summaryrefslogtreecommitdiff
path: root/src/lib/libc/net/rcmd.c (follow)
Commit message (Collapse)AuthorAgeFilesLines
* do not trash errno before using it, do not pass garbage args to getnameinfoderaadt2005-03-081-7/+10
| | | | ok henning andrushock@korovino.net
* fd leak; andrushock@korovino.netderaadt2005-03-071-2/+4
|
* remove NI_WITHSCOPEID (which is not standard)itojun2004-11-171-19/+1
|
* Do check for current pointer vs. buffer end before touching anymillert2003-09-251-5/+5
| | | | | elements in the buffer. Fixes an out of bounds access. From aaron@; OK deraadt@
* license oops; from teduderaadt2003-07-111-1/+4
|
* fix various 3/4 licenses according to "terms" filederaadt2003-06-031-9/+1
|
* more writeable -> writable by torhhenning2002-12-151-2/+2
|
* use socklen_t where needed; henning pvalchev okderaadt2002-09-061-2/+2
|
* try to use strlcpy and snprintf more; ok variousderaadt2002-05-241-3/+2
|
* more strlcpy and snprintfderaadt2002-05-221-3/+3
|
* Manual cleanup of remaining userland __P use (excluding packages maintained ↵millert2002-02-171-3/+3
| | | | outside the tree)
* Part one of userland __P removal. Done with a simple regexp with some minor ↵millert2002-02-161-4/+4
| | | | hand editing to make comments line up correctly. Another pass is forthcoming that handles the cases that could not be done automatically.
* Replace the deprecated BSD sigsetmask/sigblock/sigpause functions with their ↵millert2001-09-041-7/+9
| | | | POSIX counterparts.
* use strlcpy vs strncpy+a[len-1]='\0'. millert@ ok.lebel2001-06-271-3/+2
|
* Make sure $RSH is not the empty string.millert2001-02-101-2/+2
|
* always pass salen as separate argument, so that it can be friendly withitojun2000-02-251-20/+22
| | | | systems with pre-4.2 sockaddrs.
* do not perform sleep() every time we get ECONNREFUSED.itojun2000-02-181-6/+12
| | | | | | try all the set of addresses before go to sleep() and retry. not sure if we still need sleep() - retry logic. why is it so persistent?
* visit next addrinfo structure when rresvport_af() fails.itojun2000-01-301-4/+9
| | | | | this should fix situation like: - try to rcmd() to dual stack node from IPv4-only node/kernel
* add IPv6-ready rcmd() friends.itojun2000-01-271-76/+221
| | | | | | | | rcmd(): IPv4 only rcmd_af(): af independent ruserok(): af independent iruserok(): IPv4 only iruserok_sa(): af independent
* rresvport_af() and bindresvport_af()deraadt1999-12-161-33/+1
|
* some -Wallmillert1998-03-191-3/+2
|
* correct bounds checkderaadt1998-02-121-2/+2
|
* copyrightderaadt1998-02-121-2/+2
|
* use fgetln() instead of fgets() so that we can catch \0 in the .rhostsderaadt1998-02-111-13/+14
| | | | | file. Thanks to fc@parkone.ci.oakland.ca.us for lots of testing and diagnosis help.
* increased paranoia about .rhosts validityderaadt1998-02-111-3/+7
|
* Clean up some -Wall flowers.millert1997-07-091-1/+2
|
* if we find an unprintable char, bail totallyderaadt1997-07-061-1/+4
|
* indentderaadt1997-06-291-2/+2
|
* use bzero instead of FD_ZEROdm1997-06-041-4/+4
|
* howmany(x+1, NFDBITS)deraadt1997-06-031-2/+2
|
* malloc fd_set dynamically, for very large fd usesderaadt1997-05-281-7/+14
|
* Use in_addr_t not u_long and use in_port_t in some cases insteadmillert1997-04-051-6/+6
| | | | of u_short or u_int16_t.
* never hand over a nonresv port, even if asked. problem reported by ↵deraadt1997-01-251-2/+2
| | | | gmeinerj@ZUMtOBEL.co.at
* ignore $RSH if we are setuidmillert1996-09-051-2/+2
|
* proper typesderaadt1996-09-031-1/+2
|
* Looks at RSH envariable and calls rcmdsh() with that value if set.millert1996-09-021-2/+11
|
* i have changed this enough to put copyright on itderaadt1996-09-011-1/+3
|
* check right variablederaadt1996-08-311-2/+2
|
* avoid port 1023 morederaadt1996-08-301-7/+11
|
* call rcmdsh() if euid != 0 and port is shell/tcp. This lets usmillert1996-08-301-1/+9
| | | | get away with fewer setuid binaries.
* Fix RCS idstholo1996-08-191-7/+1
| | | | Make sure everything uses {SYS,}LIBC_SCCS properly
* seed rresvport() for aux channel with a known taken port; do not ask forderaadt1996-08-071-0/+6
| | | | next lower port because we might get it (and be predictable!)
* if port 20 incoming, kill it and try againderaadt1996-08-051-0/+9
|
* ensure bindresvport() is told to find a portderaadt1996-07-291-0/+1
|
* rresvport() in terms of bind() & bindresvport()deraadt1996-07-291-14/+12
|
* bzero sockaddr_inderaadt1996-07-221-0/+1
|
* bzero sockaddr before usederaadt1996-07-221-0/+1
|
* support # commentsderaadt1996-07-201-0/+2
|
* iruserok() can finally safely use a u_int32_t for the ip addrderaadt1996-06-101-2/+2
|
* fix long/u_int32_t botch; do gethostbyname() only if netgroups are going toderaadt1995-11-221-55/+94
| | | | | | be used and do not abort; check for A->PTR->A loop otherwise only match on #.#.#.# addresses; do not assume !0 is 1 as in ANSI C. netbsd pr#1683 led to finding these problems.
generated by cgit v1.2.3-55-g6feb (git 2.39.1) at 2025-03-22 03:41:09 +0000