
[B<-suiteB_192>]
[B<-trusted_first>]
[B<-no_alt_chains>]
[B<-use_deltas>]
[B<-verify_depth num>]
[B<-verify_email email>]
[B<-verify_hostname hostname>]
[B<-verify_ip ip>]
[B<-verify_name name>]
[B<-x509_strict>]
[B<-reconnect>]
[B<-showcerts>]
[B<-debug>]
2001-11-10 10:12:09 +08:00
[B<-msg>]
[B<-nbio_test>]
[B<-state>]
[B<-nbio>]
[B<-crlf>]
[B<-ign_eof>]
[B<-no_ign_eof>]
[B<-quiet>]
[B<-ssl3>]
[B<-tls1>]
[B<-no_ssl3>]
[B<-no_tls1>]
[B<-no_tls1_1>]
[B<-no_tls1_2>]
[B<-fallback_scsv>]
[B<-async>]
[B<-bugs>]
[B<-comp>]
[B<-no_comp>]
[B<-cipher cipherlist>]
[B<-serverpref>]
[B<-starttls protocol>]
[B<-xmpphost hostname>]
[B<-engine id>]
[B<-tlsextdebug>]
[B<-no_ticket>]
[B<-sess_out filename>]
[B<-sess_in filename>]
[B<-rand file(s)>]
[B<-serverinfo types>]
[B<-status>]
[B<-nextprotoneg protocols>]
=head1 DESCRIPTION
The B<s_client> command implements a generic SSL/TLS client which connects
to a remote host using SSL/TLS. It is a I<very> useful diagnostic tool for
SSL servers.
=head1 OPTIONS
2012-11-20 00:07:53 +08:00
In addition to the options below the B<s_client> utility also supports the
common and client only options documented in the
in the L<SSL_CONF_cmd(3)|SSL_CONF_cmd(3)/SUPPORTED COMMAND LINE COMMANDS>
manual page.
2012-11-20 00:07:53 +08:00
=over 4
=item B<-help>
Print out a usage message.
=item B<-connect host:port>
This specifies the host and optional port to connect to. If not specified
then an attempt is made to connect to the local host on port 4433.
=item B<-proxy host:port>
When used with the B<-connect> flag, the program uses the host and port
specified with this flag and issues an HTTP CONNECT command to connect
to the desired server.
=item B<-servername name>
Set the TLS SNI (Server Name Indication) extension in the ClientHello message.
=item B<-cert certname>
The certificate to use, if one is requested by the server. The default is
not to use a certificate.
=item B<-certform format>
The certificate format to use: DER or PEM. PEM is the default.
=item B<-key keyfile>
The private key to use. If not specified then the certificate file will
be used.
=item B<-keyform format>
The private format to use: DER or PEM. PEM is the default.
=item B<-pass arg>
the private key password source. For more information about the format of B<arg>
see the B<PASS PHRASE ARGUMENTS> section in L<openssl(1)>.
=item B<-verify depth>
The verify depth to use. This specifies the maximum length of the
server certificate chain and turns on server certificate verification.
Currently the verify operation continues after errors so all the problems
with a certificate chain can be seen. As a side effect the connection
will never fail due to a server certificate verify failure.
2014-04-07 20:02:10 +08:00
=item B<-verify_return_error>
Return verification errors instead of continuing. This will typically
abort the handshake with a fatal error.
=item B<-CApath directory>
The directory to use for server certificate verification. This directory
must be in "hash format", see B<verify> for more information. These are
also used when building the client certificate chain.
=item B<-CAfile file>
A file containing trusted certificates to use during server authentication
and to use when attempting to build the client certificate chain.
=item B<-no-CAfile>
Do not load the trusted CA certificates from the default file location
=item B<-no-CApath>
Do not load the trusted CA certificates from the default directory location
=item B<-dane_tlsa_domain domain>
Enable RFC6698/RFC7671 DANE TLSA authentication and specify the
TLSA base domain which becomes the default SNI hint and the primary
reference identifier for hostname checks. This must be used in
combination with at least one instance of the B<-dane_tlsa_rrdata>
option below.
When DANE authentication succeeds, the diagnostic output will include
the lowest (closest to 0) depth at which a TLSA record authenticated
a chain certificate. When that TLSA record is a "2 1 0" trust
anchor public key that signed (rather than matched) the top-most
certificate of the chain, the result is reported as "TA public key
verified". Otherwise, either the TLSA record "matched TA certificate"
at a positive depth or else "matched EE certificate" at depth 0.
=item B<-dane_tlsa_rrdata rrdata>
Use one or more times to specify the RRDATA fields of the DANE TLSA
RRset associated with the target service. The B<rrdata> value is
specied in "presentation form", that is four whitespace separated
fields that specify the usage, selector, matching type and associated
data, with the last of these encoded in hexadecimal. Optional
whitespace is ignored in the associated data field. For example:
Suppress DANE TLSA reflection when verification fails As documented both SSL_get0_dane_authority() and SSL_get0_dane_tlsa() are expected to return a negative match depth and nothing else when verification fails. However, this only happened when verification failed during chain construction. Errors in verification of the constructed chain did not have the intended effect on these functions. This commit updates the functions to check for verify_result == X509_V_OK, and no longer erases any accumulated match information when chain construction fails. Sophisticated developers can, with care, use SSL_set_verify_result(ssl, X509_V_OK) to "peek" at TLSA info even when verification fail. They must of course first check and save the real error, and restore the original error as quickly as possible. Hiding by default seems to be the safer interface. Introduced X509_V_ERR_DANE_NO_MATCH code to signal failure to find matching TLSA records. Previously reported via X509_V_ERR_CERT_UNTRUSTED. This also changes the "-brief" output from s_client to include verification results and TLSA match information. Mentioned session resumption in code example in SSL_CTX_dane_enable(3). Also mentioned that depths returned are relative to the verified chain which is now available via SSL_get0_verified_chain(3). Added a few more test-cases to danetest, that exercise the new code. Resolved thread safety issue in use of static buffer in X509_verify_cert_error_string(). Fixed long-stating issue in apps/s_cb.c which always sets verify_error to either X509_V_OK or "chain to long", code elsewhere (e.g. s_time.c), seems to expect the actual error. [ The new chain construction code is expected to correctly generate "chain too long" errors, so at some point we need to drop the work-arounds, once SSL_set_verify_depth() is also fixed to propagate the depth to X509_STORE_CTX reliably. ] Reviewed-by: Rich Salz <rsalz@openssl.org>
2016-02-08 08:07:57 +08:00
$ openssl s_client -brief -starttls smtp \
-connect smtp.example.com:25 \
-dane_tlsa_domain smtp.example.com \
-dane_tlsa_rrdata "2 1 1
B111DD8A1C2091A89BD4FD60C57F0716CCE50FEEFF8137CDBEE0326E 02CF362B" \
-dane_tlsa_rrdata "2 1 1
60B87575447DCBA2A36B7D11AC09FB24A9DB406FEE12D2CC90180517 616E8A18"
...
Suppress DANE TLSA reflection when verification fails As documented both SSL_get0_dane_authority() and SSL_get0_dane_tlsa() are expected to return a negative match depth and nothing else when verification fails. However, this only happened when verification failed during chain construction. Errors in verification of the constructed chain did not have the intended effect on these functions. This commit updates the functions to check for verify_result == X509_V_OK, and no longer erases any accumulated match information when chain construction fails. Sophisticated developers can, with care, use SSL_set_verify_result(ssl, X509_V_OK) to "peek" at TLSA info even when verification fail. They must of course first check and save the real error, and restore the original error as quickly as possible. Hiding by default seems to be the safer interface. Introduced X509_V_ERR_DANE_NO_MATCH code to signal failure to find matching TLSA records. Previously reported via X509_V_ERR_CERT_UNTRUSTED. This also changes the "-brief" output from s_client to include verification results and TLSA match information. Mentioned session resumption in code example in SSL_CTX_dane_enable(3). Also mentioned that depths returned are relative to the verified chain which is now available via SSL_get0_verified_chain(3). Added a few more test-cases to danetest, that exercise the new code. Resolved thread safety issue in use of static buffer in X509_verify_cert_error_string(). Fixed long-stating issue in apps/s_cb.c which always sets verify_error to either X509_V_OK or "chain to long", code elsewhere (e.g. s_time.c), seems to expect the actual error. [ The new chain construction code is expected to correctly generate "chain too long" errors, so at some point we need to drop the work-arounds, once SSL_set_verify_depth() is also fixed to propagate the depth to X509_STORE_CTX reliably. ] Reviewed-by: Rich Salz <rsalz@openssl.org>
2016-02-08 08:07:57 +08:00
Verification: OK
Verified peername: smtp.example.com
Suppress DANE TLSA reflection when verification fails As documented both SSL_get0_dane_authority() and SSL_get0_dane_tlsa() are expected to return a negative match depth and nothing else when verification fails. However, this only happened when verification failed during chain construction. Errors in verification of the constructed chain did not have the intended effect on these functions. This commit updates the functions to check for verify_result == X509_V_OK, and no longer erases any accumulated match information when chain construction fails. Sophisticated developers can, with care, use SSL_set_verify_result(ssl, X509_V_OK) to "peek" at TLSA info even when verification fail. They must of course first check and save the real error, and restore the original error as quickly as possible. Hiding by default seems to be the safer interface. Introduced X509_V_ERR_DANE_NO_MATCH code to signal failure to find matching TLSA records. Previously reported via X509_V_ERR_CERT_UNTRUSTED. This also changes the "-brief" output from s_client to include verification results and TLSA match information. Mentioned session resumption in code example in SSL_CTX_dane_enable(3). Also mentioned that depths returned are relative to the verified chain which is now available via SSL_get0_verified_chain(3). Added a few more test-cases to danetest, that exercise the new code. Resolved thread safety issue in use of static buffer in X509_verify_cert_error_string(). Fixed long-stating issue in apps/s_cb.c which always sets verify_error to either X509_V_OK or "chain to long", code elsewhere (e.g. s_time.c), seems to expect the actual error. [ The new chain construction code is expected to correctly generate "chain too long" errors, so at some point we need to drop the work-arounds, once SSL_set_verify_depth() is also fixed to propagate the depth to X509_STORE_CTX reliably. ] Reviewed-by: Rich Salz <rsalz@openssl.org>
2016-02-08 08:07:57 +08:00
DANE TLSA 2 1 1 ...ee12d2cc90180517616e8a18 matched TA certificate at depth 1
...
=item B<-attime>, B<-check_ss_sig>, B<-crl_check>, B<-crl_check_all>,
B<explicit_policy>, B<-extended_crl>, B<-ignore_critical>, B<-inhibit_any>,
B<-inhibit_map>, B<-issuer_checks>, B<-partial_chain>, B<-policy>,
B<-policy_check>, B<-policy_print>, B<-purpose>, B<-suiteB_128>,
B<-suiteB_128_only>, B<-suiteB_192>, B<-trusted_first>, B<-no_alt_chains>,
B<-use_deltas>, B<-verify_depth>, B<-verify_email>, B<-verify_hostname>,
B<-verify_ip>, B<-verify_name>, B<-x509_strict>
2009-04-15 23:27:03 +08:00
Set various certificate chain validation options. See the
L<verify(1)> manual page for details.
2009-04-15 23:27:03 +08:00
=item B<-reconnect>
reconnects to the same server 5 times using the same session ID, this can
be used as a test that session caching is working.
=item B<-showcerts>
display the whole server certificate chain: normally only the server
certificate itself is displayed.
=item B<-prexit>
print session information when the program exits. This will always attempt
to print out information even if the connection fails. Normally information
will only be printed out once if the connection succeeds. This option is useful
because the cipher in use may be renegotiated or the connection may fail
because a client certificate is required or is requested only after an
attempt is made to access a certain URL. Note: the output produced by this
option is not always accurate because a connection might never have been
established.
=item B<-state>
prints out the SSL session states.
=item B<-debug>
print extensive debugging information including a hex dump of all traffic.
2001-11-10 10:12:09 +08:00
=item B<-msg>
show all protocol messages with hex dump.
2012-11-20 00:37:18 +08:00
=item B<-trace>
show verbose trace output of protocol messages. OpenSSL needs to be compiled
with B<enable-ssl-trace> for this option to work.
=item B<-msgfile>
file to send output of B<-msg> or B<-trace> to, default standard output.
=item B<-nbio_test>
2000-02-04 07:23:24 +08:00
tests non-blocking I/O
=item B<-nbio>
2000-02-04 07:23:24 +08:00
turns on non-blocking I/O
=item B<-crlf>
this option translated a line feed from the terminal into CR+LF as required
by some servers.
=item B<-ign_eof>
inhibit shutting down the connection when end of file is reached in the
input.
=item B<-quiet>
2001-09-07 14:13:40 +08:00
inhibit printing of session and certificate information. This implicitly
turns on B<-ign_eof> as well.
=item B<-no_ign_eof>
shut down the connection when end of file is reached in the input.
Can be used to override the implicit B<-ign_eof> after B<-quiet>.
=item B<-psk_identity identity>
Use the PSK identity B<identity> when using a PSK cipher suite.
=item B<-psk key>
Use the PSK key B<key> when using a PSK cipher suite. The key is
given as a hexadecimal number without leading 0x, for example -psk
1a2b3c4d.
=item B<-ssl3>, B<-tls1>, B<-no_ssl3>, B<-no_tls1>, B<-no_tls1_1>, B<-no_tls1_2>
these options disable the use of certain SSL or TLS protocols. By default
the initial handshake uses a method which should be compatible with all
servers and permit them to use SSL v3 or TLS as appropriate.
Unfortunately there are still ancient and broken servers in use which
cannot handle this technique and will fail to connect. Some servers only
work if TLS is turned off.
=item B<-fallback_scsv>
Send TLS_FALLBACK_SCSV in the ClientHello.
=item B<-async>
switch on asynchronous mode. Cryptographic operations will be performed
asynchronously. This will only have an effect if an asynchronous capable engine
is also used via the B<-engine> option. For test purposes the dummy async engine
(dasync) can be used (if available).
=item B<-bugs>
there are several known bug in SSL and TLS implementations. Adding this
option enables various workarounds.
=item B<-comp>
Enables support for SSL/TLS compression.
This option was introduced in OpenSSL 1.1.0.
TLS compression is not recommended and is off by default as of
OpenSSL 1.1.0.
=item B<-no_comp>
Disables support for SSL/TLS compression.
TLS compression is not recommended and is off by default as of
OpenSSL 1.1.0.
2012-11-20 00:07:53 +08:00
=item B<-brief>
only provide a brief summary of connection parameters instead of the
normal verbose output.
=item B<-cipher cipherlist>
this allows the cipher list sent by the client to be modified. Although
the server determines which cipher suite is used it should take the first
supported cipher in the list sent by the client. See the B<ciphers>
command for more information.
=item B<-starttls protocol>
send the protocol-specific message(s) to switch to TLS for communication.
B<protocol> is a keyword for the intended protocol. Currently, the only
supported keywords are "smtp", "pop3", "imap", "ftp", "xmpp", "xmpp-server",
and "irc."
=item B<-xmpphost hostname>
This option, when used with "-starttls xmpp" or "-starttls xmpp-server",
specifies the host for the "to" attribute of the stream element.
If this option is not specified, then the host specified with "-connect"
will be used.
=item B<-tlsextdebug>
print out a hex dump of any TLS extensions received from the server.
=item B<-no_ticket>
disable RFC4507bis session ticket support.
=item B<-sess_out filename>
output SSL session to B<filename>
=item B<-sess_in sess.pem>
load SSL session from B<filename>. The client will attempt to resume a
connection from this session.
=item B<-engine id>
2009-04-15 23:27:03 +08:00
specifying an engine (by its unique B<id> string) will cause B<s_client>
to attempt to obtain a functional reference to the specified engine,
thus initialising it if needed. The engine will then be set as the default
for all available algorithms.
=item B<-rand file(s)>
a file or files containing random data used to seed the random number
generator, or an EGD socket (see L<RAND_egd(3)>).
Multiple files can be specified separated by a OS-dependent character.
The separator is B<;> for MS-Windows, B<,> for OpenVMS, and B<:> for
all others.
=item B<-serverinfo types>
a list of comma-separated TLS Extension Types (numbers between 0 and
65535). Each type will be sent as an empty ClientHello TLS Extension.
The server's response (if any) will be encoded and displayed as a PEM
file.
=item B<-status>
sends a certificate status request to the server (OCSP stapling). The server
response (if any) is printed out.
=item B<-nextprotoneg protocols>
enable Next Protocol Negotiation TLS extension and provide a list of
comma-separated protocol names that the client should advertise
support for. The list should contain most wanted protocols first.
Protocol names are printable ASCII strings, for example "http/1.1" or
"spdy/3".
Empty list of protocols is treated specially and will cause the client to
advertise support for the TLS extension but disconnect just after
receiving ServerHello with a list of server supported protocols.
2000-01-11 08:06:40 +08:00
=back
=head1 CONNECTED COMMANDS
If a connection is established with an SSL server then any data received
from the server is displayed and any key presses will be sent to the
server. When used interactively (which means neither B<-quiet> nor B<-ign_eof>
2001-09-07 14:13:40 +08:00
have been given), the session will be renegotiated if the line begins with an
B<R>, and if the line begins with a B<Q> or if end of file is reached, the
connection will be closed down.
=head1 NOTES
B<s_client> can be used to debug SSL servers. To connect to an SSL HTTP
server the command:
openssl s_client -connect servername:443
would typically be used (https uses port 443). If the connection succeeds
then an HTTP command can be given such as "GET /" to retrieve a web page.
If the handshake fails then there are several possible causes, if it is
nothing obvious like no client certificate then the B<-bugs>,
B<-ssl3>, B<-tls1>, B<-no_ssl3>, B<-no_tls1> options can be tried
in case it is a buggy server. In particular you should play with these
options B<before> submitting a bug report to an OpenSSL mailing list.
A frequent problem when attempting to get client certificates working
is that a web client complains it has no certificates or gives an empty
list to choose from. This is normally because the server is not sending
the clients certificate authority in its "acceptable CA list" when it
requests a certificate. By using B<s_client> the CA list can be viewed
and checked. However some servers only request client authentication
after a specific URL is requested. To obtain the list in this case it
is necessary to use the B<-prexit> option and send an HTTP request
for an appropriate page.
If a certificate is specified on the command line using the B<-cert>
option it will not be used unless the server specifically requests
a client certificate. Therefor merely including a client certificate
on the command line is no guarantee that the certificate works.
If there are problems verifying a server certificate then the
B<-showcerts> option can be used to show the whole chain.
2014-04-07 20:02:10 +08:00
The B<s_client> utility is a test tool and is designed to continue the
handshake after any certificate verification errors. As a result it will
accept any certificate chain (trusted or not) sent by the peer. None test
applications should B<not> do this as it makes them vulnerable to a MITM
attack. This behaviour can be changed by with the B<-verify_return_error>
option: any verify errors are then returned aborting the handshake.
=head1 BUGS
Because this program has a lot of options and also because some of
2000-01-09 09:26:43 +08:00
the techniques used are rather old, the C source of s_client is rather
hard to read and not a model of how things should be done. A typical
SSL client program would be much simpler.
The B<-prexit> option is a bit of a hack. We should really report
information whenever a session is renegotiated.
=head1 SEE ALSO
L<sess_id(1)>, L<s_server(1)>, L<ciphers(1)>
=head1 HISTORY
The -no_alt_chains options was first added to OpenSSL 1.1.0.
=cut