summaryrefslogtreecommitdiffstats
path: root/usr.bin/ssh (follow)
Commit message (Collapse)AuthorAgeFilesLines
...
* when checking that filenames sent by the server side match what thedjm2019-02-101-12/+270
| | | | | | | | client requested, be prepared to handle shell-style brace alternations, e.g. "{foo,bar}". "looks good to me" millert@ + in snaps for the last week courtesy deraadt@
* syslog when connection is dropped for attempting to run a commanddjm2019-02-101-3/+6
| | | | when ForceCommand=internal-sftp is in effect; bz2960; ok dtucker@
* Adapt code in the non-USE_PIPES codepath to the new packet API.dtucker2019-02-051-2/+2
| | | | | This code is not normally reachable since USE_PIPES is always defined. bz#2961, patch from adrian.fita at gmail com.
* fix NULL-deref crash in PKCS#11 code when attempting login to a tokendjm2019-02-041-2/+17
| | | | requiring a PIN; reported by benno@ fix mostly by markus@
* Remove obsolete "Protocol" from commented out examples.dtucker2019-02-041-2/+1
| | | | Patch from samy.mahmoudi at gmail com.
* Save connection timeout and restore for 2nd and subsequent attempts,dtucker2019-02-011-2/+3
| | | | preventing them from having no timeout. bz#2918, ok djm@
* Add authors for public domain sntrup4591761 code;markus2019-01-302-2/+23
| | | | confirmed by Daniel J. Bernstein
* add -T to usage();jmc2019-01-271-2/+2
|
* check in scp client that filenames sent during remote->local directorydjm2019-01-262-12/+43
| | | | | | | | | | | | | | copies satisfy the wildcard specified by the user. This checking provides some protection against a malicious server sending unexpected filenames, but it comes at a risk of rejecting wanted files due to differences between client and server wildcard expansion rules. For this reason, this also adds a new -T flag to disable the check. reported by Harry Sintonen fix approach suggested by markus@; has been in snaps for ~1wk courtesy deraadt@
* make ssh-keyscan return a non-zero exit status if it finds no keys.djm2019-01-261-2/+6
| | | | bz#2903
* Accept the host key fingerprint as a synonym for "yes" when acceptingdtucker2019-01-241-8/+11
| | | | | | an unknown host key. This allows you to paste a fingerprint obtained out of band into the yes/no prompt and have the client do the comparison for you. ok markus@ djm@
* Have progressmeter force an update at the beginning and end of eachdtucker2019-01-244-14/+11
| | | | | transfer. Fixes the problem recently introduces where very quick transfers do not display the progressmeter at all. Spotted by naddy@
* Check for both EAGAIN and EWOULDBLOCK. This is a no-op in OpenBSDdtucker2019-01-241-7/+5
| | | | | (they are the same value) but makes things easier in -portable where they may be distinct values. "sigh ok" deraadt@
* Always initialize 2nd arg to hpdelim2. It populates that *ONLY IF*dtucker2019-01-241-5/+5
| | | | | | | there's a delimiter. If there's not (the common case) it checked uninitialized memory, which usually passed, but if not would cause spurious failures when the uninitialized memory happens to contain "/". ok deraadt.
* Remove support for obsolete host/port syntax.dtucker2019-01-233-11/+16
| | | | | | | | | host/port was added in 2001 as an alternative to host:port syntax for the benefit of IPv6 users. These days there are establised standards for this like [::1]:22 and the slash syntax is easily mistaken for CIDR notation, which OpenSSH now supports for some things. Remove the slash notation from ListenAddress and PermitOpen. bz#2335, patch from jjelen at redhat.com, ok markus@
* Remove duplicate word. bz#2958, patch from jjelen at redhat.comdtucker2019-01-231-3/+3
|
* Remove 3 as a guess for possible generator during moduli generation.dtucker2019-01-231-3/+1
| | | | | It's not mentioned in RFC4419 and it's not possible for Sophie-Germain primes greater than 5. bz#2330, from Christian Wittenhorst , ok djm@ tb@
* Sanitize scp filenames via snmprintf. To do this we move thedtucker2019-01-235-44/+55
| | | | | | progressmeter formatting outside of signal handler context and have the atomicio callback called for EINTR too. bz#2434 with contributions from djm and jjelen at redhat.com, ok djm@
* allow auto-incrementing certificate serial number for certs signeddjm2019-01-232-7/+20
| | | | in a single commandline.
* move a bunch of global flag variables to main(); make the rest staticdjm2019-01-231-94/+82
|
* switch mainloop from select(2) to poll(2); ok deraadt@djm2019-01-231-22/+19
|
* pass most arguments to the KEX hash functions as sshbuf ratherdjm2019-01-235-63/+80
| | | | than pointer+length; ok markus@
* backoff reading messages from active connections when the input bufferdjm2019-01-221-4/+18
| | | | | is too full to read one, or if the output buffer is too full to enqueue a response; feedback & ok dtucker@
* add -m to usage(); reminded by jmc@djm2019-01-221-3/+4
|
* Correct some bugs in PKCS#11 token PIN handling at initial login,djm2019-01-221-12/+22
| | | | | | | the attempt at reading the PIN could be skipped in some cases especially on devices with integrated PIN readers. based on patch from Daniel Kucera in bz#2652; ok markus@
* Support keys that set the CKA_ALWAYS_AUTHENTICATE by requring adjm2019-01-221-26/+95
| | | | | | fresh login after the C_SignInit operation. based on patch from Jakub Jelen in bz#2638; ok markus
* Mention that configuration for the destination host is not applieddjm2019-01-222-4/+15
| | | | to any ProxyJump/-J hosts. This has confused a few people...
* Include -m in the synopsis for a few more commands that support itdjm2019-01-221-4/+10
| | | | | | Be more explicit in the description of -m about where it may be used Prompted by Jakub Jelen in bz2904
* print the full pubkey being attempted at loglevel >= debug2; bz2939djm2019-01-221-1/+17
|
* clarify: ssh-keygen -e only writes public keys, never privatedjm2019-01-221-2/+2
|
* mention the new vs. old key formats in the introduction and give somedjm2019-01-221-6/+18
| | | | hints on how keys may be converted or written in the old format.
* tweak previous;jmc2019-01-222-7/+7
|
* Forgot to add -J to the synopsis.tb2019-01-211-1/+2
|
* Add a -J option as a shortcut for -o Proxyjump= to scp(1) and sftp(1)tb2019-01-214-14/+43
| | | | | | to match ssh(1)'s interface. ok djm
* switch sntrup implementation source from supercop to libpqcrypto;djm2019-01-212-77/+79
| | | | | the latter is almost identical but doesn't rely on signed underflow to implement an optimised integer sort; from markus@
* add "-v" flags to ssh-add and ssh-pkcs11-helper to turn up debugdjm2019-01-215-14/+82
| | | | | | | | | | verbosity. Make ssh-agent turn on ssh-pkcs11-helper's verbosity when it is run in debug mode ("ssh-agent -d"), so we get to see errors from the PKCS#11 code. ok markus@
* fix reversed arguments to kex_load_hostkey(); manifested as errors indjm2019-01-211-2/+2
| | | | cert-hostkey.sh regress failures.
* forgot to cvs add this file in previous series of commits; grrrdjm2019-01-211-0/+322
|
* nothing shall escape this purgedjm2019-01-212-363/+0
|
* rename kex->kem_client_pub -> kex->client_pub now that KEM has been renameddjm2019-01-216-12/+12
| | | | | | to kexgen from markus@ ok djm@
* merge kexkem[cs] into kexgendjm2019-01-219-265/+61
| | | | from markus@ ok djm@
* pass values used in KEX hash computation as sshbuf rather thandjm2019-01-219-89/+98
| | | | | | pointer+len suggested by me; implemented by markus@ ok me
* remove kex_derive_keys_bn wrapper; no unused since the DH-like KEXdjm2019-01-212-20/+2
| | | | | | methods have moved to KEM from markus@ ok djm@
* use KEM API for vanilla ECDHdjm2019-01-2111-248/+193
| | | | from markus@ ok djm@
* use KEM API for vanilla DH KEXdjm2019-01-2112-371/+155
| | | | from markus@ ok djm@
* use KEM API for vanilla c25519 KEXdjm2019-01-2112-309/+171
|
* Add support for a PQC KEX/KEM: sntrup4591761x25519-sha512@tinyssh.orgdjm2019-01-2117-41/+1669
| | | | | | | | | using the Streamlined NTRU Prime 4591^761 implementation from SUPERCOP coupled with X25519 as a stop-loss. Not enabled by default. introduce KEM API; a simplified framework for DH-ish KEX methods. from markus@ feedback & ok djm@
* factor out kex_verify_hostkey() - again, duplicated almost exactlydjm2019-01-216-59/+27
| | | | | | across client and server for several KEX methods. from markus@ ok djm@
* factor out kex_load_hostkey() - this is duplicated in both the client anddjm2019-01-216-55/+33
| | | | | | server implementations for most KEX methods. from markus@ ok djm@
* factor out kex_dh_compute_key() - it's shared between plain DH KEX anddjm2019-01-217-163/+101
| | | | | | DH GEX in both the client and server implementations from markus@ ok djm@
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.