[sftp.c]
don't append a space after inserting a completion of a directory (i.e.
a path ending in '/') for a slightly better user experience; ok dtucker@
[servconf.c servconf.h sshd.c]
avoid run-time failures when specifying hostkeys via a relative
path by prepending the cwd in these cases; bz#1290; ok dtucker@
[monitor_fdpass.c]
avoid spinning when fd passing on nonblocking sockets by calling poll()
in the EINTR/EAGAIN path, much like we do in atomicio; ok dtucker@
[authfile.c sshconnect2.c]
Do not prompt for a passphrase if we fail to open a keyfile, and log the
reason the open failed to debug.
bz #1693, found by tj AT castaglia org, ok djm@
[ssh_config channels.c ssh.1 channels.h ssh.c]
Add a 'netcat mode' (ssh -W). This connects stdio on the client to a
single port forward on the server. This allows, for example, using ssh as
a ProxyCommand to route connections via intermediate servers.
bz #1618, man page help from jmc@, ok markus@
[channels.c ssh.1 servconf.c sshd_config.5 sshd.c channels.h servconf.h
ssh-keyscan.1 ssh-keyscan.c readconf.c sshconnect.c misc.c ssh.c
readconf.h scp.1 sftp.1 ssh_config.5 misc.h]
Remove RoutingDomain from ssh since it's now not needed. It can be
replaced with "route exec" or "nc -V" as a proxycommand. "route exec"
also ensures that trafic such as DNS lookups stays withing the specified
routingdomain. For example (from reyk):
# route -T 2 exec /usr/sbin/sshd
or inherited from the parent process
$ route -T 2 exec sh
$ ssh 10.1.2.3
ok deraadt@ markus@ stevesk@ reyk@
[sshd.c]
Afer sshd receives a SIGHUP, ignore subsequent HUPs while sshd re-execs
itself. Prevents two HUPs in quick succession from resulting in sshd
dying. bz#1692, patch from Colin Watson via Ubuntu.
[sftp.c]
Prevent sftp from derefing a null pointer when given a "-" without a
command. Also, allow whitespace to follow a "-". bz#1691, path from
Colin Watson via Debian. ok djm@ deraadt@
[mux.c sshpty.h clientloop.c sshtty.c]
quell tc[gs]etattr warnings when forcing a tty (ssh -tt), since we
usually don't actually have a tty to read/set; bz#1686 ok dtucker@
[sftp-server.c sftp-server.8]
add a 'read-only' mode to sftp-server(8) that disables open in write mode
and all other fs-modifying protocol methods. bz#430 ok dtucker@
[sftp.c]
Implement tab-completion of commands, local and remote filenames for sftp.
Hacked on and off for some time by myself, mouring, Carlos Silva (via 2009
Google Summer of Code) and polished to a fine sheen by myself again.
It should deal more-or-less correctly with the ikky corner-cases presented
by quoted filenames, but the UI could still be slightly improved.
In particular, it is quite slow for remote completion on large directories.
bz#200; ok markus@
[sshd_config.5 readconf.c ssh_config.5 scp.1 servconf.c sftp.1 ssh.1]
Rename RDomain config option to RoutingDomain to be more clear and
consistent with other options.
NOTE: if you currently use RDomain in the ssh client or server config,
or ssh/sshd -o, you must update to use RoutingDomain.
ok markus@ djm@
[ssh.c sftp.c scp.c]
When passing user-controlled options with arguments to other programs,
pass the option and option argument as separate argv entries and
not smashed into one (e.g., as -l foo and not -lfoo). Also, always
pass a "--" argument to stop option parsing, so that a positional
argument that starts with a '-' isn't treated as an option. This
fixes some error cases as well as the handling of hostnames and
filenames that start with a '-'.
Based on a diff by halex@
ok halex@ djm@ deraadt@
[sftp.c]
fix potential divide-by-zero in sftp's "df" output when talking to a server
that reports zero files on the filesystem (Unix filesystems always have at
least the root inode). From Steve McClellan at radisys, ok djm@
[sftp.c]
make passing of zero-length arguments to ssh safe by
passing "-<switch>" "<value>" rather than "-<switch><value>"
ok dtucker@, guenther@, djm@
[misc.c]
correct off-by-one in percent_expand(): we would fatal() when trying
to expand EXPAND_MAX_KEYS, allowing only EXPAND_MAX_KEYS-1 to actually
work. Note that nothing in OpenSSH actually uses close to this limit at
present. bz#1607 from Jan.Pechanec AT Sun.COM
[sftp.c]
bz#1588 change "Connecting to host..." message to "Connected to host."
and delay it until after the sftp protocol connection has been established.
Avoids confusing sequence of messages when the underlying ssh connection
experiences problems. ok dtucker@
[session.c]
Warn but do not fail if stat()ing the subsystem binary fails. This helps
with chrootdirectory+forcecommand=sftp-server and restricted shells.
bz #1599, ok djm.
[session.c]
bz#1606: error when an attempt is made to connect to a server
with ForceCommand=internal-sftp with a shell session (i.e. not a
subsystem session). Avoids stuck client when attempting to ssh to such a
service. ok dtucker@
[channels.c channels.h]
fix race condition in x11/agent channel allocation: don't read after
the end of the select read/write fdset and make sure a reused FD
is not touched before the pre-handlers are called.
with and ok djm@
[sshconnect2.c channels.c sshconnect.c]
Set close-on-exec on various descriptors so they don't get leaked to
child processes. bz #1643, patch from jchadima at redhat, ok deraadt.
[ssh_config.5 sshd.c misc.h ssh-keyscan.1 readconf.h sshconnect.c
channels.c channels.h servconf.h servconf.c ssh.1 ssh-keyscan.c scp.1
sftp.1 sshd_config.5 readconf.c ssh.c misc.c]
Allow to set the rdomain in ssh/sftp/scp/sshd and ssh-keyscan.
ok markus@
[roaming_common.c]
Do the actual suspend/resume in the client. This won't be useful until
the server side supports roaming.
Most code from Martin Forssen, maf at appgate dot com. Some changes by
me and markus@
ok markus@
[clientloop.c]
client_loop() must detect if the session has been suspended and resumed,
and take appropriate action in that case.
From Martin Forssen, maf at appgate dot com
ok markus@
set IPV6_V6ONLY for local forwarding with GatwayPorts=yes. Unify
setting IPV6_V6ONLY behind a new function misc.c:sock_set_v6only()
report and fix from jan.kratochvil AT redhat.com
[sshconnect2.c]
disallow a hostile server from checking jpake auth by sending an
out-of-sequence success message. (doesn't affect code enabled by default)
[ssh-agent.1 ssh-add.1 ssh.1]
write UNIX-domain in a more consistent way; while here, replace a
few remaining ".Tn UNIX" macros with ".Ux" ones.
pointed out by ratchov@, thanks!
ok jmc@
- markus@cvs.openbsd.org 2009/10/08 14:03:41
[sshd_config readconf.c ssh_config.5 servconf.c sshd_config.5]
disable protocol 1 by default (after a transition period of about 10 years)
ok deraadt
[sftp-batch.sh sftp-badcmds.sh sftp.sh sftp-cmds.sh sftp-glob.sh]
date: 2009/08/13 01:11:19; author: djm; state: Exp; lines: +10 -7
Swizzle options: "-P sftp_server_path" moves to "-D sftp_server_path",
add "-P port" to match scp(1). Fortunately, the -P option is only really
used by our regression scripts.
part of larger patch from carlosvsilvapt@gmail.com for his Google Summer
of Code work; ok deraadt markus
[regress/Makefile]
regression test for port number parsing. written as part of the a2port
change that went into 5.2 but I forgot to commit it at the time...
[ssh-agent.c]
fix a race condition in ssh-agent that could result in a wedged or
spinning agent: don't read off the end of the allocated fd_sets, and
don't issue blocking read/write on agent sockets - just fall back to
select() on retriable read/write errors. bz#1633 reported and tested
by "noodle10000 AT googlemail.com"; ok dtucker@ markus@
[authfd.c ssh-add.c authfd.h]
Do not fall back to adding keys without contraints (ssh-add -c / -t ...)
when the agent refuses the constrained add request. This was a useful
migration measure back in 2002 when constraints were new, but just
adds risk now.
bz #1612, report and patch from dkg AT fifthhorseman.net; ok markus@
[ssh-keygen.c]
force use of correct hash function for random-art signature display
as it was inheriting the wrong one when bubblebabble signatures were
activated; bz#1611 report and patch from fwojcik+openssh AT besh.com;
ok markus@
[sftp-client.h sftp.1 sftp-client.c sftp.c]
recursive transfer support for get/put and on the commandline
work mostly by carlosvsilvapt@gmail.com for the Google Summer of Code
with some tweaks by me; "go for it" deraadt@
[sftp.1 sftp.c]
Swizzle options: "-P sftp_server_path" moves to "-D sftp_server_path",
add "-P port" to match scp(1). Fortunately, the -P option is only really
used by our regression scripts.
part of larger patch from carlosvsilvapt@gmail.com for his Google Summer
of Code work; ok deraadt markus
[sftp.c sftp.1]
support most of scp(1)'s commandline arguments in sftp(1), as a first
step towards making sftp(1) a drop-in replacement for scp(1).
One conflicting option (-P) has not been changed, pending further
discussion.
Patch from carlosvsilvapt@gmail.com as part of his work in the
Google Summer of Code
the pty master on Solaris, since it never succeeds and can hang if large
amounts of data is sent to the slave (eg a copy-paste). Based on a patch
originally from Doke Scott, ok djm@
the setpcred call on AIX to immediately before the permanently_set_uid().
Ensures that we still have privileges when we call chroot and
pam_open_sesson. Based on a patch from David Leonard.
[readconf.h readconf.c]
Add client option UseRoaming. It doesn't do anything yet but will
control whether the client tries to use roaming if enabled on the
server. From Martin Forssen.
ok markus@
[roaming_common.c roaming.h]
It may be necessary to retransmit some data when resuming, so add it
to a buffer when roaming is enabled.
Most of this code was written by Martin Forssen, maf at appgate dot com.
ok markus@
[packet.h packet.c]
packet_bacup_state() and packet_restore_state() will be used to
temporarily save the current state ren resuming a suspended connection.
ok markus@
[monitor_wrap.c monitor_mm.c ssh-keygen.c auth2.c gss-genr.c sftp-client.c]
alphabetize includes; reduces diff vs portable and style(9).
ok stevesk djm
(Id sync only; these were already in order in -portable)
[roaming.h roaming_common.c roaming_dummy.c]
Add tags for the benefit of the sync scripts
Also: pull in the changes for 1.1->1.2 missed in the previous sync.
[sshd.c packet.c serverloop.c monitor_wrap.c clientloop.c sshconnect.c
monitor.c Added roaming.h roaming_common.c roaming_dummy.c]
Keep track of number of bytes read and written. Needed for upcoming
changes. Most code from Martin Forssen, maf at appgate dot com.
ok markus@
Also, applied appropriate changes to Makefile.in
[sshd.c packet.c serverloop.c monitor_wrap.c clientloop.c sshconnect.c
monitor.c]
Keep track of number of bytes read and written. Needed for upcoming
changes. Most code from Martin Forssen, maf at appgate dot com.
ok markus@
[sshconnect.h sshconnect.c]
Un-static ssh_exchange_identification(), part of a larger change from
Martin Forssen and needed for upcoming changes.
ok markus@
[ssh-agent.c]
My previous commit didn't fix the problem at all, so stick at my first
version of the fix presented to dtucker.
Issue notified by Matthias Barkhoff (matthias dot barkhoff at gmx dot de).
ok dtucker
[ssh-agent.c]
Fixed a possible out-of-bounds memory access if the environment variable
SHELL is shorter than 3 characters.
with input by and ok dtucker
[ssh.1]
for "Ciphers", just point the reader to the keyword in ssh_config(5), just
as we do for "MACs": this stops us getting out of sync when the lists
change;
fixes documentation/6102, submitted by Peter J. Philipp
alternative fix proposed by djm
ok markus
auth2-pubkey.c session.c openbsd-compat/bsd-cygwin_util.{c,h}
openbsd-compat/daemon.c] Remove support for Windows 95/98/ME and very old
version of Cygwin. Patch from vinschen at redhat com.
[auth2-jpake.c jpake.c jpake.h monitor_wrap.c monitor_wrap.h schnorr.c]
[sshconnect2.c]
refactor the (disabled) Schnorr proof code to make it a little more
generally useful
[regress/putty-kex.sh regress/putty-transfer.sh] Downgrade disabled
interop tests from FATAL error to a warning. Allows some interop
tests to proceed if others are missing necessary prerequisites.
[canohost.c canohost.h channels.c channels.h clientloop.c readconf.c]
[readconf.h serverloop.c ssh.c]
support remote port forwarding with a zero listen port (-R0:...) to
dyamically allocate a listen port at runtime (this is actually
specified in rfc4254); bz#1003 ok markus@
OSX provides a getlastlogxbyname function that automates the reading of
a lastlog file. Also, the pututxline function will update lastlog so
there is no need for loginrec.c to do it explicitly. Collapse some
overly verbose code while I'm in there.
If the CYGWIN environment variable is empty, the installer script
should not install the service with an empty CYGWIN variable, but
rather without setting CYGWNI entirely.
Changes to work on Cygwin 1.5.x as well as on the new Cygwin 1.7.x.
The information given for the setting of the CYGWIN environment variable
is wrong for both releases so I just removed it, together with the
unnecessary (Cygwin 1.5.x) or wrong (Cygwin 1.7.x) default setting.
[cipher.c cipher.h packet.c]
Work around the CPNI-957037 Plaintext Recovery Attack by always
reading 256K of data on packet size or HMAC errors (in CBC mode only).
Help, feedback and ok djm@
Feedback from Martin Albrecht and Paterson Kenny
[clientloop.c misc.c readconf.c readconf.h servconf.c servconf.h]
[serverloop.c ssh-keyscan.c ssh.c sshd.c]
make a2port() return -1 when it encounters an invalid port number
rather than 0, which it will now treat as valid (needed for future work)
adjust current consumers of a2port() to check its return value is <= 0,
which in turn required some things to be converted from u_short => int
make use of int vs. u_short consistent in some other places too
feedback & ok markus@
[channels.c channels.h session.c]
make Channel->path an allocated string, saving a few bytes here and
there and fixing bz#1380 in the process; ok markus@
[readconf.c]
1) use obsolete instead of alias for consistency
2) oUserKnownHostsFile not obsolete but oGlobalKnownHostsFile2 is
so move the comment.
3) reorder so like options are together
ok djm@
[kexgexs.c]
fix hash calculation for KEXGEX: hash over the original client-supplied
values and not the sanity checked versions that we acutally use;
bz#1540 reported by john.smith AT arrows.demon.co.uk
ok markus@
[channels.c]
call channel destroy callbacks on receipt of open failure messages.
fixes client hangs when connecting to a server that has MaxSessions=0
set spotted by imorgan AT nas.nasa.gov; ok markus@
[sftp.1 sftp.c]
update for the synopses displayed by the 'help' command, there are a
few missing flags; add 'bye' to the output of 'help'; sorting and spacing.
jmc@ suggested replacing .Oo/.Oc with a single .Op macro.
ok jmc@
[channels.c servconf.c]
channel_print_adm_permitted_opens() should deal with all the printing
for that config option. suggested by markus@; ok markus@ djm@
dtucker@
[readconf.c]
don't leave junk (free'd) pointers around in Forward *fwd argument on
failure; avoids double-free in ~C -L handler when given an invalid
forwarding specification; bz#1539 report from adejong AT debian.org
via Colin Watson; ok markus@ dtucker@
[sftp.c]
Deal correctly with failures in remote stat() operation in sftp,
correcting fail-on-error behaviour in batchmode. bz#1541 report and
fix from anedvedicky AT gmail.com; ok markus@
[clientloop.c]
The ~C escape handler does not work correctly for multiplexed sessions -
it opens a commandline on the master session, instead of on the slave
that requested it. Disable it on slave sessions until such time as it
is fixed; bz#1543 report from Adrian Bridgett via Colin Watson
ok markus@
[clientloop.c]
we have to use the recipient's channel number (RFC 4254) for
SSH2_MSG_CHANNEL_SUCCESS/SSH2_MSG_CHANNEL_FAILURE messages,
otherwise we trigger 'Non-public channel' error messages on sshd
systems with clientkeepalive enabled; noticed by sturm; ok djm;
[packet.c]
packet_disconnect() on padding error, too. should reduce the success
probability for the CPNI-957037 Plaintext Recovery Attack to 2^-18
ok djm@
[servconf.c]
USE_AFS not referenced so remove #ifdef. fixes sshd -T not printing
kerberosgetafstoken. ok dtucker@
(Id sync only, we still want the ifdef in portable)
[readconf.c]
because parse_forward() is now used to parse all forward types (DLR),
and it malloc's space for host variables, we don't need to malloc
here. fixes small memory leaks.
previously dynamic forwards were not parsed in parse_forward() and
space was not malloc'd in that case.
ok djm@
[auth.h auth2.c monitor.c monitor.h monitor_wrap.c monitor_wrap.h]
[readconf.c readconf.h servconf.c servconf.h ssh2.h ssh_config.5]
[sshconnect2.c sshd_config.5 jpake.c jpake.h schnorr.c auth2-jpake.c]
[Makefile.in]
Add support for an experimental zero-knowledge password authentication
method using the J-PAKE protocol described in F. Hao, P. Ryan,
"Password Authenticated Key Exchange by Juggling", 16th Workshop on
Security Protocols, Cambridge, April 2008.
This method allows password-based authentication without exposing
the password to the server. Instead, the client and server exchange
cryptographic proofs to demonstrate of knowledge of the password while
revealing nothing useful to an attacker or compromised endpoint.
This is experimental, work-in-progress code and is presently
compiled-time disabled (turn on -DJPAKE in Makefile.inc).
"just commit it. It isn't too intrusive." deraadt@
[ssh.1 ssh.c]
Add -y option to force logging via syslog rather than stderr.
Useful for daemonised ssh connection (ssh -f). Patch originally from
and ok'd by markus@
[sshconnect2.c]
Repair strnvis() buffersize of 4*n+1, with termination gauranteed by the
function.
spotted by des@freebsd, who commited an incorrect fix to the freebsd tree
and (as is fairly typical) did not report the problem to us. But this fix
is correct.
ok djm
[sshd.8]
do not give an example of how to chmod files: we can presume the user
knows that. removes an ambiguity in the permission of authorized_keys;
ok deraadt
[compat.c compat.h nchan.c ssh.c]
only send eow and no-more-sessions requests to openssh 5 and newer;
fixes interop problems with broken ssh v2 implementations; ok djm@
[servconf.c]
do not try to print options that have been compile-time disabled
in config test mode (sshd -T); report from nix-corp AT esperi.org.uk
ok dtucker@
[auth2-hostbased.c]
strip trailing '.' from hostname when HostbasedUsesNameFromPacketOnly=yes
report and patch from res AT qoxp.net (bz#1200); ok markus@
[sftp.c]
increase number of piplelined requests so they properly fill the
(recently increased) channel window. prompted by rapier AT psc.edu;
ok markus@
[ssh-keygen.c]
Change "ssh-keygen -F [host] -l" to not display random art unless
-v is also specified, making it consistent with the manual and other
uses of -l.
ok grunk@
[auth1.c auth2.c]
Make protocol 1 MaxAuthTries logic match protocol 2's.
Do not treat the first protocol 2 authentication attempt as
a failure IFF it is for method "none".
Makes MaxAuthTries' user-visible behaviour identical for
protocol 1 vs 2.
ok dtucker@
[packet.c scp.c serverloop.c sftp-client.c ssh-agent.c ssh-keyscan.c]
[sshd.c] Explicitly handle EWOULDBLOCK wherever we handle EAGAIN, on
some platforms (HP nonstop) it is a distinct errno;
bz#1467 reported by sconeu AT yahoo.com; ok dtucker@
[ssh.1 ssh.c]
When forking after authentication ("ssh -f") with ExitOnForwardFailure
enabled, delay the fork until after replies for any -R forwards have
been seen. Allows for robust detection of -R forward failure when
using -f (similar to bz#92); ok dtucker@
[auth2-none.c auth2.c]
Make protocol 2 MaxAuthTries behaviour a little more sensible:
Check whether client has exceeded MaxAuthTries before running
an authentication method and skip it if they have, previously it
would always allow one try (for "none" auth).
Preincrement failure count before post-auth test - previously this
checked and postincremented, also to allow one "none" try.
Together, these two changes always count the "none" auth method
which could be skipped by a malicious client (e.g. an SSH worm)
to get an extra attempt at a real auth method. They also make
MaxAuthTries=0 a useful way to block users entirely (esp. in a
sshd_config Match block).
Also, move sending of any preauth banner from "none" auth method
to the first call to input_userauth_request(), so worms that skip
the "none" method get to see it too.
[auth-rsa.c auth.c auth2-pubkey.c auth.h]
Merge duplicate host key file checks, based in part on a patch from Rob
Holland via bz #1348 . Also checks for non-regular files during protocol
1 RSA auth. ok djm@
[sshd_config sshd_config.5 sshd.8 servconf.c]
increase default size of ssh protocol 1 ephemeral key from 768 to 1024
bits; prodded by & ok dtucker@ ok deraadt@
[regress/Makefile regress/test-exec.sh regress/conch-ciphers.sh]
very basic regress test against Twisted Conch in "make interop"
target (conch is available in ports/devel/py-twisted/conch);
ok markus@
[readconf.c readconf.h ssh.1 ssh_config.5 sshconnect.c]
Move SSH Fingerprint Visualization away from sharing the config option
CheckHostIP to an own config option named VisualHostKey.
While there, fix the behaviour that ssh would draw a random art picture
on every newly seen host even when the option was not enabled.
prodded by deraadt@, discussions,
help and ok markus@ djm@ dtucker@
[dh.c dh.h moduli.c]
when loading moduli from /etc/moduli in sshd(8), check that they
are of the expected "safe prime" structure and have had
appropriate primality tests performed;
feedback and ok dtucker@
[sftp-client.c sftp-server.c]
allow the sftp chmod(2)-equivalent operation to set set[ug]id/sticky
bits. Note that this only affects explicit setting of modes (e.g. via
sftp(1)'s chmod command) and not file transfers. (bz#1310)
ok deraadt@ at c2k8
- martynas@cvs.openbsd.org 2008/06/21 07:46:46
[sftp.c]
use optopt to get invalid flag, instead of return value of getopt,
which is always '?'; ok djm@
[channels.c channels.h session.c]
don't call isatty() on a pty master, instead pass a flag down to
channel_set_fds() indicating that te fds refer to a tty. Fixes a
hang on exit on Solaris (bz#1463) in portable but is actually
a generic bug; ok dtucker deraadt markus
[ssh.1]
Explain the use of SSH fpr visualization using random art, and cite the
original scientific paper inspiring that technique.
Much help with English and nroff by jmc@, thanks.
[monitor.c]
Clear key options in the monitor on failed authentication, prevents
applying additional restrictions to non-pubkey authentications in
the case where pubkey fails but another method subsequently succeeds.
bz #1472, found by Colin Watson, ok markus@ djm
Darren Tucker
Tim Rice
Damien Miller