- Copy stable/10 (r259064) to releng/10.0 as part of the
10.0-RELEASE cycle.
- Update __FreeBSD_version [1]
- Set branch name to -RC1

[1] 10.0-CURRENT __FreeBSD_version value ended at '55', so
start releng/10.0 at '100' so the branch is started with
a value ending in zero.

Approved by: re (implicit)
Sponsored by: The FreeBSD Foundation

Copy head (r256279) to stable/10 as part of the 10.0-RELEASE cycle.

Approved by: re (implicit)
Sponsored by: The FreeBSD Foundation

Check the return error of set[ug]id. While this can never fail in the
current version of FreeBSD, this isn't guarenteed by the API. Custom
security modules, or future implementations of the setuid and setgid
may fail.

PR: bin/172289
PR: bin/172290
PR: bin/172291
Submittud by: Erik Cederstrand <erik@cederstrand.dk>
Discussed by: freebsd-security
Approved by: cperciva
MFC after: 1 week

Use timeclear() instead of home grown version.

MFC after: 2 weeks

Small style fixes:

- ANSIfy prototypes.
- Remove unneeded whitespace.
- Add const keyword to function where it can be used.

- improve readability

Reviewed by: trhodes, keramida
MFC after: 3 days

- rename the RETURN VALUES section to EXIT STATUS
- not bumping a date as this is not a real content change

Approved by: ru
MFC after: 3 days

Commit IPv6 support for FAST_IPSEC to the tree.
This commit includes all remaining changes for the time being including
user space updates.

Submitted by: bz
Approved by: re

Force the alignment of the chars arrays, as they are casted later to
structs.
gcc 4.2 doesn't do it by default, and that results in unaligned access on
arm.

Remove alpha left-overs.

Print packet loss figures with one decimal place. ping6 already does
this, and OpenBSD and NetBSD pings do it too. This is primarily useful
for comparing low levels of packet loss.

Add a new feature to ping(8) - possibility to specify maximum
wait time for a packet. This allows to:

- Count number of packets received before and after specified
time.
- Shorten time of execution of 'ping -c 1' scripts.

Submitted by: Lytochkin Boris <lytboris gmail.com>

Sync usage() with SYNOPSIS.

-mdoc sweep.

Implement a new feature for ping(8) - sweeping pings. In a sweeping
ping ICMP payload of packets being sent is increased with given step.
Sweeping pings are useful for testing problematic channels, MTU
issues or traffic policing functions in networks.

PR: bin/82625
Submitted by: Chris Hellberg <chellberg juniper.net> (with some cleanups)

o Store timestamp in network byte order.
o Remove an assumption sizeof(struct timeval) == 8 (this is not
true on sparc64).

Reviewed by: imp, -hackers
Obtained from: NetBSD (rev. 1.75)
MT5 after: 1 month

Avoid using void pointers in additive expressions.

PR: 56653

Remove advertising clause from University of California Regent's license,
per letter dated July 22, 1999.

Approved by: core, imp

Don't turn off the regular SIGINFO status information. The use of
the NOKERNINFO flag only marginally de-clutters the output and
has a number of unwanted side effects:
o The kernel info might be what you want to see
o ^T is left non-functional if ping is killed non-cleanly
o "ping -q foo &" gets suspended on tty output

Encouraged by: bde

style.Makefile(5):
Use WARNS?= instead of WARNS=.

Certain ICMP error replies cause ping to perform a reverse DNS
lookup on an IP address from the packet (such as the IP that sent
a TTL exceeded error). If the DNS lookup takes a long time, ^C will
appear to be ineffective since the SIGINT handler just sets a flag
and returns. Work around this by exiting immediately on receipt of
a second SIGINT when DNS lookups are enabled.

PR: bin/4696
MFC after: 1 week

mdoc(7): Use the new feature of the .In macro.

o Rename local variables, do not shadow global declarations.

o Kill MINICMPLEN, there is ICMP_MINLEN already.

o Be ready to get a reply with length up to IP_MAXPACKET.
o Warn when recieved packet length is not equal to length of the
packet we sent out. Idea from NetBSD.
o Fit the dump of packet with wrong data to 80 columns (from NetBSD).

Comments from: bde

The current recommended default time to live (TTL) for the Internet
Protocol (IP) is 64 [RFC791, RFC1122].

Prodded by: Igor Kucherenko <kivvy@sunbay.com>
Obtained from: http://www.iana.org/assignments/ip-parameters

Fix SIGINFO handling: do not print incorrect timing statistics
when the data for it is not available.

Founded by: ping -s0 localhost; press ^T

Use __FBSDID() to quiet GCC 3.3 warnings.

o Skip timestamp part when checking payload.

PR: bin/50909
Submitted by: Barney Wolff <barney@lab.databus.com>

Deal with a case where the returned packed was smaller than the
transmitted packet (because the remote host stripped off our icmp_data).

Submitted by: Maxim Konovalov <maxim@macomnet.ru>

Back out support for RFC3514.

RFC3514 poses an unacceptale risk to compliant systems.

IP_EVIL -> IP_EF

Implement support for RFC 3514 (The Security Flag in the IPv4 Header).
(See: ftp://ftp.rfc-editor.org/in-notes/rfc3514.txt)

This fulfills the host requirements for userland support by
way of the setsockopt() IP_EVIL_INTENT message.

There are three sysctl tunables provided to govern system behavior.

net.inet.ip.rfc3514:

Enables support for rfc3514. As this is an
Informational RFC and support is not yet widespread
this option is disabled by default.

net.inet.ip.hear_no_evil

If set the host will discard all received evil packets.

net.inet.ip.speak_no_evil

If set the host will discard all transmitted evil packets.

The IP statistics counter 'ips_evil' (available via 'netstat') provides
information on the number of 'evil' packets recieved.

For reference, the '-E' option to 'ping' has been provided to demonstrate
and test the implementation.

Default data length should be 56, not 64.

- Make -M mask and -M time exclusive.
- Correct some problems with packet construction.

+--------+------------+----------+-------------+---------+
| | | | | |
| IP Hdr | MINICMPLEN | phdr_len | TIMEVAL_LEN | payload |
| | | | | |
+--------+------------+----------+-------------+---------+
| | | |
|<- IP ->|<------- ICMP -------->|<------ datalen ------>|

My previous changes tried to mess around with 'datalen' instead of
modifying 'phdr_len'.

I'm including this nice ASCII diagram (from Maxim) to further clarify things
in CVS history.

Submitted by: Maxim Konovalov <maxim@macomnet.ru>

Fixed an errx format error in rev.1.81. This should have been fatal
when WARNS was increased recently, but __printf0like() has been
temporarily disabled for 8 months.

Fixed related style bugs (disordered declaraction and silly type for
maxpayload -- assume 16-bit ints like the rest of ping.c).

Demangle the usage message at the source level.

Submitted by: bde

Fix and sync SYNOPSIS and usage().

Prodded by: bde

Update ping to be WARNS=2 compliant.

Reviewed by: -audit (no objections ~1mo)
Approved by: nectar

Spelling, grammar fixes.

Remove documentation of old '-M' flag.
Re-order.

Support ICMP_TSTAMP/ICMP_TSTAMPREPLY.

This alters the meaning of the '-M' flag.

Sort options.

Call fill() after maxpayload has been initialized.

Reviewed by: maxim

Send ICMP_MASKREQ packets when the '-M' option is specified.

Better error handling for -s.

Submitted by: bde

Do not constantify maximum payload size. It is 65467 with -R
(record route), and 65507 without it.

Reviewed by: silence on -net
Proposed by: bde

I am going to MFC rev.1.77 - 1.81 ping.c and rev.1.39 and 1.40 ping.8:

MFC after: 6 months

style(9) sweep.

Submitted by: bde

Anti-magic: s/255/MAXTTL/

Implement -D (do not fragment) and -z (TOS) options.

PR: bin/29164
Reviewed by: silence of -net
Obtained from: OpenBSD

Print strict source routing option.

Reviewed by: silence on -net
Proposed by: David Wang <dsw@juniper.net>

english(4) police.

Spelling: s/then/than/ where appropriate.

Fix LSRR option length check: it has to be less or equal remained header's
length minus sizeof(struct ip).

MFC after: 1 week

Add an -o option to exit after receiving one reply. This can be used
to test whether a link is live.

PR: 38573
Submitted by: David Taylor <davidt@yadt.co.uk>
Obtained from: NetBSD

Commit "unrelated style fixes" part of Bruce's patch (regarding
bcopy/memcpy) seperately.

Submitted by: bde

Modify previous commit to solve the real problem that made gcc think
the timestamp was aligned. ie: Use a void * instead of struct timeval *
which gcc assumes will be aligned. Go back to memcpy().

Submitted by: bde

In reference to previous commit: use err(3).

Noticed by: bde

Check for FD_SET overrun.

Add the -A flag to the usage line.

Submitted by: Yutaka KAWASE <yutaka@mailhost.net>
MFC after: 3 days

Fix the broken "avoid unaligned data" fix. The problem is that the builtin
gcc memcpy "knows" about types that are supposed to be actually already
aligned and triggers alignment errors doing the memcpy itself.
"Fix" this by changing it to a bcopy(). In this case, we had:
struct timeval *tp;
struct timeval tv1;
memcpy(&tv1,tp,sizeof(tv1));
.. and since gcc *knows* that a pointer to a timeval is longword aligned
and that tv1 is longword aligned, then it can use an inline that assumes
alignment. The following works too:
cp = (char *)tp;
memcpy(&tv1,cp,sizeof(tv1));
Simply casting (char *)tp for the memcpy doesn't work. :-(
This affected different 64 bit platforms in different ways and depends
a lot on gcc as well. I've seen this on alpha and ia64 at least, although
alpha isn't doing it right now.

The .Nm utility

o Change almost all magic numbers to the appropriate constants.
o Fix zero payloading, unbreak ping -s 0.
o Increase socket recieve buffer, ping -s 65467 is working now.

Submitted by: anti-magic sweep based on kris's patch
Reviewed by: bde, silence on -audit
MFC after: 2 months

More strict ip options parsing.

Reviewed by: bde (style), silence on -audit
MFC after: 2 months

WARNS fixes.

Submitted by: kris
Reviewed by: bde
Approved by: ru
MFC after: 2 weeks

style(9) cleanup.

Submitted by: kris (an early version of this patch)
Reviewed by: bde
Approved by: ru
MFC after: 2 weeks

Style(9) local vars after 'register' removal.

Requested by: bde

Remove 'register' keyword.
It does not help modern compilers, and some may take some hit from it.
(I also found several functions that listed *every* of its 10 local vars with
"register" -- just how many free registers do people think machines have?)

Use network byte order for the ICMP sequence number. This is only
significant when monitoring packets on another system, since
otherwise the ICMP sequence number is only used by the ping client.

Document ping(8) reaction to SIGINFO.

PR: doc/33639
Reviewed by: ru
Approved by: ru
MFC after: 1 week

Fixed two bugs with the "-l preload" option:

- first ping after a preload was sent undelayed
- we could send more than -c packets in preload

PR: bin/32354

Default to WARNS=2.
Binary builds that cannot handle this must explicitly set WARNS=0.

Reviewed by: mike

The -A option (beep when packets are dropped) didn't work quite
right; after a single packet was dropped it beeped after every
transmission.

Change its implementation to only output a bell when there is an
increase in the maximum value of the number of packets that were
sent but not yet received. This has the benefit that even for very
long round-trip times, ping -A will do roughly the right thing
after a few inital false-positives.

Reviewed by: ru

Set BINOWN=root explicitly for setuid root binaries.

This is not "useless", as one may have non-default
setting for BINOWN in make.conf, and we still want
these to be installed setuid root in this case.

Use CFLAGS, not COPTS, in the Makefile. bsd.prog.mk conveniently adds
COPTS towards the end of final CFLAGS so that it can be used to
override Makefile and other defaults. Using it in Makefiles risks
having options set using it clobbered when somebody uses it on the
command line.

Approved by: bde

mdoc(7) police:

Avoid using parenthesis enclosure macros (.Pq and .Po/.Pc) with plain text.
Not only this slows down the mdoc(7) processing significantly, but it also
has an undesired (in this case) effect of disabling hyphenation within the
entire enclosed block.

Remove whitespace at EOL.

mdoc(7) police: removed HISTORY info from the .Os call.

Ian Dowse writes:

The original code was certainly broken; it knows that whereto is
to be used for a sockaddr_in, so it should be declared as such.
To support multiple protocols, there is also a sockaddr_storage
struct that can be used; I don't think struct sockaddr is supposed
to be used anywhere other than for casts and pointers.

Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
MFC after: 3 weeks

Fix unaligned access faults on alpha.

This one is strange and goes against my rusty compiler knowledge.

The global declaration

struct sockaddr whereto;

produces for both i386 && alpha:

.comm whereto,16,1

which means common storage, byte aligned. Ahem. I though structs
were supposed to be ALDOUBLE always? I mean, w/o pragma packed?

Later on, this address is coerced to:

to = (struct sockaddr_in *)&whereto;

Up until now, we've been fine on alpha because the address
just ended up aligned to a 4 byte boundary. Lately, though,
it end up as:

0000000120027b0f B whereto

And, tra la, you get unaligned access faults. The solution I picked, in
lieu of understanding what the compiler was doing, is to put whereto
as a union of a sockaddr and sockaddr_in. That's more formally correct
if somewhat awkward looking.

Modify ping so that it increases the send socket buffer size if the
user runs with privilege, allowing the sending of icmp packets with
larger size (up to 48k, the default receive buffer size in ping),
which is useful for network driver development testing, as well
as experimentation with fragmentation.

Reviewed by: wpaul

Add a -A option to ping which beeps when packets are lost.

PR: 11818
Submitted by: Marc Evans marc@destek.net

- Backout botched attempt to introduce MANSECT feature.
- MAN[1-9] -> MAN.

Set the default manual section for sbin/ to 8.

Add new option -m which allows the user to set IP TTL.

Prepare for mdoc(7)NG.

mdoc(7) police: use the new features of the Nm macro.

Format string paranoia

Use CMSG_LEN() to compute ping SCM_TIMESTAMP cmsg_len.
Without this, cmsg_len size should be smaller than necessary on alpha.
(Though, no obvious error was seen on ping time value on beast.)

CMSG_XXX macros alignment fixes to follow RFC2292.

Approved by: jkh

Submitted by: Partly from tech@openbsd
Reviewed by: itojun

Remove single-space hard sentence breaks. These degrade the quality
of the typeset output, tend to make diffs harder to read and provide
bad examples for new-comers to mdoc.

ping -S option was missing from IPsec merge, so enable it again.

Approved by: jkh

Submitted by: Matthew Reimer <mreimer@vpop.net>

Brucify:
add an upper limit to -t
match the types of return values and the variables they are stuffed in
make the man page and usage() a little more consistantly ugly
less obfuscation.

Submitted by: adrian, billf

Add '-t X' which times out after pinging for X seconds.

Submitted by: adrian

Don't try to include ipsec support if we are building for the install
floppy image.

libipsec and IPsec related apps. (and some KAME related man pages)

Reviewed by: freebsd-arch, cvs-committers
Obtained from: KAME project

Make last commit bde-compliant:
- correct indentation
- change data types for consistency with the rest of ping.c
- create new variable separate from "answer" for clarity

Fix aliasing bug causing in_cksum() to fail on odd packet sizes
due to compiler optimization.

PR: bin/13292
Suggested by: wollman

$Id$ -> $FreeBSD$

Re-add -Wmissing-prototypes to `COPTS'. I mistook -W for -Wno in this case.

-Wmissing-prototypes is unnecessary here. There are no warnings produced
with this option disabled.

Make flood ping flood again. It hasn't worked since 2.2.6. Make the "-l"
option work as described in the manual, not as another flood ping variant.

Once discussed to death with: bde

More egcs warning fixes:
o main returns int not void
o use braces to avoid potentially ambiguous else

Note: The fix to natd is potentially functional in nature since I used
the indentation as the right thing rather than the struct semantics.
Someone more familiar with the code should double check me on this one.

Reviewed by: obrien and chuckr

Finish documenting -S src-addr change.

Forgotten by: me
Forgotten for: Far Too Long
Gently Reminded by: Ruslan Ermilov <ru@ucb.crimea.ua>
Submitted by: Ruslan Ermilov <ru@ucb.crimea.ua>

Allow PINGing from any address on multihomed hosts

In the words of the submitter:
"The patch below allows to ping from any address on the multihomed host.
The man page is also updated, the text was cutted from traceroute(8)."

Submitted by: Ruslan Ermilov
PR: 6832

Fixed non-use of .Nm and hard line breaks in rev.1.16-1.17:

Remove useless `BINOWN=root' now that it is the default.

Disable gcc's builtin memcpy for alpha since it doesn't cope with unaligned
regions properly and this triggers an unaligned access trap.

Updated manual page. Removed description of (now defunct) -c restrictions.

Remove -c restrictions from previous commit.

(well tested at BEST): -i option can now take FP values (e.g. -i 0.1),
extremely useful for networking testing. Other options secured from
user-level D.O.S. attacks. -f, -s now root-only. -i wait times < 1.0
root-only. -c count limited to 100 and defaults to 16 when ping run
by non-root user.

Correct use of .Nm. Remove unused #includes. Use warn(). cosmetic in usage()
string.
Translate 0 sec 1000000 usec to 1 sec 0 usec.

Fix an alignment problem on alpha by doing a bytewise copy.

Use select() timeouts instead of SIGALRM to schedule packet transmission.
Fixes bin/6649 and removes the last abusive signal handler.
Use SO_TIMESTAMP to get the kernel to timestamp packets on reception.
Fixes bin/5658 and provides slightly better accuracy.
Explicitly zero and terminate the IP options when using -R.

PR: bin/5658
PR: bin/6649

Revert the previous fix. As it turns out Warner Losh is
working a better fix.

Attempt to stop another DoS attack related to ping flooding.

PR: 6649
Submitted by: Jason Young <doogie@forbidden-donut.anet-stl.com>

When a timeval is stored at the beginning of icmp_data, the fields
are unaligned for access by the alpha, so copy the value to a variable
that is aligned.

When checking the returned data, be careful to avoid confusing the
size of the icmp header with the size of a timeval. On i386 these
are both 8, but on alpha, a timeval is 16 bytes. This means that
a packet sent from an alpha contains 48 bytes of data, not 56 like
on i386.

If ping write fails with short packet count, the
error message prints the two numbers backwards.
PR: 6313
Reviewed by: phk
Submitted by: Archie Cobbs

Add aback in a line a accidentally killed in my last commit.
Submitted by: Bill Fenner

Make this compile (and seemingly work).

A fairly rare possible buffer overflow:
Theo fixed this and tqbf reported it
Obtained from: OpenBSD

.Sh AUTHOR -> .Sh AUTHORS. Use .An/.Aq.

Minor style nit noticed by bde.

Fixed DPADD.

Add an ntohs() and format an IP address with inet_ntoa() when
printing the details of a received ICMP packet.

PR: bin/3766
Submitted by: denny1@home.com (Denton Gentry)

don't lose track of how many packets we've sent
if we get 'host unreachable'. (or any other errors than ENOBUFS)
makes -c work again even if you can't get there..
This really needs a rework..

Cleaned up revisions 1.22 and 1.23.
Fixed minor bugs in revisions 1.12 and 1.23 (variables assigned to in signal
handlers weren't declared as volatile).

Calculate and print out the standard deviation of the round trip times.
This isn't necessarily the best statistic, but it is by far the easiest to
calculate. Update the man page to be more explicit about precisely which
statistics are printed out. Revert some of jmg's bogus man page changes from
rev 1.11.

Fix a problem introduced with a recent change that caused a hang with
unreachable hosts. Note that most of this consists of telling SIGINT
and SIGALRM to interrupt the system call, instead of restarting them.
Also try to get rid of some potential races Bruce didn't like; hopefully
they aren't a problem (potential or otherwise) now.

Reviewed by: julian

don't count packets as being transmitted when we know that the sendto()
failed (e.g. ENOBUFS)

ping called printf() from a signal handler..
this is a NO-NO

re-arange to just set a "please die immediatly" flag in the signal handler
and handle this in the normal thread.

also handle ping -f better on slow links by backing off a bit when
we get a ENOBUFFS from the sendto().

Fix typo introduced in 1.17.

Submitted by: "Philippe Charnier" <charnier@xp11.frmug.org>

Check the size of the IP address reutrned from gethostbyname2, per PR
2578 from Julian A. Likely not strictly needed, but it doesn't hurt
and protects ping against possible buffer overflows if the resolver
were to return large IP addresses.

fix man page to use marcos instead of parens, quotes and braces

plus other minor fixes

1) Minor nits in the usage message from bde. I took out the hard tabs and
used spaces to align the second line under the program name.
2) Cache uid after call to setuid(getuid()) so we don't waste a system call
for each packet with a call to getuid for the -v case.
3) Update manual to reflect new restriction on -l from last delta.

Suggested by: bde, Bill Fenner

1) Fix usage to match convention and manual. Fix manual to match usage.
2) Must be root to run preload (OpenBSD ping.c 1.8)
3) Don't print all replies unless verbose and root (from idea in
OpenBSD ping.c 1.10 and 1.11) to avoid leaking information available
only to root.
4) Remove unused h: from option string to getopt.
5) Make the compiler happy with exit(0) (Lite-2?)

Reviewed by: Dan Cross <tenser@spitfire.ecsel.psu.edu>

Good candidate for 2.2 and 2.1 (as are many of the 1.17 changes).

Make ping -Wall clean (except for one warning).
Do a better job of argument parsing.
Don't permit ping -f to a multicast address (very antisocial).
Don't permit -L, -I, -T options with unicast addresses.
Ensure that we ask for only AF_INET addresses (should close PR#2584).
Return <sysexits.h> error codes for failures. Document this.
Fix man page to identify the author and put sections in correct order.

Revert $FreeBSD$ to $Id$

Reviewed by: Gary Jennejohn <gj@freebsd.org>
Submitted by: Bruce Murphy <packrat@iinet.net.au>

Add '-a' audible flag, so terminal will beep upon receipt of a reply
packet. Useful for debugging ethernet runs, among other things.

Make the long-awaited change from $Id$ to $FreeBSD$

This will make a number of things easier in the future, as well as (finally!)
avoiding the Id-smashing problem which has plagued developers for so long.

Boy, I'm glad we're not using sup anymore. This update would have been
insane otherwise.

Add multicast options -I (source interface), -T (set ttl), -L (no loopback).
They were all lowercase in the original, but our ping already uses -i and
-l so I made them all uppercase.

Obtained from: Multicast release 3.5

Cleaned up recent SIGINFO changes.

Simplified average calculations. This also fixes potential overflow errors
and minor rounding inconsistencies (always round to nearest now).

Don't lose more SIGINFOs than necessary.

Print \r at the start of the status message. This clears the dots that
are sometimes left by ping -f.

Reviewed by: sef

Fix typo in last patch. empty!=emtpy

Change how SIGINFO is handled -- set a flag (of sig_atomic_t), and check
that flag on every iteration of the loop. This avoids calling fprintf
inside a signal handler, which is always somewhat icky.

Reviewed by: bde

Fix math in SIGINFO printer.

Move the terminal handling code to a later location; this allows
the signal handlers to be set up; it also means that a failed name lookup
won't mess up the nokerninfo setting of the tty.

Reviewed by: pst, nate

Fixed backgrounding of ping, which was broken by the ^T changes in rev.1.3.

delete doubled words, e.g.: "the the" -> "the"

Print out ICMP errors that are responses to our own packets by default.
Turn this behavior off using '-Q'. This makes '-v' useless other than as
an ICMP-sniffer, which tcpdump is better at anyway.

Print out another couple of ICMP messages, and fix the printing of the
original packet (mostly byte order problems).

Drop setuid ASAP, to minimize code executed as root.

Reviewed by: pst

Limit the risk of `buf' overrun in ping.c when printing hostnames.

Note, this is not really a security risk, because the buffer in question
is a static variable in the data segment and not on the stack, and hence
cannot subert the flow of execution in any way. About the worst case was
that if you pinged a long hostname, ping could coredump.

Pointed out on: bugtraq (listserv@netspace.org)

Remove trailing whitespace.

In past releases of FreeBSD, I have used the exit status of "ping" to
determine whether a connection to a given machine is up or not.

In FreeBSD 2.0 (and therefore, I assume, BSD 4.4) the exit code of ping
is always zero, even if no packets were received.

I would like to propose the following change to /usr/src/sbin/ping/ping.c
to restore this useful behaviour:

Submitted by: Denis Fortin

Reviewed by: jkh

Print out summary information on receipt of SIGINFO; also, stop the
kernel printing of information and restore it on exit. Now, it needs
an option to be quiet. ;)

Changed output formatting to accurately represent the precision.

Convert to our man installation style. Also fixed long-standing bug
in `fastboot'/`fasthalt' in which the interpreter would hang around
after `reboot' or `halt' is run, causing an irritating ``Killed'' message.

This commit was generated by cvs2svn to compensate for changes in r1558,
which included commits to RCS files with non-trunk default branches.