Age | Commit message (Collapse) | Author |
|
This patch allows extra callback options to be passed to the module
implementing the SSH callback module behaviour.
A module implementing the SSH key callback API is used to customize
the handling of public key. This patch allows extra callback options
to be passed to the module implementing the SSH callback module
behaviour.
The key_cb option has been changed:
{key_cb, atom()} -> {key_cb, key_cb()}
Where:
key_cb() :: atom() | {atom(), list()}
The callback options, if specified, is made available to the callback
module via the options passed to it under the key 'key_cb_private'.
More details and some backgorund is available here[1].
[1]: http://erlang.org/pipermail/erlang-patches/2015-November/004800.html
|
|
* maint:
Updated OTP version
Update release notes
ssh: vsn.mk updated
ssh: New test cases for SSH_MSG_KEX_DH_GEX_REQUEST_OLD
ssh: update existing testcases
ssh: Improve group selection
ssh: implemented server side SSH_MSG_KEX_DH_GEX_REQUEST_OLD for putty client
Conflicts:
OTP_VERSION
|
|
* maint-18:
Updated OTP version
Update release notes
ssh: vsn.mk updated
ssh: New test cases for SSH_MSG_KEX_DH_GEX_REQUEST_OLD
ssh: update existing testcases
ssh: Improve group selection
ssh: implemented server side SSH_MSG_KEX_DH_GEX_REQUEST_OLD for putty client
Conflicts:
lib/ssh/src/ssh_transport.erl
lib/ssh/test/ssh_protocol_SUITE.erl
lib/ssh/vsn.mk
|
|
|
|
|
|
* maint:
ssh: testcases for starting daemon with given fd
ssh: Make it possible for more than one daemon started with option fd
|
|
* hans/ssh/fd_more/OTP-12966:
ssh: testcases for starting daemon with given fd
ssh: Make it possible for more than one daemon started with option fd
|
|
* maint:
ssh: fix dialyzer reported error
ssh: renegotiate test group for AES_GCM
ssh: documentation updates
ssh: AEAD_AES_(128|256)_GCM and aes(128|256)[email protected] crypto
ssh: refactor packet reception
|
|
|
|
|
|
* maint:
ssh: Add env var info printout to ssh_algorithms_SUTE:init_per_suite
ssh: more public keys added to test suite
|
|
|
|
|
|
* maint:
ssh: changes after doc review
ssh: Document pwdfun
ssh: make corrections of keyboard-interactive client
ssh: enable users to give option keyboard_interact_fun
ssh: pwdfun/4 and simple tests
|
|
* hans/ssh/pwdfun/OTP-13055:
ssh: changes after doc review
ssh: Document pwdfun
ssh: make corrections of keyboard-interactive client
ssh: enable users to give option keyboard_interact_fun
ssh: pwdfun/4 and simple tests
Conflicts:
lib/ssh/doc/src/ssh.xml
|
|
* maint:
ssh: correct a bad doc xref
public_key: renamed priv/ssh_moduli -> priv/moduli
ssh, public_key: updates after doc review
ssh: changes after doc review
ssh: extend 'dh_gex_limits' to server side
ssh: option dh_gex_groups with tag ssh_moduli_file
ssh, public_key: random selection of diffie-hellman moduli
|
|
OTP-13066
|
|
Makes this existing option also accept a file in openssh format
OTP-13052
|
|
Also tool (public_key:gen_moduli_hrl) to convert an openssh moduli file to erlang format.
|
|
Also solves OTP-13053
|
|
|
|
Includes a ssh_transport:pack/3 function for generating invalid packets
|
|
* maint:
ssh: Adjust the test ssh_renegotiate_SUITE:rekey_limit
ssh: fix spelling error pulic -> public
|
|
- Remove random length padding to make the data volume deterministic
- Increase data volume limit for re-keying because the kexinit message has grown
|
|
|
|
* maint:
ssh: testcases for bad service names
|
|
|
|
* maint:
ssh: dynamic check of supported algos in ssh_to_openssh_SUITE
ssh: delete now obsolete test case group 'hardening_tests'
|
|
|
|
Thoose tests are already moved to other test suites. However, the init_per_group and end_per_group clauses were accidently left
|
|
* maint:
ssh: Removed testcases from ssh_to_openssh_SUITE
ssh: test suites probes ssh client
ssh: running ssh_algorithms_SUITE tests in parallel
ssh: wait for subsystem exit in ssh_connection_SUITE:max_channels_option
|
|
They are covered in ssh_algorithms_SUITE
|
|
Previously we called the ssh client with the -Q option to query about supported algorithms. Since old clients do not support this and at least one lies, we now set up a fake server at let the ssh client start negotiation instead. Much more robust hopefully.
|
|
|
|
|
|
* maint:
ssh: Implemented ssh_transport:ecdh_validate_public_key (partly)
ssh: dialyzer fix
public_key: add/update -spec for ssh functions
ssh, public_key: use pubkey encode/decode in app public_key
ssh, public_key: Change EC Public Key representation to what was intended
|
|
|
|
Conflicts:
OTP_VERSION
|
|
Conflicts:
lib/ssh/vsn.mk
|
|
|
|
|
|
|
|
It actually counts the number of subsystem alive. Allocating a channel does not consume any resources
(except some cpu cycles), but the subsystem start spawns processes.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* maint:
ssh: dh_gex defautl values increased. Groups added
|