=========================
This is a major release, containing a number of new features as
well as a large internal re-factoring.
Potentially-incompatible changes
--------------------------------
* sshd(8): UseDNS now defaults to 'no'. Configurations that match
against the client host name (via sshd_config or authorized_keys)
may need to re-enable it or convert to matching against addresses.
New Features
------------
* Much of OpenSSH's internal code has been re-factored to be more
library-like. These changes are mostly not user-visible, but
have greatly improved OpenSSH's testability and internal layout.
* Add FingerprintHash option to ssh(1) and sshd(8), and equivalent
command-line flags to the other tools to control algorithm used
for key fingerprints. The default changes from MD5 to SHA256 and
format from hex to base64.
Fingerprints now have the hash algorithm prepended. An example of
the new format: SHA256:mVPwvezndPv/ARoIadVY98vAC0g+P/5633yTC4d/wXE
Please note that visual host keys will also be different.
* ssh(1), sshd(8): Experimental host key rotation support. Add a
protocol extension for a server to inform a client of all its
available host keys after authentication has completed. The client
may record the keys in known_hosts, allowing it to upgrade to better
host key algorithms and a server to gracefully rotate its keys.
The client side of this is controlled by a UpdateHostkeys config
option (default off).
* ssh(1): Add a ssh_config HostbasedKeyType option to control which
host public key types are tried during host-based authentication.
* ssh(1), sshd(8): fix connection-killing host key mismatch errors
when sshd offers multiple ECDSA keys of different lengths.
* ssh(1): when host name canonicalisation is enabled, try to
parse host names as addresses before looking them up for
canonicalisation. fixes bz#2074 and avoiding needless DNS
lookups in some cases.
* ssh-keygen(1), sshd(8): Key Revocation Lists (KRLs) no longer
require OpenSSH to be compiled with OpenSSL support.
* ssh(1), ssh-keysign(8): Make ed25519 keys work for host based
authentication.
* sshd(8): SSH protocol v.1 workaround for the Meyer, et al,
Bleichenbacher Side Channel Attack. Fake up a bignum key before
RSA decryption.
* sshd(8): Remember which public keys have been used for
authentication and refuse to accept previously-used keys.
This allows AuthenticationMethods=publickey,publickey to require
that users authenticate using two _different_ public keys.
* sshd(8): add sshd_config HostbasedAcceptedKeyTypes and
PubkeyAcceptedKeyTypes options to allow sshd to control what
public key types will be accepted. Currently defaults to all.
* sshd(8): Don't count partial authentication success as a failure
against MaxAuthTries.
* ssh(1): Add RevokedHostKeys option for the client to allow
text-file or KRL-based revocation of host keys.
* ssh-keygen(1), sshd(8): Permit KRLs that revoke certificates by
serial number or key ID without scoping to a particular CA.
* ssh(1): Add a "Match canonical" criteria that allows ssh_config
Match blocks to trigger only in the second config pass.
* ssh(1): Add a -G option to ssh that causes it to parse its
configuration and dump the result to stdout, similar to "sshd -T".
* ssh(1): Allow Match criteria to be negated. E.g. "Match !host".
* The regression test suite has been extended to cover more OpenSSH
features. The unit tests have been expanded and now cover key
exchange.
Bugfixes
* ssh-keyscan(1): ssh-keyscan has been made much more robust again
servers that hang or violate the SSH protocol.
* ssh(1), ssh-keygen(1): Fix regression bz#2306: Key path names were
being lost as comment fields.
* ssh(1): Allow ssh_config Port options set in the second config
parse phase to be applied (they were being ignored). bz#2286
* ssh(1): Tweak config re-parsing with host canonicalisation - make
the second pass through the config files always run when host name
canonicalisation is enabled (and not whenever the host name
changes) bz#2267
* ssh(1): Fix passing of wildcard forward bind addresses when
connection multiplexing is in use; bz#2324;
* ssh-keygen(1): Fix broken private key conversion from non-OpenSSH
formats; bz#2345.
* ssh-keygen(1): Fix KRL generation bug when multiple CAs are in
use.
* Various fixes to manual pages: bz#2288, bz#2316, bz#2273
Portable OpenSSH
* Support --without-openssl at configure time
Disables and removes dependency on OpenSSL. Many features,
including SSH protocol 1 are not supported and the set of crypto
options is greatly restricted. This will only work on systems
with native arc4random or /dev/urandom.
Considered highly experimental for now.
* Support --without-ssh1 option at configure time
Allows disabling support for SSH protocol 1.
* sshd(8): Fix compilation on systems with IPv6 support in utmpx; bz#2296
* Allow custom service name for sshd on Cygwin. Permits the use of
multiple sshd running with different service names.
Checksums:
==========
- SHA1 (openssh-6.8.tar.gz) = 99903c6ca76e0a2c044711017f81127e12459d37
- SHA256 (openssh-6.8.tar.gz) = N1uzVarFbrm2CzAwuDu3sRoszmqpK+5phAChP/QNyuw=
- SHA1 (openssh-6.8p1.tar.gz) = cdbc51e46a902b30d263b05fdc71340920e91c92
- SHA256 (openssh-6.8p1.tar.gz) = P/ZM5z7hJEgLW/dnuYMNfTwDu8tqvnFrePAZLDfOFg4=
Please note that the PGP key used to sign releases was recently rotated.
The new key has been signed by the old key to provide continuity. It is
available from the mirror sites as RELEASE_KEY.asc.
Reporting Bugs:
===============
- Please read http://www.openssh.com/report.html
Security bugs should be reported directly to openssh@openssh.com
OpenSSH is brought to you by Markus Friedl, Niels Provos, Theo de Raadt,
Kevin Steves, Damien Miller, Darren Tucker, Jason McIntyre, Tim Rice and
Ben Lindstrom.
* update protocol bind implementations to use/expect sockaddr *
instead of mbuf *
* introduce sockaddr_big struct for storage of addr data passed via
sys_bind; sockaddr_big is of sufficient size and alignment to
accommodate all addr data sizes received.
* modify sys_bind to allocate sockaddr_big instead of using an mbuf.
* bump kernel version to 7.99.9 for change to pr_bind() parameter type.
Patch posted to tech-net@
http://mail-index.netbsd.org/tech-net/2015/03/15/msg005004.html
The choice to use a new structure sockaddr_big has been retained since
changing sockaddr_storage size would lead to unnecessary ABI change. The
use of the new structure does not preclude future work that increases
the size of sockaddr_storage and at that time sockaddr_big may be
trivially replaced.
Tested by mrg@ and myself, discussed with rmind@, posted to tech-net@
has VM_MAXUSER_ADDRESS set to 2GB. Fixes "fatal error: had to relocate PCH"
from g++ building wxGTK28 or wxGTK30.
For details see https://gcc.gnu.org/bugzilla/show_bug.cgi?id=58379
thanks to Martin Husemann for the hint.
While the rump kernel and local clients are by definition in the same
host vmspace, there are subtle differences in how in-kernel code works
in case accessing the kernel vmspace or a user process vmspace.
Problem discovered by riastradh's "read(fd, NULL, 1)" test.
