Nate Willams [Mon, 14 Feb 1994 21:13:59 +0000 (21:13 +0000)]
From: piero@strider.st.dsi.unimi.it (Piero Serini)
Subject: man reboot patch
Date: Mon, 14 Feb 1994 19:47:50 +0100 (MET)
reboot_i386.8 references [the U word].
Removed.
Nate Willams [Mon, 14 Feb 1994 19:22:34 +0000 (19:22 +0000)]
From: Jim Babb <babb@sedhps01.mdc.com>
Subject: Bug Fix for ft.c - please commit
Date: Mon, 14 Feb 94 10:22:54 CST
This fixes a bug where the system can crash if the tape is used
after the floppy has been used.
Rod Grimes [Mon, 14 Feb 1994 13:09:14 +0000 (13:09 +0000)]
Update Rob Shady's email address.
Rod Grimes [Mon, 14 Feb 1994 10:02:13 +0000 (10:02 +0000)]
I yelled, I warned and now I am SCREAMING. Add -D__FreeBSD__ to CFLAGS+=
so that you can bootstrap a 1.0.2 system. Please oh please no one but
no one add any more #ifdef __FreeBSD__ stuff UNTIL 1.1 has SHIPPED!!!
Joerg Wunsch [Sun, 13 Feb 1994 21:41:19 +0000 (21:41 +0000)]
Added my mail address
Jordan K. Hubbard [Sun, 13 Feb 1994 20:53:11 +0000 (20:53 +0000)]
New dlopen, dlclose, etc al. For the new ld changes.
Jordan K. Hubbard [Sun, 13 Feb 1994 20:50:21 +0000 (20:50 +0000)]
These are my rewritten ENTRY macros for adding in .type information
for the new ld (and with a little bit of general cleanup).
Jordan K. Hubbard [Sun, 13 Feb 1994 20:47:18 +0000 (20:47 +0000)]
This is the new link.h from Paul K. Keeping history is even probably a bad
idea considering the old version (but I haven't time for attic surgery - this
is just a caution). This is part of (and the major reason for) the new ld
changes.
Jordan K. Hubbard [Sun, 13 Feb 1994 20:44:09 +0000 (20:44 +0000)]
A fast private malloc for ld.so
Jordan K. Hubbard [Sun, 13 Feb 1994 20:43:13 +0000 (20:43 +0000)]
This is Paul K's latest set of ld changes. A commit was necessary at this
late stage due to the fact that link.h was copyright Sun Microsystems.
This version of ld sync's us up with NetBSD's ld and supports compatablily
with NetBSD's -[zZ] flags (which we had reversed). Compiling with this
new ld will give you RRS warnings for libraries which do not contain .type
infomation - these wsarnings are harmless and will go away as soon as you
recompile your libraries (cd /usr/src; make libraries).
Jordan K. Hubbard [Sun, 13 Feb 1994 20:37:32 +0000 (20:37 +0000)]
Use -Z instead of -z to build kernels. This is consistent with NetBSD's
usage of the -[Zz] flags (we're now in sync). Part of the ld update
changes.
Andrey Chernov [Sun, 13 Feb 1994 20:29:50 +0000 (20:29 +0000)]
Make window 8-bit clean
Jordan K. Hubbard [Sun, 13 Feb 1994 18:15:00 +0000 (18:15 +0000)]
Add Julian Stacey's ports changes (don't affect anything else)
Andrey Chernov [Sun, 13 Feb 1994 18:00:14 +0000 (18:00 +0000)]
1) Fix bug: VTIME/VMIN not setted at all!
2) Remove _R from Makefile
3) Start 8-bit cleanup (not completed)
Andrey Chernov [Sun, 13 Feb 1994 17:21:31 +0000 (17:21 +0000)]
3 occurances of "ttyout" chaged to
"ttywai", "ttyfls", "ttyout" for better debugging and diagnostics.
Guido van Rooij [Sun, 13 Feb 1994 15:03:03 +0000 (15:03 +0000)]
Fixed the reported place of the '*' when write perm is denied in the
manpage.
Christoph Robitschko [Sun, 13 Feb 1994 14:48:21 +0000 (14:48 +0000)]
Use KERNBASE in address calculation.
Christoph Robitschko [Sun, 13 Feb 1994 14:31:17 +0000 (14:31 +0000)]
print error message for nonexistent screensaver.
David Greenman [Sun, 13 Feb 1994 13:35:03 +0000 (13:35 +0000)]
Corrected my email address
Andrey Chernov [Sun, 13 Feb 1994 09:52:48 +0000 (09:52 +0000)]
Fixed bugs list:
1) Several places with comparation of int and unsigned (fails,
if unsigned equal zero);
2) All windows not started from top of the screen refreshes
incorrectly (I mean here not visual looking, but
minimize-output-chars refresh technique).
UCB often forget to add win->begy to index;
3) Uncommented #ifdef DEBUG in refresh dumps core, address
beyond array ranges;
4) clsp in refresh.c (number of last space on curscr) calculates
incorrectly, so CE don't work properly;
5) Litlle prototype cleanup.
David Greenman [Sun, 13 Feb 1994 08:29:33 +0000 (08:29 +0000)]
Fixed bug in handling of COW - the original code was bogus and it was
only accidental that it worked. Also, don't cache non-managed pages.
Jordan K. Hubbard [Sun, 13 Feb 1994 03:05:00 +0000 (03:05 +0000)]
Add some personal acks.
Jordan K. Hubbard [Sun, 13 Feb 1994 02:59:49 +0000 (02:59 +0000)]
Put in proper address information for Poul-Henning Kamp.
Andrey Chernov [Sun, 13 Feb 1994 00:16:23 +0000 (00:16 +0000)]
Man page update from wilko@idca.tds.philips.nl
Andrey Chernov [Sun, 13 Feb 1994 00:04:42 +0000 (00:04 +0000)]
Handle cua (dialout) at the same manner as tty, i.e. print
port number instead of "cu"
Jordan K. Hubbard [Sat, 12 Feb 1994 23:51:38 +0000 (23:51 +0000)]
From: Piero Serini <piero@strider.st.dsi.unimi.it>
Message-Id: <
199402111717.SAA05326@strider.st.dsi.unimi.it>
Subject: Re: cpio bug ?
Quoting from J Wunsch:
> | From the man page:
> |
> | -l, --link
> | Link files instead of copying them, when possible.
>
> (Usable only with the -p option.)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
This is not in cpio man page. Someone please add it.
Garrett Wollman [Sat, 12 Feb 1994 21:39:59 +0000 (21:39 +0000)]
Correctly describe only the process flags that we support.
Andreas Schulz [Sat, 12 Feb 1994 21:00:27 +0000 (21:00 +0000)]
Deleted the private clean target in the Makefile, put a CLEANFILES
target into the file instead.
Andreas Schulz [Sat, 12 Feb 1994 20:03:27 +0000 (20:03 +0000)]
Changed a wrong description for the location of the time command.
Failure found by "Julian Howard Stacey".
Andreas Schulz [Sat, 12 Feb 1994 19:53:34 +0000 (19:53 +0000)]
Added the version file to the CLEANFILE target in the Makefile.
Removed the version file in the source-tree.
Andrey Chernov [Fri, 11 Feb 1994 21:41:44 +0000 (21:41 +0000)]
Some fixes from Joerg
Guido van Rooij [Fri, 11 Feb 1994 21:14:52 +0000 (21:14 +0000)]
ps works inside ddb now.
Guido van Rooij [Fri, 11 Feb 1994 21:14:28 +0000 (21:14 +0000)]
Now 'ps' works inside ddb, just like the man page says. Included is a
very simple more version to actuallyt view what ps is telling you.
Nate Willams [Fri, 11 Feb 1994 12:21:22 +0000 (12:21 +0000)]
Documented 'hanging keyboard' problems with pccons and temporary
solution until syscons becomes the default console driver.
Nate Willams [Fri, 11 Feb 1994 12:12:51 +0000 (12:12 +0000)]
Changed NetBSD -> FreeBSD
Nate Willams [Fri, 11 Feb 1994 12:08:14 +0000 (12:08 +0000)]
Piddly little changes to update who did what.
Nate Willams [Fri, 11 Feb 1994 12:02:35 +0000 (12:02 +0000)]
Added third parameter to wdwait which contains the amount of time to
'wait' based on suggestion by Rod in -hackers.
No functional change as all the TIMEOUTS are identical currently, but
this will probably change.
Andrey Chernov [Fri, 11 Feb 1994 11:00:13 +0000 (11:00 +0000)]
rm# for 5.25 driver changed per Bruce suggestion
Andrey Chernov [Fri, 11 Feb 1994 10:55:12 +0000 (10:55 +0000)]
com.4 restored back
Nate Willams [Fri, 11 Feb 1994 09:58:06 +0000 (09:58 +0000)]
Removed cut-here line from bottom on man-page.
Martin Renters [Thu, 10 Feb 1994 22:16:40 +0000 (22:16 +0000)]
Change to make diskless_cfg compile on HPUX
Martin Renters [Thu, 10 Feb 1994 22:09:24 +0000 (22:09 +0000)]
Documentation on how to set up diskless client on FreeBSD
Andrey Chernov [Thu, 10 Feb 1994 15:27:59 +0000 (15:27 +0000)]
Enable ISO8859-* and koi8-r charsets, inspired by Poul-Henning Kamp patch
Andrey Chernov [Thu, 10 Feb 1994 11:40:40 +0000 (11:40 +0000)]
Add many missed capabilities to cons25
Add missed SF/SR to pc3
David Greenman [Thu, 10 Feb 1994 11:24:32 +0000 (11:24 +0000)]
When hexdumping a RIP entry, type cast the bytes to unsigned first. Not
doing so caused them to be sign-extended to an int (which really messed
things up).
David Greenman [Thu, 10 Feb 1994 11:04:48 +0000 (11:04 +0000)]
Patch from Michael Galassi (nerd@percy.rain.com):
The problem as I saw it was that a route was being added to the routing
tables in the kernel even in the case where the interface pointer in
the route structure was a null, this was fine until an atempt was made
to remove this route at which point a zero derefernce was "commited" and
the inevitable core was dumped. The easy solution turns out to be to
check for a null interface pointer and not add the route. The only time
I've observed the problem is when adding a route to the local end of a
point to point connection (SLIP or PPP).
Andrey Chernov [Thu, 10 Feb 1994 10:44:08 +0000 (10:44 +0000)]
Add floppy entries for new fd driver
(from Poul-Henning Kamp) slightly fixed
Andrey Chernov [Thu, 10 Feb 1994 10:17:58 +0000 (10:17 +0000)]
Add insert/delete line capabilities to match
termcap pc3 entry and syscons & codrv
David Greenman [Thu, 10 Feb 1994 09:17:57 +0000 (09:17 +0000)]
Fixed bug in decoding of RIP packets. RIP data is in "osockaddr" format;
the fix is to import the structures from protocols/routed.h and change
the sockaddr to osockaddr. We can't change the generic include file
because too many other things rely on this in order to deal with the
kernel routing socket (which is sockaddr format).
David Greenman [Thu, 10 Feb 1994 08:08:37 +0000 (08:08 +0000)]
Patches from John Dyson to fix hangs under certain paging loads.
David Greenman [Thu, 10 Feb 1994 08:04:07 +0000 (08:04 +0000)]
Fixed previous bungled patch - variable declaration was missing.
David Greenman [Thu, 10 Feb 1994 03:03:44 +0000 (03:03 +0000)]
Patch from John Dyson:
a pv chain was being traversed while interrupts were
fully enabled in pmap_remove_all ... this is bogus, and
has been fixed in pmap.c. (sorry for adding the splimp)
Jordan K. Hubbard [Thu, 10 Feb 1994 02:02:38 +0000 (02:02 +0000)]
__386BSD__ -> __FreeBSD__
Still doesn't print disk I/O stats correctly.
Jordan K. Hubbard [Thu, 10 Feb 1994 01:57:49 +0000 (01:57 +0000)]
__386BSD__ -> __FreeBSD__
Still looking for more problems..
Andrey Chernov [Thu, 10 Feb 1994 01:20:28 +0000 (01:20 +0000)]
Fix size of il (interleave) array.
Andrey Chernov [Thu, 10 Feb 1994 01:15:01 +0000 (01:15 +0000)]
Add interleave > 1 support by Poul-Henning Kamp.
Chack valid range of sectors per track.
David Greenman [Thu, 10 Feb 1994 00:20:26 +0000 (00:20 +0000)]
Patch from John Dyson:
On slow, memory poor machines with lot's of networking activity it was
possible to deplete all of main memory with mbufs and mbuf clusters. The
fix for this is to limit the amount of memory that can be used to 1/16
of total memory.
David Greenman [Thu, 10 Feb 1994 00:15:51 +0000 (00:15 +0000)]
Patch from John Dyson:
Under high load conditions with large numbers of fork/exec's occurring
while paging heavily, processes would core dump. Removing a very minor
optimization seems to fix the problem.
Andrey Chernov [Wed, 9 Feb 1994 23:49:45 +0000 (23:49 +0000)]
Remove nonexistent com driver manpage.
Describe some additional features of sio.
Andrey Chernov [Wed, 9 Feb 1994 19:42:36 +0000 (19:42 +0000)]
1) UCB code to scroll portion of screen with al/dl works only for fullscreen
and only if AL/DL (parameters) defined too, fix this ugly bug.
2) Add ability to scroll portion of screen via change scroll region
(cs), save/restore cursor (sc,rc) and scroll forward/backward.
It really helps don't redraw whole screen on poor vt100 without
al/dl.
Andrey Chernov [Wed, 9 Feb 1994 16:55:52 +0000 (16:55 +0000)]
Since new curses have ability to scroll screen parts,
old xscroll hack not needed even on slow displays.
Don't forget update your termcaps with recently commited,
al/dl capability (insert/delete lines) added to bring scrolling
to work.
Andrey Chernov [Wed, 9 Feb 1994 16:23:21 +0000 (16:23 +0000)]
Here simple script demonstrating, why it is bad choice to strip
libraries on install, many autoconfig pgms can't search names!
#!/bin/sh
cp /usr/lib/libc.so.1.0 /tmp/libc.so.1.0
cd /tmp
nm -pg libc.so.1.0 | grep index
strip libc.so.1.0
nm -pg libc.so.1.0 | grep index
Nate Willams [Wed, 9 Feb 1994 09:25:31 +0000 (09:25 +0000)]
My Kudos!
David Greenman [Wed, 9 Feb 1994 07:03:10 +0000 (07:03 +0000)]
Fixes from John Dyson:
vm_page.c
vm_page_deactivate had a window that interrupts were
not disabled, but should have been.
vm_object.c
in 3 places, a page was being protected VM_PROT_NONE, which
if the page is pager mapped, can cause a system failure
because a pager mapping could disappear. Now, we block when
a page is busy.
swap_pager.c
a deadlock could occur because a process could be waiting on
a swap clean entry, and that case was not handled correctly.
now, user processes dont call swap_pager_clean and the
normal user process wakes-up the pagedaemon when no swap clean
entries are available.
Garrett Wollman [Wed, 9 Feb 1994 06:22:58 +0000 (06:22 +0000)]
Correct some errors in what Nate wrote.
Nate Willams [Wed, 9 Feb 1994 05:35:57 +0000 (05:35 +0000)]
Commented on what the mse0, mcd0, and lpt/lpa devices were used for.
(No functional changes)
Rod Grimes [Wed, 9 Feb 1994 05:26:16 +0000 (05:26 +0000)]
Upgrade to 1.1 release level, add procfs entry on fstab creation.
Nate Willams [Wed, 9 Feb 1994 05:25:23 +0000 (05:25 +0000)]
Added problems with if_ep driver hanging under heavy loads and a warning
about syscons not compiling with the 1.0.2 compiler and a workaround for
it.
Rod Grimes [Wed, 9 Feb 1994 03:40:59 +0000 (03:40 +0000)]
Add obj/ in front of freefall.cf so that the install works.
Rod Grimes [Wed, 9 Feb 1994 02:38:15 +0000 (02:38 +0000)]
Remove ${DESTDIR} from destination of symbolic links, it creates
really screwy links when you do this and are building releases using
DESTDIR:
skyrsh# ls -lag map3270 termcap
lrwxrwxr-x 1 root wheel 41 Feb 6 21:20 map3270 -> /a/one-off/filesys/usr/share/misc/map3270
lrwxrwxr-x 1 root wheel 41 Feb 6 21:20 termcap -> /a/one-off/filesys/usr/share/misc/termcap
skyrsh#
Rod Grimes [Wed, 9 Feb 1994 02:29:31 +0000 (02:29 +0000)]
Remove /usr/X386/lib and /usr/local/lib from the default locations,
these are handled by /etc/rc and this was causing errors on a new
install as these places do NOT exsist by default!
Rod Grimes [Wed, 9 Feb 1994 01:57:37 +0000 (01:57 +0000)]
A real good idea...
>From: "Chris G. Demetriou" <cgd@sun-lamp.cs.berkeley.edu>
Update of /b/source/CVS/src/etc
In directory sun-lamp.cs.berkeley.edu:/usr/src/etc
Modified Files:
master.passwd
Log Message:
disable toor by default
Rod Grimes [Wed, 9 Feb 1994 01:50:39 +0000 (01:50 +0000)]
New file to list known problems with the current set of sources.
Garrett Wollman [Wed, 9 Feb 1994 01:25:40 +0000 (01:25 +0000)]
Forgot to include backward-compatibility data file (puts links in
for old names like US/Eastern -> America/New_York).
Rod Grimes [Wed, 9 Feb 1994 00:58:12 +0000 (00:58 +0000)]
Update for new COPYRIGHT.* files.
Rod Grimes [Wed, 9 Feb 1994 00:53:24 +0000 (00:53 +0000)]
Rename the 386BSD copyright file to COPYRIGHT.386BSD and put in a new
COPYRIGHT.FreeBSD created by Garrett Wollman.
This is a more complete copyright notice.
Andrey Chernov [Tue, 8 Feb 1994 21:57:34 +0000 (21:57 +0000)]
Make fmt 8-bit clean
David Greenman [Tue, 8 Feb 1994 12:58:44 +0000 (12:58 +0000)]
From: Dave Matthews <dave@prlng.co.uk>
Description:
The integer overflow instruction (into) and the interrupt instruction with
value 4 (int #4) both give rise to SIGBUS signals rather than SIGFPE. The
problem is that overflow is a trap not a fault (unlike the BOUND instruction).
Rod Grimes [Tue, 8 Feb 1994 09:37:28 +0000 (09:37 +0000)]
One more reference to fdXa removed.
Run fdisk -u after disklabel just incase we wrote block 0, this should fix
the bogus partition table when the install is to the whole disk.
David Greenman [Tue, 8 Feb 1994 09:26:04 +0000 (09:26 +0000)]
Fixed bugs in stack grow code, and moved it back into a seperate function
like it was originally. Also added back call to "grow" in sendsig now
that this routine actually works.
Rod Grimes [Tue, 8 Feb 1994 08:58:41 +0000 (08:58 +0000)]
Add PERSONAL.ACKS to file list.
Use freefall.cf as sendmail prototype file, it is more realistic than the
tcpproto.cf file for a FreBSD system. Fix so that obj dir is created in
sendmail/cf/cf as to not polute the source tree and to have the Makefile
in there do the right things.
Remove all the extra /dev/fd0?* entries on the floppies, they where using
up all the inodes and are not needed at this time.
Temporarily remove the floppy target from release: untilit is
fixed.
This file has lots more work coming, but to get the 1.1 BETA out I am
going to hand craft the floppies :-(.
Rod Grimes [Tue, 8 Feb 1994 08:45:10 +0000 (08:45 +0000)]
Update to use /dev/fd0 instead of /dev/fd0a.
Rod Grimes [Tue, 8 Feb 1994 08:34:57 +0000 (08:34 +0000)]
>From curt@watson.ibm.com Mon Feb 7 13:28:20 1994
Date: Mon, 7 Feb 94 16:27:55 EST
From: "Curt McDowell (914-784-6016 T/L 863)" <curt@watson.ibm.com>
Subject: Re: Bugs in "cal"
Jeffrey,
I obtained v5.2 of cal.c and found that all but one of the bugs
have been fixed. The remaining bug causes junk to be output at the end
of each line when printing out the calendar for a single month:
Andrey Chernov [Tue, 8 Feb 1994 08:29:33 +0000 (08:29 +0000)]
Add al,dl,AL,DL (insert/delete lines (multiple)) to pc3 entry
Andrey Chernov [Tue, 8 Feb 1994 06:46:47 +0000 (06:46 +0000)]
Some cosmetique cleanup VMIN/VTIME code
Garrett Wollman [Tue, 8 Feb 1994 03:34:15 +0000 (03:34 +0000)]
Calculate percentages using floating-point math, to avoid integer overflow
giving bogus results when top >= LONG_MAX/100. Use unsigned long for
parameters.
David Greenman [Tue, 8 Feb 1994 03:08:06 +0000 (03:08 +0000)]
Fixes from John Dyson to fix out-of-memory hangs and other problems (such
as increased swap space usage) related to (incorrectly) paging out the
page tables.
Jordan K. Hubbard [Mon, 7 Feb 1994 23:59:46 +0000 (23:59 +0000)]
Prevent talk from dumping core if TERM is not set - jkh
Andrey Chernov [Mon, 7 Feb 1994 23:47:51 +0000 (23:47 +0000)]
Partially undo prevoius fix from Ian Taylor, VTIME must be 0 here
Andrey Chernov [Mon, 7 Feb 1994 22:37:18 +0000 (22:37 +0000)]
Add Moscow timezone.
Andrew Moore [Mon, 7 Feb 1994 22:12:45 +0000 (22:12 +0000)]
Tentative fdc patches...
Andrey Chernov [Mon, 7 Feb 1994 19:53:25 +0000 (19:53 +0000)]
udb variable can't be static, because netstat wants it
Andrey Chernov [Mon, 7 Feb 1994 18:37:21 +0000 (18:37 +0000)]
Change all siointr (expect in Vsio, of course) to comintr1
This saves two commands per each direct comintr1 call,
allows some strict check in siointr and also
fix bug with suspended output on multiport cards
David Greenman [Mon, 7 Feb 1994 15:46:22 +0000 (15:46 +0000)]
Fixed bug in probe that was causing it to always probe true.
Andrey Chernov [Mon, 7 Feb 1994 15:40:38 +0000 (15:40 +0000)]
Add "wdx: not found" diagnostic missed in previous commits
Rod Grimes [Mon, 7 Feb 1994 10:42:04 +0000 (10:42 +0000)]
Remove the ft driver, it causes instant reboots during attatch, and this
was new code added at the last minute!
Rod Grimes [Mon, 7 Feb 1994 09:20:42 +0000 (09:20 +0000)]
Use MAN8 to pick up the right manual page, no need for SRCS= on a single
source file.
Correct .Nm macro call (probably a typo).
Rod Grimes [Mon, 7 Feb 1994 09:07:48 +0000 (09:07 +0000)]
The last fix applied broke this very bad! Please test things before commiting:
skyrsh# make depend
make: don't know how to make klm_prot_xdr.c. Stop
skyrsh# make all
make: don't know how to make klm_prot_xdr.c. Stop
skyrsh#
Rod Grimes [Mon, 7 Feb 1994 08:40:16 +0000 (08:40 +0000)]
Manual is in section 8, you MUST say MAN8= for this to work. Add $Id$
David Greenman [Mon, 7 Feb 1994 07:44:42 +0000 (07:44 +0000)]
Fixed calculation of physmem when the special MAXMEM kernel config overide
is used. This bug caused the buffer cache to be WAY too big when memory
was being restricted - resulting in hangs and other out of memory problems.