]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
27 years agoMake error checking less zealous to handle devices like /dev/null
John Birrell [Tue, 1 Apr 1997 22:49:58 +0000 (22:49 +0000)]
Make error checking less zealous to handle devices like /dev/null
which don't provide a non-blocking interface.

This is a short term "fix" which changes a half-lose to a half-win.
The thread that accesses a device that does not provide a non-blocking
interface will block for its time slice.

A medium term solution would be to use rfork. A long-term solution
would be some sort of kernel thread/SMP implementation.

27 years agoAdd parentheses to make blocking mode work.
John Birrell [Tue, 1 Apr 1997 22:44:18 +0000 (22:44 +0000)]
Add parentheses to make blocking mode work.

27 years agoBe more specific as to which flags may not be turned off when the
Mike Pritchard [Tue, 1 Apr 1997 20:41:04 +0000 (20:41 +0000)]
Be more specific as to which flags may not be turned off when the
system is running in secure mode.

Obtained from: NetBSD PR# 3299

27 years agoUse MAXHOSTNAMELEN, as opposed to an arbitrary number, and ensure
Joerg Wunsch [Tue, 1 Apr 1997 20:39:59 +0000 (20:39 +0000)]
Use MAXHOSTNAMELEN, as opposed to an arbitrary number, and ensure
correct termination if it overflows.

Closes PR # misc/2982.

Submitted by: Drew Derbyshire <ahd@kew.com>

27 years agoAdd od(4). Closes PR# 3170.
Mike Pritchard [Tue, 1 Apr 1997 19:52:37 +0000 (19:52 +0000)]
Add od(4).  Closes PR# 3170.

Submitted by: Shunsuke Akiyama <akiyama@jp.freebsd.org>

27 years agoAdd an entry for the newer Philips CDD26xx family CD-R drives. They
Joerg Wunsch [Tue, 1 Apr 1997 19:28:03 +0000 (19:28 +0000)]
Add an entry for the newer Philips CDD26xx family CD-R drives.  They
come out with PHILIPS as the vendor string, as opposed to IMS for the
older ones.

Submitted by: "Brian N. Handy" <handy@sag.space.lockheed.com>

27 years agoGrammar fix.
Mike Pritchard [Tue, 1 Apr 1997 19:09:20 +0000 (19:09 +0000)]
Grammar fix.

27 years agoXref the new issetugid(2) man page where appropriate.
Mike Pritchard [Tue, 1 Apr 1997 18:58:38 +0000 (18:58 +0000)]
Xref the new issetugid(2) man page where appropriate.

Also add additional set*id(2) xrefs where appropriate.

27 years agoDon't xref setregid(2) twice - one of them should be setreuid(2).
Mike Pritchard [Tue, 1 Apr 1997 18:50:56 +0000 (18:50 +0000)]
Don't xref setregid(2) twice - one of them should be setreuid(2).

27 years agoFix one very minor mdoc problem. Gentle enough, Peter? :-)
Mike Pritchard [Tue, 1 Apr 1997 18:45:57 +0000 (18:45 +0000)]
Fix one very minor mdoc problem.  Gentle enough, Peter? :-)

27 years agoRemove a doubled word. Closes PR# 3163.
Mike Pritchard [Tue, 1 Apr 1997 18:28:25 +0000 (18:28 +0000)]
Remove a doubled word.  Closes PR# 3163.

Submitted by:  Kazuo Horikawa <horikawa@jp.freebsd.org>

27 years agoFix a minor grammar problem.
Mike Pritchard [Tue, 1 Apr 1997 18:06:33 +0000 (18:06 +0000)]
Fix a minor grammar problem.

27 years agoMLINK uu_lockerr(3).
Mike Pritchard [Tue, 1 Apr 1997 17:44:58 +0000 (17:44 +0000)]
MLINK uu_lockerr(3).

27 years agoMdoc police.
Mike Pritchard [Tue, 1 Apr 1997 17:44:31 +0000 (17:44 +0000)]
Mdoc police.

27 years agoMake mount_nfs use reserved ports by default.. Mounts already use
Guido van Rooij [Tue, 1 Apr 1997 17:20:17 +0000 (17:20 +0000)]
Make mount_nfs use reserved ports by default.. Mounts already use
a reserved port, so why not the nfs rpc's themselves?
With user allowed mounts, this perhaps needs a closer look, but
on the other hand, a user could already specify the flag.
If normal users should not be able to use resserved ports, the kernel
should check for the flag at mount time.

27 years agoRemoved a wrong comment of mine.
Bruce Evans [Tue, 1 Apr 1997 16:22:31 +0000 (16:22 +0000)]
Removed a wrong comment of mine.

Removed unused #includes.

27 years agoFixed gratuitous ANSIisms.
Bruce Evans [Tue, 1 Apr 1997 16:13:31 +0000 (16:13 +0000)]
Fixed gratuitous ANSIisms.

Removed unused declarations.

27 years agoFixed gratuitous ANSIisms.
Bruce Evans [Tue, 1 Apr 1997 15:22:59 +0000 (15:22 +0000)]
Fixed gratuitous ANSIisms.

Removed trailing newline from panic messages.

27 years agoUse __i386__ instead of i386 in ifdefs.
Bruce Evans [Tue, 1 Apr 1997 15:10:38 +0000 (15:10 +0000)]
Use __i386__ instead of i386 in ifdefs.

Don't compile unused (debugging?) functions.

27 years agoFixed gratuitous ANSIisms.
Bruce Evans [Tue, 1 Apr 1997 14:31:06 +0000 (14:31 +0000)]
Fixed gratuitous ANSIisms.

27 years agoDon't fail when the vfs.nfs.nfs_privport sysctl doesn't exist
Bruce Evans [Tue, 1 Apr 1997 14:15:30 +0000 (14:15 +0000)]
Don't fail when the vfs.nfs.nfs_privport sysctl doesn't exist
(presumably because the kernel is old).  Moved the declaration of a
variable realated to this sysctl outside of an unrelated ifdef.

Not fixed:
- this sysctl is badly named (nfs occurs twice).
- it's silly to have for FreeBSD in FreeBSD code, especially when
  only half of the FreeBSD-dependent code is ifdefed.

27 years agoSimplified install rule.
Bruce Evans [Tue, 1 Apr 1997 13:53:16 +0000 (13:53 +0000)]
Simplified install rule.

27 years agoUse OID_AUTO instead of magic number for the Lite2 sysctl debug.busyprt.
Bruce Evans [Tue, 1 Apr 1997 13:05:34 +0000 (13:05 +0000)]
Use OID_AUTO instead of magic number for the Lite2 sysctl debug.busyprt.

Removed declaration of vfs_unmountroot() again.

Staticized vgonel().

27 years agoconsistency update.
Adam David [Tue, 1 Apr 1997 12:34:13 +0000 (12:34 +0000)]
consistency update.
since we have "unsizeable" also have "sizeable"

27 years agoSubmitted by: Sergei Chechetkin <csl@whale.sunbay.crimea.ua>
David Nugent [Tue, 1 Apr 1997 11:54:06 +0000 (11:54 +0000)]
Submitted by: Sergei Chechetkin <csl@whale.sunbay.crimea.ua>

Fix incorrect bracket nesting. Closes PR#3144.

27 years agoUse OID_AUTO instead of magic number for the old sysctl debug.rcluster.
Bruce Evans [Tue, 1 Apr 1997 11:48:30 +0000 (11:48 +0000)]
Use OID_AUTO instead of magic number for the old sysctl debug.rcluster.
The magic number conflicted with the rotting disabled one in ext2fs for
debug.doasyncfree.

Removed messy debugging variable/constant/sysctl debug.doreallocblks.
Lite2 removed it, and we don't use the code that it controls.

27 years agoUse OID_AUTO instead of magic number for old sysctl debug.elf_trace. The
Bruce Evans [Tue, 1 Apr 1997 10:41:48 +0000 (10:41 +0000)]
Use OID_AUTO instead of magic number for old sysctl debug.elf_trace.  The
magic number conflicted with the one for the Lite2 sysctl debug.busyprt.

Staticized some variables.

Removed unused #includes.

27 years agoUse OID_AUTO instead of a magic number for Lite2 sysctl debug.lockf_debug.
Bruce Evans [Tue, 1 Apr 1997 10:30:06 +0000 (10:30 +0000)]
Use OID_AUTO instead of a magic number for Lite2 sysctl debug.lockf_debug.

Removed unused #includes.

27 years agoFixed commented-out Lite2 sysctl debug.lockpausetime.
Bruce Evans [Tue, 1 Apr 1997 10:18:51 +0000 (10:18 +0000)]
Fixed commented-out Lite2 sysctl debug.lockpausetime.

Removed unused #includes.

27 years agoSync from 2.2
Jordan K. Hubbard [Tue, 1 Apr 1997 09:53:32 +0000 (09:53 +0000)]
Sync from 2.2

27 years agoRemoved potentially harmful garbage <vm/lock.h> and fixed bogus
Bruce Evans [Tue, 1 Apr 1997 08:39:07 +0000 (08:39 +0000)]
Removed potentially harmful garbage <vm/lock.h> and fixed bogus
use of it.  It was actually harmless because the use was null due
to fortuitous include orders and identical (wrong) idempotency
macros.

27 years agoRemoved nested include of <ufs/ufs/dir.h>. Use the pre-Lite2 hack of
Bruce Evans [Tue, 1 Apr 1997 08:02:00 +0000 (08:02 +0000)]
Removed nested include of <ufs/ufs/dir.h>.  Use the pre-Lite2 hack of
defining doff_t both here and in <ufs/ufs/dir.h> so that this file
is independent of <ufs/ufs/dir.h>.  It still has old prerequisites
<sys/param.h> and <ufs/ufs/quota.h>, and a new Lite2 prerequisite of
<sys/lock.h>, sigh.

This might fix lsof, which was broken by namespace pollution giving
conflicting definitions of DIRBLKSIZ.

27 years agoFixed missing DPADD. Removed now-redundant SRCS.
Bruce Evans [Tue, 1 Apr 1997 07:43:14 +0000 (07:43 +0000)]
Fixed missing DPADD.  Removed now-redundant SRCS.

27 years agoFixed DPADD.
Bruce Evans [Tue, 1 Apr 1997 07:37:28 +0000 (07:37 +0000)]
Fixed DPADD.

27 years agoAdded pop3pw amandaidx amidxtape
Paul Traina [Tue, 1 Apr 1997 06:22:22 +0000 (06:22 +0000)]
Added pop3pw amandaidx amidxtape

27 years agoCorrect the name of a header file that is referenced.
John Polstra [Tue, 1 Apr 1997 04:04:09 +0000 (04:04 +0000)]
Correct the name of a header file that is referenced.

27 years agoFix the typo.
Hiroyuki Hanai [Tue, 1 Apr 1997 02:38:01 +0000 (02:38 +0000)]
Fix the typo.
Submitted by: Nishika <nishika@cheerful.com>

27 years agoFix the output format for numbers >= 1E99.
Jean-Marc Zucconi [Tue, 1 Apr 1997 02:32:41 +0000 (02:32 +0000)]
Fix the output format for numbers >= 1E99.
Closes PR bin/648.

27 years agoCatch up with the original.
Hiroyuki Hanai [Tue, 1 Apr 1997 02:32:28 +0000 (02:32 +0000)]
Catch up with the original.

27 years agoAdd the -DPedantic flag. This is used in wref.c only.
Jean-Marc Zucconi [Tue, 1 Apr 1997 02:27:11 +0000 (02:27 +0000)]
Add the -DPedantic flag. This is used in wref.c only.

27 years agoMerge in bugfixes from 2.2
Paul Traina [Tue, 1 Apr 1997 02:08:07 +0000 (02:08 +0000)]
Merge in bugfixes from 2.2

27 years agoUse mediaClose instead of shutdown
Paul Traina [Tue, 1 Apr 1997 02:00:42 +0000 (02:00 +0000)]
Use mediaClose instead of shutdown

27 years agoAdd portmapflags and syslogd flags.
Paul Traina [Tue, 1 Apr 1997 01:00:35 +0000 (01:00 +0000)]
Add portmapflags and syslogd flags.
Reviewed by: jkh

27 years agoRemove the syslog stuff, and allow various return values
Brian Somers [Mon, 31 Mar 1997 22:51:00 +0000 (22:51 +0000)]
Remove the syslog stuff, and allow various return values
in uu_lock().  Add uu_lockerr() for turning the results of
uu_lock into something printable.  Remove bogus section in man page
about race conditions allowing both processes to get the lock.
Include libutil.h and use uu_lock() correctly  where it should.

Suggested by: ache@freebsd.org

27 years agoAdded new file bsd-family-tree
Andreas Klemm [Mon, 31 Mar 1997 20:18:55 +0000 (20:18 +0000)]
Added new file bsd-family-tree
Forgotten by: wosch ;-)

27 years agoAdd a note in BUGS section that shar(1) cannot handle files
Wolfram Schneider [Mon, 31 Mar 1997 17:58:27 +0000 (17:58 +0000)]
Add a note in BUGS section that shar(1) cannot handle files
without a newline ('\n') as last character.

27 years agoFix the mount_mfs case from the last cleanup. The code was (ab)using
Peter Wemm [Mon, 31 Mar 1997 16:43:16 +0000 (16:43 +0000)]
Fix the mount_mfs case from the last cleanup.  The code was (ab)using
it's internal malloc() implementation to try and avoid overstepping it's
resource limits (yuk!).  Remain using libc's malloc(), but check the
resource limits right before trying to malloc the ramdisk space and leave
some spare memory for libc.  In Andrey's words, the internal malloc
was "true evil"..  Among it's sins is it's ability to allocate less memory
than asked for and still return success.  stdio would just love that. :-)

Reviewed by: ache

27 years agoGeneral cleanup and update to specify exactly how our implementation
Peter Wemm [Mon, 31 Mar 1997 16:36:14 +0000 (16:36 +0000)]
General cleanup and update to specify exactly how our implementation
passes on the status across fork/exec.

The previous version had some typos, referred to itself as link(2) in
one place :-), and didn't really match openbsd's implementation either.

Now that I've mentioned typos, hopefully our Typo Police and Xref Police
will be gentle with me. :-)

27 years agoSteal issetugid man page from OpenBSD. Needs work..
Peter Wemm [Mon, 31 Mar 1997 15:59:14 +0000 (15:59 +0000)]
Steal issetugid man page from OpenBSD.  Needs work..

Obtained from: OpenBSD

27 years agoAdd libc hook for issetugid()
Peter Wemm [Mon, 31 Mar 1997 15:43:22 +0000 (15:43 +0000)]
Add libc hook for issetugid()

27 years agoImplement code for an OpenBSD-style issetuigid().
Peter Wemm [Mon, 31 Mar 1997 15:13:33 +0000 (15:13 +0000)]
Implement code for an OpenBSD-style issetuigid().

This is valueable for library code which needs to be able to find out
whether the current process is or *was* set[ug]id at some point in the
past, and may have a "tainted" execution environment.  This is especially
a problem with the trend to immediately revoke privs at startup and regain
them for critical sections.  One problem with this is that if a cracker
is able to compromise the program while it's still got a saved id, the
cracker can direct the program to regain the privs.  Another problem is
that the user may be able to affect the program in some other way (eg:
setting resolver host aliases) and the library code needs to know when it
should disable these sorts of features.

Reviewed by: ache
Inspired by: OpenBSD (but with a different implementation)

27 years agoRegenerate (added issetugid)
Peter Wemm [Mon, 31 Mar 1997 14:55:05 +0000 (14:55 +0000)]
Regenerate (added issetugid)

27 years agoissetugid is now implemented rather than reserved
Peter Wemm [Mon, 31 Mar 1997 14:52:50 +0000 (14:52 +0000)]
issetugid is now implemented rather than reserved

27 years agoDo not set the P_SUGID flag in seteuid()/setegid() if nothing has changed.
Peter Wemm [Mon, 31 Mar 1997 13:47:00 +0000 (13:47 +0000)]
Do not set the P_SUGID flag in seteuid()/setegid() if nothing has changed.
(for supporting issetugid())

Reviewed by: ache  (as was the previous commit, rev 1.29)

27 years agoDo not set the P_SUGID flag in seteuid()/setegid() if nothing has changed.
Peter Wemm [Mon, 31 Mar 1997 13:41:49 +0000 (13:41 +0000)]
Do not set the P_SUGID flag in seteuid()/setegid() if nothing has changed.
(for supporting issetugid())

27 years agoFully implement the clause in Appendix B.4.2.2 from Posix 1003.1
Peter Wemm [Mon, 31 Mar 1997 13:36:46 +0000 (13:36 +0000)]
Fully implement the clause in Appendix B.4.2.2 from Posix 1003.1
that allows traditional BSD setuid/setgid behavior.

The only visible difference should be that a non-root setuid program
(eg: inn's "rnews" program) that is setuid to news, can completely
"become" uid news. (ie: setuid(geteuid())  This was allowed in
traditional 4.2/4.3BSD and is now "blessed" by Posix as a special
case of "appropriate privilige".

Also, be much more careful with the P_SUGID flag so that we can use it
for issetugid() - only set it if something changed.

Reviewed by: ache

27 years agoMake setgroups(0, xxx) behave as it does on SYSV, namely clear the groups
Peter Wemm [Mon, 31 Mar 1997 13:21:37 +0000 (13:21 +0000)]
Make setgroups(0, xxx) behave as it does on SYSV, namely clear the groups
vector except for the egid in groups[0].  There is a risk that programs
that come from SYSV/Linux that expect this to work and don't check for
error returns may accidently pass root's groups on to child processes.

We now do what is least suprising (to non BSD programs/programmers) in
this scenario, and nothing is changed for programs written with BSD groups
rules in mind.

Reviewed by: ache

27 years agoActivate the -h flag which tells chown/chgrp to work on the symlink itself
Peter Wemm [Mon, 31 Mar 1997 13:03:49 +0000 (13:03 +0000)]
Activate the -h flag which tells chown/chgrp to work on the symlink itself
using lchown().  Most of the code was already here, the option was
recognised but ignored for SYSV/POSIX.2(?) compatability.

27 years agoToast dummy lchown(), it's in libc now.
Peter Wemm [Mon, 31 Mar 1997 12:56:00 +0000 (12:56 +0000)]
Toast dummy lchown(), it's in libc now.

27 years agoDeclare lchown()..
Peter Wemm [Mon, 31 Mar 1997 12:52:52 +0000 (12:52 +0000)]
Declare lchown()..

27 years agoCreate lchown() in libc and document it as a variation of chown().
Peter Wemm [Mon, 31 Mar 1997 12:37:35 +0000 (12:37 +0000)]
Create lchown() in libc and document it as a variation of chown().

27 years agoIn accept1(), falloc() is called after the process has awoken, but prior
David Greenman [Mon, 31 Mar 1997 12:30:01 +0000 (12:30 +0000)]
In accept1(), falloc() is called after the process has awoken, but prior
to removing the connection from the queue. The problem here is that
falloc() may block and this would allow another process to accept the
connection instead. If this happens to leave the queue empty, then the
system will panic with an "accept: nothing queued".

Also changed a wakeup() to a wakeup_one() to avoid the "thundering herd"
problem on new connections in Apache (or any other application that has
multiple processes blocked in accept() for the same socket).

27 years agoCode to do lchown(2), copied from chown(2) except it's NOFOLLOW in ND_INIT
Peter Wemm [Mon, 31 Mar 1997 12:21:37 +0000 (12:21 +0000)]
Code to do lchown(2), copied from chown(2) except it's NOFOLLOW in ND_INIT
instead of FOLLOW.

27 years agoRegenerate
Peter Wemm [Mon, 31 Mar 1997 12:17:45 +0000 (12:17 +0000)]
Regenerate

27 years agoReserve 252 (poll, first in OpenBSD)
Peter Wemm [Mon, 31 Mar 1997 12:14:48 +0000 (12:14 +0000)]
Reserve 252 (poll, first in OpenBSD)
Reserve 253 (issetugid, as in OpenBSD)
Allocate 254 for lchown(2)

27 years agoTreat symlinks as first class citizens with their own uid/gid rather than
Peter Wemm [Mon, 31 Mar 1997 12:02:53 +0000 (12:02 +0000)]
Treat symlinks as first class citizens with their own uid/gid rather than
as shadows of their containing directory.  This should solve the problem
of users not being able to delete their symlinks from /tmp once and for
all.

Symlinks do not have modes though, they are accessable to everything that
can read the directory (as before).  They are made to show this fact at
lstat time (they appear as mode 0777 always, since that's how the the
lookup routines in the kernel treat them).

More commits will follow, eg: add a real lchown() syscall and man pages.

27 years agoChanged the way that the exec image header is read to be filesystem-
David Greenman [Mon, 31 Mar 1997 11:11:26 +0000 (11:11 +0000)]
Changed the way that the exec image header is read to be filesystem-
centric rather than VM-centric to fix a problem with errors not being
detectable when the header is read.
Killed exech_map as a result of these changes.
There appears to be no performance difference with this change.

27 years agoThis commit was generated by cvs2svn to compensate for changes in r24434,
David E. O'Brien [Mon, 31 Mar 1997 09:37:59 +0000 (09:37 +0000)]
This commit was generated by cvs2svn to compensate for changes in r24434,
which included commits to RCS files with non-trunk default branches.

27 years agoVirgin import of GNU cpio v2.4.2.
David E. O'Brien [Mon, 31 Mar 1997 09:37:59 +0000 (09:37 +0000)]
Virgin import of GNU cpio v2.4.2.

27 years agoSmall formatting change to interval() to align columns so that:
Marc G. Fournier [Mon, 31 Mar 1997 09:36:59 +0000 (09:36 +0000)]
Small formatting change to interval() to align columns so that:

toronto       up  5+18:58,     0 users,  load 0.00, 0.00, 0.00
zeus          up 109+13:53,     0 users,  load 0.55, 0.28, 0.15

looks like:

toronto       up   5+18:58,     0 users,  load 0.00, 0.00, 0.00
zeus          up 109+13:53,     0 users,  load 0.55, 0.28, 0.15

27 years agoSynchronize with sys/i386/conf/GENERIC revision 1.89.
KATO Takenori [Mon, 31 Mar 1997 08:27:05 +0000 (08:27 +0000)]
Synchronize with sys/i386/conf/GENERIC revision 1.89.

27 years agoRemove ${MAKEFLAGS} on recommendation by Bruce, mainly because the reason
Marc G. Fournier [Mon, 31 Mar 1997 05:30:16 +0000 (05:30 +0000)]
Remove ${MAKEFLAGS} on recommendation by Bruce, mainly because the reason
for adding it was so that -j3 carried through on a make -j3 world, but found
at least one circumstance where it breaks  'make depend'

27 years agoRemove MAKEFLAGS example from /etc/make.conf...
Marc G. Fournier [Mon, 31 Mar 1997 05:26:39 +0000 (05:26 +0000)]
Remove MAKEFLAGS example from /etc/make.conf...

Bruce points out that it breaks things if someone doesn't do a 'make depend',
but I just found out that it breaks 'make depend' itself :(

27 years agoFixed wording of previous change.
Bruce Evans [Mon, 31 Mar 1997 05:18:27 +0000 (05:18 +0000)]
Fixed wording of previous change.

Obtained from: fgets.3
Guided by: ISO C standard

27 years agocompare return value from getopt against -1 rather than EOF, per the final
Warner Losh [Mon, 31 Mar 1997 05:11:47 +0000 (05:11 +0000)]
compare return value from getopt against -1 rather than EOF, per the final
posix standard on the topic.

27 years agoAdded my name/address as requested by Wolfram Schneider.
John Birrell [Mon, 31 Mar 1997 04:58:47 +0000 (04:58 +0000)]
Added my name/address as requested by Wolfram Schneider.

27 years agoRevert my last few changes. They were bogus. Replaced them with
Warner Losh [Mon, 31 Mar 1997 04:51:13 +0000 (04:51 +0000)]
Revert my last few changes.  They were bogus.  Replaced them with
the original text plus a statement saying that if strftime fails,
the results are undefined.

Requested a long time ago by: bde

27 years agoCatch up with the original.
Hiroyuki Hanai [Mon, 31 Mar 1997 02:40:10 +0000 (02:40 +0000)]
Catch up with the original.

27 years agoThe Unix system family tree, BSD part. This is a local copy of
Wolfram Schneider [Mon, 31 Mar 1997 00:06:03 +0000 (00:06 +0000)]
The Unix system family tree, BSD part. This is a local copy of
http://www.de.freebsd.org/de/ftp/unix-stammbaum

27 years agoAdd/document MAKEFLAGS into /etc/make.conf, using -j3 as example
Marc G. Fournier [Sun, 30 Mar 1997 23:43:41 +0000 (23:43 +0000)]
Add/document MAKEFLAGS into /etc/make.conf, using -j3 as example

27 years agoAdd ${MAKEFLAGS} to ${MAKE} in _SUBDIRUSE so that something like
Marc G. Fournier [Sun, 30 Mar 1997 23:39:39 +0000 (23:39 +0000)]
Add ${MAKEFLAGS} to ${MAKE} in _SUBDIRUSE so that something like
'make -j3 world' works

Jordan points out that this may not be the only place this is required to be
added, but so far, its the only one I've found to break -j3

27 years agoUse srandomdev() now
Andrey A. Chernov [Sun, 30 Mar 1997 23:33:29 +0000 (23:33 +0000)]
Use srandomdev() now
Add range checking for invalid denominator values

27 years agoUse srandomdev() now
Andrey A. Chernov [Sun, 30 Mar 1997 23:13:59 +0000 (23:13 +0000)]
Use srandomdev() now
Fix error with begin variable calculation

27 years agoImplement the `detach' command for remote GDB. It gets you back at DDB.
Joerg Wunsch [Sun, 30 Mar 1997 16:03:00 +0000 (16:03 +0000)]
Implement the `detach' command for remote GDB.  It gets you back at DDB.

27 years agoMove uucplock into libutil and create a manual page.
Brian Somers [Sun, 30 Mar 1997 12:12:20 +0000 (12:12 +0000)]
Move uucplock into libutil and create a manual page.

27 years agoJKH says:
David E. O'Brien [Sun, 30 Mar 1997 11:01:43 +0000 (11:01 +0000)]
JKH says:

   Change "Found end of tape.  Load next tape ..." messages to say
   "volume" instead of tape.  Running cpio off of /dev/fd0 and having
   it say "give me the next tape" is kind of ludicrous.. :-)

27 years agodon't lookup rsh(1) at bogus pathnames
David E. O'Brien [Sun, 30 Mar 1997 10:56:38 +0000 (10:56 +0000)]
don't lookup rsh(1) at bogus pathnames

Submitted by: joerg via old gnu/usr.bin/cpio v2.3

27 years agoDon't set the umask until after we have processed the arguments
David E. O'Brien [Sun, 30 Mar 1997 10:50:01 +0000 (10:50 +0000)]
Don't set the umask until after we have processed the arguments
and opened the archive file.  This allows "cpio -o -O output_file"
to create the output file with the callers proper umask.
Closed PR# 1391

Add setlocale LC_ALL (from ache).

27 years agoSupport dummy lchown() so that cpio understands 4.4 symbolic links.
David E. O'Brien [Sun, 30 Mar 1997 10:45:44 +0000 (10:45 +0000)]
Support dummy lchown() so that cpio understands 4.4 symbolic links.

27 years ago- Correct the recommended option for "find" from "-depth" to "-d".
David E. O'Brien [Sun, 30 Mar 1997 10:42:21 +0000 (10:42 +0000)]
- Correct the recommended option for "find" from "-depth" to "-d".
- Fix gross spelling and typographical errors pointed out by Keith Bostic.
- Mention -l, --link is only usable with "-p".

Obtained from: old gnu/usr.bin/cpio v2.3.

27 years agoOutput a zero rdev except for bdevs, cdevs, fifos and sockets. This
David E. O'Brien [Sun, 30 Mar 1997 10:34:16 +0000 (10:34 +0000)]
Output a zero rdev except for bdevs, cdevs, fifos and sockets.  This
stops regular files with unrepresentable rdevs from being rejected
and makes the output independent of unpreservable metadata.

Don't output a file if the major, minor or totality of its rdev would be
truncated.  Print a message about the skipped files to stderr but don't
report the error in the exit status.  cpio's abysmal error handling doesn't
allow continuing after an error, and the rdev checks had to be misplaced
to avoid the problem of returning an error code from routines that return
void.

Minor numbers are limited to 21 bits in pax's ustar format and to 18
bits in archives created by gnu tar (gnu tar wastes 3 bits for padding).
pax's and cpio's ustar format is incompatible with gnu tar's ustar
format for other reasons (see cpio/README).

Submitted by: bde via old gnu/usr.bin/cpio v2.3.

27 years agoFix typo.
David E. O'Brien [Sun, 30 Mar 1997 10:32:26 +0000 (10:32 +0000)]
Fix typo.

27 years agoAdd an example README.FreeBSD. The tcl one you tend to refer to since tcl
David E. O'Brien [Sun, 30 Mar 1997 09:55:11 +0000 (09:55 +0000)]
Add an example README.FreeBSD.  The tcl one you tend to refer to since tcl
is the example is quite spartan.

As pointed out by Bruce there are *three* different filenames being used:

cvs             FREEBSD.README
groff           FREEBSD-upgrade
libgmp          FREEBSD-upgrade
libpcap         FREEBSD-upgrade
tcpdump         FREEBSD-upgrade
traceroute      FREEBSD-upgrade
tcl             README.FreeBSD

The handbook states "README.FreeBSD".  Perhaps this should be changed?

27 years agoFix mount call for devfs.
Poul-Henning Kamp [Sun, 30 Mar 1997 09:22:41 +0000 (09:22 +0000)]
Fix mount call for devfs.

Submitted by: bde

27 years agoFix embarrasing typo that survived because I applied the wrong patch-
Poul-Henning Kamp [Sun, 30 Mar 1997 09:08:00 +0000 (09:08 +0000)]
Fix embarrasing typo that survived because I applied the wrong patch-
file.
Noticed by: Eric Jones <ejon@psa.pencom.com>

27 years agoI goofed, Makefile lchown.c don't belong here.
David E. O'Brien [Sun, 30 Mar 1997 08:19:38 +0000 (08:19 +0000)]
I goofed, Makefile lchown.c don't belong here.

27 years agoReplace ctime by strftime %c to use national representation.
David E. O'Brien [Sun, 30 Mar 1997 00:23:39 +0000 (00:23 +0000)]
Replace ctime by strftime %c to use national representation.
Use isprint() for Localization.

Submitted by: ache via gnu/usr.bin/cpio v2.3

27 years agoFreeBSD BSD-style Makefile for contributed CPIO source.
David E. O'Brien [Sat, 29 Mar 1997 23:11:22 +0000 (23:11 +0000)]
FreeBSD BSD-style Makefile for contributed CPIO source.

27 years agoRodney Grimes's dummy lchown() so that cpio understands 4.4 symbolic links.
David E. O'Brien [Sat, 29 Mar 1997 23:10:24 +0000 (23:10 +0000)]
Rodney Grimes's dummy lchown() so that cpio understands 4.4 symbolic links.
   (taken from old gnu/usr.bin/cpio)

Reviewed by: phk

27 years agoDon't need tcexparg.c on FreeBSD.
David E. O'Brien [Sat, 29 Mar 1997 23:05:27 +0000 (23:05 +0000)]
Don't need tcexparg.c on FreeBSD.