]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
22 years agoThis commit was generated by cvs2svn to compensate for changes in r86218,
wollman [Fri, 9 Nov 2001 19:00:37 +0000 (19:00 +0000)]
This commit was generated by cvs2svn to compensate for changes in r86218,
which included commits to RCS files with non-trunk default branches.

22 years agoUpdated timezone database from Arthur Olson.
wollman [Fri, 9 Nov 2001 19:00:37 +0000 (19:00 +0000)]
Updated timezone database from Arthur Olson.

PR: conf/31851

22 years agoAdd share/zoneinfo/North_Dakota in preparation for tzdata update.
wollman [Fri, 9 Nov 2001 18:58:07 +0000 (18:58 +0000)]
Add share/zoneinfo/North_Dakota in preparation for tzdata update.

22 years agoAdd terminating % so the last fortune can be found.
dillon [Fri, 9 Nov 2001 18:24:09 +0000 (18:24 +0000)]
Add terminating % so the last fortune can be found.
Add a fortune.

22 years agoBack out switching from libreadline to libedit:
ache [Fri, 9 Nov 2001 17:09:42 +0000 (17:09 +0000)]
Back out switching from libreadline to libedit:
1) It is capabilities degradation: f.e. libedit can't process ~/.inputrc files
2) It have no sense to try to un-GNU GNU program.

22 years ago- Put missing prototype for rcmd() in <unistd.h>.
ru [Fri, 9 Nov 2001 15:19:25 +0000 (15:19 +0000)]
- Put missing prototype for rcmd() in <unistd.h>.
- Clean up the manpage.
- style(9) rcmdsh.c.

Committed from: BSDCon/EU 2k+1 terminal room

22 years ago* Make sure we increment pm_stats.resident_count in pmap_enter_quick
dfr [Fri, 9 Nov 2001 13:25:14 +0000 (13:25 +0000)]
* Make sure we increment pm_stats.resident_count in pmap_enter_quick
* Re-organise RID allocation so that we don't accidentally give a RID
  to two different processes. Also randomise the order to try to reduce
  collisions in VHPT and TLB. Don't allocate RIDs for regions which are
  unused.
* Allocate space for VHPT based on the size of physical memory. More
  tuning is needed here.
* Add sysctl instrumentation for VHPT - see sysctl vm.stats.vhpt
* Fix a bug in pmap_prefault() which prevented it from actually adding
  pages to the pmap.
* Remove ancient dead debugging code.
* Add DDB commands for examining translation registers and region
  registers.

The first change fixes the 'free/cache page %p was dirty' panic which I
have been seeing when the system is put under moderate load. It also
fixes the negative RSS values in ps which have been confusing me for a
while.

With this set of changes the ia64 port is reliable enough to build its
own kernels, even with a 20-way parallel build. Next stop buildworld.

22 years agoRaise SIGILL for General Exceptions - its closer to the correct meaning.
dfr [Fri, 9 Nov 2001 13:11:17 +0000 (13:11 +0000)]
Raise SIGILL for General Exceptions - its closer to the correct meaning.

22 years agoReserve more space for phys_avail. Really need to be more careful about
dfr [Fri, 9 Nov 2001 13:09:57 +0000 (13:09 +0000)]
Reserve more space for phys_avail. Really need to be more careful about
overflowing phys_avail.

22 years agoTeach DDB about branch registers.
dfr [Fri, 9 Nov 2001 13:08:25 +0000 (13:08 +0000)]
Teach DDB about branch registers.

22 years agoDefine PS and VE fields of region register correctly.
dfr [Fri, 9 Nov 2001 13:07:44 +0000 (13:07 +0000)]
Define PS and VE fields of region register correctly.

22 years agoThe algorithm that computes the tables used in the BM search algorithm sometimes
dcs [Fri, 9 Nov 2001 10:17:44 +0000 (10:17 +0000)]
The algorithm that computes the tables used in the BM search algorithm sometimes
access an array beyond it's length. This only happens in the last iteration of
a loop, and the value fetched is not used then, so the bug is a relatively
innocent one. Fix this by not fetching any value on the last iteration of said
loop.

Submitted by: MKI <mki@mozone.net>
MFC after: 1 week

22 years agoo Try to do 3.3V support better for the 6722 and 6729/30.
imp [Fri, 9 Nov 2001 07:33:54 +0000 (07:33 +0000)]
o Try to do 3.3V support better for the 6722 and 6729/30.
o Bite the bullet and create controller types for the 6729 and also for
  the 673x.  Rename the 672x to 6722.
o Define minimal extended register info (just register 0xa for reading VS[12]).

# I think the last version may have broken 673x controllers, but this should
# fix them.

Tested on the 6722, but not the 6729.

Ideas from: Chiharu Shibata-san's article in bsd-nomads:15866

22 years agoFix the 3.3V support for Cirrus Logic CL-PD6710. This appears to work
imp [Fri, 9 Nov 2001 06:26:57 +0000 (06:26 +0000)]
Fix the 3.3V support for Cirrus Logic CL-PD6710.  This appears to work
on my CL-PD6722, but won't work on the CL-PD6729.  The latter two need
more sophisticated detection of 3.3V cards than I'm up to at the
moment.  Also, only a few of the ISA chipsets that support 3.3V will
likely work at the moment.

It appears that for 3.3V cards we must detect them and adjust the
pwr.vcc value from 50 to 33.  Give a strong hint to automatically
power up the card for PD_POWER cards.

This makes my SMC 2602W (the 3.3V version of the 2632W) work on my
Fujitsu Stylistic 500.

SMC 3.3V card donated by: Ryan Losh
Thanks to: bsd-nomads for reviews of past 3.3V code

22 years agoForced commit to make diffs easier
imp [Fri, 9 Nov 2001 05:24:51 +0000 (05:24 +0000)]
Forced commit to make diffs easier

22 years agoImplement os_boot_rendez. Application processors are initialized
marcel [Fri, 9 Nov 2001 05:18:45 +0000 (05:18 +0000)]
Implement os_boot_rendez. Application processors are initialized
and brought to a point where kernel specific initializations can
be done. That will be the next step...

22 years agoo nvi failed to correctly handle EINTR returned from tcsetattr(), exiting
rwatson [Fri, 9 Nov 2001 02:23:05 +0000 (02:23 +0000)]
o nvi failed to correctly handle EINTR returned from tcsetattr(), exiting
  instead of recovering, which happens in particular during a rapid series
  of SIGWINCH's.  This change forces nvi to loop on the call in the event
  that the call is interrupted.

Interestingly, I submitted this bug report in 1998, and a solution was
posted shortly thereafter by Matthias Buelow; it's been sitting in the
PR database ever since.

Note: this takes this file off the vendor branch.  If and when we find
a vendor for this code, the fix should be given back to them.

PR: bin/8438
Submitted by: Matthias Buelow <mkb@altair.mayn.de>
MFC after: 1 month

22 years agoNote that the manpage is incorrect about the vector argument.
peter [Fri, 9 Nov 2001 01:01:50 +0000 (01:01 +0000)]
Note that the manpage is incorrect about the vector argument.

22 years agoThis commit was generated by cvs2svn to compensate for changes in r86198,
peter [Thu, 8 Nov 2001 23:15:13 +0000 (23:15 +0000)]
This commit was generated by cvs2svn to compensate for changes in r86198,
which included commits to RCS files with non-trunk default branches.

22 years ago<repo surgery - force revision change>
peter [Thu, 8 Nov 2001 23:15:13 +0000 (23:15 +0000)]
<repo surgery - force revision change>

22 years agoActually, arc4random() returns a uint32_t, not an int. Use the correct
wollman [Thu, 8 Nov 2001 22:01:35 +0000 (22:01 +0000)]
Actually, arc4random() returns a uint32_t, not an int.  Use the correct
constant, just in case uint32_t turns into a `short' ten years from now.
If this is MFC'd it will be necessary to hard-code the constant since
-stable doesn't have UINT32_MAX.

22 years agofix jot -r on 64-bit platforms by teaching it that arc4random outputs
gallatin [Thu, 8 Nov 2001 20:49:24 +0000 (20:49 +0000)]
fix jot -r on 64-bit platforms by teaching it that arc4random outputs
an int, not a long

Submitted by: Paul Herman <pherman@frenchfries.net>
PR#: alpha/31859

22 years agoPlacemark an interrupt race in -current which is currently protected by
dillon [Thu, 8 Nov 2001 18:09:18 +0000 (18:09 +0000)]
Placemark an interrupt race in -current which is currently protected by
Giant.  -stable will get spl*() fixes for the race.

Reported by: Rob Anderson <rob@isilon.com>
MFC after: 0 days

22 years agoDon't allocate a zero byte segment.
mikeh [Thu, 8 Nov 2001 16:47:05 +0000 (16:47 +0000)]
Don't allocate a zero byte segment.

PR: bin/11900
MFC after: 2 weeks

22 years agoAdd support for Intel's i820/i840/i845/i850/i860 chipset.
kuriyama [Thu, 8 Nov 2001 16:03:23 +0000 (16:03 +0000)]
Add support for Intel's i820/i840/i845/i850/i860 chipset.

Submitted by: nork@cityfujisawa.ne.jp (Norikatsu Shigemura)
PR: kern/31559, kern/31825
MFC after: 1 week

22 years agoo General style improvemnts.
rwatson [Thu, 8 Nov 2001 15:31:19 +0000 (15:31 +0000)]
o General style improvemnts.

Submitted by: bde

22 years agoo Trim trailing whitespace from kern_mib.c, as suggested by bde. Good
rwatson [Thu, 8 Nov 2001 15:20:00 +0000 (15:20 +0000)]
o Trim trailing whitespace from kern_mib.c, as suggested by bde.  Good
  grief.

22 years agoleverage fixed fifo implementation (rev 1.56 of
alfred [Thu, 8 Nov 2001 10:37:32 +0000 (10:37 +0000)]
leverage fixed fifo implementation (rev 1.56 of
src/sys/fs/fifofs/fifo_vnops.c) to serve locks better, my previous
workaround for lack of decent fifo system wasn't cutting it,
particularly the kernel would send a message down the fifo and
immediately close it, this would lead to delayed unlock requests
being seen by the lockd causing all sorts of badness.

Basically, don't reopen the fifo, just select(2) on it.

22 years agouse syslog rather than printf so we can see what's going on in debug mode
alfred [Thu, 8 Nov 2001 10:34:21 +0000 (10:34 +0000)]
use syslog rather than printf so we can see what's going on in debug mode

22 years agostyle fixes, fix line wrap and function declaration
alfred [Thu, 8 Nov 2001 10:33:41 +0000 (10:33 +0000)]
style fixes, fix line wrap and function declaration

22 years agoSwitch behavior of fifos to more closely match what goes on in other OSes.
alfred [Thu, 8 Nov 2001 10:28:32 +0000 (10:28 +0000)]
Switch behavior of fifos to more closely match what goes on in other OSes.
Basically FIFOs become a real pain to abuse as a rendevous point without
this change because you can't really select(2) on them because they always
return ready even though there is no writer (to signal EOF).

Obtained from: BSD/os

22 years agoo Replace reference to 'struct proc' with 'struct thread' in 'struct
rwatson [Thu, 8 Nov 2001 02:13:18 +0000 (02:13 +0000)]
o Replace reference to 'struct proc' with 'struct thread' in 'struct
  sysctl_req', which describes in-progress sysctl requests.  This permits
  sysctl handlers to have access to the current thread, permitting work
  on implementing td->td_ucred, migration of suser() to using struct
  thread to derive the appropriate ucred, and allowing struct thread to be
  passed down to other code, such as network code where td is not currently
  available (and curproc is used).

o Note: netncp and netsmb are not updated to reflect this change, as they
  are not currently KSE-adapted.

Reviewed by: julian
Obtained from: TrustedBSD Project

22 years agoWe need to operate on struct thread now, not struct proc.
obrien [Thu, 8 Nov 2001 01:49:06 +0000 (01:49 +0000)]
We need to operate on struct thread now, not struct proc.

Submitted by: rwatson

22 years agoFor what its worth, sync up the type of ps_arg_cache_max (unsigned long)
peter [Thu, 8 Nov 2001 00:24:48 +0000 (00:24 +0000)]
For what its worth, sync up the type of ps_arg_cache_max (unsigned long)
with the sysctl type (signed long).

22 years agokern.ps_arg_max_cache is a long, not an int. I believe this is half of
peter [Thu, 8 Nov 2001 00:23:06 +0000 (00:23 +0000)]
kern.ps_arg_max_cache is a long, not an int.  I believe this is half of
what broke ps on ia64.  It probably also broke on alpha, but the fallback
method of using lseek/read on /proc/*/mem to read ps_strings seems to
work there.  It doesn't on ia64 yet.

22 years agoTraditionally work in the -traditional case.
obrien [Wed, 7 Nov 2001 23:14:31 +0000 (23:14 +0000)]
Traditionally work in the -traditional case.

Submitted by: joerg

22 years agoAdd quirk for Minolta 2330 Zoom digital camera.
kbyanc [Wed, 7 Nov 2001 23:11:13 +0000 (23:11 +0000)]
Add quirk for Minolta 2330 Zoom digital camera.

Submitted by: Jan Stocker <Jan.Stocker@t-online.de>

22 years agoDon't call setvar() with the VTEXTFIXED flag. A copy is created by
tegge [Wed, 7 Nov 2001 23:00:05 +0000 (23:00 +0000)]
Don't call setvar() with the VTEXTFIXED flag.  A copy is created by
setvar() and passed to setvareq(). When the VTEXTFIXED flag is set,
that copy is never freed, causing a memory leak.

PR: 31533
Submitted by: maxim@macomnet.ru

22 years agoFix vendor ID's.
obrien [Wed, 7 Nov 2001 22:12:52 +0000 (22:12 +0000)]
Fix vendor ID's.

22 years agoAdd quirk for Nikon Coolpix 995.
kbyanc [Wed, 7 Nov 2001 21:47:49 +0000 (21:47 +0000)]
Add quirk for Nikon Coolpix 995.

Submitted by: Jos Vissers <jos@tunix.nl>

22 years agoCatch up on old release notes: fxp(4) dynamic standby mode workaround,
bmah [Wed, 7 Nov 2001 20:30:12 +0000 (20:30 +0000)]
Catch up on old release notes:  fxp(4) dynamic standby mode workaround,
keyinfo(1) in C.

MFC noted:  fxp(4) interrupt coalescing/packet bundling.

22 years agorcsid[]->__FBSDID
obrien [Wed, 7 Nov 2001 19:55:16 +0000 (19:55 +0000)]
rcsid[]->__FBSDID

22 years agoo When "-n" is used with watch, it disables the use of <control-X> to
rwatson [Wed, 7 Nov 2001 19:15:19 +0000 (19:15 +0000)]
o When "-n" is used with watch, it disables the use of <control-X> to
  change terminals being watched.  This change makes watch pass the
  <control-X> through to the terminal if it's not being intercepted--
  previously, the keypress would simply be dropped.

22 years agoInitialize first to -1; this will create a (nearly) empty file on failure,
fenner [Wed, 7 Nov 2001 17:53:25 +0000 (17:53 +0000)]
Initialize first to -1; this will create a (nearly) empty file on failure,
instead of looping until the disk is full.  This kind of failure can
especially happen when a version of awk that doesn't support POSIX
character classes is used.

Submitted by: David Wolfskill <david@catwhisker.org>

22 years agoNMBCLUSTERS, not NMBCLUSTER.
dd [Wed, 7 Nov 2001 14:36:50 +0000 (14:36 +0000)]
NMBCLUSTERS, not NMBCLUSTER.

Submitted by: markus ryynnen <markus.ryynanen@kolumbus_nospam.fi>

22 years agoFix printf format bugs introduced in rev 1.34 for printing times.
peter [Wed, 7 Nov 2001 02:51:25 +0000 (02:51 +0000)]
Fix printf format bugs introduced in rev 1.34 for printing times.
quad_t cannot be printed with %lld on 64 bit systems.

Dont waste cpu to round user and system times up to long long, it is
highly improbable that a process will have accumulated 68 years of
user or system cpu time (not wall clock time) before a reboot or
process restart.

22 years agoRemove a few more debugging bits and turn on twiddle output while reading
jhb [Wed, 7 Nov 2001 01:20:33 +0000 (01:20 +0000)]
Remove a few more debugging bits and turn on twiddle output while reading
from the CD.  This turns off the dual console output to COM1.

22 years agoUpdate the nsswitch.conf -> host.conf generator to handle criteria,
fenner [Wed, 7 Nov 2001 00:33:56 +0000 (00:33 +0000)]
Update the nsswitch.conf -> host.conf generator to handle criteria,
 continuation lines, extra whitespace, and to use the last matching
 line in the file.  This syncs the host.conf generation with how
 the nsswitch.conf is parsed.
Only print " host.conf" instead of a multi-line message, since this
 happens on every boot.

22 years agoDon't ignore unknown characters. The previous code treated a line like:
fenner [Wed, 7 Nov 2001 00:05:26 +0000 (00:05 +0000)]
Don't ignore unknown characters.  The previous code treated a line like:

hosts:!!!!!!!!@@@@@$%^&*()()*$(files{}{}|||++!)(dns

exactly the same as:

hosts: files dns

Recover from parse errors by looking for the end of a line; this
allows entries without errors to still be parsed even if there is
an erroneous entry earlier in the file.

22 years agoFix bug in scsi_read_write() where it might use 6-byte commands when
kbyanc [Tue, 6 Nov 2001 23:50:33 +0000 (23:50 +0000)]
Fix bug in scsi_read_write() where it might use 6-byte commands when
10/12-byte-specific flags where specified.

Reviewed by: ken
MFC after: 1 day

22 years agoAdd AMD766 OHCI USB controller. Remove unused VENDORID lists.
peter [Tue, 6 Nov 2001 23:44:03 +0000 (23:44 +0000)]
Add AMD766 OHCI USB controller.  Remove unused VENDORID lists.

22 years agoFixed the behavior when there is no inbound policy for the ipsec
ume [Tue, 6 Nov 2001 22:45:29 +0000 (22:45 +0000)]
Fixed the behavior when there is no inbound policy for the ipsec
tunneled packet.
When there is no suitable inbound policy for the packet of the ipsec
tunnel mode, the kernel never decapsulate the tunneled packet
as the ipsec tunnel mode even when the system wide policy is "none".
Then the kernel leaves the generic tunnel module to process this
packet.  If there is no rule of the generic tunnel, the packet
is rejected and the statistics will be counted up.

Obtained from: KAME
MFC after: 1 week

22 years ago- Change the f_isdir field of struct file to be a flags field and two
jhb [Tue, 6 Nov 2001 22:31:10 +0000 (22:31 +0000)]
- Change the f_isdir field of struct file to be a flags field and two
  new flags: F_ROOTDIR and F_RR (Rock Ridge present).
- Cache the SUSP LEN_SKP parameter in struct file as well.
- If we open() '/', then force a read of the directory's contents so we
  can examine the directory record of '.' to see if Rock Ridge is present.
- If Rock Ridge extensions are present, lookup Rock Ridge names in
  readdir().

22 years agoFixes:
olgeni [Tue, 6 Nov 2001 21:34:11 +0000 (21:34 +0000)]
Fixes:

  * least -> list
  * recommendations -> recommended

22 years agoFixes:
olgeni [Tue, 6 Nov 2001 21:32:02 +0000 (21:32 +0000)]
Fixes:

  * acceptible -> acceptable
  * partiton -> partition
  * bootdisk -> boot disk
  * harddisk -> hard disk
  * it's -> its
  * bandwith -> bandwidth
  * website -> web site

22 years agoTypo: compatability -> compatibility
olgeni [Tue, 6 Nov 2001 21:27:30 +0000 (21:27 +0000)]
Typo: compatability -> compatibility

22 years agoFixes:
olgeni [Tue, 6 Nov 2001 21:26:13 +0000 (21:26 +0000)]
Fixes:

  * compatability -> compatibility
  * El Torrito -> El Torito
  * repectively -> respectively

22 years agowebsite -> web site
olgeni [Tue, 6 Nov 2001 21:24:59 +0000 (21:24 +0000)]
website -> web site

22 years agoFix typo:
olgeni [Tue, 6 Nov 2001 21:23:28 +0000 (21:23 +0000)]
Fix typo:

  * compatability -> compatibility

22 years agoFix typos:
olgeni [Tue, 6 Nov 2001 21:21:33 +0000 (21:21 +0000)]
Fix typos:

  * initialise -> initialize
  * maintenence -> maintenance
  * manufacured -> manufactured
  * accomodate -> accommodate

22 years agoFix typos:
olgeni [Tue, 6 Nov 2001 21:17:25 +0000 (21:17 +0000)]
Fix typos:

  * intially -> initially
  * accomodate -> accommodate

22 years agoAdd emul.c and in_cksum.c, make the OpenFirmware console driver optional,
tmm [Tue, 6 Nov 2001 20:25:44 +0000 (20:25 +0000)]
Add emul.c and in_cksum.c, make the OpenFirmware console driver optional,
reorder a bit.

22 years agoAdd a special OpenFirmware entry point for terminating the kernel (in
tmm [Tue, 6 Nov 2001 20:22:18 +0000 (20:22 +0000)]
Add a special OpenFirmware entry point for terminating the kernel (in
this case, the firmware trap table needs to be restored). Make use of
it in cpu_halt() and cpu_reset(), and make cpu_reset() reboot the kernel
that was used previously insead of behaving like cpu_halt().
Add a shutdown_final event handler that turns the power off if requested.

22 years agoAdd code to emulate unimplemented (non-fp) instructions and to fixup
tmm [Tue, 6 Nov 2001 20:13:44 +0000 (20:13 +0000)]
Add code to emulate unimplemented (non-fp) instructions and to fixup
unaligned accesses, and instr.h, which contrains definitions for the
sparc64 instruction set (partly from NetBSD).
Make use of some definitions from instr.h in db_disasm.c.

22 years agoo Cache the process's struct prison so as to create a more visually
rwatson [Tue, 6 Nov 2001 20:09:33 +0000 (20:09 +0000)]
o Cache the process's struct prison so as to create a more visually
  appealing code structure.  In particular, s/req->p->p_ucred->cr_prison/pr/

Requested by: imp, jhb, jake, other hangers on

22 years agoAdd optimized implementations of in_cksum_skip() and related functions
tmm [Tue, 6 Nov 2001 20:05:01 +0000 (20:05 +0000)]
Add optimized implementations of in_cksum_skip() and related functions
for sparc64.

22 years agoFix the intial setup of the stray interrupt handler (it takes a struct
tmm [Tue, 6 Nov 2001 19:59:21 +0000 (19:59 +0000)]
Fix the intial setup of the stray interrupt handler (it takes a struct
*intr_vec as argument now, not the vector number).

22 years ago- Add a simple SUSP parser.
jhb [Tue, 6 Nov 2001 19:59:19 +0000 (19:59 +0000)]
- Add a simple SUSP parser.
- Use the SUSP parser to detect Rock Ridge (RRIP) extensions and to look
  up alternate names when opening files.

22 years agoo Remove a tab missed in the previous whitespace commit.
rwatson [Tue, 6 Nov 2001 19:58:43 +0000 (19:58 +0000)]
o Remove a tab missed in the previous whitespace commit.

22 years agoo Remove double-indentation of sysctl_kern_securelvl. This change is
rwatson [Tue, 6 Nov 2001 19:56:58 +0000 (19:56 +0000)]
o Remove double-indentation of sysctl_kern_securelvl.  This change is
  consistent with the one other function in the file, and prevents long
  lines in up-coming changes.  This nominally pulls kern_mib.c a little
  further down the long path to style(9) compliance.

22 years agoFix typo: s/dirver/driver/
bmah [Tue, 6 Nov 2001 19:48:22 +0000 (19:48 +0000)]
Fix typo:  s/dirver/driver/

22 years agoAdd missing include for <ctype.h>
green [Tue, 6 Nov 2001 19:26:51 +0000 (19:26 +0000)]
Add missing include for <ctype.h>

Make sockaddr printing code unbad.

Always copy in sizeof(sockaddr_un) bytes for an AF_UNIX sockaddr,
despite what the length may be.

22 years agoSwitch to using ANSI function declarations and add missing function
jhb [Tue, 6 Nov 2001 17:13:05 +0000 (17:13 +0000)]
Switch to using ANSI function declarations and add missing function
prototypes.  I'm tired of getting stupid bugs from changing function
parameters and not getting warnings from the compiler when I goof it up.

22 years agoCorrectly unlock the target process if /proc/$foo/mem is open()ed by
green [Tue, 6 Nov 2001 17:00:40 +0000 (17:00 +0000)]
Correctly unlock the target process if /proc/$foo/mem is open()ed by
another process which cannot p_candebug() it.  The bug was introduced
in rev. 1.100.

Approved by: des

22 years agoFix tab damage in rev 1.326.
obrien [Tue, 6 Nov 2001 16:15:47 +0000 (16:15 +0000)]
Fix tab damage in rev 1.326.

22 years agoAdd S4BIOS sleep (BIOS hibernation) and DSDT overriding support.
iwasaki [Tue, 6 Nov 2001 15:00:30 +0000 (15:00 +0000)]
Add S4BIOS sleep (BIOS hibernation) and DSDT overriding support.
 - Add S4BIOS sleep implementation.  This will works well if MIB
   hw.acpi.s4bios is set (and of course BIOS supports it and hibernation
   is enabled correctly).
 - Add DSDT overriding support which is submitted by takawata originally.
   If loader tunable acpi_dsdt_load="YES" and DSDT file is set to
   acpi_dsdt_name (default DSDT file name is /boot/acpi_dsdt.aml),
   ACPI CA core loads DSDT from given file rather than BIOS memory block.
   DSDT file can be generated by iasl in ports/devel/acpicatools/.
 - Add new files so that we can add our proposed additional code to Intel
   ACPI CA into these files temporary.  They will be removed when
   similar code is added into ACPI CA officially.

22 years agoRemove /usr/src/sbin/pc98 from srcdirs entry. It is not exist.
nyan [Tue, 6 Nov 2001 12:34:07 +0000 (12:34 +0000)]
Remove /usr/src/sbin/pc98 from srcdirs entry. It is not exist.
Add special entry for the fdisk command.

22 years agoMFi386: sys/boot/i386/loader/main.c revision 1.25
nyan [Tue, 6 Nov 2001 12:31:56 +0000 (12:31 +0000)]
MFi386: sys/boot/i386/loader/main.c revision 1.25

22 years agoMFi386: sys/boot/i386/libi386/Makefile revision 1.21
nyan [Tue, 6 Nov 2001 12:31:05 +0000 (12:31 +0000)]
MFi386: sys/boot/i386/libi386/Makefile revision 1.21

22 years agoThis file was part of something grander, which doesn't seem to be
phk [Tue, 6 Nov 2001 07:26:21 +0000 (07:26 +0000)]
This file was part of something grander, which doesn't seem to be
around anymore.

22 years agoThis file is a 2.2 vintage pre-CAM file of no current value.
phk [Tue, 6 Nov 2001 07:24:54 +0000 (07:24 +0000)]
This file is a 2.2 vintage pre-CAM file of no current value.

22 years agouse cp instead of mv to get the right name of the yacc-generated
assar [Tue, 6 Nov 2001 04:30:15 +0000 (04:30 +0000)]
use cp instead of mv to get the right name of the yacc-generated
files, mv fails badly with parallel makes

Submitted by: Vincent Poy <vince@oahu.WURLDLINK.NET>

22 years agoMore drive-by stylification, and some clarifications
des [Tue, 6 Nov 2001 03:28:21 +0000 (03:28 +0000)]
More drive-by stylification, and some clarifications

22 years agoMore drive-by stylification
des [Tue, 6 Nov 2001 02:40:09 +0000 (02:40 +0000)]
More drive-by stylification

22 years agoDrive-by stylification + reintroduce set_border().
des [Tue, 6 Nov 2001 02:38:09 +0000 (02:38 +0000)]
Drive-by stylification + reintroduce set_border().

22 years agosyscons' set_border() is now named sc_set_border(), so there is no longer
des [Tue, 6 Nov 2001 02:36:26 +0000 (02:36 +0000)]
syscons' set_border() is now named sc_set_border(), so there is no longer
a naming conflict.

22 years ago- Fixes non-zero'd out sin_zero field problem so that the padding
arr [Tue, 6 Nov 2001 00:48:01 +0000 (00:48 +0000)]
- Fixes non-zero'd out sin_zero field problem so that the padding
  is used as it is supposed to be.

Inspired by: PR #31704
Approved by: jdp
Reviewed by: jhb, -net@

22 years agoDocument the reset command. I would have thought it blatently obvious,
peter [Mon, 5 Nov 2001 23:46:10 +0000 (23:46 +0000)]
Document the reset command.  I would have thought it blatently obvious,
but here it is:  reset - Hard reset the system

22 years agoAdd quirk entry for Olympus E-100RS digital camera. This and the existing
kbyanc [Mon, 5 Nov 2001 23:33:18 +0000 (23:33 +0000)]
Add quirk entry for Olympus E-100RS digital camera. This and the existing
quirk regarding the C- series makes me suspect that all Olympus models have
the same quirks, but I cannot prove it.

Submitted by: Bernd Walter <ticso@cicely8.cicely.de>

22 years agoo No need to set values to 0 when we utilize M_ZERO
arr [Mon, 5 Nov 2001 22:27:46 +0000 (22:27 +0000)]
o No need to set values to 0 when we utilize M_ZERO

Approved by: peter

22 years agoUpdate.
phk [Mon, 5 Nov 2001 22:07:44 +0000 (22:07 +0000)]
Update.

22 years agoThis is long since unused.
phk [Mon, 5 Nov 2001 22:06:33 +0000 (22:06 +0000)]
This is long since unused.

22 years agoThis simple shell-script tries to find *.c files under /sys
phk [Mon, 5 Nov 2001 22:06:09 +0000 (22:06 +0000)]
This simple shell-script tries to find *.c files under /sys
which are not in some way used or referenced somewhere.

22 years agoGC the a.out support in DDB, nothing anywhere would pull this
phk [Mon, 5 Nov 2001 21:55:42 +0000 (21:55 +0000)]
GC the a.out support in DDB, nothing anywhere would pull this
file into a build.

22 years ago- Include machine/md_var.h to get rid of cpu_reset() warning. (-Wall)
arr [Mon, 5 Nov 2001 21:50:55 +0000 (21:50 +0000)]
- Include machine/md_var.h to get rid of cpu_reset() warning. (-Wall)

Approved by: jhb

22 years agoGC userconfig after Peter axed it 15 months ago.
phk [Mon, 5 Nov 2001 21:46:35 +0000 (21:46 +0000)]
GC userconfig after Peter axed it 15 months ago.

22 years agoAdd a note explaining why CLOCKS_PER_SEC isn't 1000000.
dwmalone [Mon, 5 Nov 2001 21:30:43 +0000 (21:30 +0000)]
Add a note explaining why CLOCKS_PER_SEC isn't 1000000.
While I'm here, "of a second" does not belong to CLOCKS_PER_SEC.

PR: 30297
Submitted by: Bernd Luevelsmeyer <bdluevel@heitec.net>

22 years ago3.5 years ago Wollman wrote:
phk [Mon, 5 Nov 2001 21:25:02 +0000 (21:25 +0000)]
3.5 years ago Wollman wrote:
   "[...] and removes the hostcache code from standard kernels---the
   code that depends on it is not going to happen any time soon,
   I'm afraid."
Time to clean up.

22 years agoAlso remove the .h files for the old rp driver.
phk [Mon, 5 Nov 2001 21:23:24 +0000 (21:23 +0000)]
Also remove the .h files for the old rp driver.

22 years agoGarbage call!
phk [Mon, 5 Nov 2001 21:10:09 +0000 (21:10 +0000)]
Garbage call!

These files have been obsoleted by compiler technology (such as it is
for gcc) for quite some time.