]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
25 years ago- When aborting an interrupt pipe, just wait for 1ms for the completion
Nick Hibma [Sun, 30 May 1999 12:49:39 +0000 (12:49 +0000)]
- When aborting an interrupt pipe, just wait for 1ms for the completion
of the current interrupt trasaction.
- Do not schedule the next interrupt transaction if the pipe is being
aborted or the last round of the interrupt transaction ended with error.

Submitted by: Kazutaka YOKOTA <yokota@zodiac.mech.utsunomiya-u.ac.jp>

25 years agoWhen the mouse is being disconnected, do the followings:
Nick Hibma [Sun, 30 May 1999 12:48:49 +0000 (12:48 +0000)]
When the mouse is being disconnected, do the followings:
- Call ums_disable() to abort the pipe.
- Do not wake up processes which has been waiting or polling for mouse
  data.  It won't be available anymore.

Submitted by: Kazutaka YOKOTA <yokota@zodiac.mech.utsunomiya-u.ac.jp>

25 years agoSimplistic pnp support. Needs more ids to be added.
Doug Rabson [Sun, 30 May 1999 11:14:39 +0000 (11:14 +0000)]
Simplistic pnp support. Needs more ids to be added.

25 years agoNo support for pnp yet.
Doug Rabson [Sun, 30 May 1999 11:12:30 +0000 (11:12 +0000)]
No support for pnp yet.

25 years agoNo support for pnp devices yet.
Doug Rabson [Sun, 30 May 1999 11:10:10 +0000 (11:10 +0000)]
No support for pnp devices yet.

25 years agoAllow up to 8 ports, 4 memory regions and two irqs and drqs.
Doug Rabson [Sun, 30 May 1999 11:05:39 +0000 (11:05 +0000)]
Allow up to 8 ports, 4 memory regions and two irqs and drqs.

25 years ago* Use ISA_DELETE_RESOURCE to cancel memory resource if driver clears maddr.
Doug Rabson [Sun, 30 May 1999 11:04:31 +0000 (11:04 +0000)]
* Use ISA_DELETE_RESOURCE to cancel memory resource if driver clears maddr.
* Don't allow new-style isapnp devices through the compat interfaces.

25 years agoDon't include isa_dma.h if we have already seen isavar.h
Doug Rabson [Sun, 30 May 1999 11:03:10 +0000 (11:03 +0000)]
Don't include isa_dma.h if we have already seen isavar.h

25 years ago* Add ivars for ISA pnp.
Doug Rabson [Sun, 30 May 1999 11:02:17 +0000 (11:02 +0000)]
* Add ivars for ISA pnp.
* Move isa_dma* declarations to isavar.h.
* Add a method ISA_DELETE_RESOURCE() to the ISA interface.
* Tidy up include protection defines.

25 years agoIn pci_alloc_resource() only check start and end to see if its a default.
Doug Rabson [Sun, 30 May 1999 10:54:31 +0000 (10:54 +0000)]
In pci_alloc_resource() only check start and end to see if its a default.

25 years agoActivate/deactivate resources by calling the method, not through the
Doug Rabson [Sun, 30 May 1999 10:50:57 +0000 (10:50 +0000)]
Activate/deactivate resources by calling the method, not through the
resource manager automatic handling of RF_ACTIVE.

25 years agoUse resource allocation apis to get ports.
Doug Rabson [Sun, 30 May 1999 10:49:25 +0000 (10:49 +0000)]
Use resource allocation apis to get ports.

25 years agoTypo: "can't boot 'boot'" should be "can't boot 'kernel'"
Doug Rabson [Sun, 30 May 1999 10:45:03 +0000 (10:45 +0000)]
Typo: "can't boot 'boot'" should be "can't boot 'kernel'"

25 years ago* Add a function devclass_create() which looks up the named devclass and
Doug Rabson [Sun, 30 May 1999 10:27:11 +0000 (10:27 +0000)]
* Add a function devclass_create() which looks up the named devclass and
  creates it if it doesn't exist.
* Rename resource_list_remove() to resource_list_delete() for consistency.

25 years agoAddendum to 1.155. Verify the existence of the object before checking
Alan Cox [Sun, 30 May 1999 01:12:19 +0000 (01:12 +0000)]
Addendum to 1.155.  Verify the existence of the object before checking
its reference count.

25 years agoWhen memory mapping a PCI device (like XFree86 does) on an XP1000, we must
Andrew Gallatin [Sat, 29 May 1999 19:47:51 +0000 (19:47 +0000)]
When memory mapping a PCI device (like XFree86 does) on an XP1000, we must
use ALPHA_PHYS_TO_K0SEG(offset) rather than just plain offet.  I've verified
that this does not break other platforms (I've tested an AlphaStation 200
and a Personal Workstation 500au with this patch).

As to why this works, well.. Its black magic as far as I know.  I obtained
this hack from Myricom, who in turn, obtained it from Compaq engineers.
Without this hack, XFree86 cannot talk to a PCI graphics card.

Reviewed by: Doug Rabson <dfr@nlsystems.com>
Obtained from:  feldy@myri.com (Bob Felderman)

25 years agoFix my DOC2k CHRDEV goof
Poul-Henning Kamp [Sat, 29 May 1999 19:36:55 +0000 (19:36 +0000)]
Fix my DOC2k CHRDEV goof

25 years agoA small change to allow automatic detection of the boot
Andrew Gallatin [Sat, 29 May 1999 19:23:20 +0000 (19:23 +0000)]
A small change to allow automatic detection of the boot
disk on Compaq XP1000's where Disk zero is dka0.0.0.6.1,
eg bus #1, slot #6.

Reviewed by: Doug Rabson <dfr@nlsystems.com>

25 years ago"Tn DNS" -> ".Tn DNS"
Nik Clayton [Sat, 29 May 1999 13:08:54 +0000 (13:08 +0000)]
"Tn DNS" -> ".Tn DNS"

PR:             docs/10474
Submitted by:   Yoshiteru Kageyama <yt-kage@cb3.so-net.ne.jp>

25 years agoFix a couple of typos related to wfd.
Nik Clayton [Sat, 29 May 1999 13:08:12 +0000 (13:08 +0000)]
Fix a couple of typos related to wfd.

PR:             docs/10624
Submitted by:   Joe Abley <jabley@buddha.clear.net.nz>

25 years agoFix typo which causes an unnecessary line break.
Nik Clayton [Sat, 29 May 1999 13:02:00 +0000 (13:02 +0000)]
Fix typo which causes an unnecessary line break.

PR:             docs/11561
Submitted by:   Norihiro Kumagai <kuma@nk.rim.or.jp>

25 years agoDescription of LOCKLEAF said "inode" when it should have said "vnode".
Nik Clayton [Sat, 29 May 1999 13:00:40 +0000 (13:00 +0000)]
Description of LOCKLEAF said "inode" when it should have said "vnode".

PR:             docs/11596
Submitted by:   Robert Watson <robert@fledge.watson.org>

25 years agoNote that using the dirent.d_type member is non-portable, and might
Nik Clayton [Sat, 29 May 1999 12:59:51 +0000 (12:59 +0000)]
Note that using the dirent.d_type member is non-portable, and might
fail on some filesystems.

PR:             docs/11645
Submitted by:   Harold Gutch <logix@foobar.franken.de>

25 years agoFix the typos, and add new entries (like the /etc/periodic/ and /etc/mail/
Nik Clayton [Sat, 29 May 1999 12:58:44 +0000 (12:58 +0000)]
Fix the typos, and add new entries (like the /etc/periodic/ and /etc/mail/
directories).

PR:             docs/11695
Submitted by:   Chris Costello <chris@calldei.com>

25 years ago$Header$ -> $Id$
Kris Kennaway [Sat, 29 May 1999 08:16:34 +0000 (08:16 +0000)]
$Header$ -> $Id$
Manpage formatting/grammatical fixes.
Check return value of malloc()
Add sccsid[], other minor source code cleanups

Submitted by: Philippe Charnier

25 years agoXref mountd in text.
Kris Kennaway [Sat, 29 May 1999 08:14:00 +0000 (08:14 +0000)]
Xref mountd in text.

Submitted by: Philippe Charnier

25 years agoManpage cleanup, move $Id$ to #ifndef lint, remove unused includes,
Kris Kennaway [Sat, 29 May 1999 08:12:38 +0000 (08:12 +0000)]
Manpage cleanup, move $Id$ to #ifndef lint, remove unused includes,
grammatical fixes.

Submitted by: Philippe Charnier

25 years agoStop the TSC from being used as timecounter on K5/step0 machines.
Poul-Henning Kamp [Sat, 29 May 1999 06:57:55 +0000 (06:57 +0000)]
Stop the TSC from being used as timecounter on K5/step0 machines.

25 years agoDon't put CVS files in ports. It's kinda useful for cvs updating
Jordan K. Hubbard [Sat, 29 May 1999 00:42:05 +0000 (00:42 +0000)]
Don't put CVS files in ports.  It's kinda useful for cvs updating
a ports tree which was installed initially with the system later,
but this is probably not the general case (user CVSups the repository
rather than the checked-out bits) and it's penalizing everyone else
with excessive inode consumption.

25 years agoFix support for the PNIC II. Earlier I had assumed that the PNIC II was
Bill Paul [Fri, 28 May 1999 18:43:25 +0000 (18:43 +0000)]
Fix support for the PNIC II. Earlier I had assumed that the PNIC II was
similar to the PNIC I (supported by the pn driver). In fact, it's really
a Macronix 98715A with wake on LAN support added. According to LinkSys,
the PNIC II was jointly developed by Lite-On and Macronis. I get the
feeling Macronix did most of the work. (The datasheet has the Macronix
logo on it, and is in fact nearly identical to the 98715 datasheet, except
for the extra wake on LAN registers.) In any case, the PNIC II works just
fine with the Macronix driver.

The changes are:

- Move PCI ID for the PNIC II from the pn driver to the mx driver.
- Mention PNIC II support in mx.4.
- Mention PNIC II support in RELNOTES.TXT and HARDWARE.TXT.

25 years agoFixed glitches (jumps) of about 1/HZ seconds for the i8254 timecounter.
Bruce Evans [Fri, 28 May 1999 14:08:59 +0000 (14:08 +0000)]
Fixed glitches (jumps) of about 1/HZ seconds for the i8254 timecounter.
The old version only worked right when the time was read strictly
more often than every 1/HZ seconds, but we only guarantee reading
it every (1/HZ + epsilon) seconds.  Part of rev.1.126-1.127 attempted
to fix this but didn't succeed.  Detect counter rollover using the
heuristic from the old version of microtime() with additional
complications for supporting calls from fast interrupt handlers.
This works provided i8254 interrupts are not delayed by more than
1/(2*HZ) seconds.

This needs more comments, and cleanups for the SMP case, and more
testing of the SMP case before it is merged into RELENG_3.

Tested by: jhay

25 years agoFixes from OpenBSD:
Kris Kennaway [Fri, 28 May 1999 13:23:27 +0000 (13:23 +0000)]
Fixes from OpenBSD:

Add $Id$
Removes temp file exploits in gzexe (predictable filenames)
Fixes some typos
Fixes a buffer overrun in gzip -S
Don't prepend /usr/local/bin to path in scripts
Correct location of /usr/share/misc/termcap

Obtained from: OpenBSD

25 years agoDon't call disable_intr() when interrupts are already disabled, since
Bruce Evans [Fri, 28 May 1999 13:23:21 +0000 (13:23 +0000)]
Don't call disable_intr() when interrupts are already disabled, since
disable_intr() does non-recursive locking in the SMP case.  This should
fix cy-driver-related panics when SMP is configured.

Broken in: rev.1.73 (3.1 and -current)

25 years agoUse .Dq instead of ``'' in manpage
Kris Kennaway [Fri, 28 May 1999 12:47:31 +0000 (12:47 +0000)]
Use .Dq instead of ``'' in manpage
Use optimal blocksize for rm -P, instead of always using 8192-byte blocks
to overwrite the file.

Obtained from: OpenBSD

25 years agoSet FTP_PASSIVE_MODE=YES by default in the default login class.
Jordan K. Hubbard [Fri, 28 May 1999 11:07:16 +0000 (11:07 +0000)]
Set FTP_PASSIVE_MODE=YES by default in the default login class.

25 years agoAdd new Bt848/Bt878 driver options.
Roger Hardiman [Fri, 28 May 1999 10:27:22 +0000 (10:27 +0000)]
Add new Bt848/Bt878 driver options.
(Eventually I expect to move these into the man page)

25 years agoremote the bt848/bt878 comments
Roger Hardiman [Fri, 28 May 1999 09:55:59 +0000 (09:55 +0000)]
remote the bt848/bt878 comments

Suggested by: Bruce

25 years agoDocument device_add_child_ordered().
Doug Rabson [Fri, 28 May 1999 09:38:25 +0000 (09:38 +0000)]
Document device_add_child_ordered().

25 years agoAllow a sio port with its flags set to DEBUGGER+LLCONSOLE to be used for
Doug Rabson [Fri, 28 May 1999 09:37:11 +0000 (09:37 +0000)]
Allow a sio port with its flags set to DEBUGGER+LLCONSOLE to be used for
gdb. This allows the sio probe for that port to be disabled which stops
it from confusing the debugger.

25 years ago* Change device_add_child_after() to device_add_child_ordered() which is
Doug Rabson [Fri, 28 May 1999 09:25:16 +0000 (09:25 +0000)]
* Change device_add_child_after() to device_add_child_ordered() which is
  easier to use and more flexible.
* Change BUS_ADD_CHILD to take an order argument instead of a place.
* Define a partial ordering for isa devices so that sensitive devices are
  probed before non-sensitive ones.

25 years agoDon't forget to free the mbufs that get processed by
Brian Somers [Fri, 28 May 1999 08:03:24 +0000 (08:03 +0000)]
Don't forget to free the mbufs that get processed by
mp_Assemble().
Leak spotted by: louqi

25 years agoAlways output a linefeed when we've decided whether to boot
Brian Somers [Fri, 28 May 1999 08:01:52 +0000 (08:01 +0000)]
Always output a linefeed when we've decided whether to boot
or display a loader prompt; either we've said
"Booting [%s]..." or we've received a non-line-feed character
and need a '\n' anyway.

25 years agopmap_object_init_pt:
Alan Cox [Fri, 28 May 1999 05:38:56 +0000 (05:38 +0000)]
pmap_object_init_pt:
The size of vm_object::memq is vm_object::resident_page_count,
not vm_object::size.

25 years agofix End key name
Andrey A. Chernov [Fri, 28 May 1999 04:52:29 +0000 (04:52 +0000)]
fix End key name

25 years agodon't include Makefile.inc twice
Andrey A. Chernov [Fri, 28 May 1999 04:43:35 +0000 (04:43 +0000)]
don't include Makefile.inc twice

25 years agoAdd description on configuration options and the EXAMPLE section.
Kazutaka YOKOTA [Fri, 28 May 1999 04:33:20 +0000 (04:33 +0000)]
Add description on configuration options and the EXAMPLE section.

25 years agoadd libhistory
Andrey A. Chernov [Fri, 28 May 1999 04:30:05 +0000 (04:30 +0000)]
add libhistory

25 years agoupgrade to 4.0
Andrey A. Chernov [Fri, 28 May 1999 04:26:18 +0000 (04:26 +0000)]
upgrade to 4.0

25 years agoAvoid the creation of unnecessary shadow objects.
Alan Cox [Fri, 28 May 1999 03:39:44 +0000 (03:39 +0000)]
Avoid the creation of unnecessary shadow objects.

25 years agodescribe process
Andrey A. Chernov [Fri, 28 May 1999 02:20:07 +0000 (02:20 +0000)]
describe process

25 years agomerge
Andrey A. Chernov [Fri, 28 May 1999 02:17:15 +0000 (02:17 +0000)]
merge

25 years agoAdd missing -s option.
Guy Helmer [Fri, 28 May 1999 02:09:46 +0000 (02:09 +0000)]
Add missing -s option.

PR: docs/10378

25 years agoThis commit was generated by cvs2svn to compensate for changes in r47558,
Andrey A. Chernov [Fri, 28 May 1999 01:59:23 +0000 (01:59 +0000)]
This commit was generated by cvs2svn to compensate for changes in r47558,
which included commits to RCS files with non-trunk default branches.

25 years agoreadline v4.0
Andrey A. Chernov [Fri, 28 May 1999 01:59:23 +0000 (01:59 +0000)]
readline v4.0

25 years agoWhere the fstab page discusses filesystem options, add a reference
Guy Helmer [Fri, 28 May 1999 01:54:38 +0000 (01:54 +0000)]
Where the fstab page discusses filesystem options, add a reference
to the '-o' flag in the mount(8) and filesystem specific
mount pages where information can be obtained about the available
options.

PR: docs/10108

25 years agoAdd the missing cd9660-specific options to the -o flag. While I'm here,
Guy Helmer [Fri, 28 May 1999 01:16:14 +0000 (01:16 +0000)]
Add the missing cd9660-specific options to the -o flag.  While I'm here,
add the missing -j flag to the synopsys.

PR: docs/1667

25 years agoA forgotten bit of my ST6600 commit
Andrew Gallatin [Thu, 27 May 1999 22:03:31 +0000 (22:03 +0000)]
A forgotten bit of my ST6600 commit

25 years agoHandle unterminated quoted strings without dup'ing the last character
Brian Somers [Thu, 27 May 1999 16:51:47 +0000 (16:51 +0000)]
Handle unterminated quoted strings without dup'ing the last character
with an out-by-one-memcpy.

25 years agoIn tunclose():
Brian Somers [Thu, 27 May 1999 13:18:28 +0000 (13:18 +0000)]
In tunclose():
  Delete all routes if IFF_RUNNING is set, irrespective of IFF_UP
  Unset IFF_RUNNING.

25 years agoFixed usage message to almost match the man page. The old -d option and
Bruce Evans [Thu, 27 May 1999 12:58:41 +0000 (12:58 +0000)]
Fixed usage message to almost match the man page.  The old -d option and
the new -i option were missing.

Fixed style bugs in previous commit:
(1) initialisation of a local variable in its declaration.
(2) inconsistency of (1) with style of nearby code.
(3) disorder of declaration for (1).
(4) a line longer than 80 characters.
(5) bitrot in the printf() -> err() cleanups to help bloat the line in (4).

25 years agoPrint some diagnostic messages for the PCI-ISA bridge,
Kazutaka YOKOTA [Thu, 27 May 1999 12:26:50 +0000 (12:26 +0000)]
Print some diagnostic messages for the PCI-ISA bridge,
if bootverbose > 0.

Reviewed by: dfr

25 years agoAdded net.inet.tcp.path_mtu_discovery variable which when set to 0
David Greenman [Thu, 27 May 1999 12:24:21 +0000 (12:24 +0000)]
Added net.inet.tcp.path_mtu_discovery variable which when set to 0
(default 1) disables PMTUD globally. Although PMTUD can be disabled in
the standard case by locking the MTU on a static route (including the
default route), this method doesn't work in the face of dynamic routing
protocols like gated.

25 years agoMade net.inet.ip.intr_queue_maxlen writeable.
David Greenman [Thu, 27 May 1999 12:20:33 +0000 (12:20 +0000)]
Made net.inet.ip.intr_queue_maxlen writeable.

25 years agoMove BX PCI-PCI (AGP) bridge from chip_match to pcib_match
Roger Hardiman [Thu, 27 May 1999 10:59:40 +0000 (10:59 +0000)]
Move BX PCI-PCI (AGP) bridge from chip_match to pcib_match
so it is reported correctly in dmesg.

Add 440 LX and 440 BX to the descriptions as these are the names
in common usage

25 years agoDo a clean-up pass on error/warning messages.
Jordan K. Hubbard [Thu, 27 May 1999 10:32:50 +0000 (10:32 +0000)]
Do a clean-up pass on error/warning messages.

25 years agoAdd some comments, optimise out a few strlen()s and
Brian Somers [Thu, 27 May 1999 09:50:10 +0000 (09:50 +0000)]
Add some comments, optimise out a few strlen()s and
remove a few extraneous assignments.

25 years agoCall tty_Offline() from tty_Cooked() and tty_Free(), just in
Brian Somers [Thu, 27 May 1999 08:42:49 +0000 (08:42 +0000)]
Call tty_Offline() from tty_Cooked() and tty_Free(), just in
case ppp is abending and hasn't called physical_Offline()
already.

25 years agoMove the code for tweaking interface flags into one function.
Brian Somers [Thu, 27 May 1999 08:42:17 +0000 (08:42 +0000)]
Move the code for tweaking interface flags into one function.

25 years agoFix an embarrasing typo in device_add_child_after(). I can't understand
Doug Rabson [Thu, 27 May 1999 07:18:41 +0000 (07:18 +0000)]
Fix an embarrasing typo in device_add_child_after(). I can't understand
how this hasn't caused problems before.

Submitted by: Kazutaka YOKOTA <yokota@zodiac.mech.utsunomiya-u.ac.jp>

25 years agoAdd support for ASound Gold card using the ALS120
Roger Hardiman [Thu, 27 May 1999 06:12:40 +0000 (06:12 +0000)]
Add support for ASound Gold card using the ALS120
chipset (a vibra 16x clone)
Identified automatically by its PnP ID

Approved by: Luigi

25 years agoThere are six bit fields in the flags now, not four.
Guy Helmer [Thu, 27 May 1999 02:41:06 +0000 (02:41 +0000)]
There are six bit fields in the flags now, not four.

PR: i386/10328

25 years agoAdd support for the DEC_ST6600 family of alphas. This includes the
Andrew Gallatin [Wed, 26 May 1999 23:22:14 +0000 (23:22 +0000)]
Add support for the DEC_ST6600 family of alphas.  This includes the
Compaq XP1000, AlphaServer DS20, AlphaServer DS10, and DP264

This has been tested *only* on XP1000's.  I'll be interested to hear from
owners of other types of DEC_ST6600 alphas.

I'd like to thank Don Rice of Compaq for providing the documentation required
to support this platform on FreeBSD.  I'd also like to thank Doug Rabson for newbus,
and for helping me get a multiple hoses working with newbus.

Reviewed by: Doug Rabson <dfr@nlsystems.com>

25 years agoRemove redundant redefinition of vtophys(). This is already in if_pnreg.h
Andrew Gallatin [Wed, 26 May 1999 23:08:04 +0000 (23:08 +0000)]
Remove redundant redefinition of vtophys().  This is already in if_pnreg.h

25 years agoForgotten in previous commit:
Andrew Gallatin [Wed, 26 May 1999 23:05:23 +0000 (23:05 +0000)]
Forgotten in previous commit:
Allow chipset drivers to specify the direct-mapped DMA window's mask in
preparation for tsunami support.  Previous chipsets' direct-mapped DMA
mask was always 1024*1024*1024.  The Tsunami chipset needs it to be
2*1024*1024*1024

Reviewed by: Doug Rabson <dfr@nlsystems.com>

25 years agoAllow chipset drivers to specify the direct-mapped DMA window's mask in
Andrew Gallatin [Wed, 26 May 1999 23:01:57 +0000 (23:01 +0000)]
Allow chipset drivers to specify the direct-mapped DMA window's mask in
preparation for tsunami support.  Previous chipsets' direct-mapped DMA
mask was always 1024*1024*1024.  The Tsunami chipset needs it to be
2*1024*1024*1024

These changes should not affect the i386 port

Reviewed by: Doug Rabson <dfr@nlsystems.com>

25 years agoTwo small tweaks:
Bill Paul [Wed, 26 May 1999 22:56:22 +0000 (22:56 +0000)]
Two small tweaks:

- Clear the IFF_OACTIVE flag when al_txeof() runs down the last TX mbuf chain.
- Mark the workaround for the transmitter stalling bug with
  #ifdef AL_TX_STALL_WAR/#endif.

25 years agoIgnore subvendor and subdevice ids for all ai7880 parts. These chips
Justin T. Gibbs [Wed, 26 May 1999 16:59:17 +0000 (16:59 +0000)]
Ignore subvendor and subdevice ids for all ai7880 parts.  These chips
only support 'mirroring' the vendor and device ids, so we don't
lose any information.  Certain revisions of the aic7880 will not
perform the mirroring so to match all possiblities would double
the number of table entries.  This change also allows us to match
things like the 2944B which I missed in the original table.

25 years agoUpdated options for the Bt848/Bt878 driver
Roger Hardiman [Wed, 26 May 1999 13:24:35 +0000 (13:24 +0000)]
Updated options for the Bt848/Bt878 driver
This includes the BKTR_430_FX_MODE and BKTR_SIS_VIA_MODE options which make
Bt878/879 cards work better on 430FX and old SiS/VIA/OPTi boards

25 years agoIdentify Intel 440 LX chipset motherboards
Roger Hardiman [Wed, 26 May 1999 13:14:24 +0000 (13:14 +0000)]
Identify Intel 440 LX chipset motherboards

25 years agoAdd TDC 3800 to quirk list for fixed mode device.
Matt Jacob [Tue, 25 May 1999 23:10:54 +0000 (23:10 +0000)]
Add TDC 3800 to quirk list for fixed mode device.
PR:11866
Submitted by: Helge Oldach hmo@sep.hamburg.com

25 years agoRun xpt_action at splsoftcam() so that it cannot be re-entered by our SWI.
Justin T. Gibbs [Tue, 25 May 1999 20:17:29 +0000 (20:17 +0000)]
Run xpt_action at splsoftcam() so that it cannot be re-entered by our SWI.
The XPT doesn't have a problem with this itself, but some controllers
drivers may have been caught off guard by the old behavior.

XPT_CONT_TARGET_IO is also a valid ccb type for cam_periph_unmapmem.

25 years agoTurn the 'ahc dump eeprom' flag into a real option.
Justin T. Gibbs [Tue, 25 May 1999 20:15:54 +0000 (20:15 +0000)]
Turn the 'ahc dump eeprom' flag into a real option.

25 years agoTurn the 'dump eeprom' flag into a real option.
Justin T. Gibbs [Tue, 25 May 1999 20:15:41 +0000 (20:15 +0000)]
Turn the 'dump eeprom' flag into a real option.

25 years agoThe 1542 cards do not allow adapter commands to be queued while mailbox
Justin T. Gibbs [Tue, 25 May 1999 20:15:20 +0000 (20:15 +0000)]
The 1542 cards do not allow adapter commands to be queued while mailbox
commands are outstanding.  You'd think they'd just clear the IDLE bit,
but alas, no.  Delay until all pending mailbox commands have completed
in aha_cmd to work around this.

Report sync rates correctly on Fast Adaptec cards.  Clones may still be
reported incorrectly since there is no documenation on how they report
extended sync values.

Clean up some unused fields in the aha softc.

25 years agoAll cards using aic789X chips use the new eeprom format.
Justin T. Gibbs [Tue, 25 May 1999 20:12:32 +0000 (20:12 +0000)]
All cards using aic789X chips use the new eeprom format.
Corrects bogus negotiation values on aic7890 based controllers.

PR: 11872

25 years agoCompaq "Intelligent Disk Array" driver.
Julian Elischer [Tue, 25 May 1999 19:45:17 +0000 (19:45 +0000)]
Compaq "Intelligent Disk Array" driver.
Mark Dawson holds teh copyright on this and has releases from
Compaq to allow him to do so..

Not functional in 4.0 yet but being checked in to allow the functional
3.x version to be branched at this point.

25 years agoFrom our i386 counterpart:
Dmitrij Tejblum [Tue, 25 May 1999 19:32:53 +0000 (19:32 +0000)]
From our i386 counterpart:
|revision 1.13
|date: 1995/09/15 23:49:23;  author: davidg;  state: Exp;  lines: +15 -2
|Check for page being resident when doing I/O with /dev/kmem and return
|EFAULT if it is not resident. This prevents the system from manufacturing
|a zero-fill page for unused but allocated areas of the kernel's VM.

25 years agoSpell vp0 correctly. Add id.
Tim Vanderhoek [Tue, 25 May 1999 17:34:07 +0000 (17:34 +0000)]
Spell vp0 correctly.  Add id.

PR: docs/10500 (Masaru Tsunoda [3]masaru@pokemon.to)

25 years agoFix the read/write permission checks in cam_periph_mapmem(). If we're
Kenneth D. Merry [Tue, 25 May 1999 17:10:04 +0000 (17:10 +0000)]
Fix the read/write permission checks in cam_periph_mapmem().  If we're
writing, we want to be able to read the buffer.  If we're reading, we want
to be able to write to the buffer.

PR: kern/11870
Submitted by: Andrew Mobbs <amobbs@allstor-sw.co.uk>

25 years agoAdd OPTi 82c822 host to PCI bridge
Roger Hardiman [Tue, 25 May 1999 15:56:10 +0000 (15:56 +0000)]
Add OPTi 82c822 host to PCI bridge

This is an old OPTi chipset.
If you use a Bt878 card with this chipset, be sure to enable
the SIS/VIA chipset compatiblity mode workaround.

Tested By: Ben Laurie <ben@algroup.co.uk>

25 years agoHandle ".if ${T} > 1 || ${T} < 3" and friends correctly.
Tim Vanderhoek [Tue, 25 May 1999 13:45:08 +0000 (13:45 +0000)]
Handle ".if ${T} > 1 || ${T} < 3" and friends correctly.

Reported-by: asami

25 years agoRename 849 to 849A (the correct name for the IC)
Roger Hardiman [Tue, 25 May 1999 12:44:40 +0000 (12:44 +0000)]
Rename 849 to 849A  (the correct name for the IC)

25 years agoAdd support for the Bt878/Bt879's Intel 430 FX and
Roger Hardiman [Tue, 25 May 1999 12:43:40 +0000 (12:43 +0000)]
Add support for the Bt878/Bt879's Intel 430 FX and
SIS/VIA/ OPTi chipset PCI bus workarounds.

These make the Bt878/879 chips stabler on certain
older and non-intel motherboards.

Use options BKTR_430_FX_MODE
or  options BKTR_SIS_VIA_MODE
to enable these modes.

Also rename 849 to 849A

25 years agoTypo.
Pierre Beyssac [Tue, 25 May 1999 11:52:30 +0000 (11:52 +0000)]
Typo.

PR: docs/11832
Submitted by: Ralf S. Engelschall <rse@engelschall.com>

25 years agoAdd a -i localname switch (restricted to root) to specify a different
Peter Wemm [Tue, 25 May 1999 11:14:33 +0000 (11:14 +0000)]
Add a -i localname switch (restricted to root) to specify a different
local "login" name for rcmd(3).  This is particularly useful for things
like portslave and other packages with terminal server functionality
where a login can either run ppp locally or get shunted off to another
box via rlogin depending on radius authentication etc.  Quite often the
local box doesn't even have accounts, so a flag such as this is needed.
Obviously this is restricted to callers with uid == 0.

25 years agoMerge with sys/i386/isa/wd.c revision 1.192 and 1.193.
KATO Takenori [Tue, 25 May 1999 10:41:14 +0000 (10:41 +0000)]
Merge with sys/i386/isa/wd.c revision 1.192 and 1.193.

Submitted by: Takahashi Yoshihiro <nyan@wyvern.cc.kogakuin.ac.jp>

25 years agoSync with sys/i386/i386/userconfig.c revision 1.145.
KATO Takenori [Tue, 25 May 1999 10:32:10 +0000 (10:32 +0000)]
Sync with sys/i386/i386/userconfig.c revision 1.145.

25 years agoSync with sys/i386/conf/options.i386 revision 1.114.
KATO Takenori [Tue, 25 May 1999 10:30:58 +0000 (10:30 +0000)]
Sync with sys/i386/conf/options.i386 revision 1.114.

25 years agoBack the last change out.
Brian Somers [Tue, 25 May 1999 10:30:52 +0000 (10:30 +0000)]
Back the last change out.

It upsets the calculations in sys.c (switch(maj)).

25 years agoAllow booting from "ad"
Brian Somers [Tue, 25 May 1999 09:32:00 +0000 (09:32 +0000)]
Allow booting from "ad"
Reviewed by: Søren