]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
30 years agoUpdated to latest 1.1.5.1+ version
sos [Wed, 17 Aug 1994 08:51:59 +0000 (08:51 +0000)]
Updated to latest 1.1.5.1+ version

Reviewed by:
Submitted by:

30 years agoUpdated to latest version from 1.1.5.1+
sos [Wed, 17 Aug 1994 08:51:10 +0000 (08:51 +0000)]
Updated to latest version from 1.1.5.1+
Reviewed by:
Submitted by:

30 years agoIt wasn't put in during the initial import.
sef [Wed, 17 Aug 1994 05:36:45 +0000 (05:36 +0000)]
It wasn't put in during the initial import.

Reviewed by: Sean Eric Fagan

30 years agoBuild vi (nvi), now that it's checked in.
sef [Tue, 16 Aug 1994 23:58:00 +0000 (23:58 +0000)]
Build vi (nvi), now that it's checked in.
Submitted by: Sean Eric Fagan

30 years agoThis commit was generated by cvs2svn to compensate for changes in r2081,
sef [Tue, 16 Aug 1994 23:36:45 +0000 (23:36 +0000)]
This commit was generated by cvs2svn to compensate for changes in r2081,
which included commits to RCS files with non-trunk default branches.

30 years agoLatest public release of nvi, from Keith Bostic. I hope I got this
sef [Tue, 16 Aug 1994 23:36:45 +0000 (23:36 +0000)]
Latest public release of nvi, from Keith Bostic.  I hope I got this
right ;).
Reviewed by: Sean Eric Fagan

30 years agoReenable the makedb stuff; it works fine.
jkh [Tue, 16 Aug 1994 22:44:28 +0000 (22:44 +0000)]
Reenable the makedb stuff; it works fine.
Submitted by: jkh

30 years agoWhoops, nuke libobjc until we have it too.
jkh [Tue, 16 Aug 1994 02:55:14 +0000 (02:55 +0000)]
Whoops, nuke libobjc until we have it too.
Submitted by: jkh

30 years agoAdd option KBD_RESET_FAIL_OK so that GCC 2.6 doesn't break my
wollman [Mon, 15 Aug 1994 22:41:33 +0000 (22:41 +0000)]
Add option KBD_RESET_FAIL_OK so that GCC 2.6 doesn't break my
console.

30 years agoPlug already known security hole. (Brought over from 1.1.5):
guido [Mon, 15 Aug 1994 20:06:13 +0000 (20:06 +0000)]
Plug already known security hole. (Brought over from 1.1.5):
Fixed security problem with telnetd, which allowed
        telnet -l -hcert.org localhost
to change the user's host in utmp.
Thanks to Matthew Green <mrgreen@@mame.mu.oz.au> for showing me this one.

Reviewed by: karl, guido
Submitted by: mrgreen@@mame.mu.oz.au

30 years agoPlug security hole that was already fixed in 1.1. It prevents
guido [Mon, 15 Aug 1994 19:44:50 +0000 (19:44 +0000)]
Plug security hole that was already fixed in 1.1. It prevents
user from specifying their hostname when rlogin()-ing in
(using rlogin -f-h<host>)

Reviewed by:
Submitted by:

30 years agoDon't bother with XNSrouted and trsp, nobody uses either. If anybody
wollman [Mon, 15 Aug 1994 03:19:25 +0000 (03:19 +0000)]
Don't bother with XNSrouted and trsp, nobody uses either.  If anybody
decides they want to, they can change the Makefile.

30 years agoEnable use of the RTC chip for the statistical clock. While this does
wollman [Mon, 15 Aug 1994 03:15:20 +0000 (03:15 +0000)]
Enable use of the RTC chip for the statistical clock.  While this does
not provide the full accuracy of a randomized statistical clock, it does
provide greater accuracy than the previous method, while not significantly
increasing overhead.  It also provides profiling support at 1024 Hz.

You must re-compile config before making a new kernel, or you will end
up with unresolved symbols.

Reviewed uy: Bruce evans said it worked for him.

30 years agoGenerate vector.h line for statclock. (See next commit for details.)
wollman [Mon, 15 Aug 1994 03:09:29 +0000 (03:09 +0000)]
Generate vector.h line for statclock.  (See next commit for details.)

30 years ago`init' needs crypt() for insecure consoles.
wollman [Mon, 15 Aug 1994 02:45:37 +0000 (02:45 +0000)]
`init' needs crypt() for insecure consoles.

30 years agoSubmitted by: Bruce Evans
ats [Sun, 14 Aug 1994 23:32:45 +0000 (23:32 +0000)]
Submitted by: Bruce Evans
Delete the ifdef GPL_EMULATE case here and made the padding work for
both types of emulators so that there is no longer a need to compile
ps and friends new if you are using the GPL math emulator instead the
normal one.

30 years agoReviewed by:
ats [Sun, 14 Aug 1994 21:06:00 +0000 (21:06 +0000)]
Reviewed by:
Added code to check for an adaptec 1542B Version 3.20 Board. This was the
first board that supports >1Gb drives and has the extended bios. So we
need also to disable the exbios like it is done for the 1542C/CF boards.

30 years agocc -> cc26 for now.
jkh [Sun, 14 Aug 1994 16:53:33 +0000 (16:53 +0000)]
cc -> cc26 for now.
Submitted by: jkh

30 years agoKill an errant libg++ entry.
jkh [Sun, 14 Aug 1994 16:52:44 +0000 (16:52 +0000)]
Kill an errant libg++ entry.
Submitted by: jkh

30 years agoMake keywords.h get properly built.
jkh [Sun, 14 Aug 1994 06:57:34 +0000 (06:57 +0000)]
Make keywords.h get properly built.
Submitted by: jkh

30 years agoPulled over missing ftape.h from 1.1.5
paul [Sun, 14 Aug 1994 01:52:48 +0000 (01:52 +0000)]
Pulled over missing ftape.h from 1.1.5

Reviewed by:
Submitted by: Paul Richards

30 years agoChanged the include <sound/ulaw.h> to be <i386/isa/sound/ulaw.h> so
paul [Sun, 14 Aug 1994 01:52:02 +0000 (01:52 +0000)]
Changed the include <sound/ulaw.h> to be <i386/isa/sound/ulaw.h> so
it compiles.

Reviewed by:
Submitted by: Paul Richards

30 years agoThe tcp/ip interfaces needed longer timeouts, works fine under 2.0.
phk [Sun, 14 Aug 1994 01:46:28 +0000 (01:46 +0000)]
The tcp/ip interfaces needed longer timeouts, works fine under 2.0.

30 years agoMore stuff from the latest curses. Really minor this time.
sef [Sat, 13 Aug 1994 23:23:53 +0000 (23:23 +0000)]
More stuff from the latest curses.  Really minor this time.

Reviewed by: Sean Eric Fagan

30 years agoBrought the 2.0 libcurses up-to-date with the current 4.4 stuff, as
sef [Sat, 13 Aug 1994 23:15:38 +0000 (23:15 +0000)]
Brought the 2.0 libcurses up-to-date with the current 4.4 stuff, as
distributed in keith bostic's nvi (got his permission first).  Most changes
are cosmetic, but a few errors (mostly in tty..c) were cleared up.

Reviewed by: Sean Eric Fagan

30 years agoThis is the first attempt to drag the top level Makefile Rod sent me
jkh [Sat, 13 Aug 1994 22:47:28 +0000 (22:47 +0000)]
This is the first attempt to drag the top level Makefile Rod sent me
into 2.0.  It will surely need more work.
Submitted by: jkh

30 years agoFix conditional-compilation mixup, pointed out by Paul Richards.
wollman [Sat, 13 Aug 1994 17:45:09 +0000 (17:45 +0000)]
Fix conditional-compilation mixup, pointed out by Paul Richards.

30 years agoMade the kernel compile cleanly with gcc 2.6.0. Thanks go to Bruce
dg [Sat, 13 Aug 1994 14:21:58 +0000 (14:21 +0000)]
Made the kernel compile cleanly with gcc 2.6.0. Thanks go to Bruce
Evans for suggesting a method to detect various versions of gcc.

30 years ago Fixed problem with returning -1 on error when the return value is a
dg [Sat, 13 Aug 1994 14:00:26 +0000 (14:00 +0000)]
   Fixed problem with returning -1 on error when the return value is a
long long. Done by plugging both eax and edx with -1. This will clobber
edx unnecessarily when the return value is only 32bit...though probably
always an okay thing to do, it could stand a better fix.
   This was the cause of gawk being broken (boy was THAT ever a subtle
bug!!!).

30 years agoChange .0 man extents to .1, as reported by Julian Elischer. Whoops,
jkh [Sat, 13 Aug 1994 05:00:39 +0000 (05:00 +0000)]
Change .0 man extents to .1, as reported by Julian Elischer.  Whoops,
thought I'd gotten this one the first time around.
Reviewed by:
Submitted by:

30 years agoChange all #includes to follow the current Berkeley style. Some of these
wollman [Sat, 13 Aug 1994 03:50:34 +0000 (03:50 +0000)]
Change all #includes to follow the current Berkeley style.  Some of these
``changes'' are actually not changes at all, but CVS sometimes has trouble
telling the difference.

This also includes support for second-directory compiles.  This is not
quite complete yet, as `config' doesn't yet do the right thing.  You can
still make it work trivially, however, by doing the following:

rm /sys/compile
mkdir /usr/obj/sys/compile
ln -s M-. /sys/compile
cd /sys/i386/conf
config MYKERNEL
cd ../../compile/MYKERNEL
ln -s /sys @
rm machine
ln -s @/i386/include machine
make depend
make

30 years agoSubmitted by: Poul-Henning Kamp <phk@login.dkuug.dk>
phk [Sat, 13 Aug 1994 00:37:03 +0000 (00:37 +0000)]
Submitted by: Poul-Henning Kamp <phk@login.dkuug.dk>

Added my if_lp TCP/IP driver to lpt.c.
I have (surprise) not been able to test it on a 2.0 machine yet.  Connect the
machines with a parallel "lap-link" cable, and get rates from 35 kbyte/sec
up to 75 kbyte/sec.  (when ftp'ing foo: dev/zero -> bar:/dev/null).

The same lpt.c file should compile under 1.1.5.1 without problems.

I think we should promote this feature when we dump 2.0 on the expectant
public, because it provides a cheap and efficient way to move data to and
from notebooks &c.  It is not a replacement for ethernet, but a cheap
substitute sometimes.

30 years agoRemove the scary bits from telnetd - no more encryption left here.
csgr [Fri, 12 Aug 1994 23:00:04 +0000 (23:00 +0000)]
Remove the scary bits from telnetd - no more encryption left here.
Only crypt.c in libc remains.
Reviewed by: Geoff Rehmet

30 years agoThis commit was generated by cvs2svn to compensate for changes in r2050,
csgr [Fri, 12 Aug 1994 22:53:46 +0000 (22:53 +0000)]
This commit was generated by cvs2svn to compensate for changes in r2050,
which included commits to RCS files with non-trunk default branches.

30 years agoMove the telnetd with encryption to src/secure
csgr [Fri, 12 Aug 1994 22:53:46 +0000 (22:53 +0000)]
Move the telnetd with encryption to src/secure
This needs tidying up and having makefiles sorted out later.
Reviewed by: Geoff Rehmet

30 years agoMove the telnetd with encryption to src/secure
csgr [Fri, 12 Aug 1994 22:53:46 +0000 (22:53 +0000)]
Move the telnetd with encryption to src/secure
This needs tidying up and having makefiles sorted out later.
Reviewed by: Geoff Rehmet

30 years agoThe big crypt removal - make libtelnet exportable.
csgr [Fri, 12 Aug 1994 22:41:29 +0000 (22:41 +0000)]
The big crypt removal - make libtelnet exportable.
Securedist can be sorted out later - getting these bits exportable
is top priority.
The libtelnet with encryption has been moved to src/secure/lib.
It will either become part of libsecure, or or be made available
under another name, once the securedist strategy has been completely
worked out.
Submitted by: Geoff Rehmet

30 years agoLibtelnet with encryption goes walkies to src/secure/lib
csgr [Fri, 12 Aug 1994 22:03:01 +0000 (22:03 +0000)]
Libtelnet with encryption goes walkies to src/secure/lib
Reviewed by: Geoff Rehmet

30 years agoLibtelnet with encryption goes walkies to src/secure/lib
csgr [Fri, 12 Aug 1994 22:03:01 +0000 (22:03 +0000)]
Libtelnet with encryption goes walkies to src/secure/lib
Reviewed by: Geoff Rehmet

30 years agoThis commit was generated by cvs2svn to compensate for changes in r2047,
csgr [Fri, 12 Aug 1994 22:03:01 +0000 (22:03 +0000)]
This commit was generated by cvs2svn to compensate for changes in r2047,
which included commits to RCS files with non-trunk default branches.

30 years agowhen making test programs, look for libdescrypt, not libcrypt
csgr [Fri, 12 Aug 1994 21:55:04 +0000 (21:55 +0000)]
when making test programs, look for libdescrypt, not libcrypt
Submitted by: Geoff Rehmet

30 years agoFix afterinstall rule for NOSHARED case
csgr [Fri, 12 Aug 1994 21:12:37 +0000 (21:12 +0000)]
Fix afterinstall rule for NOSHARED case
Submitted by: Geoff Rehmet

30 years ago1) don't make bdes yet
csgr [Fri, 12 Aug 1994 21:02:31 +0000 (21:02 +0000)]
1) don't make bdes yet
2) fix .include in secure/lib/Makefile.inc
3) fix afterinstall rule in libcrypt/Makefile
Submitted by: Geoff Rehmet

30 years agoAdded conditionals to make this compile cleanly in FreeBSD 2.0.
dg [Fri, 12 Aug 1994 11:42:37 +0000 (11:42 +0000)]
Added conditionals to make this compile cleanly in FreeBSD 2.0.

30 years agoTook out junk to save the old value of BLOCKSIZE as it is unnecessary.
dg [Fri, 12 Aug 1994 10:26:27 +0000 (10:26 +0000)]
Took out junk to save the old value of BLOCKSIZE as it is unnecessary.
Remove CFLAGS=-g from Makefile.

30 years agoThe last commit was bogus...the putenv doesn't affect the parent process,
dg [Fri, 12 Aug 1994 10:23:49 +0000 (10:23 +0000)]
The last commit was bogus...the putenv doesn't affect the parent process,
so the BLOCKSIZE doesn't need to be preserved.
Also initialized the flags variables, and used 1k instead of 1024 for
BLOCKSIZE.

30 years agoThe last commit was bogus...the putenv doesn't affect the parent process,
dg [Fri, 12 Aug 1994 10:19:05 +0000 (10:19 +0000)]
The last commit was bogus...the putenv doesn't affect the parent process,
so the BLOCKSIZE doesn't need to be preserved.

30 years agoImplemented the -k flag more properly...don't destroy the user's setting
dg [Fri, 12 Aug 1994 07:26:52 +0000 (07:26 +0000)]
Implemented the -k flag more properly...don't destroy the user's setting
of the BLOCKSIZE environment variable.

30 years agoNew ethernet device driver from Matt Thomas:
dg [Fri, 12 Aug 1994 06:51:12 +0000 (06:51 +0000)]
New ethernet device driver from Matt Thomas:

This driver supports all the DEC EtherWORKS III NICs (DE203, DE204,
and DE205) and the later DEC EtherWORKS II NICs (DE200, DE201, DE202,
DE422).  DEPCA-style boards prior to the DE200 have not been tested
and may not work.

Submitted by: Matt Thomas (thomas@lkg.dec.com)

30 years agoRemoved some unnecessary code that sets the link layer address. This should
dg [Fri, 12 Aug 1994 06:36:51 +0000 (06:36 +0000)]
Removed some unnecessary code that sets the link layer address. This should
be removed from the other drivers, too, as it is already done at a higher level
in the kernel.

30 years agoAdd missing ifr variable declaration that I forgot when adding MTU
dg [Fri, 12 Aug 1994 06:06:19 +0000 (06:06 +0000)]
Add missing ifr variable declaration that I forgot when adding MTU
ioctl support.

30 years agoImplement -k flag.
dg [Fri, 12 Aug 1994 02:02:22 +0000 (02:02 +0000)]
Implement -k flag.

30 years agoswapinfo(1) command from 1.1.5. This version has been modified to work
dg [Fri, 12 Aug 1994 01:49:30 +0000 (01:49 +0000)]
swapinfo(1) command from 1.1.5. This version has been modified to work
with the new libkvm interfaces.

30 years agoMake it set-uid until procfs is fixed.
wollman [Thu, 11 Aug 1994 20:11:00 +0000 (20:11 +0000)]
Make it set-uid until procfs is fixed.

30 years agoMake `ps' set-uid root so that it can read the same set of arguments
wollman [Thu, 11 Aug 1994 20:06:55 +0000 (20:06 +0000)]
Make `ps' set-uid root so that it can read the same set of arguments
that old `ps'es did.  I'm not too thrilled about this, but I'm not
enough of an FS person to hack procfs so that /proc/xxx/mem is readable
by members of group `kmem'.  If this is done, then `ps' can go back to
being set-gid kmem.

30 years agoAdd rotattion of cron log to weekly.
guido [Thu, 11 Aug 1994 17:23:29 +0000 (17:23 +0000)]
Add rotattion of cron log to weekly.

Reviewed by:
Submitted by: guido

30 years agoUse nulls to pad proctitle rather than spaces...makes ps(1) output more
dg [Thu, 11 Aug 1994 13:40:58 +0000 (13:40 +0000)]
Use nulls to pad proctitle rather than spaces...makes ps(1) output more
readable.

30 years agoMade kvm routines use procfs to get out process data such as argument
dg [Thu, 11 Aug 1994 13:38:23 +0000 (13:38 +0000)]
Made kvm routines use procfs to get out process data such as argument
strings.

30 years agoChange outb() as per Bruce's instructions so that it doesn't explicitly
jkh [Thu, 11 Aug 1994 02:26:08 +0000 (02:26 +0000)]
Change outb() as per Bruce's instructions so that it doesn't explicitly
try to pass its argument in the ax register.
Reviewed by:
Submitted by:

30 years agoCorrect and add proper man page extents.
jkh [Thu, 11 Aug 1994 01:46:41 +0000 (01:46 +0000)]
Correct and add proper man page extents.
Submitted by: jkh

30 years agoFix man extents from .0 to .1
jkh [Thu, 11 Aug 1994 01:42:28 +0000 (01:42 +0000)]
Fix man extents from .0 to .1
Submitted by: jkh

30 years agoBruce Evans is right - this shouldn't touch /etc *at all* and I should
jkh [Thu, 11 Aug 1994 01:39:23 +0000 (01:39 +0000)]
Bruce Evans is right - this shouldn't touch /etc *at all* and I should
have nuked it, not fixed it.  No longer install /etc/localtime.
Reviewed by:
Submitted by:

30 years agoUse the right man page.
jkh [Thu, 11 Aug 1994 01:38:23 +0000 (01:38 +0000)]
Use the right man page.
Submitted by: jkh

30 years agoAdd renice's correct man page to the Makefile.
jkh [Thu, 11 Aug 1994 01:33:22 +0000 (01:33 +0000)]
Add renice's correct man page to the Makefile.
Submitted by: jkh

30 years agoSomewhere in the shuffle this lost its man page. Use the 1.1.5
jkh [Thu, 11 Aug 1994 01:06:29 +0000 (01:06 +0000)]
Somewhere in the shuffle this lost its man page.  Use the 1.1.5
one.
Submitted by: jkh

30 years agoAdd a preventative rm of /etc/localtime in the install rule just in
jkh [Thu, 11 Aug 1994 00:54:30 +0000 (00:54 +0000)]
Add a preventative rm of /etc/localtime in the install rule just in
case it's a link (in which case the subsequent install will fall over).
Submitted by: jkh

30 years agoChange a .0 to a .5.
jkh [Thu, 11 Aug 1994 00:45:38 +0000 (00:45 +0000)]
Change a .0 to a .5.
Submitted by: jkh

30 years agoPut the cons25 entries for syscons back.
jkh [Thu, 11 Aug 1994 00:39:39 +0000 (00:39 +0000)]
Put the cons25 entries for syscons back.
Submitted by: jkh

30 years agoFor Pentium machines, use a faster version of microtime with 8 usec
wollman [Thu, 11 Aug 1994 00:28:24 +0000 (00:28 +0000)]
For Pentium machines, use a faster version of microtime with 8 usec
resolution (can probably be improved somewhat).  Other machines take
a three-instruction hit if I586_CPU is defined, none otherwise.

30 years agotermcap.0 -> termcap.5
jkh [Thu, 11 Aug 1994 00:24:33 +0000 (00:24 +0000)]
termcap.0 -> termcap.5
Submitted by: jkh

30 years agoCorrect me.0 -> me.7
jkh [Thu, 11 Aug 1994 00:11:44 +0000 (00:11 +0000)]
Correct me.0 -> me.7
Submitted by: jkh

30 years agoTell Pentium users their CPU speed. (More changes to make use of this
wollman [Wed, 10 Aug 1994 23:28:33 +0000 (23:28 +0000)]
Tell Pentium users their CPU speed.  (More changes to make use of this
to come later.)

30 years agoInstall procfs(5).
wollman [Wed, 10 Aug 1994 20:04:57 +0000 (20:04 +0000)]
Install procfs(5).

30 years agoOops, fix typo.
wollman [Wed, 10 Aug 1994 20:03:54 +0000 (20:03 +0000)]
Oops, fix typo.

30 years agoInitial manual page for the new procfs.
wollman [Wed, 10 Aug 1994 20:03:06 +0000 (20:03 +0000)]
Initial manual page for the new procfs.

30 years agoInitialize lockf pointer. I missed this when I made NFS use the generic
dg [Wed, 10 Aug 1994 19:48:23 +0000 (19:48 +0000)]
Initialize lockf pointer. I missed this when I made NFS use the generic
advlock mechanism, and not doing so results in random system crashes.

30 years agoGimme my `-k' option back!
wollman [Wed, 10 Aug 1994 19:25:34 +0000 (19:25 +0000)]
Gimme my `-k' option back!

30 years agoGimme back my `-k' option!
wollman [Wed, 10 Aug 1994 19:21:07 +0000 (19:21 +0000)]
Gimme back my `-k' option!
Also document `-t' option in the usage message.

30 years agoRemove kerberosIV until it works or is replace with eBones.. It's
jkh [Wed, 10 Aug 1994 19:19:35 +0000 (19:19 +0000)]
Remove kerberosIV until it works or is replace with eBones..  It's
hosing my build worlds..
Submitted by: jkh

30 years agoAdded example of how to tell various FreeBSD versions apart.
wollman [Wed, 10 Aug 1994 06:40:36 +0000 (06:40 +0000)]
Added example of how to tell various FreeBSD versions apart.

30 years agoOops, forgot to cvs add this file.
wollman [Wed, 10 Aug 1994 06:27:35 +0000 (06:27 +0000)]
Oops, forgot to cvs add this file.

30 years agoMake it easier for programs to figure out what revision of FreeBSD they
wollman [Wed, 10 Aug 1994 06:25:07 +0000 (06:25 +0000)]
Make it easier for programs to figure out what revision of FreeBSD they
are running under.  Here's how to bootstrap (order is important):

1) Re-compile gcc (just the driver is all you need).
2) Re-compile libc.
3) Re-compile your kernel.  Reboot.
4) cd /usr/src/include; make install

You can now detect the compilation environment with the following code:

#if !defined(__FreeBSD__)
#define __FreeBSD_version 199401
#elif __FreeBSD__ == 1
#define __FreeBSD_version 199405
#else
#include <osreldate.h>
#endif

You can determine the run-time environment by calling the new C library
function getosreldate(), or by examining the MIB variable kern.osreldate.

For the time being, the release date is defined as 199409, which we have
already established as our target.

30 years agoDefine __FreeBSD__ to be 2.
wollman [Wed, 10 Aug 1994 06:12:51 +0000 (06:12 +0000)]
Define __FreeBSD__ to be 2.

30 years agoChange bogus --I to -I to make compilation happen again.
jkh [Wed, 10 Aug 1994 04:55:19 +0000 (04:55 +0000)]
Change bogus --I to -I to make compilation happen again.
Submitted by: jkh

30 years agoHandle NMI's in accordance with data in van Gilluwe book.
wollman [Wed, 10 Aug 1994 04:39:52 +0000 (04:39 +0000)]
Handle NMI's in accordance with data in van Gilluwe book.

30 years agoEliminate non-fatal error message so Jordan doesn't get confused.
wollman [Wed, 10 Aug 1994 04:25:19 +0000 (04:25 +0000)]
Eliminate non-fatal error message so Jordan doesn't get confused.

30 years agoSome programs (like GNU configure programs) depend on the output of
wollman [Wed, 10 Aug 1994 03:53:33 +0000 (03:53 +0000)]
Some programs (like GNU configure programs) depend on the output of
`uname -s' to be something reasonable (traditionally, `i386') rather
than `PC-Class'.  Make it so.

30 years agoAdd back in CPU detection copde from 1.1.5. As an added bonus, the
wollman [Wed, 10 Aug 1994 03:51:18 +0000 (03:51 +0000)]
Add back in CPU detection copde from 1.1.5.  As an added bonus, the
hw.model MIB variable is now declared correctly.

30 years agoFixed vm_page_deactivate to deal with getting called with a page that's
dg [Wed, 10 Aug 1994 03:09:37 +0000 (03:09 +0000)]
Fixed vm_page_deactivate to deal with getting called with a page that's
not on any queue. This is an old patchkit days fix.

Reviewed by: John Dyson and David Greenman
Submitted by: originally by Paul Mackerras

30 years agoMake the install targets obey the INSTALLFLAGS requests in the new
wollman [Wed, 10 Aug 1994 02:48:08 +0000 (02:48 +0000)]
Make the install targets obey the INSTALLFLAGS requests in the new
Makefiles.

DANGER WILL ROBINSON!

This will cause repeat installs of certain programs, such as `init' and
`rcp', to fail unless one of the two conditions is met:

1) You are in single-user mode.
2) Your security level is set to 0 or -1.

If you have compiled a kernel from the latest sources, your kernel
security level is set to -1 by default, which will keep `init' from
fiddling with it.  You can increase it, but not decrease it, from the
command line with the command `sysctl -w kern.securelevel=<new value>'.
I believe that -1 is the most appropriate value to use while we are still
developing the code, although when we ship it should be changed back to 0.

See init(8) for more information.

30 years agoChange default security level to -1, so that users don't get bitten by
wollman [Wed, 10 Aug 1994 02:41:09 +0000 (02:41 +0000)]
Change default security level to -1, so that users don't get bitten by
upcoming makefile change.

30 years agoAdd back set_rpc_grouplistsize(), so mount_nfs compiles again. Also
wollman [Wed, 10 Aug 1994 02:25:22 +0000 (02:25 +0000)]
Add back set_rpc_grouplistsize(), so mount_nfs compiles again.  Also
fixed incipient bug wrt gid_t versus int.

30 years agoNuke GNU pr.
jkh [Wed, 10 Aug 1994 00:24:25 +0000 (00:24 +0000)]
Nuke GNU pr.
Reviewed by:
Submitted by:

30 years agoFixed typo.
wollman [Tue, 9 Aug 1994 22:44:12 +0000 (22:44 +0000)]
Fixed typo.

30 years agoAdd (substantially re-written) support for /etc/host.conf, and reintegrated
wollman [Tue, 9 Aug 1994 20:23:14 +0000 (20:23 +0000)]
Add (substantially re-written) support for /etc/host.conf, and reintegrated
1.1.5 support for YP, fixing a bug in 1.1.5 that prevented YP from ever
working reliably.  (I'm amazed that there were no bug reports.)

IWBRNI someone could write a host.conf(5) manual page.  Please look at
the code before doing so; this version is somewhat more flexible in the
format of its input.

30 years agoInstall secure/lib/libcrypt as libdescrypt, and symlink it to
csgr [Tue, 9 Aug 1994 18:52:52 +0000 (18:52 +0000)]
Install secure/lib/libcrypt as libdescrypt, and symlink it to
libcrypt.  There may be a little modification neede to this makefile once
we start working on tidy make world's.
Submitted by: geoff.

30 years agoThe password scrambler now becomes libscrypt, and libcrypt is
csgr [Tue, 9 Aug 1994 18:49:04 +0000 (18:49 +0000)]
The password scrambler now becomes libscrypt, and libcrypt is
a symlink to it. (The real libcrypt will be installed as libdescrypt.)
Submitted by: Geoff.

30 years agoMake man page installation work for ``unattached'' and kernel pages.
wollman [Tue, 9 Aug 1994 18:41:38 +0000 (18:41 +0000)]
Make man page installation work for ``unattached'' and kernel pages.

30 years agoMake examples install into /usr/share/examples. Like with includes, you
wollman [Tue, 9 Aug 1994 18:25:38 +0000 (18:25 +0000)]
Make examples install into /usr/share/examples.  Like with includes, you
have a choice between SHARED=copies and SHARED=symlinks.  The default
is to copy.

I have also added a /usr/share/examples/etc directory, where I hope to
have all sample configuration files which in real use go into /etc installed.
(This way, if the user really screws the real one up, they can always go
back to a known-working distribution copy, even if they don't have sources.)

30 years agoThis commit was generated by cvs2svn to compensate for changes in r1984,
csgr [Tue, 9 Aug 1994 17:07:27 +0000 (17:07 +0000)]
This commit was generated by cvs2svn to compensate for changes in r1984,
which included commits to RCS files with non-trunk default branches.