CVS log for src/usr.sbin/rpcbind/util.c

Up to [cvs.NetBSD.org] / src / usr.sbin / rpcbind

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

Include explicitly <rump/rump_syscallshotgun.h> for previous indirect users
via <rump/rump.h>.

Sync with HEAD

Sync with HEAD

Pull up following revision(s) (requested by christos in ticket #1167):

	usr.sbin/rpcbind/util.c: revision 1.23

- fix wrong size allocation that triggers buffer overflow
- remove unneeded casts and assertions

XXX: pullup-8

- fix wrong size allocation that triggers buffer overflow
- remove unneeded casts and assertions
XXX: pullup-8

Synch with HEAD

Handle memory allocations and error exit codes consistently.

merge FreeBSD changes:
- fixes CVE-2015-7236
- adds -h hostip to bind, -6 for only ipv6 access, -a for abort gracefully
- documents -w (warmstart)
XXX: should fix warmstart file to go to /var/run instead of /tmp

merge RUMP changes.

perror -> warn

CID 1018907: Freeifaddrs

Rebase to HEAD as of a few days ago.

sync with head.

for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.

this commit was splitted into small chunks to avoid
a limitation of cvs.  ("Protocol error: too many arguments")

use new scopeid functions

fix -Wsign-compare issues

sync with head.

Remove clause 3 and 4 from TNF licenses

sync with HEAD

Depracate ND_DEBUG and BIND_DEBUG in favour of RPCBIND_DEBUG.
Add commented out entries for RPCBIND_DEBUG and SVC_RUN_DEBUG in Makefile.

- WARNS = 4
- De __P
- While starting up, send messages to stderr not syslog.
- Fix error messages
- Don't play with umask and forget to restore it on error

Coverity CID 783: Add an assertion for a NULL deref.

Support MKINET6=no building after the -Wuninitalized fix.
Noted by Jukka Salmi in email.

appease gcc -Wuninitialized

null commit

poll.h, not sys/poll.h

pullup usr.sbin/rpcbind/util.c 1.5 (bouyer)

Move routines which fill in tbuf in addrmerge(), making sure tbuf is always
initialised.

Pull up revisions 1.6-1.7 (requested by fvdl):
  Fix some minor security-related issues.

For the local case, return serv_uaddr in addrmerge, not the address
by which the client contacted rpcbind.

Also fix a missing \n in a debug printf.

From Martin Blapp, mb@imp.ch

For an AF_LOCAL connection, don't attempt to do the address merge dance.
While here, fix some memory leaks in error cases.

Move routines which fill in tbuf in addrmerge(). The fixes in 1.4 pick a new
interface but don't fill in tbuf which results in clients getting garbage back
in part of the answer. This makes sure there's something in tbuf.

At some point this whole logic should be looked at, as just picking a random
local up interface for non-local requests doesn't feel right.

Pull up version 1.4:
	When merging addresses, and no direct match is found, pick the first
	configured interface that is preferably not local or point to point.
	This appears to be what Solaris does as well.

When merging addresses, and no direct match is found, pick the first
configured interface that is preferably not local or point to point.
This appears to be what Solaris does as well.

Pull up revision 1.3:
  - Fix memory leaks.

Sync w/ netbsd-1-5-base.

file util.c was added on branch minoura-xpg4dl on 2000-06-22 18:01:12 +0000

Make sure to free the netconfig entry in addrmerge().

Need <string.h> for LP64 systems.

rpcbind, the successor to portmap. Taken from Sun TI-RPC source code,
then modified. This still contains some ugly special casing for
procedures in the security section, but I haven't found a nicer way
so far.