]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
24 years agoFix a bunch of typos and spelling errors.
mpp [Fri, 6 Aug 1999 15:43:53 +0000 (15:43 +0000)]
Fix a bunch of typos and spelling errors.

PR: misc/12965
Partially Submitted by: Patrick Seal <patseal@hyperhost.net>

24 years agostrfile(8) incorrectly states that the output files generated
mpp [Fri, 6 Aug 1999 15:40:58 +0000 (15:40 +0000)]
strfile(8) incorrectly states that the output files generated
will have the suffix ".out" appended to them.  They are acutally
created as outputfile.dat.

PR: docs/12968
Submitted by: Patrick Seal <patseal@hyperhost.net>

24 years agoFix integer overflow for files containing more than 4GB by using
sheldonh [Fri, 6 Aug 1999 14:37:13 +0000 (14:37 +0000)]
Fix integer overflow for files containing more than 4GB by using
u_quad_int instead of u_long for counters. (NetBSD's rev 1.15 - 1.18)

Deprecate register. (NetBSD's rev 1.13)

The diffs from NetBSD were not applied verbatim, because we don't care
about NO_QUAD right now.

PR: 12959
Reported by: Nicholas Barnes <nb@ravenbrook.com>
Obtained from: NetBSD

24 years agoupdating isdn4bsd to beta version 0.83
hm [Fri, 6 Aug 1999 14:05:10 +0000 (14:05 +0000)]
updating isdn4bsd to beta version 0.83

24 years agoDefine IF_MAXMTU and IF_MINMTU and don't allow MTUs with
brian [Fri, 6 Aug 1999 13:53:03 +0000 (13:53 +0000)]
Define IF_MAXMTU and IF_MINMTU and don't allow MTUs with
out-of-range values.

``comparison is always 0'' warnings are silly !

Ok'd by: wollman, dg
Advised by: bde

24 years agoSet the close-on-exec flag when we lock the file we're editing.
brian [Fri, 6 Aug 1999 08:34:42 +0000 (08:34 +0000)]
Set the close-on-exec flag when we lock the file we're editing.
This prevents any background sub-command executed from inheriting
the descriptor & lock (and making vi think that someone else is
editing the file when it re-edits).

Remembered from: An OpenBSD commit message from May '99

24 years agoFixed printf format errors. One of them fixes the garbage output for
bde [Fri, 6 Aug 1999 07:03:25 +0000 (07:03 +0000)]
Fixed printf format errors.  One of them fixes the garbage output for
sa -km reported in the PR.

PR: 4582

24 years agoMention that CCD requires partitions of type FS_BSDFFS.
jkoshy [Fri, 6 Aug 1999 06:25:07 +0000 (06:25 +0000)]
Mention that CCD requires partitions of type FS_BSDFFS.

PR: docs/12985
Submitted by: Kiril Mitev <kiril@ideaglobal.com>

24 years agoPerhaps I should have looked at the whole file rather than just the delta.
peter [Fri, 6 Aug 1999 06:22:43 +0000 (06:22 +0000)]
Perhaps I should have looked at the whole file rather than just the delta.
Use -eq rather than -ne.

Obtained from: OpenBSD (millert, etc/rc rev 1.104)

24 years agoI don't think `customify' is a real word.
peter [Fri, 6 Aug 1999 06:20:19 +0000 (06:20 +0000)]
I don't think `customify' is a real word.

24 years agoDon't use /tmp/_motd in the aftermath of the chflags(2) revelations.
peter [Fri, 6 Aug 1999 06:16:06 +0000 (06:16 +0000)]
Don't use /tmp/_motd in the aftermath of the chflags(2) revelations.

Obtained from: OpenBSD (millert etc/rc rev 1.102 and hugh)

24 years agoRemove empty "SEE ALSO" section.
jkoshy [Fri, 6 Aug 1999 05:51:07 +0000 (05:51 +0000)]
Remove empty "SEE ALSO" section.

PR: docs/12990
Submitted by: Chris Costello <chris@calldei.com>

24 years agoTypo police. 'Acknowledgement' is spelled correctly so it was left untouched.
dwhite [Fri, 6 Aug 1999 04:19:27 +0000 (04:19 +0000)]
Typo police.  'Acknowledgement' is spelled correctly so it was left untouched.

PR: docs/12986
Submitted by: Seth <seth@freebie.dp.ny.frb.org>

24 years agoIf the clock is put back, ensure that we don't end up dividing by
brian [Fri, 6 Aug 1999 01:34:03 +0000 (01:34 +0000)]
If the clock is put back, ensure that we don't end up dividing by
zero when calculating our throughput

24 years agoAdd usr.sbin/elf2exe to the build-tools target on the Alpha. This
jdp [Fri, 6 Aug 1999 00:58:41 +0000 (00:58 +0000)]
Add usr.sbin/elf2exe to the build-tools target on the Alpha.  This
fixes the make buildworld breakage in sys/boot/arc/loader.

24 years agoSync with alc's revision 1.125 of i386/i386/vm_machdep.c. This
jdp [Thu, 5 Aug 1999 23:38:13 +0000 (23:38 +0000)]
Sync with alc's revision 1.125 of i386/i386/vm_machdep.c.  This
fixes the kernel build breakage.

24 years agoAdd a local distribution under the "costum distribution" menu.
sos [Thu, 5 Aug 1999 20:04:56 +0000 (20:04 +0000)]
Add a local distribution under the "costum distribution" menu.
This allows to add local additions on install time..

24 years agoMake the newfs parameters a global option.
phk [Thu, 5 Aug 1999 19:50:26 +0000 (19:50 +0000)]
Make the newfs parameters a global option.

The default is still "-b 8192 -f 1024" but my experiments show that
"-b 16384 -f 4096 -c 100" is a more sensible value for modern
disksizes.

24 years agoAdd RCS IDs to those files without them.
deischen [Thu, 5 Aug 1999 12:15:30 +0000 (12:15 +0000)]
Add RCS IDs to those files without them.
Fix copyrights (s/REGENTS/AUTHOR).

Suggested by: tg
Approved by: jb

24 years agoFix thread initialization to allow for the case where stdio file
deischen [Thu, 5 Aug 1999 12:08:10 +0000 (12:08 +0000)]
Fix thread initialization to allow for the case where stdio file
descriptors are not opened.

PR: bin/12853
Reviewed by: jb

24 years agoAdd access(2) to the list of recognized syscalls.
des [Thu, 5 Aug 1999 12:03:50 +0000 (12:03 +0000)]
Add access(2) to the list of recognized syscalls.

24 years ago- ignore the user-supplied umask, set it to 0022 by default
ru [Thu, 5 Aug 1999 11:48:47 +0000 (11:48 +0000)]
- ignore the user-supplied umask, set it to 0022 by default
- install /etc/wall_cmos_clock with 0444 mode rather than 0666
- -Wall inspection

PR: 12967

24 years agoo Obsolete the undocumented ``set weight'' command.
brian [Thu, 5 Aug 1999 10:32:16 +0000 (10:32 +0000)]
o Obsolete the undocumented ``set weight'' command.
o If we're using RADIUS and the RADIUS mtu is less than our
  peers mru/mrru, reduce our mtu to this value for NetBSD too.
o Make struct throughput's sample period dynamic and tweak the ppp
  version number to reflect the extra stuff being passed through
  the local domain socket as a result (MP mode).
o Measure the current throughput based on the number of samples actually
  taken rather than on the full sample period.
o Keep the throughput statisics persistent while being passed to
  another ppp invocation through the local domain socket.
o When showing throughput statistics after the timer has stopped, use
  the stopped time for overall calculations, not the current time.
  Also show the stopped time and how long the current throughput has
  been sampled for.
o Use time() consistently in throughput.c
o Tighten up the ``show bundle'' output.
o Introduce the ``set bandwidth'' command.
o Rewrite the ``set autoload'' command.  It now takes three arguments
  and works based on a rolling bundle throughput average compared against
  the theoretical bundle bandwidth over a given period (read: it's now
  functional).

24 years agoStay on the serial console if installed that way.
jkh [Thu, 5 Aug 1999 10:07:09 +0000 (10:07 +0000)]
Stay on the serial console if installed that way.

Requested by: max

24 years agoBackup existing init(8) as /sbin/init.bak.
ru [Thu, 5 Aug 1999 09:13:57 +0000 (09:13 +0000)]
Backup existing init(8) as /sbin/init.bak.

PR: 12976
Submitted by: Peter Jeremy <peter.jeremy@alcatel.com.au>

24 years agoI was clearly smoking crack when I committed the change to allow for
jkh [Thu, 5 Aug 1999 08:37:31 +0000 (08:37 +0000)]
I was clearly smoking crack when I committed the change to allow for
architecture-specific docs on the boot floppy and CDROM.  Do it right this
time.

Noticed by: jhay

24 years agocommit these files after updating syscalls.master.
jkh [Thu, 5 Aug 1999 08:27:08 +0000 (08:27 +0000)]
commit these files after updating syscalls.master.

24 years agoRe-commit these files after updating syscalls.master (in the proper order
jkh [Thu, 5 Aug 1999 08:26:27 +0000 (08:26 +0000)]
Re-commit these files after updating syscalls.master (in the proper order
this time).

Pointed out by: bde

24 years agoMove syscall 180 back to where it was before and fix the
jkh [Thu, 5 Aug 1999 08:18:45 +0000 (08:18 +0000)]
Move syscall 180 back to where it was before and fix the
incorrect comment which led me to move it in the first place.

24 years ago- Don't assume that SCSI ID numbers of HDD units are contiguous. That
kato [Thu, 5 Aug 1999 03:19:37 +0000 (03:19 +0000)]
- Don't assume that SCSI ID numbers of HDD units are contiguous.  That
  is, don't assume that SCSI ID corresponds to a unit number of da
  device.  Unit number of da device is provided by 2nd stage loader
  and 3rd stage loader now use it.
- Fix drive letter to display.

Submitted by: IMAI Takeshi <take-i@ceres.dti.ne.jp>

24 years agoProvides discontinuous unit number support to make external FDD
kato [Thu, 5 Aug 1999 03:12:20 +0000 (03:12 +0000)]
Provides discontinuous unit number support to make external FDD
bootable on 1 FDD PC98 machines.  (When an external FDD unit is
installed, unit numbers become discontinuous.)

Submitted by: IMAI Takeshi <take-i@ceres.dti.ne.jp>

24 years agoTurn docs back on - ache was very quick.
jkh [Thu, 5 Aug 1999 01:28:00 +0000 (01:28 +0000)]
Turn docs back on - ache was very quick.

24 years agoMore committage for the arla pioctl syscall.
jkh [Thu, 5 Aug 1999 00:44:11 +0000 (00:44 +0000)]
More committage for the arla pioctl syscall.

Reminded by: green

24 years agoFix typo which would have caused MTRR support on non-SMP systems to
msmith [Wed, 4 Aug 1999 20:24:49 +0000 (20:24 +0000)]
Fix typo which would have caused MTRR support on non-SMP systems to
behave in an utterly random fashion.

Submitted by: gibbs

24 years agoReserve a syscall for the arla folks. I'm assuming that since syscalls.c
jkh [Wed, 4 Aug 1999 20:04:25 +0000 (20:04 +0000)]
Reserve a syscall for the arla folks.  I'm assuming that since syscalls.c
and init_sysent.c are checked into CVS, I should also commit the regenerated
copies even though they're built by syscalls.master.  Correct?  Bruce? :)

24 years agoThe ARLA folks don't need a bdev after all - remove it.
jkh [Wed, 4 Aug 1999 19:55:37 +0000 (19:55 +0000)]
The ARLA folks don't need a bdev after all - remove it.

24 years ago"Linux emulator" -> "Linux image activator"
nik [Wed, 4 Aug 1999 18:58:37 +0000 (18:58 +0000)]
"Linux emulator" -> "Linux image activator"

PR:             docs/12882
Submitted by:   des

24 years agoFix fd race conditions (during shared fd table usage.) Badfileops is
green [Wed, 4 Aug 1999 18:53:50 +0000 (18:53 +0000)]
Fix fd race conditions (during shared fd table usage.) Badfileops is
now used in f_ops in place of NULL, and modifications to the files
are more carefully ordered. f_ops should also be set to &badfileops
upon "close" of a file.

This does not fix other problems mentioned in this PR than the first
one.

PR: 11629
Reviewed by: peter

24 years agoAdd various network cards that have been new-busified and made into KLDs
green [Wed, 4 Aug 1999 17:44:07 +0000 (17:44 +0000)]
Add various network cards that have been new-busified and made into KLDs
(thanks Bill!)

Remove lkm.

24 years agoTurn docs off by default again - somebody has broken the docproj port.
jkh [Wed, 4 Aug 1999 17:30:00 +0000 (17:30 +0000)]
Turn docs off by default again - somebody has broken the docproj port.

24 years agoCorrection: "ans" -> "and."
green [Wed, 4 Aug 1999 17:29:33 +0000 (17:29 +0000)]
Correction: "ans" -> "and."

24 years agoLog hostname settings/changes.
obrien [Wed, 4 Aug 1999 15:12:27 +0000 (15:12 +0000)]
Log hostname settings/changes.

Submitted by: John Baldwin <jobaldwi@vt.edu>

24 years agoBack out rev 1.2. The correct type of host-name is `X'. There was a problem
obrien [Wed, 4 Aug 1999 15:02:28 +0000 (15:02 +0000)]
Back out rev 1.2.  The correct type of host-name is `X'.  There was a problem
with options.c which was fixed in ISC's version 2.0 (rev 1.1.1.2 --> 1.1.1.3).

I have tested host-name with both types `X' and `t' and things work fine
either way.  I would prefer to match the offical sources when easily possible.

PR: 12205
Submitted by: John Baldwin <jobaldwi@vt.edu>

24 years agoDon't over-allocate and over-copy shorter NFSv2 filehandles and then
peter [Wed, 4 Aug 1999 14:41:39 +0000 (14:41 +0000)]
Don't over-allocate and over-copy shorter NFSv2 filehandles and then
correct the pointers afterwards.

It's kinda bogus that we generate a 24 (?) byte filehandle (2 x int32
fsid and 16 byte VFS fhandle) and pad it out to 64 bytes for NFSv3 with
garbage.  The whole point of NFSv3's variable filehandle length was
to allow for shorter handles, both in memory and over the wire.  I plan
on taking a shot at fixing this shortly.

24 years agoDon't probe if pci_cfgopen() fails to find pci hardware, like we used to
peter [Wed, 4 Aug 1999 13:38:24 +0000 (13:38 +0000)]
Don't probe if pci_cfgopen() fails to find pci hardware, like we used to
to.  This might have caused interesting things on non-PCI hardware if
PCI was compiled in.

24 years agoMFS: use logger(1) instead of plain echo(1).
obrien [Wed, 4 Aug 1999 05:46:11 +0000 (05:46 +0000)]
MFS: use logger(1) instead of plain echo(1).
(some how these local changes made it out of my local tree into the
repository, so lets do it right)

24 years agoSupport the setting of hostname via DHCP.
obrien [Wed, 4 Aug 1999 05:34:33 +0000 (05:34 +0000)]
Support the setting of hostname via DHCP.

Submitted by: myself & Ted Lemon <mellon@isc.org>

24 years agoo Typo in prior version kept it from compiling (blush).
imp [Wed, 4 Aug 1999 04:52:18 +0000 (04:52 +0000)]
o Typo in prior version kept it from compiling (blush).

Noticed by: Nobody!

o Add comment about why we restrict chflags to root for devices.
o nit noticed by bde wrt return values.

24 years agobrucify:
imp [Tue, 3 Aug 1999 17:07:04 +0000 (17:07 +0000)]
brucify:
o use suser_xxx rather than suser to support JAIL code.
o KNF comment convention
o use vp->type rather than vaddr.type and eliminate call to
  VOP_GETATTR.  Bruce says that vp->type is valid at this
  point.

Submitted by: bde.

Not fixed:
o return (value)
o Comment needs to be longer and more explicit.  It will be after
  the advisory.

24 years agoReorganise ppp's usage to avoid some mandoc limitations.
brian [Tue, 3 Aug 1999 16:14:38 +0000 (16:14 +0000)]
Reorganise ppp's usage to avoid some mandoc limitations.
Suggested by: wollman

24 years agoAdd a sample entry for Exim, in preparation for the upcoming behaviour
sheldonh [Tue, 3 Aug 1999 14:52:46 +0000 (14:52 +0000)]
Add a sample entry for Exim, in preparation for the upcoming behaviour
change in the port, where TCP Wrapper support will become the default.

Requested by: markm

24 years agoBetter comment on ntpdate_enable - if I have misunderstood when
eivind [Tue, 3 Aug 1999 10:11:02 +0000 (10:11 +0000)]
Better comment on ntpdate_enable - if I have misunderstood when
ntpdate was run (I thought it was cron'ed), other people probably do
too.

24 years agoTypo
wosch [Tue, 3 Aug 1999 08:15:02 +0000 (08:15 +0000)]
Typo

Submitted by:  Christian Weisgerber <naddy@mips.rhein-neckar.de>

24 years agoMinor tweak for last commit: insert extra delay between issuing master
wpaul [Mon, 2 Aug 1999 21:57:57 +0000 (21:57 +0000)]
Minor tweak for last commit: insert extra delay between issuing master
reset and RX/TX resets.

24 years agoMention that ospf is a possible filter protocol.
brian [Mon, 2 Aug 1999 21:45:36 +0000 (21:45 +0000)]
Mention that ospf is a possible filter protocol.

24 years agoOnly allow root to set file flags on devices.
imp [Mon, 2 Aug 1999 21:34:46 +0000 (21:34 +0000)]
Only allow root to set file flags on devices.

24 years agoPerform an RX reset and TX reset in xl_reset() along with the master
wpaul [Mon, 2 Aug 1999 21:06:16 +0000 (21:06 +0000)]
Perform an RX reset and TX reset in xl_reset() along with the master
reset command.

I observed some anomalous behavior while testing a 3c905C with a
Dell PowerEdge 4300/500 dual PIII 500Mhz system. The NIC would seem
to work correctly most of the time but would sometimes fail to receive
certain packets, in particular NFS create requests. I could mount
an NFS filesystem from the PowerEdge and do an ls on it, but trying
to do a "touch foo" would hang. Monitoring traffic from another host
revealed that the client was properly sending an NFS create request
but the server was not receiving it. It *did* receive it when I
ran the same test with an Intel fxp card.

I don't understand the exact mechanics of this strange behavior, but
resetting the receiver and transmitter seems to get rid of it. I used
to perform an RX and TX reset in xl_init(), but stopped doing it there
because on 3c905B and later cards this causes the autoneg session to
restart, which would lead to the NIC waiting a long time before exchanging
traffic after being brought up the first time. Apparently the receiver
and transmitter resets should be performed at least once when initializing
the card.

Hopefully this will cure problems that people have been having with the
3c905C -- this was the only strange behavior that I have observed with
the 3c905C so far which does not appear with the 3c905B or 3c905.

24 years agoThe dev returned here is what is found in the st_dev field.
rvb [Mon, 2 Aug 1999 20:55:11 +0000 (20:55 +0000)]
The dev returned here is what is found in the st_dev field.
This should not be further translated ... hence the 0.

24 years agoI'm not sure how this wasn't in the last commit but anyhow...
mdodd [Mon, 2 Aug 1999 20:01:40 +0000 (20:01 +0000)]
I'm not sure how this wasn't in the last commit but anyhow...
'int irq' -> 'struct irq_node *irq'

24 years agoMake this build if there is an obj directory.
dfr [Mon, 2 Aug 1999 19:16:42 +0000 (19:16 +0000)]
Make this build if there is an obj directory.

24 years agoDon't trust the segment limits that the BIOS reports; instead give it
msmith [Mon, 2 Aug 1999 18:46:34 +0000 (18:46 +0000)]
Don't trust the segment limits that the BIOS reports; instead give it
as much space as possible to avoid lossage due to sloppy BIOS
programming.

24 years agoAdd missing "path mnemonics" as given in Table 4-3 of rev. 1.2 of the
se [Mon, 2 Aug 1999 18:26:28 +0000 (18:26 +0000)]
Add missing "path mnemonics" as given in Table 4-3 of rev. 1.2 of the
ARC Specification.

24 years agoRemove accidental 8bit character in error message.
se [Mon, 2 Aug 1999 17:50:07 +0000 (17:50 +0000)]
Remove accidental 8bit character in error message.
Add checks for malloc failure and add error exit in case of malloc failure.
Add rcsid.

Submitted by: charnier

24 years agoFix typo and use Fx macro instead of literal FreeBSD.
se [Mon, 2 Aug 1999 17:45:21 +0000 (17:45 +0000)]
Fix typo and use Fx macro instead of literal FreeBSD.

Submitted by: charnier

24 years agoUse err[x]/warn[x] and improve error reporting
brian [Mon, 2 Aug 1999 16:59:28 +0000 (16:59 +0000)]
Use err[x]/warn[x] and improve error reporting
Submitted by: charnier

24 years ago#ifdef IPPROTO_OSPFIGP before expecting it to be defined.
brian [Mon, 2 Aug 1999 15:29:19 +0000 (15:29 +0000)]
#ifdef IPPROTO_OSPFIGP before expecting it to be defined.
This unbreaks OpenBSD.

24 years agoAdd -n option to print numeric user and group IDs instead of names
sheldonh [Mon, 2 Aug 1999 14:55:58 +0000 (14:55 +0000)]
Add -n option to print numeric user and group IDs instead of names
in a long (-l) listing.

MFC-jockies should make sure that bde's concerns regarding the number
of digits required to represent a uid_t and the use of snprintf
on the associated PR have been addressed before going wild.

PR: 12866
Reported by: Philip Kizer <pckizer@nostrum.com>
Obtained from: NetBSD

24 years agoFilter ospf and igmp separately.
brian [Mon, 2 Aug 1999 11:53:16 +0000 (11:53 +0000)]
Filter ospf and igmp separately.

Kind-of submitted by: phk

24 years agoAdd new ports category "ftp".
asami [Mon, 2 Aug 1999 11:33:19 +0000 (11:33 +0000)]
Add new ports category "ftp".

24 years agoAdd ports-ftp.
asami [Mon, 2 Aug 1999 11:31:04 +0000 (11:31 +0000)]
Add ports-ftp.

24 years agoUse Byacc rather than Bison.
obrien [Mon, 2 Aug 1999 10:02:07 +0000 (10:02 +0000)]
Use Byacc rather than Bison.

24 years agoClear flags on psuedo terminals before trying to chmod and chown them.
imp [Mon, 2 Aug 1999 05:42:44 +0000 (05:42 +0000)]
Clear flags on psuedo terminals before trying to chmod and chown them.

24 years agoEnglish fixes: consistent spacing after periods, "userland", not "user land",
hoek [Mon, 2 Aug 1999 04:00:33 +0000 (04:00 +0000)]
English fixes: consistent spacing after periods, "userland", not "user land",
               other typos, ~four grammar gnits, an ironic case of incorrect
               parallelization, bad capitalization, an incorrect use of the
               infamous slash ('/'), and an unclear sentence.

24 years agoBruce noted that the use of err(), fixed to errx() in the last commit,
hoek [Mon, 2 Aug 1999 03:18:17 +0000 (03:18 +0000)]
Bruce noted that the use of err(), fixed to errx() in the last commit,
did not specify an exit code.  This implies the use of either a hand-
rolled err() (Bruce's suggestion) or a random error code (my suggestion),
both of which are against the style guidelines.  This commit specifies
the correct error code (implicitly).  This also changes the error message
to be a little more helpful.

24 years agoMove the specification of EDGE/LEVEL triggered interrupts to
mdodd [Sun, 1 Aug 1999 22:57:09 +0000 (22:57 +0000)]
Move the specification of EDGE/LEVEL triggered interrupts to
eisa_add_intr() which now takes an additional arguement (one of
EISA_TRIGGER_LEVEL or EISA_TRIGGER_EDGE).

The flag RR_SHAREABLE has no effect when passed to
bus_alloc_resource(dev, SYS_RES_IRQ, ...) in an EISA device context as
the eisa_alloc_resource() call (bus_alloc_resource method) now deals
with this flag directly, depending on the device ivars.

This change does nothing more than move all the 'shared = inb(foo + iobsse)'
nonesense to the device probe methods rather than the device attach.

Also, print out 'edge' or 'level' in the IRQ announcement message.

Reviewed by: dfr

24 years agoMention that EXIT_SUCCESS and EXIT_FAILURE are preferred over
hoek [Sun, 1 Aug 1999 20:46:45 +0000 (20:46 +0000)]
Mention that EXIT_SUCCESS and EXIT_FAILURE are preferred over
sysexits.h-values as exit codes for portable programs.

24 years agoTypo of sample command line.
sada [Sun, 1 Aug 1999 20:31:57 +0000 (20:31 +0000)]
Typo of sample command line.

Submitted by: Seiichi Satomi <ssatomi@rr.iij4u.or.jp>

24 years agoRichard the Lionheart was crowned on 3rd September 1189, not the 27th
nik [Sun, 1 Aug 1999 19:54:02 +0000 (19:54 +0000)]
Richard the Lionheart was crowned on 3rd September 1189, not the 27th
February.

If you do a web search for "lionheart crowned" you'll get lots of
conflicting information.  Some sites say 3rd September, while others
say 27th February.  Most of the "27th February" crowd seem to take their
information from other incarnations of this file on other operating
systems.

After a very pleasant afternoon spent lunching with my girlfriend's
parents, I availed myself of their extensive reference library.

You'd be surprised how hard it is to get concrete information about this.
The _Encyclopedia Brittanica_ doesn't mention the date, only the year, as
does _Brewer's Dictionary of Phrase and Fable_, as do all the other printed
sources I tried.  One of them even said July 7th 1189!  Microsoft's (yeah,
so sue me) Encarta '95 has quite a comprehensive entry, but again, no
day and month information

In desperation, I tried the web once more, and finally stumbled upon
http://www.btinternet.com/~timeref/hsttime2.htm.  This revealed that
Henry II died on 6th July 1189 (presumably the source of the 7th July
entry in another reference), and that Richard was crowned on 3rd
September.

Best of all, this site gives references.  So if any of you have a copy of
_The Life and Times of Richard I_, John Gillingham, pub. George Weidenfeld
and Nicholson Limited, 1974, then you can confirm this for yourselves.

For completenesses sake, I tried to find an ISBN number for the above
book.  But Amazon and Barnes and Noble don't appear to stock it (although
it looks like a revised version, by the same author, is due out in October
1999, in case anyone's interested).

PR:             docs/10488
Submitted by:   solon@macaulay.demon.co.uk

24 years agoAdd ipfw recompile and bpfilter -> bpf changes
imp [Sun, 1 Aug 1999 18:58:45 +0000 (18:58 +0000)]
Add ipfw recompile and bpfilter -> bpf changes

24 years agoNote need to recompile pccardd with latest kernel.
imp [Sun, 1 Aug 1999 18:19:03 +0000 (18:19 +0000)]
Note need to recompile pccardd with latest kernel.

24 years agoo Add additional printfs for error cases when we can't attach the device.
imp [Sun, 1 Aug 1999 18:12:51 +0000 (18:12 +0000)]
o Add additional printfs for error cases when we can't attach the device.
o Add field to dev_desc for the size of the io port range.  This isn't
  used yet in the committed sources, but will make the transition easier
  in the future.

If you build this into your kernel, you will need to rebuild pccardd.

24 years agoo Add "debuglevel" keyword to config file.
imp [Sun, 1 Aug 1999 18:04:26 +0000 (18:04 +0000)]
o Add "debuglevel" keyword to config file.
o Document debug level keyword
o Implement debug level:
o For most of the diagnostic messages, change them from #ifdef DEBUG
  to if (debuglevel > 0).
o Add a couple more diagnostic messages that weren't present before
o Fix a couple of excessively long lines.
Reviewed by: hosokawa-san

o Start to implement the stopgap kludge for -current's pccard code by passing
  the length of the i/o range.  If DEV_DESC_HAS_SIZE is defined, we'll set
  the size.  This is done as an ifdef so that I can generate patches
  against the kernel more easily.
o Add preliminary support for tweaking sleep times, but leave it
  disabled until a good range of values can be established.

Didn't fix: logmsg problem noted by Nate.

24 years ago Make ipfw's logging more dynamic. Now, log will use the default limit
green [Sun, 1 Aug 1999 16:57:24 +0000 (16:57 +0000)]
  Make ipfw's logging more dynamic. Now, log will use the default limit
_or_ you may specify "log logamount number" to set logging specifically
the rule.
   In addition, "ipfw resetlog" has been added, which will reset the
logging counters on any/all rule(s). ipfw resetlog does not affect
the packet/byte counters (as ipfw reset does), and is the only "set"
command that can be run at securelevel >= 3.
   This should address complaints about not being able to set logging
amounts, not being able to restart logging at a high securelevel,
and not being able to just reset logging without resetting all of the
counters in a rule.

24 years agoClarify that "-w" ignores horizontal white space changes.
obrien [Sun, 1 Aug 1999 16:07:09 +0000 (16:07 +0000)]
Clarify that "-w" ignores horizontal white space changes.

24 years agoSync with sys/i386/conf/majors.i386 revision 1.80.
kato [Sun, 1 Aug 1999 16:03:31 +0000 (16:03 +0000)]
Sync with sys/i386/conf/majors.i386 revision 1.80.

24 years agoOnly use Bison to generate cc1plus's parse code.
obrien [Sun, 1 Aug 1999 15:44:51 +0000 (15:44 +0000)]
Only use Bison to generate cc1plus's parse code.

24 years agoCleanup - diked out a heap of old KLM stuff that isn't needed anymore.
newton [Sun, 1 Aug 1999 12:51:06 +0000 (12:51 +0000)]
Cleanup - diked out a heap of old KLM stuff that isn't needed anymore.

24 years agoRemove inappropriate references to the wrong name for the bootstrap
grog [Sun, 1 Aug 1999 06:40:11 +0000 (06:40 +0000)]
Remove inappropriate references to the wrong name for the bootstrap
manager, thus killing two birds with one stone.

Objected-to-by:       bde
      rnordier

24 years agoMove the memory access behavior information provided by madvise
alc [Sun, 1 Aug 1999 06:05:09 +0000 (06:05 +0000)]
Move the memory access behavior information provided by madvise
from the vm_object to the vm_map.

Submitted by: dillon

24 years agopmap_object_init_pt:
alc [Sat, 31 Jul 1999 23:02:57 +0000 (23:02 +0000)]
pmap_object_init_pt:
Verify that object != NULL.

24 years agoFixed $Id$.
bde [Sat, 31 Jul 1999 22:29:17 +0000 (22:29 +0000)]
Fixed $Id$.

Removed POSIX.1/NetBSD markup (braces) for NAME_MAX, etc.  We don't
define this.  Most FreeBSD man pages hard-code the limits; in fact,
utimes.2 recently became the only file in libc/sys/*.2 that mentions
NAME_MAX.  There probably should be mandoc macros for this.

24 years agoRemoved a duplicate reference to System V.4.
bde [Sat, 31 Jul 1999 22:14:46 +0000 (22:14 +0000)]
Removed a duplicate reference to System V.4.

24 years agoFixed syntax error in previous commit.
bde [Sat, 31 Jul 1999 22:10:17 +0000 (22:10 +0000)]
Fixed syntax error in previous commit.

24 years agoFIxed disordering in previous commit. Fixed some old disorder.
bde [Sat, 31 Jul 1999 22:00:09 +0000 (22:00 +0000)]
FIxed disordering in previous commit.  Fixed some old disorder.

24 years agoAdd a -E flag, similar to -e (overide variables from environment) except
hoek [Sat, 31 Jul 1999 20:53:02 +0000 (20:53 +0000)]
Add a -E flag, similar to -e (overide variables from environment) except
that -E only operates for a specified variable.  Useful since the -e option
will often pull-in many unwanted variable overrides (esp. in a make world
situation).  Uses include overriding BINOWN (which cannot be done by normal
methods or through abuses of MAKEFLAGS) or likely for ports to honour CFLAGS
(provided they're running on a system whose make(1) has this option).

24 years agoPrint an error message on illegal numerical arguments.
hoek [Sat, 31 Jul 1999 20:40:23 +0000 (20:40 +0000)]
Print an error message on illegal numerical arguments.

Submitted by: bin/9349 (slightly modified) Assar Westerlund <assar@sics.se>

24 years agoWhen printing an error message after command execution fails, don't assume
hoek [Sat, 31 Jul 1999 20:38:22 +0000 (20:38 +0000)]
When printing an error message after command execution fails, don't assume
the only possible error is "command not found".  This makes debugging of
messages such as "/bin/sh: not found" less common...

24 years agoRemove explicit testing of the NOFSCHG variable. Its functionality (in this
hoek [Sat, 31 Jul 1999 20:32:31 +0000 (20:32 +0000)]
Remove explicit testing of the NOFSCHG variable.  Its functionality (in this
file) is now superceeded by INSTALLFLAGS_EDIT in bsd.{prog,kmod,lib}.mk.

24 years agoAdd variable INSTALLFLAGS_EDIT used to remove arbitrary INSTALLFLAGS.
hoek [Sat, 31 Jul 1999 20:27:33 +0000 (20:27 +0000)]
Add variable INSTALLFLAGS_EDIT used to remove arbitrary INSTALLFLAGS.

Specifically intended for removing -fschg ("INSTALLFLAGS_EDIT=:S/schg/uchg/")
this makes the NOFSCHG flag redundant.  NOFSCHG will still be honoured by
bsd.lib.mk but is valid for buildworld only.  NOFSCHG is still implemented in
the old way (ie. _not_ ".if NOFSCHG then { INSTALLFLAGS_EDIT+=:S/schg/,/ }"
to emphasize the fact that NOFSCHG is only supported in a limited
fashion and for buildworld.

The interface and implementation are such that future use of flags such
as sappnd can also be easily removed or altered (perhaps to uappnd).

This commit brought to you by the letters B, D, and E, and the numbers six,
one, thirteen, and three.

24 years agoAdd INSTALL.TXT back.
jkh [Sat, 31 Jul 1999 19:15:29 +0000 (19:15 +0000)]
Add INSTALL.TXT back.