Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
|
|
Sun's SKIP support added.
|
|
|
|
master NIS server. I gave yp_bind_host an extra argument to select
TCP/UDP transport.
|
|
|
|
|
|
|
|
|
|
Added explicit dependencies on yacc generated files. From Luke Mewburn
<lukem@supp.cpr.itg.telecom.com.au>. Fixes PR #2032.
|
|
Fixed Typo. Closes PR 2001 from Thorsten Frueauf <frueauf@ira.uka.de>
Keep them coming!
|
|
Spelling correction
|
|
Fix dangling 'The'; From Mike Long <mike.long@analog.com>
|
|
fix pr/1226 from David Gluss: allow rdate to use adjtime(2)
|
|
Be explicit about the fact that ip-up and ip-down are executed with
a null environment. Closes PR #1866 by John M Vinopal.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
in xdr_ypresp_key_val() invert the order of loading the key and value
(rpcgen says [1] stat, [2] key, [3] value, but the data on the wire
comes in as [1] stat, [2] val, [3] key... you can't fix this in yp.x
because if you invert the ordr in yp.x not only does it invert the order
in ypserv_xdr.c, but it also inverts it in the structure in yp.h which
cancels out the inversion in ypserv_xdr.c. i guess this means it
is not possible to have a yp.x that descrives the YP protocol properly?).
***NOTE*** this file is also needed because some of the YP functions in
libc are broken. For example, compare src/lib/libc/yp/xdryp.c's version
of xdr_domainname() to the one in this file generated by rpcgen. Note
that xdryp.c calls:
if (!xdr_string(xdrs, &objp, YPMAXDOMAIN)) { /* xdryp.c */
while ypserv_xdr.c (generated by rpcgen) calls:
if (!xdr_string(xdrs, objp, YPMAXDOMAIN)) { /* rpcgen */
(note the lack of "&" in the second line). Clearly libc is wrong and
needs to be fixed (XXX). (Of course changing libc could break other
YP stuff, which is why I haven't attemped it... :) ) When libc is
fixed, it should be possible to phase out this file, I think.
|
|
- convert to new yplog() function and clean up log message formats
- remove transp args from all _svc functions since you can get it
from rqstp->rq_xprt, and rpcgen seems to favor this format
- delete all the calls to svc_sendreply() and svc_freeargs() as
this can be done in one place in ypserv.c (except for the "all"
request which forks off a child process)
- avoid possible sprintf() buffer overflows by using snprintf()
- use MAXPATHLEN, rather than a hardcoded 255.
- eliminate un-needed static variables
- avoid possible off by one buffer overflow in tid/prog/port and use
snprintf() just to be sure.
- random code cleanup and simplification
|
|
- use queue.h rather than do by hand
- fix up logging to use new yplog
- fix a few off by one errors
- don't double malloc the DBM structure
- avoid possible sprintf buffer overflows
- random code cleanup
|
|
highlights:
- fixed yp first call (uses ypreq_nokey rather than ypreq_key, despite
what yp.x says)
- cleanup of looging
- removed SVCXPRT args to "local" since you can get them from svc_req
and this makes it look more like rpcgen's output
- use new yplog()
- add call to db init
|
|
|
|
|
|
|
|
|
|
|
|
|
|
- don't store null in the database
- use MAXPATHLEN
- avoid sprintf() buffer overflows
- minor cleanups
|
|
(made it more printf like)
|
|
(we don't store nulls in the db [any more])
|
|
|
|
|
|
Cross-reference fsinfo(8) and mk-amd-map(8), as per PR #1057.
From David Brownlee <abs@mono.city.ac.uk>.
|
|
(James Jegers; PR#592).
|