Matthew Dillon [Thu, 17 Mar 2005 21:41:16 +0000 (21:41 +0000)]
Ignore additional stty control characters like ^Y which interfere with
the editor's operation.
Reported-by: Jasse Jansson <jasse@hornet.ac>
David Rhodus [Thu, 17 Mar 2005 20:59:34 +0000 (20:59 +0000)]
Testing a new commits monitoring script.
David Rhodus [Thu, 17 Mar 2005 20:53:55 +0000 (20:53 +0000)]
Unbreak the pkg_add -r which the last commit broke.
Matthew Dillon [Thu, 17 Mar 2005 17:28:46 +0000 (17:28 +0000)]
Clean up a number of caching edge cases in NFS, rework the code to be
a bit more readable, document some bits, and fix some cache coherency
detection issues. The caching cleanups should allow the NFS client to
retain more of the NFS cache when doing complex operations on a file.
* Properly check and update the mtime using WCC records in the NFS response.
This record gives us the 'before' and 'after' mtime. The 'before' mtime
must match our existing idea of the mtime, if it doesn't we flag the
nfsnode as having been modified by the server. Our notion of the mtime
is then set to the 'after time. This was not being done properly for
several edge cases.
This required extending the nfsm macros a bit in order to be able to tell
loadattrcache how to handle the mtime data. This also required rearranging
(really fixing) the sequence in nfs_open(), nfs_write(), etc.
* Rearrange the flags a bit. NSIZECHANGED -> NRMODIFIED (nfsnode modified
by server), NMODIFIED -> NLMODIFIED (nfsnode modified by client). Do
not clear NRMODIFIED until we have actually invalidated the cache (this
fixes a problem where programs using mmap() were not properly clearing
the cache after a file was modified on the server).
* Don't code NRMODIFIED as an exception to NLMODIFIED. Recode the flags so
they (mostly) operate in tandem.
* When appending to a file, use nfs_flush() instead of nfs_vinvalbuf().
There is no need to destroy our data cache for the file. This makes
appends considerably more efficient.
* Hopefully fix the last problem associated with attribute timeouts.
* Clear the attribute cache when a file is opened for write in nfs_open()
BEFORE doing other checks rather then after.
* Document some of the nastier cache coherency hacks.
Joerg Sonnenberger [Thu, 17 Mar 2005 15:39:34 +0000 (15:39 +0000)]
Fix warnings. Raise WARNS to 6.
Submitted-by: Samuel J. Greear <sjg@evilcode.net>
Add some explicit XXX comments about the signal handlers,
they have to be rewritten.
Joerg Sonnenberger [Thu, 17 Mar 2005 15:09:01 +0000 (15:09 +0000)]
Use a common source for the string to integer conversion. Add
strtoimax and strtoumax.
Inspired-by: wide char versions.
Sascha Wildner [Thu, 17 Mar 2005 08:22:38 +0000 (08:22 +0000)]
Now that make no longer needs it, remove the ispc98 sysctl.
YONETANI Tomokazu [Thu, 17 Mar 2005 07:11:15 +0000 (07:11 +0000)]
Now that const'ification of users of dktypenames also const'ified
most of users of fstypenames, finish the last bit of const'ification.
Max Okumoto [Thu, 17 Mar 2005 02:34:31 +0000 (02:34 +0000)]
Remove pc98 support as a gift to Sascha Wildner :-)
Matthew Dillon [Thu, 17 Mar 2005 02:17:54 +0000 (02:17 +0000)]
Close a possible security hole by using strvis() when displaying td_comm.
Matthew Dillon [Thu, 17 Mar 2005 02:01:12 +0000 (02:01 +0000)]
Fix a potential security hole by outputing non-printable characters related
to file names or program commands as '?'.
Max Okumoto [Wed, 16 Mar 2005 22:50:06 +0000 (22:50 +0000)]
patch-7.142
modifier_C()
Update vp->ptr instead of using temp variable.
Max Okumoto [Wed, 16 Mar 2005 22:49:51 +0000 (22:49 +0000)]
patch-7.141
modifier_S()
Update vp->ptr instead of using temp variable.
Max Okumoto [Wed, 16 Mar 2005 22:49:01 +0000 (22:49 +0000)]
patch-7.140
- Removed comment about assuming that Var_Parse() is always
passed '$' as first char. Sometimes is is the last character
of a conditional (e.g., defined(DragonFly), etc.)
- Clean up some comments.
- Moved common code out of if-statement.
Max Okumoto [Wed, 16 Mar 2005 22:48:47 +0000 (22:48 +0000)]
patch-7.139
Split out sysVvarsub() from ParseModifier()
Max Okumoto [Wed, 16 Mar 2005 22:48:31 +0000 (22:48 +0000)]
patch-7.138
- renamed freePtr to freeResult.
Max Okumoto [Wed, 16 Mar 2005 22:48:15 +0000 (22:48 +0000)]
patch-7.137
- Cleanup comments
- Moved code from bottom of ParseModifier() in to calling function.
Max Okumoto [Wed, 16 Mar 2005 22:47:47 +0000 (22:47 +0000)]
patch-7.136
Clean up handling of VAR_JUNK.
Matthew Dillon [Wed, 16 Mar 2005 22:17:59 +0000 (22:17 +0000)]
Fix a bugaboo in the last commit. We cannot safely modify n_mtime based
on the server indicating that it has changed without also flushing the
vnode's buffers. Otherwise we end up with stale data in the buffer cache
and the client unaware that it is stale. Since we do not want to flush
the buffers in the getattr code we instead simply mark the nfs node
NSIZECHANGED which will flush the buffers on the next read.
Chris Pressey [Wed, 16 Mar 2005 21:26:01 +0000 (21:26 +0000)]
Raise WARNS to 6.
Chris Pressey [Wed, 16 Mar 2005 21:21:34 +0000 (21:21 +0000)]
Raise WARNS to 6:
- Rename local variables to avoid conflict with globals.
Max Okumoto [Wed, 16 Mar 2005 20:03:00 +0000 (20:03 +0000)]
FreeBSD-Date: 2005/03/16 08:04:45
FreeBSD-Date: 2005/03/16 08:06:28
FreeBSD-Date: 2005/03/16 12:22:15
FreeBSD-Date: 2005/03/16 16:02:09
FreeBSD-Date: 2005/03/16 16:11:11
Author: harti
o Get rid of another bunch of Lst_ForEach in favour of LST_FOREACH
and simplify code accordingly.
o Fix a comment.
o Replace another bunch of Lst_ForEachs by LST_FOREACHs and simplify
code.
o Replace Lst_ForEach by LST_FOREACH.
o Now that there are no users of Lst_ForEach and Lst_ForEachFrom are
left delete these two macros and all the associated stuff.
YONETANI Tomokazu [Wed, 16 Mar 2005 17:54:59 +0000 (17:54 +0000)]
Constify all users of dktypenames to unbreak buildworld.
Reported-by: walt
David Xu [Wed, 16 Mar 2005 14:30:31 +0000 (14:30 +0000)]
Add types prgregset_t and psaddr_t which will be used by
thread debugging code.
Chris Pressey [Wed, 16 Mar 2005 08:02:05 +0000 (08:02 +0000)]
Style(9):
- Spell NULL as NULL, and '\0' as '\0'.
- Explicitly check non-booleans against 0, '\0', or NULL.
Chris Pressey [Wed, 16 Mar 2005 07:30:33 +0000 (07:30 +0000)]
Style(9):
- Braces around multiline blocks.
- Spell NULL as NULL, not 0.
- Don't treat pointers and ints as booleans; use explicit tests.
- Minor whitespace adjustments.
Joerg Sonnenberger [Wed, 16 Mar 2005 07:22:07 +0000 (07:22 +0000)]
Remove redundant declarations. Convert to ANSI C function declarations.
Convert dates to ISO style.
Chris Pressey [Wed, 16 Mar 2005 07:16:14 +0000 (07:16 +0000)]
Style(9):
- Braces around multiline blocks.
- Space after 'switch'.
- 'errno' is not a boolean, use explicit test.
- Add explanatory comment for an issue which has caused confusion at
least once in the past.
Chris Pressey [Wed, 16 Mar 2005 07:08:04 +0000 (07:08 +0000)]
Style(9):
- Braces around multiline block.
- Space after 'switch'.
- stat(2), lstat(2), link(2), unlink(2) and symlink(2) don't return
boolean values; use explicit checks against 0.
Joerg Sonnenberger [Wed, 16 Mar 2005 06:56:20 +0000 (06:56 +0000)]
Add some more path definitions as used by the Citrus framework.
Joerg Sonnenberger [Wed, 16 Mar 2005 06:54:41 +0000 (06:54 +0000)]
Add Citrus files not conflicting with the current rune implementation.
Obtained-from: NetBSD / Citrus Project
Joerg Sonnenberger [Wed, 16 Mar 2005 06:42:34 +0000 (06:42 +0000)]
More Citrus code.
Obtained-from: NetBSD
Joerg Sonnenberger [Wed, 16 Mar 2005 06:38:26 +0000 (06:38 +0000)]
More stuff for Citrus.
Chris Pressey [Wed, 16 Mar 2005 06:35:48 +0000 (06:35 +0000)]
We have _DIAGASSERT now; might as well enable it here.
Chris Pressey [Wed, 16 Mar 2005 06:33:11 +0000 (06:33 +0000)]
Followup to previous commit: use uint8_t instead of u_char, for a real
guarantee of byte-ness.
Pointed-out-by: joerg
Joerg Sonnenberger [Wed, 16 Mar 2005 06:29:03 +0000 (06:29 +0000)]
Make dktypnames const.
Joerg Sonnenberger [Wed, 16 Mar 2005 06:27:17 +0000 (06:27 +0000)]
s/index/idx. Add an explicit default case to teach GCC that idx is
used correctly.
Chris Pressey [Wed, 16 Mar 2005 06:26:49 +0000 (06:26 +0000)]
De-spaghettify:
- fail() consists of two statements and is called only once; inline it.
- The point of finish() is to reap the child, so call it reap().
- Make the child's pid local to main(), and pass it to reap().
- done() is always called after reap(), so calling done() from inside
reap() is not necessary; let main() handle it instead.
Joerg Sonnenberger [Wed, 16 Mar 2005 06:24:36 +0000 (06:24 +0000)]
Fix some warnings.
Joerg Sonnenberger [Wed, 16 Mar 2005 06:19:09 +0000 (06:19 +0000)]
idx is always used, shut GCC up by initialising it to 0.
Chris Pressey [Wed, 16 Mar 2005 06:18:56 +0000 (06:18 +0000)]
Raise WARNS to 6.
Joerg Sonnenberger [Wed, 16 Mar 2005 06:13:24 +0000 (06:13 +0000)]
Fix warnings.
Chris Pressey [Wed, 16 Mar 2005 06:08:06 +0000 (06:08 +0000)]
When dumping the contents of a struct sockaddr with an unknown address
family, dump individual bytes (u_chars) instead of u_shorts. This
avoids endianness issues while also simplifying the code.
In-discussion-with: joerg
Joerg Sonnenberger [Wed, 16 Mar 2005 05:27:27 +0000 (05:27 +0000)]
WARNS=6.
Submitted-by: Samuel J. Greear <sjg@evilcode.net>
Joerg Sonnenberger [Wed, 16 Mar 2005 05:19:11 +0000 (05:19 +0000)]
Raise WARNS to 6:
- correctly initialize globals completely
- don't shadow globals
- change some locals to match the type in the ng headers
Submitted-by: Samuel J. Greear <sjg@evilcode.net>
Chris Pressey [Wed, 16 Mar 2005 04:47:00 +0000 (04:47 +0000)]
When rendering the routename() or netname() of a struct sockaddr
with an unknown address family, use a more straightforward loop
for dumping the contents of sa_data as hex:
- start at sa_data, end at sa + sa_len (NOT sa_data + sa_len);
- don't make assumptions about the size of u_short.
Inspired-by-patch-by: Sepherosa Ziehau <sepherosa_at_softhome_dot_net>
Joerg Sonnenberger [Wed, 16 Mar 2005 04:45:07 +0000 (04:45 +0000)]
Pass const ** as second argument to the command functions,
explicitly cast this away in calls to getopt. Change some
local variables to const char * as well, to match this.
Chris Pressey [Wed, 16 Mar 2005 04:18:35 +0000 (04:18 +0000)]
Raise WARNS to 6:
- Rename local variables to avoid conflicts with globals;
- Make local functions static;
- Add missing function prototypes;
- Remove stale function prototypes and struct declarations;
- Un-nest nested extern declarations;
- Constify some char *'s where appropriate;
- Resolve comparisons between signed and unsigned;
- Copy some strings into work buffers instead of returning
a const char * from a function defined to return a char *.
Submitted-by: Sepherosa Ziehau <sepherosa_at_softhome_dot_net>
Some-tweaks-by: cpressey
Max Okumoto [Wed, 16 Mar 2005 00:15:21 +0000 (00:15 +0000)]
patch-7.135
Don't pass endc since we can calucluate it in the function that
uses it.
Max Okumoto [Wed, 16 Mar 2005 00:15:04 +0000 (00:15 +0000)]
patch-7.134
Cleanup calculation of consumed.
Max Okumoto [Wed, 16 Mar 2005 00:14:43 +0000 (00:14 +0000)]
patch-7.133
ParseModifier()
- substitute in delim
- Use shorter variable name patt instead of pattern.
- Move calc of consumed into subroutine.
Max Okumoto [Wed, 16 Mar 2005 00:14:18 +0000 (00:14 +0000)]
patch-7.132
VarParseLong()
- update value in vp->ptr instead of using a temp variable.
ParseRestEnd()
- Reduce number of explict parameter passed bu passing them
via struct VarParser.
Max Okumoto [Wed, 16 Mar 2005 00:13:56 +0000 (00:13 +0000)]
patch-7.131
ParseRestEnd()
- Removed consumed as a input param to ParseRestEnd(), since
it can be calculated.
Max Okumoto [Wed, 16 Mar 2005 00:13:38 +0000 (00:13 +0000)]
patch-7.130
ParseRestEnd(), ParseRestModifier(), ParseModifier()
- Reduce number of explict parameter passed bu passing them
via struct VarParser.
Max Okumoto [Wed, 16 Mar 2005 00:13:23 +0000 (00:13 +0000)]
patch-7.129
VarParseLong()
- Reduce number of explict parameter passed bu passing them
via struct VarParser.
- Change loop termination from endc to '\0'
Max Okumoto [Wed, 16 Mar 2005 00:13:04 +0000 (00:13 +0000)]
patch-7.128
Var_Parse()
Add temp variable so calculations can be performed before
we return.
VarParse()
Caclucate consumed via (vp->ptr - vp->input)
VaraParseShort()
Reduce number of explict parameter passed bu passing them
via struct VarParser.
Max Okumoto [Wed, 16 Mar 2005 00:09:46 +0000 (00:09 +0000)]
Var_Parse(), VarParse()
- Reduce the number of explict parameter passed to VarParse()
by passing them via struct VarParser.
Max Okumoto [Tue, 15 Mar 2005 23:40:48 +0000 (23:40 +0000)]
FreeBSD-Date: 2005/03/15 14:28:39
Author: harti
Get rid of a number of Lst_ForEach calls in favour of LST_FOREACH
and simplify the printing functions.
Max Okumoto [Tue, 15 Mar 2005 23:36:57 +0000 (23:36 +0000)]
FreeBSD-Date: 2005/03/15 14:25:24
Author: harti
Yet another version of passing the last line to ParseFinishLine().
It turns out that some ports use the obscure feature of spreading
a dependency block across multiple include files. While this seems
bad style, allow it for now and call said function only at end of
all input to process the really last line of everything.
Max Okumoto [Tue, 15 Mar 2005 23:31:15 +0000 (23:31 +0000)]
Get rid of another Lst_ForEach in favour of LST_FOREACH. Get rid
of the now unused struct LstSrc and collapes two functions into one.
FreeBSD-Date: 2005/03/14 17:04:58
Author: harti
Max Okumoto [Tue, 15 Mar 2005 23:27:19 +0000 (23:27 +0000)]
Move the creation of a Src structure into its own function.
FreeBSD-Date: 2005/03/14 12:04:20
Author: harti
Matthew Dillon [Tue, 15 Mar 2005 20:53:43 +0000 (20:53 +0000)]
src/games WARNS 6 cleanups.
Submitted-by: Craig Dooley <xlnxminusx@gmail.com>
Matthew Dillon [Tue, 15 Mar 2005 20:42:14 +0000 (20:42 +0000)]
Bring in some CAM bug fixes from FreeBSD.
Submitted-by: Xin LI <delphij@frontfree.net>
Obtained-from: FreeBSD
Sascha Wildner [Tue, 15 Mar 2005 19:16:17 +0000 (19:16 +0000)]
Remove the euc(4) and utf2(4) man pages. They are now in section 5.
David Xu [Tue, 15 Mar 2005 15:09:22 +0000 (15:09 +0000)]
Fix incorrect comment.
David Xu [Tue, 15 Mar 2005 12:27:43 +0000 (12:27 +0000)]
Add following prototypes:
pthread_mutex_timedlock
pthread_rwlock_timedrdlock
pthread_rwlock_timedwrlock
David Xu [Tue, 15 Mar 2005 12:04:14 +0000 (12:04 +0000)]
Add prototypes for pthread_condattr_getclock and pthread_condattr_setclock.
David Xu [Tue, 15 Mar 2005 12:01:51 +0000 (12:01 +0000)]
unconstify a modifiable parameter.
David Xu [Tue, 15 Mar 2005 11:26:34 +0000 (11:26 +0000)]
Fix comments.
David Xu [Tue, 15 Mar 2005 11:24:23 +0000 (11:24 +0000)]
Add clockid to _thr_umtx_wait, so that clockid attribute of condition
variable can be supported.
Matthew Dillon [Tue, 15 Mar 2005 02:45:59 +0000 (02:45 +0000)]
flesh out a tiny bit so people at least know what the help key is.
Matthew Dillon [Tue, 15 Mar 2005 02:43:18 +0000 (02:43 +0000)]
Add 'mined' to the bin build and add a minimal manual page for it.
Matthew Dillon [Tue, 15 Mar 2005 02:32:57 +0000 (02:32 +0000)]
Support XTERM function key escape sequences.
Matthew Dillon [Tue, 15 Mar 2005 02:25:25 +0000 (02:25 +0000)]
Make ^L redraw the screen, like people expect.
Make ^Z tty-stop the editor, like people expect (but don't try to
deal with window resizes).
Matthew Dillon [Tue, 15 Mar 2005 02:13:15 +0000 (02:13 +0000)]
Bring in PicoBSD changes from the FreeBSD Attic.
Prepend the Minix LICENSE in all source files.
Add a few other random defines... slightly different from the way FreeBSD
did it.
Matthew Dillon [Tue, 15 Mar 2005 01:56:24 +0000 (01:56 +0000)]
Bring in the minix editor 'mined', in its original form except for necessary
DragonFly CVS IDs at the top of each file. Bring the LICENSE file in from
http://www.cs.vu.nl/pub/minix/LICENSE.
Joerg Sonnenberger [Tue, 15 Mar 2005 01:54:28 +0000 (01:54 +0000)]
Stop building old GDB.
Joerg Sonnenberger [Tue, 15 Mar 2005 01:54:58 +0000 (01:54 +0000)]
GC
Matthew Dillon [Tue, 15 Mar 2005 01:12:23 +0000 (01:12 +0000)]
Implement CLOCK_MONOTONIC using getnanouptime(), which in DragonFly is
guarenteed to be monotonically increasing.
Joerg Sonnenberger [Mon, 14 Mar 2005 23:55:41 +0000 (23:55 +0000)]
Instead of using the non-standard conforming %+ format string,
use nl_langinfo to query the default format string.
Joerg Sonnenberger [Mon, 14 Mar 2005 15:32:51 +0000 (15:32 +0000)]
Use NXCC to build make_hash and make_keys.
Joerg Sonnenberger [Mon, 14 Mar 2005 15:15:14 +0000 (15:15 +0000)]
Move SHLIB_MAJOR / SHLIB_MINOR assignment up into Makefile.inc.
Joerg Sonnenberger [Mon, 14 Mar 2005 14:46:44 +0000 (14:46 +0000)]
Merge r1.2 from NetBSD:
anonymous union between chlen and _UTF7StatePrive has
compilation problem with gcc295.
this union attempt to make mbsinit(3) handle multibyte state correctly,
but it's useless as far as we use utf-7 only iconv interface.
so i eliminate ctype feature.
patch contributed by Joerg Sonnenberg(who porting Citrus to DragonFlyBSD).
and yamt-san gave me advice, thanks a lot.
Joerg Sonnenberger [Mon, 14 Mar 2005 14:26:16 +0000 (14:26 +0000)]
Back out switch to getopt_long implementation of getopt, it breaks
too much. This will be done (more carefully) later.
Max Okumoto [Mon, 14 Mar 2005 13:31:36 +0000 (13:31 +0000)]
Convert a couple of other uses of Lst_ForEach to LST_FOREACH and
simplify code.
FreeBSD-Date: 2005/03/14 08:29:58
Author: harti
Joerg Sonnenberger [Mon, 14 Mar 2005 12:48:57 +0000 (12:48 +0000)]
Add '+' to argument list of find to prevent getopt from reordering
the arguments.
Joerg Sonnenberger [Mon, 14 Mar 2005 12:36:25 +0000 (12:36 +0000)]
Correctly return -1 for "-" as argument as required by SUS.
Fix a typo where specifying a single invalid short option could
result in an infinite loop.
Joerg Sonnenberger [Mon, 14 Mar 2005 11:55:33 +0000 (11:55 +0000)]
SUS says that getopt() should return -1 immediately if the argument is "-".
Correctly handle this case. We formerly depended on the (undocumented)
behaviour of getopt to return this as '-' option.
Eirik Nygaard [Mon, 14 Mar 2005 09:14:57 +0000 (09:14 +0000)]
Fix nested comments. Vim added another */ while I copy/pasted some of the lines
there.
Submitted by: Chris Pressey <cpressey@catseye.mine.nu>
YONETANI Tomokazu [Sun, 13 Mar 2005 22:22:42 +0000 (22:22 +0000)]
Change the behavior of -fstype so as not to terminate the process
when statfs() failed(meaning that it couldn't type a filesystem),
but report false and continue. This fixes locate.updatedb program
failing on a mounted-but-unavailable mountpoint unless such a
mountpoint is explicitly specified in PRUNEPATHS.
Reported-by: Rumko <rumcic (at) gmail.com>
Ok: Matt
Matthew Dillon [Sun, 13 Mar 2005 22:17:40 +0000 (22:17 +0000)]
Fix a couple of NFS client side caching snafus. First, we have to update
np->n_mtime if the file has not being modified locally, otherwise n_mtime
becomes stale and both the dynamic cache timeout calculation breaks and
stat() calls checking for the last modified time break. This was causing
indeterminant operation with make for obvious reasons.
The Second problem was that we were not checking for an attrstamp of 0
(indicating a timed out entry), resulting in another incorrect timeout
calculation.
Matthew Dillon [Sun, 13 Mar 2005 21:33:48 +0000 (21:33 +0000)]
Add a function that returns an approximate time_t for realtime for
routines which do not need the exactness of microtime().
Liam J. Foy [Sun, 13 Mar 2005 21:17:47 +0000 (21:17 +0000)]
- Remove sys/stat.h, sys/signal.h, varargs.h and ctype.h
- Issue the correct amount of space for hostname
- Use strtol to parse integers
- We want to use signal.h, not sys/signal.h
- Use STDIN_FILENO over '0'. More informative.
- Static functions
- Remove unnecessary cast
- Style(9)
Ok: Joerg
Joerg Sonnenberger [Sun, 13 Mar 2005 20:44:38 +0000 (20:44 +0000)]
Another man page which was moved to lib/libc/gen.
Joerg Sonnenberger [Sun, 13 Mar 2005 20:20:22 +0000 (20:20 +0000)]
Use libc's getopt_long, I've been building and using ports for a while with
this. Also move th USE_RC_SUBR check into the shared section, just in case
we forget this for an override port.
Joerg Sonnenberger [Sun, 13 Mar 2005 20:13:41 +0000 (20:13 +0000)]
Use getopt_long's getopt implementation.
Joerg Sonnenberger [Sun, 13 Mar 2005 19:18:29 +0000 (19:18 +0000)]
Move ctype man page from locale to gen, they are locale-sensitive, but
not locale-specific. Move description of euc, gb18030, gbk, utf2 and utf8
to share/man/man5, they are not device drivers.
Joerg Sonnenberger [Sun, 13 Mar 2005 19:05:23 +0000 (19:05 +0000)]
Don't build groff as buildtool, it's not needed anymore.
Joerg Sonnenberger [Sun, 13 Mar 2005 18:42:39 +0000 (18:42 +0000)]
Don't create some directories which will now be empty.
Joerg Sonnenberger [Sun, 13 Mar 2005 18:05:53 +0000 (18:05 +0000)]
Don't build / install the various historic troff papers to
/usr/share/doc. Don't install the KAME implementation details,
the document is outdated too.