]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
27 years agoadd dtterm
Torsten Blum [Mon, 13 Jan 1997 12:37:05 +0000 (12:37 +0000)]
add dtterm

Obtained-From: Xinside's CDE for FreeBSD with Thomas Roell's permission

27 years agoSynchronize with sys/i386/i386/userconfig.c revision 1.80.
KATO Takenori [Mon, 13 Jan 1997 12:32:09 +0000 (12:32 +0000)]
Synchronize with sys/i386/i386/userconfig.c revision 1.80.

27 years agoFix BB profiling.
Poul-Henning Kamp [Mon, 13 Jan 1997 12:15:29 +0000 (12:15 +0000)]
Fix BB profiling.

Submitted by: bde

27 years agoMerge the en version change made in 1.189 -> 1.190.
Masafumi Max NAKANE [Mon, 13 Jan 1997 12:02:43 +0000 (12:02 +0000)]
Merge the en version change made in 1.189 -> 1.190.

27 years agoAdd Luigi Rizzo, who was inadvertently left out.
Thomas Gellekum [Mon, 13 Jan 1997 11:16:09 +0000 (11:16 +0000)]
Add Luigi Rizzo, who was inadvertently left out.

27 years agoBack out previous fix. It was bogus.
Jordan K. Hubbard [Mon, 13 Jan 1997 10:55:49 +0000 (10:55 +0000)]
Back out previous fix.  It was bogus.

27 years agoMerge the en version changes made in 1.188 -> 1.189.
Masafumi Max NAKANE [Mon, 13 Jan 1997 10:15:51 +0000 (10:15 +0000)]
Merge the en version changes made in 1.188 -> 1.189.

27 years agoCorrect a problem which broke +DISPLAY file handling.
Jordan K. Hubbard [Mon, 13 Jan 1997 10:14:26 +0000 (10:14 +0000)]
Correct a problem which broke +DISPLAY file handling.

27 years agoInstall a dir file if one doesn't exist. This seems the most likely
Jordan K. Hubbard [Mon, 13 Jan 1997 09:54:38 +0000 (09:54 +0000)]
Install a dir file if one doesn't exist.  This seems the most likely
place to do it, I guess, and it's the utility which falls over without one.

27 years agoAdd Simon Marlow for his cgoban port.
Thomas Gellekum [Mon, 13 Jan 1997 08:51:04 +0000 (08:51 +0000)]
Add Simon Marlow for his cgoban port.

27 years agoDocument effect of login class capabilities.
David Nugent [Mon, 13 Jan 1997 06:52:24 +0000 (06:52 +0000)]
Document effect of login class capabilities.

27 years agoMake su login_cap savvy.
David Nugent [Mon, 13 Jan 1997 06:39:19 +0000 (06:39 +0000)]
Make su login_cap savvy.
As with login(1), LOGIN_CAP_AUTH is not yet enabled since we don't
yet have authorisation modules.

27 years agoJust 'warn' the user that they are mixing dedicated drivers and generic
Nate Williams [Mon, 13 Jan 1997 05:22:40 +0000 (05:22 +0000)]
Just 'warn' the user that they are mixing dedicated drivers and generic
code if the option 'LINT_PCCARD_HACK' is specified (which should only
be used in LINT).

Demanded by: bde

27 years agoAdded options 'LINT_PCCARD_HACK' which (will very soon) allow LINT to
Nate Williams [Mon, 13 Jan 1997 05:21:09 +0000 (05:21 +0000)]
Added options 'LINT_PCCARD_HACK' which (will very soon) allow LINT to
compile again.  The code to protect users from combining the dedicated
PCCARD drivers and the generic code is a warning if the above option
is included in the config file.

Demanded by: bde

27 years agoTurn this into legal texinfo. Since I don't know what the author
Jordan K. Hubbard [Mon, 13 Jan 1997 05:02:49 +0000 (05:02 +0000)]
Turn this into legal texinfo.  Since I don't know what the author
intended (and clearly, neither did the author :) I've just used the
bullet list attribute everywhere.

27 years ago - Add [?] key and corresponding helptext/pager.
Mike Smith [Mon, 13 Jan 1997 04:13:15 +0000 (04:13 +0000)]
 - Add [?] key and corresponding helptext/pager.
   (helptext from Philippe Regnauld)
 - Make introfunc() work with serial terminals.
   (submitted by Jean-Marc Zucconi)
 - Eliminate excessive statusline redraw during screen updates.
   (requested by Peter Wemm)
 - Some trivial output formatting dinks.

27 years agoBump libutil.so version (2.1 -> 2.2) since a whole heap of new functions
Peter Wemm [Mon, 13 Jan 1997 03:05:04 +0000 (03:05 +0000)]
Bump libutil.so version (2.1 -> 2.2) since a whole heap of new functions
were added with the login class stuff.  This is needed since libutil.so.2.1
is what is used in RELENG_2_2 and well into the release cycle.  We only
bump once per release cycle as needed.

27 years agoImplement minimal login class support (ie: does a setusercontext()).
Peter Wemm [Mon, 13 Jan 1997 02:52:30 +0000 (02:52 +0000)]
Implement minimal login class support (ie: does a setusercontext()).
Enabled by defining LOGIN_CAP in Makefile, on by default.

27 years agoReport the reason why select unexpectedly failed..
Peter Wemm [Mon, 13 Jan 1997 02:34:56 +0000 (02:34 +0000)]
Report the reason why select unexpectedly failed..

27 years agoSkip non-existent subdirectories. The diff is rather large because
Satoshi Asami [Mon, 13 Jan 1997 02:13:19 +0000 (02:13 +0000)]
Skip non-existent subdirectories.  The diff is rather large because
the original logic went into a section of code assuming some
incarnation is there, but it's basically a "test -d" fix.  Closes PR
ports/2082.

Reviewed by: max ("although I didn't test it, it looks fine")

27 years agoSort cross references.
Wolfram Schneider [Mon, 13 Jan 1997 00:25:51 +0000 (00:25 +0000)]
Sort cross references.

27 years agoAdd RCS id's since these files have left the default branch.
Steve Price [Sun, 12 Jan 1997 21:29:55 +0000 (21:29 +0000)]
Add RCS id's since these files have left the default branch.

27 years agoThis doesn't change any functionality, it simple makes yacc(1)
Steve Price [Sun, 12 Jan 1997 21:16:25 +0000 (21:16 +0000)]
This doesn't change any functionality, it simple makes yacc(1)
compile -Wall clean.

27 years agoHandle linuxdoc docs with a declaration like:
John Fieber [Sun, 12 Jan 1997 20:59:49 +0000 (20:59 +0000)]
Handle linuxdoc docs with a declaration like:

  <!doctype linuxdoc system>

which is the standard practice used for the linux HOWTO documents.

27 years agoUse the RTLD_NOW symbol, now that it is defined in <dlfcn.h>.
John Polstra [Sun, 12 Jan 1997 19:59:26 +0000 (19:59 +0000)]
Use the RTLD_NOW symbol, now that it is defined in <dlfcn.h>.

27 years agoConvert to mdoc format.
John Polstra [Sun, 12 Jan 1997 19:55:49 +0000 (19:55 +0000)]
Convert to mdoc format.
Add descriptions of RTLD_LAZY and RTLD_NOW.
Correct the synopsis to agree with the actual function prototypes.
Add clarifications of a few things.
Clean up the wording in a few places.

27 years agoFix the following old problem:
Joerg Wunsch [Sun, 12 Jan 1997 19:38:32 +0000 (19:38 +0000)]
Fix the following old problem:

  date: 1994/10/09 07:37:18;  author: davidg;  state: Exp;  lines: +7 -1
  #if 0'd out the meat of the swap code until I get a chance to rewrite it.

...mainly by stealing the code from pstat(8).

27 years agoAdd definitions of RTLD_LAZY and RTLD_NOW.
John Polstra [Sun, 12 Jan 1997 18:37:46 +0000 (18:37 +0000)]
Add definitions of RTLD_LAZY and RTLD_NOW.

27 years agoFix calendar so that you can run it like:
Mike Pritchard [Sun, 12 Jan 1997 18:35:14 +0000 (18:35 +0000)]
Fix calendar so that you can run it like:

calendar -t 0101 -f file

Previously calendar's time processing routine directly
modified the "0101" argument" which confused getopt.
The time routines now make a copy of the argument
to mess with.

27 years agoAdd "lib/csu/i386" to the "includes" target, so that <dlfcn.h> is
John Polstra [Sun, 12 Jan 1997 18:32:53 +0000 (18:32 +0000)]
Add "lib/csu/i386" to the "includes" target, so that <dlfcn.h> is
installed before it is needed.
Noticed by: David Conran <lucifer@maths.uq.oz.au>

27 years agoAdd commented out example entry for imap4
Andrey A. Chernov [Sun, 12 Jan 1997 17:55:16 +0000 (17:55 +0000)]
Add commented out example entry for imap4

27 years agoAdd comments about variables and targets. Incomplete, see '???'.
Wolfram Schneider [Sun, 12 Jan 1997 16:20:13 +0000 (16:20 +0000)]
Add comments about variables and targets. Incomplete, see '???'.

27 years agoAdd a general comment about this include file.
Wolfram Schneider [Sun, 12 Jan 1997 16:12:40 +0000 (16:12 +0000)]
Add a general comment about this include file.

27 years agoI added both features Peter Wemm requested. First, the new option
Joshua Peck Macdonald [Sun, 12 Jan 1997 14:43:22 +0000 (14:43 +0000)]
I added both features Peter Wemm requested.  First, the new option
--forceentry=TEXT will override any entries contained in the info
file.  Second, I made it recognize that "gmp" and "gmp.info" are
the same when looking for whether the info file being installed is
already installed.

27 years agoinclude ctype.h to pick isprint macro
Andrey A. Chernov [Sun, 12 Jan 1997 13:34:13 +0000 (13:34 +0000)]
include ctype.h to pick isprint macro

27 years agoMerge the en version changes made in 1.174 -> 1.188.
Masafumi Max NAKANE [Sun, 12 Jan 1997 13:32:48 +0000 (13:32 +0000)]
Merge the en version changes made in 1.174 -> 1.188.

27 years agoMerge the en version changes made in 1.45 -> 1.47.
Masafumi Max NAKANE [Sun, 12 Jan 1997 13:31:38 +0000 (13:31 +0000)]
Merge the en version changes made in 1.45 -> 1.47.

27 years agoMove some lines to make the list in alphabetical order by the entity name.
Masafumi Max NAKANE [Sun, 12 Jan 1997 13:29:43 +0000 (13:29 +0000)]
Move some lines to make the list in alphabetical order by the entity name.

27 years agoAdd OpenBSD support. Change some comments to reflect the fact. Also:
Satoshi Asami [Sun, 12 Jan 1997 12:37:48 +0000 (12:37 +0000)]
Add OpenBSD support.  Change some comments to reflect the fact.  Also:

(1) Change commented out MAINTAINER to FreeBSD_MAINTAINER and
    OpenBSD_MAINTAINER.  These are not comments anymore, so we may
    even use it in the future.

(2) Instead of the ".if ${OPSYS} = "NetBSD" hack, use ".if exists()"
    to find the location of md5 an tar.  Play similar trick for fetch
    (OpenBSD uses /usr/bin/ftp which groks http: addresses).

This commit includes most of the changes made in 1.242 (although many
of them are done differently after more discussion).  One thing that
is conspicuously missing is NOMANCOMPRESS, which has been postponed
until Warner figures out what exactly the situation is on the OpenBSD
ports paradigm.  (In a nutshell, we can't just define NOMANCOMPRESS in
this file even if uncompressed manpages is the default for OpenBSD,
because that will take away the ability of individual users to select
manpage compression.)

Reviewed by: imp@openbsd.org

27 years agoTypo fix, and translation improvement.
Masafumi Max NAKANE [Sun, 12 Jan 1997 12:17:43 +0000 (12:17 +0000)]
Typo fix, and translation improvement.
Submitted by: Yoshiaki Uchikawa <yoshiaki@kt.rim.or.jp>

27 years ago(1) New variable, SCRIPTS_ENV, which is passed to scripts/configure
Satoshi Asami [Sun, 12 Jan 1997 11:48:26 +0000 (11:48 +0000)]
(1) New variable, SCRIPTS_ENV, which is passed to scripts/configure
    and scripts/{pre,post}-* as environment vars.  Also, if BATCH is
    set, "BATCH=yes" is automatically added to SCRIPTS_ENV.
    (Requested by: max)

(2) The INSTALL_* macros are added to SCRIPTS_ENV and MAKE_ENV as
    BSD_INSTALL_*.  (Requested by: obrien)

(3) New variable MOTIF_ONLY, which will only build ports with
    REQUIRES_MOTIF defined.  This doesn't do dependencies right (what
    if the depended port doesn't need Motif) yet.

(4) Try not to clean the same port twice in clean-depends when (for
    instance) it's defined in both BUILD_DEPENDS and INSTALL_DEPENDS.
    Note that it won't check chained dependencies so you may still see
    the same port cleaned multiple times, but checking that far will
    surely make this run slower than the un-"optimized" case so I left
    it as is.  (Requested by: jkh)

(5) Ignore *.rej files in patches/ directory in addition to *~ and
    *.orig.

27 years agoMerge the en version changes made in 1.4 -> 1.5.
Masafumi Max NAKANE [Sun, 12 Jan 1997 11:28:09 +0000 (11:28 +0000)]
Merge the en version changes made in 1.4 -> 1.5.
(No actual change in the translation.)

27 years agoAdd more protection when running as init and don't do stupid things.
Jordan K. Hubbard [Sun, 12 Jan 1997 11:13:43 +0000 (11:13 +0000)]
Add more protection when running as init and don't do stupid things.
Fix looping in fixit cdrom selection by adding a missing break.
Move unmount of cdrom down to proper place.

27 years agoAdded two useful URL's.
David E. O'Brien [Sun, 12 Jan 1997 10:07:52 +0000 (10:07 +0000)]
Added two useful URL's.

27 years agoUpdated the Pentium chipset descriptions as suggested by Rodney Grimes.
David E. O'Brien [Sun, 12 Jan 1997 09:36:52 +0000 (09:36 +0000)]
Updated the Pentium chipset descriptions as suggested by Rodney Grimes.

Submitted by: Rodney Grimes

27 years ago*gulp* break this file off the vendor branch in order to fix the
Peter Wemm [Sun, 12 Jan 1997 08:28:56 +0000 (08:28 +0000)]
*gulp* break this file off the vendor branch in order to fix the
info entry.  It's a real shame that install-info can't override these.

I'll fix the repository for this (and the readline commit as well) once
this has been fixed properly...  Looking at the install-info source gives
me a headache. :-(  This should be enough to get the tree to build again.

27 years agoAck. When I moved the getsockopt() calls, I forgot to move the lines
Bill Paul [Sun, 12 Jan 1997 08:18:17 +0000 (08:18 +0000)]
Ack. When I moved the getsockopt() calls, I forgot to move the lines
that initialize the getsockopt() args (type and len) too.

27 years agoPatch up the readline info building. This one is a bit of a problem
Peter Wemm [Sun, 12 Jan 1997 08:10:55 +0000 (08:10 +0000)]
Patch up the readline info building.  This one is a bit of a problem
since the source name is not the same as the texinfo name so we have to
use SRCS=.  This means we can't build two info sets in the same directory
so I've split it.

27 years agoDon't conflict with INFODIR?=${SHAREDIR}/info from the system Makefiles
Peter Wemm [Sun, 12 Jan 1997 07:36:43 +0000 (07:36 +0000)]
Don't conflict with INFODIR?=${SHAREDIR}/info from the system Makefiles

27 years agoone more tweak to make it fly without INFO defined. (like bsd.prog.mk)
Peter Wemm [Sun, 12 Jan 1997 07:30:23 +0000 (07:30 +0000)]
one more tweak to make it fly without INFO defined. (like bsd.prog.mk)

27 years agoMake this slightly less bogus. I'm not 100% sure what is correct here,
Peter Wemm [Sun, 12 Jan 1997 07:23:58 +0000 (07:23 +0000)]
Make this slightly less bogus.  I'm not 100% sure what is correct here,
but replacing the "dir" file unconditionally isn't it.  During the course
of development, if .info files go away from the sources, nothing removes
them from /usr/share/info, this is the same as system binaries etc.
Removing the entire index isn't helpful, because you've got to reinstall
the entire tree to get it back again.  bsd.info.mk has a reference to
/usr/share/info/dir-tmpl, I wonder if it once created dir if needed?

27 years agoMake this (hopefully) build now. There was no dir entry.
Peter Wemm [Sun, 12 Jan 1997 07:18:05 +0000 (07:18 +0000)]
Make this (hopefully) build now.  There was no dir entry.

27 years agooops, missed a ${BINDIR} -> ${INFODIR}. Also remove the slash between
Peter Wemm [Sun, 12 Jan 1997 07:15:24 +0000 (07:15 +0000)]
oops, missed a ${BINDIR} -> ${INFODIR}.  Also remove the slash between
${DESTDIR}/${INFODIR}.

27 years agoBack out the last backout. This should work now that the bsd.info.mk
Peter Wemm [Sun, 12 Jan 1997 07:10:57 +0000 (07:10 +0000)]
Back out the last backout.  This should work now that the bsd.info.mk
spammed revisions have been restored.

27 years agorecover spammed revisions 1.20 and 1.21. This fixes the breaks in the
Peter Wemm [Sun, 12 Jan 1997 07:05:07 +0000 (07:05 +0000)]
recover spammed revisions 1.20 and 1.21.  This fixes the breaks in the
builds when INFO is commented out.

Pointed out by: bde

27 years agoFix path to readline docs, since it's now in contrib
Peter Wemm [Sun, 12 Jan 1997 06:33:27 +0000 (06:33 +0000)]
Fix path to readline docs, since it's now in contrib

27 years agoRemove yacc from the lib-tools target since it is now
Steve Price [Sun, 12 Jan 1997 05:33:24 +0000 (05:33 +0000)]
Remove yacc from the lib-tools target since it is now
back to its original form.

Requested by: jkh

27 years agoDisconnect the doc subdir as well, bsd.info.mk freaks out if there
Peter Wemm [Sun, 12 Jan 1997 05:03:29 +0000 (05:03 +0000)]
Disconnect the doc subdir as well, bsd.info.mk freaks out if there
is no ${INFO} variable set.. :-(

27 years agoDisconnect the doc subdir until somebody fixes this. I don't see how this
Peter Wemm [Sun, 12 Jan 1997 05:01:29 +0000 (05:01 +0000)]
Disconnect the doc subdir until somebody fixes this.  I don't see how this
could possibly have worked during the testing before the last commit. :-(

27 years agoRevert the '-o output_file_name' option changes.
Steve Price [Sun, 12 Jan 1997 04:04:52 +0000 (04:04 +0000)]
Revert the '-o output_file_name' option changes.

27 years agoRevert changes to make the .y.c rule 'make -j n' safe.
Steve Price [Sun, 12 Jan 1997 04:01:25 +0000 (04:01 +0000)]
Revert changes to make the .y.c rule 'make -j n' safe.

27 years ago- Save parent pid and check current pid against it upon entry to
Bill Paul [Sun, 12 Jan 1997 02:48:09 +0000 (02:48 +0000)]
- Save parent pid and check current pid against it upon entry to
  terminate(). This makes doubly sure we don't trip the SIGTERM handler
  in a child process. Suggested by: phk.

- It's int main(argc, argv), not void main(argc, argv), gosh darn it.

27 years agoAdd missing RETURN VALUES/ERRORS sections.
Mike Pritchard [Sun, 12 Jan 1997 00:38:36 +0000 (00:38 +0000)]
Add missing RETURN VALUES/ERRORS sections.

27 years agoCorrect typos and spelling errors.
John Polstra [Sun, 12 Jan 1997 00:19:14 +0000 (00:19 +0000)]
Correct typos and spelling errors.

27 years agoAdd support for the LD_BIND_NOW environment variable. If it is set to a
John Polstra [Sun, 12 Jan 1997 00:16:36 +0000 (00:16 +0000)]
Add support for the LD_BIND_NOW environment variable.  If it is set to a
nonempty string, then function calls are relocated at program start-up
rather than lazily.  This variable is standard on Sun and SVR4 systems.

The dlopen() function now supports both lazy and immediate binding, as
determined by its "mode" argument, which can be either 1 (RTLD_LAZY) or
2 (RTLD_NOW).  I will add defines of these symbols to <dlfcn.h> as soon
as I've done a little more checking to make sure they won't cause
collisions or bootstrapping problems that would break "make world".

The "LD_*" environment variables which alter dynamic linker behavior are
now treated as unset if they are set to the empty string.  This agrees
with the standard SVR4 conventions for the dynamic linker.

Add a work-around for programs compiled with certain buggy versions of
crt0.o.  The buggy versions failed to set the "crt_ldso" member of the
interface structure.  This caused certain error messages from the
dynamic linker to begin with "(null)" instead of the pathname of the
dynamic linker.

27 years agoSet LD_TRACE_LOADED_OBJECTS to "1" instead of to "". The dynamic linker
John Polstra [Sun, 12 Jan 1997 00:10:57 +0000 (00:10 +0000)]
Set LD_TRACE_LOADED_OBJECTS to "1" instead of to "".  The dynamic linker
now treats empty "LD_*" environment variables as if they were unset, per
the standard SVR4 conventions.

27 years agoAdd support for the LD_BIND_NOW environment variable. If it is set to a
John Polstra [Sun, 12 Jan 1997 00:09:11 +0000 (00:09 +0000)]
Add support for the LD_BIND_NOW environment variable.  If it is set to a
nonempty string, then function calls are relocated at program start-up
rather than lazily.  This variable is standard on Sun and SVR4 systems.

The dlopen() function now supports both lazy and immediate binding, as
determined by its "mode" argument, which can be either 1 (RTLD_LAZY) or
2 (RTLD_NOW).  I will add defines of these symbols to <dlfcn.h> as soon
as I've done a little more checking to make sure they won't cause
collisions or bootstrapping problems that would break "make world".

27 years agoRename the DIAGNOSTICS sections in several man pages
Mike Pritchard [Sun, 12 Jan 1997 00:09:02 +0000 (00:09 +0000)]
Rename the DIAGNOSTICS sections in several man pages
to RETURN VALUES like they should be.

27 years agoSet the "crt_ldso" member of the crt-to-ld.so interface structure. This
John Polstra [Sat, 11 Jan 1997 23:59:34 +0000 (23:59 +0000)]
Set the "crt_ldso" member of the crt-to-ld.so interface structure.  This
was apparently overlooked at the time the member was added.  Its absence
causes some error messages from the dynamic linker to begin with
"(null):" instead of with the pathname of the dynamic linker as they
should.

I am also adding a work-around to the dynamic linker, to cope with
legacy binaries that were built with older versions of crt0.

27 years agoRemove the EINVAL error from the ERRORS sections that
Mike Pritchard [Sat, 11 Jan 1997 23:56:32 +0000 (23:56 +0000)]
Remove the EINVAL error from the ERRORS sections that
say is means that a pathname had the high-order bit
set, since this is no longer an error.

27 years agoCorrect entity entry for Ricardo AG, make him the primary contact
Jordan K. Hubbard [Sat, 11 Jan 1997 23:54:15 +0000 (23:54 +0000)]
Correct entity entry for Ricardo AG, make him the primary contact
for gallery entries.

27 years agoWhen we changed pmap_protect to support adding the writeable
John Dyson [Sat, 11 Jan 1997 23:50:42 +0000 (23:50 +0000)]
When we changed pmap_protect to support adding the writeable
attribute to a page range, we forgot to set the PG_WRITEABLE
flag in the vm_page_t.  This fixes that problem.

27 years agoInitialize pcic_imask with SWI_MASK to interference from timeout routines.
Nate Williams [Sat, 11 Jan 1997 23:40:08 +0000 (23:40 +0000)]
Initialize pcic_imask with SWI_MASK to interference from timeout routines.

Suggested by: bde

27 years agoMinor mdoc fixes in msync.2 and munmap.2.
Mike Pritchard [Sat, 11 Jan 1997 23:33:18 +0000 (23:33 +0000)]
Minor mdoc fixes in msync.2 and munmap.2.

Add RETURN VALUES and ERRORS sections to mincore.

Closes PR# 1493.

27 years agoForgot a .El macro.
Mike Pritchard [Sat, 11 Jan 1997 23:26:44 +0000 (23:26 +0000)]
Forgot a .El macro.

27 years agoAdd RETURN VALUES and ERRORS sections.
Mike Pritchard [Sat, 11 Jan 1997 23:20:29 +0000 (23:20 +0000)]
Add RETURN VALUES and ERRORS sections.

27 years agoSome mdoc cleanup. Also added a RETURN VALUES and ERRORS
Mike Pritchard [Sat, 11 Jan 1997 22:51:27 +0000 (22:51 +0000)]
Some mdoc cleanup.  Also added a RETURN VALUES and ERRORS
section.

Part of PR# 1493.

27 years agoRemove the comment about file names having to
Mike Pritchard [Sat, 11 Jan 1997 22:31:47 +0000 (22:31 +0000)]
Remove the comment about file names having to
be 7 bit clean, since it isn't true anymore.

Part of PR# 1493.

27 years agoPublic identifiers for Docbook 3.0
John Fieber [Sat, 11 Jan 1997 21:54:40 +0000 (21:54 +0000)]
Public identifiers for Docbook 3.0

27 years agoAdd docbook 3.0
John Fieber [Sat, 11 Jan 1997 21:53:52 +0000 (21:53 +0000)]
Add docbook 3.0

27 years agoRemove docbook 2.4.1
John Fieber [Sat, 11 Jan 1997 21:52:22 +0000 (21:52 +0000)]
Remove docbook 2.4.1

27 years agoBring stat(2) into line with what is now actually in
Mike Pritchard [Sat, 11 Jan 1997 21:45:57 +0000 (21:45 +0000)]
Bring stat(2) into line with what is now actually in
stat.h.  Also add a little blurb regarding st_mtime &
friends clarifiying how they are defined in a
non-_POSIX_SOURCE envorinment.  Closes PR# 1089.

27 years agoDocument the fact that the administrator may have
Mike Pritchard [Sat, 11 Jan 1997 20:17:05 +0000 (20:17 +0000)]
Document the fact that the administrator may have
to change syslog's config file in order for all of
ftpd's log messages to be displayed by syslogd.

Closes PR# 1559.

27 years agoChange Brian Tao's mail address from taob@io.org to
Mike Pritchard [Sat, 11 Jan 1997 20:06:03 +0000 (20:06 +0000)]
Change Brian Tao's mail address from taob@io.org to
taob@risc.org as requested (It looks like Nate forgot or
didn't know about this one, or I just jumped the gun
and got to it before he did :-).

Submitted by: Brian Tao <taob@risc.org>

27 years agoAdd a couple of xrefs. Closes PR# 1280.
Mike Pritchard [Sat, 11 Jan 1997 19:58:11 +0000 (19:58 +0000)]
Add a couple of xrefs.  Closes PR# 1280.

27 years agoAdd a couple of xrefs.
Mike Pritchard [Sat, 11 Jan 1997 19:57:47 +0000 (19:57 +0000)]
Add a couple of xrefs.

27 years agoAdd a couple of additional xrefs.
Mike Pritchard [Sat, 11 Jan 1997 19:57:27 +0000 (19:57 +0000)]
Add a couple of additional xrefs.

27 years agoMinor mdoc cleanup.
Mike Pritchard [Sat, 11 Jan 1997 19:47:07 +0000 (19:47 +0000)]
Minor mdoc cleanup.

27 years agoAdd a blurb describing the fact that ls will print the
Mike Pritchard [Sat, 11 Jan 1997 19:24:22 +0000 (19:24 +0000)]
Add a blurb describing the fact that ls will print the
year in place of the hour/minute fields if the time is
more than 6 months in the past or future.

Also some minor mdoc cleanup.

27 years agoMake ls include the year when displaying times that are more than
Mike Pritchard [Sat, 11 Jan 1997 19:15:53 +0000 (19:15 +0000)]
Make ls include the year when displaying times that are more than
6 months into the future.  Closes PR# 1657.

Submitted by: Sakari Jalowaara <sja.home.tekla.fi>

27 years agoWhoops, typo.
Nate Williams [Sat, 11 Jan 1997 18:32:27 +0000 (18:32 +0000)]
Whoops, typo.

27 years agoUpdate the PCIC controller's imask with individual slot IRQ's.
Nate Williams [Sat, 11 Jan 1997 18:23:20 +0000 (18:23 +0000)]
Update the PCIC controller's imask with individual slot IRQ's.

Assuming that the intr_mask[] was updated by changing the maskptrs (the
existing update_intr_masks() function will not work) this code was
written so the PCIC controller insertion/removal events will not
interrupt the card IRQ handler events.

Some possible scenarios:
+ Card is removed during IRQ handler:

  - PCIC card handler is allowed to interrupt
  - card removal event is called, removing the driver and data structures
  * card interrupt handler continues w/out driver, data structures, and hardware

OR (the code just committed)

  * card IRQ handler has no hardware to read/write to, but has code and
    data to run on (XXX- Assume it completes and doesn't spin forever)
  - PCIC card handler unloads the card driver

The current situation at least leaves the card interrupt handlers the
drivers and data structures to work with although the hardware can't be
guaranteed.

Reviewed by: bde

27 years agoRemoved magic # 16 and replaced with ICU_LEN.
Nate Williams [Sat, 11 Jan 1997 17:59:38 +0000 (17:59 +0000)]
Removed magic # 16 and replaced with ICU_LEN.

27 years agoNow that all of the IRQ's should be allocated by the time this
Nate Williams [Sat, 11 Jan 1997 17:54:57 +0000 (17:54 +0000)]
Now that all of the IRQ's should be allocated by the time this
pccard_configure() is called (except for LKm's :( ), build a list of
unassigned interrupts for the PCCARD code to use.

Reviewed by: bde

27 years agoMoved pccard_configure() to the end of the configure() list. This
Nate Williams [Sat, 11 Jan 1997 17:41:36 +0000 (17:41 +0000)]
Moved pccard_configure() to the end of the configure() list.  This
avoids problems with the PCIC controller grabbing an interrupt that
another card needs.

Closes PR: kernel/2405

Reviewed by: bde

27 years agoCorrect two bugs:
Bill Paul [Sat, 11 Jan 1997 17:16:18 +0000 (17:16 +0000)]
Correct two bugs:

- If a child receives a SIGTERM, it will call the terminate() function
  and end up doing the shutdown procedurs that should really only be
  done by the parent. Set the SIGTERM behavior back to SIG_DLT in the
  child after fork()ing.

- If the parent fails to read data back from the child because the
  child has exited, it will call rpc_received() with bogus tdata that
  can cause the parent to SEGV. Make handle_children() detect this
  condition correctly and handle it sanely.

*sigh* Another 2.2 candidate.

27 years agoAdded 'xuser' class entry point for X users (who need more resources).
David Nugent [Sat, 11 Jan 1997 16:08:53 +0000 (16:08 +0000)]
Added 'xuser' class entry point for X users (who need more resources).
Increased default procs/file handles so that man will work in more
situations. Other suggestions welcome, btw.

27 years agoAdd -DNOINFO to MK_FLAGS, that's okay right? The comment says
Joshua Peck Macdonald [Sat, 11 Jan 1997 10:53:01 +0000 (10:53 +0000)]
Add -DNOINFO to MK_FLAGS, that's okay right?  The comment says
MK_FLAGS contains -DNOMAN and -DNOPROFILE for speed, so I assume
NOINFO is okay as well.

Add gnu/usr.bin/texinfo to build-tools so everything does the right
thing.

27 years agoIn an attempt to make the new info rules bootstrap themselves, make
Joshua Peck Macdonald [Sat, 11 Jan 1997 10:51:36 +0000 (10:51 +0000)]
In an attempt to make the new info rules bootstrap themselves, make
the rules conditional on NOINFO not being defined, just like NOMAN.

27 years agoThis is purely a bandaid to lower the risk of running out of slots
Peter Wemm [Sat, 11 Jan 1997 09:21:29 +0000 (09:21 +0000)]
This is purely a bandaid to lower the risk of running out of slots
for the local interfaces.  It's not uncommon to have >32 interfaces
these days.  This is a temporary patch until it's fixed to use sysctl
or whatever.