summaryrefslogtreecommitdiff
path: root/usr.bin/nc
AgeCommit message (Collapse)Author
2015-09-25avoid trailing .Ns, reduce .Xo and .Sm, drop redundant .BkIngo Schwarze
2015-09-13display negotiated TLS version and cipher suite in verbose mode.Bob Beck
ok jsing@
2015-09-12Adapt to just committed libtls api changeBob Beck
2015-09-12use SOCK_CLOEXEC instead of fnctl; ok guenther beck jsingTheo de Raadt
2015-09-12fix previous;Jason McIntyre
2015-09-11spaces found during a readTheo de Raadt
2015-09-11Add TLS suppport to nc. Provides a useful little test and script tool.Bob Beck
ok jsing@ bluhm@
2015-09-08Netcat could hang during write(2) although poll(2) reports that theAlexander Bluhm
socket is writeable. This happens because netcat tries to write more than the low water mark of the socket write buffer. With a non-blocking socket you may get a short write, otherwise it blocks. The latter could cause a total hang of the netcat process depending on the upper protocol. So make the network connection non-blocking. OK claudio@ millert@
2015-09-03synchronize synopsis and usage.Igor Sobrado
2015-07-26remove unused variableCharles Longeau
ok tedu@
2015-03-26The code in socks.c writes multiple times in a row to a socket. If theTobias Stoeckmann
socket becomes invalid between these calls (e.g. connection closed), write will throw SIGPIPE. With this patch, SIGPIPE is ignored so we can handle write's -1 return value (errno will be EPIPE). Ultimately, it leads to program exit, too -- but with nicer error message. :) with input by and ok djm
2015-03-26Fix error message in case of write failure.Tobias Stoeckmann
ok djm
2015-03-26Check for short writes in fdpass(). Clean up while at it.Tobias Stoeckmann
ok djm
2015-03-26Sync camel case of "ProxyUseFdpass" with ssh_config(5).Tobias Stoeckmann
ok djm
2015-02-14Support for nc -T on IPv6 addresses.Jeremie Courreges-Anglas
ok sthen@
2014-10-30my mistake. we already did increase buffers to 16k; increasing to 64kTed Unangst
would be the next stage of embiggening. restore 16k.
2014-10-30rework the poll loop to poll in both directions so it doesn't get stuckTed Unangst
if one pipe stalls out. from a diff by Arne Becker. (buffer size left alone for now)
2014-10-26POLLIN is not guaranteed to be set in revents for EOF so check forTodd C. Miller
POLLHUP too. OK deraadt@
2014-10-24Remove unnecessary include: netinet/in_systm.h is not needed by theseLawrence Teo
programs. ok deraadt@ millert@
2014-07-20Make sure the correct errno is reported by warn* or err* and notPhilip Guenther
the errno of an intervening cleanup operation like close/unlink/etc. Diff from Doug Hogan (doug (at) acyclic.org)
2014-06-10stick with 16k buffers for a little while to avoid bufferbloat.Ted Unangst
atomicio writing out 64k in one direction will cause traffic in the other direction to stall until it's complete. discussion with deraadt
2014-06-10increase buffer size to 64k, and actually use it. ok deraadtTed Unangst
from John-Mark Gurney
2014-05-20Use errc/warnc to simplify code.Philip Guenther
Also, in 'ftp', always put the error message last, after the hostname/ipaddr. ok jsing@ krw@ millert@
2014-03-12Unbreak nc -6 -l. Don't retrieve and thus later set the routing tableJeremie Courreges-Anglas
unless -V is passed (intent of the previous commit), and use SOL_SOCKET instead of IPPROTO_IP to set the rtable in local_listen(). ok sthen@
2014-02-26Once more, the default routing table id is inherited from the processClaudio Jeker
like for any other process as well. OK by many
2014-02-10one of the examples needs -N to work again;Jason McIntyre
the paper trail appears to be: reported in feebsd pr docs/185353 by rol robert-eckardt de fix suggested by peter wemm diff submitted to tech by allan jude
2013-10-26Only use setsockopt(..SO_RTABLE..) if the -V flag is given to nc/telnet,Stuart Henderson
same style as traceroute6 (change to int and use -1 as a flag, so rtable 0 can still be used as an explicit parameter).
2013-10-21Oups, a little bit overzealous. If we "route -T4 exec telnet foo", thenPeter Hessler
make sure telnet runs in rdomain 4 as expected. Same for nc.
2013-10-21Set the requested rdomain on the socket instead of only on the IPv4 part.Peter Hessler
Also, switching to rdomain zero is a legit option so we should allow that.
2013-08-20escape not-full-stop-dot at eol;Jason McIntyre
2013-08-20add -F to usage() and help();Jason McIntyre
ok djm
2013-08-20add -F flag to enabled fd-pass mode: establish connection and passDamien Miller
connected socket to stdout. This is useful in proxy mode to establish a connection for use by ssh in conjunction with its new ProxyUseFDPass option; ok markus@
2013-07-16use .Mt for email addresses; from Jan Stary <hans at stare dot cz>; ok jmc@Ingo Schwarze
2013-04-29use FD_CLOEXEC instead of 1; from David HillOkan Demirmen
ok otto
2013-03-20Don't shutdown nc(1)'s network socket when stdin closes. Matches *Hobbit*'sStuart Henderson
original netcat and GNU netcat; revert to old behaviour with the new -N flag if needed. After much discussion with otto deraadt tedu and Martin Pelikan. ok deraadt@
2013-03-12All accept() errors are considered fatal, until someone gives a differentTheo de Raadt
reason. No code changed, just documenting it...
2012-12-04remove some unnecessary sys/param.h inclusionsTheo de Raadt
2012-07-07Allow UDP server to receive datagrams from multiple socket pairs with -kChristiano F. Haesbaert
flag. Prompted by a question from dsp at 2f30 dot org, diff from Lazarom Koromil with a few tweaks by me, many thanks. ok mikeb@ nicm@ haesbaert@
2012-07-07Report incoming connections when -l is specified with -v.Christiano F. Haesbaert
From Ricky Zhou with a few tweaks by me. ok henning@ haesbaert@
2012-04-01error out if accept() failsTheo de Raadt
2012-03-17remove IP_JUMBO, SO_JUMBO, and RTF_JUMBO.David Gwynne
no objection from mcbride@ krw@ markus@ deraadt@
2012-03-08fix format strings for size_tMarc Espie
ok okan@
2012-02-09Update comments. ok deraadt@lum
2012-02-09Remove unused protocol value check. ok brynet@lum
2012-02-07Expand the CAVEAT section to more fully describe the behaviour oflum
the -uz combination. This diff stems from a diff by brynet@ sent to tech@ help from jmc@ regarding wording.
2011-10-04change -w to apply to the connection as well. manpage bit from jmc@Federico G. Schwindt
nicm@ ok.
2011-09-17Standarize the ToS option across nc/ping/traceroute so that they'llChristiano F. Haesbaert
accept the same values as pf.conf. It accepts decimal, hexadecimal and the dscp/tos keywords. The ping option was ripped of in SMALL. ok mcbride@ sthen@
2011-06-21Convert SO_RTABLE's protocol level to the SOL_SOCKET; ok claudioMike Belopuhov
2011-02-12fix from pr 6207. a bit more of an explanation: we write the correctOkan Demirmen
number of bits when connecting via a SOCKS 5 proxy over ipv6, but we also need to read the same number depending on the received address type. this issue is not noticeable with ssh's SOCKS 5 support since it always set the address type as ipv4. this fixes connections via SOCKS 5 proxies which set their address type as ipv6 when using ipv6. after review with, and ok, nicm@
2011-01-09Minor tweaks to nc(1) man page and usage.Jeremy Evans
OK jmc@, nicm@, tedu@