When getpwnam(3) reaches out to YP, it calls clntudp_create(3) with a
pre-initialized ypconnect(2) socket. That calls clntudp_bufcreate(),
which contains code checking if the socket and address are

When getpwnam(3) reaches out to YP, it calls clntudp_create(3) with a
pre-initialized ypconnect(2) socket. That calls clntudp_bufcreate(),
which contains code checking if the socket and address are configured..
If not, socket(2) is called, or an address allocation is performed via
the portmapper (which calls a whole lot more code).

Split clnt_udp.c into two .c files (which will compile as seperate .o
files), and create a new libc-private clntudp_bufcreate_simple() function
which skips the socket and address work.

Result: In most static binaries, this reduces the text segment by
~100K, and removes 5-7 system call stubs -- which might matter for
non-pledged binaries with otherwise lack socket(2).
ok millert jmatthew

show more ...

1) The yp_bind/yp_unbind and internal _yp_dobind/_yp_unbind sequences shared
dom_binding structs between threads, which is unsafe -- example, dom_vers
signalled retry events, and structs+socket would

1) The yp_bind/yp_unbind and internal _yp_dobind/_yp_unbind sequences shared
dom_binding structs between threads, which is unsafe -- example, dom_vers
signalled retry events, and structs+socket would get deallocated in _yp_unbind.
Change all yp_first (and similar) functions to understand that _yp_dobind now
provides a private dom_binding and socket, which must be released using
_yp_unbind. Use similar methods in the one-step yp_all function.
2) domainname caching in get* is not neccessary now that the domainname cannot
change relative to ypconnect(2)'s decisions.
Many fields in dom_binding struct become unused, so delete them.
ok jmatthew, also tested by miod

show more ...

avoid use after free
ok deraadt@

ypbinding should not be intrude to application namespace.
spotted by guenther

Rather than opening the binding file, checking for advisory lock, reading a
piece of it for the address, opening a socket, and providing the address to
the RPC clnt layer.. do all these steps with th

Rather than opening the binding file, checking for advisory lock, reading a
piece of it for the address, opening a socket, and providing the address to
the RPC clnt layer.. do all these steps with the magic system call ypconnect(2)
which performs these steps without other system calls, and provides a socket
which is not readily abuseable for other purposes.
ok jmatthew, miod

show more ...

Stop publicly declaring _yp_dobind() and struct dom_binding, closing out
a rant Theo wrote 24 years ago. Mark __ypexclude_{add,is,free}() as hidden

"get off my lawn!" deraadt@

_yp_bind(), _ypbindlist, _yp_domain, and _yplib_timeout are no longer
exported, so declare them as hidden to avoid pointless GOT relocations

ok millert@ deraadt@

remove the "YP server for domain %s not responding, still trying" code.
For years, it talked to stderr. That was wrong. Then it was converted
to opening /dev/tty, which is also wrong (pledge says s

remove the "YP server for domain %s not responding, still trying" code.
For years, it talked to stderr. That was wrong. Then it was converted
to opening /dev/tty, which is also wrong (pledge says so). Upon
reconsideration people in these more modern times have adapted to all
sorts of services not being available, so axe the alert and retry
silently.
ok beck

show more ...

Wrap _yp_check() and _yp_unbind() so internal calls go direct and they're weak

Wrap <rpcsvc/ypclnt.h> so that calls go direct and the symbols are all weak

rather than assuming stderr is alive, open + dprintf to /dev/tty
ok guenther millert

Move to the <limits.h> universe.
review by millert, binary checking process with doug, concept with guenther

change _yp_domain[] to HOST_NAME_MAX+1, and re-arrange those who include
it to get limits.h early enough

use calloc, from Benjamin Baier

Use PATH_MAX, NAME_MAX and LOGIN_NAME_MAX not MAXPATHNAMELEN,
MAXNAMLEN or MAXLOGNAME where possible. OK deraadt@

use FD_CLOEXEC instead of 1; from David Hill

ok otto

improve yp_bind(3) error reporting after clnttcp_create(3) failure:
map ECONNREFUSED -> YPERR_YPBIND, ENOMEM -> YPERR_RESRC, else YPERR_YPERR
while here, malloc(3) failure should raise YPERR_RESRC, n

improve yp_bind(3) error reporting after clnttcp_create(3) failure:
map ECONNREFUSED -> YPERR_YPBIND, ENOMEM -> YPERR_RESRC, else YPERR_YPERR
while here, malloc(3) failure should raise YPERR_RESRC, not YPERR_YPERR
ok millert@

show more ...

Check snprintf(3) return value for error or truncation.
Mostly path construction, where truncation could be bad.

ok and input from deraadt@ millert@ ray@

- remove rcsid
- put all the YPMATCHCACHE code into a #ifdef, proper
- remove all extra header files that are not needed

okay deraadt@

remove terms 3 & 4 on my own software

ansi

more pid_t use

use strlcpy vs strncpy+a[len-1]='\0'. millert@ ok.

More <sys/file.h> vs. <fcntl.h> and open() flags fixes.

possible buf oflow