]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
24 years agoCatch up to the fact that block devices are toast.
Jordan K. Hubbard [Wed, 15 Dec 1999 08:33:56 +0000 (08:33 +0000)]
Catch up to the fact that block devices are toast.
Teach about the afd driver.
Teach new char dev for ad driver.
Make ownerships correct.

Submitted by: jhb

24 years agoAdd alloc/release resource. Need to connect functionality later.
Warner Losh [Wed, 15 Dec 1999 06:57:18 +0000 (06:57 +0000)]
Add alloc/release resource.  Need to connect functionality later.

24 years agoEnable INET6 by default.
Yoshinobu Inoue [Wed, 15 Dec 1999 05:13:22 +0000 (05:13 +0000)]
Enable INET6 by default.
This should be OK on non INET6 enabled kernel.

24 years agoAdd the description and example of sbc for non-PnP cards.
Seigo Tanimura [Wed, 15 Dec 1999 05:10:00 +0000 (05:10 +0000)]
Add the description and example of sbc for non-PnP cards.

Noticed by: Kentaro Inagaki <inagaki@tg.rim.or.jp>

24 years ago-Wall fixes.
Bill Fumerola [Wed, 15 Dec 1999 04:28:10 +0000 (04:28 +0000)]
-Wall fixes.

24 years agoadd human readable output (-h and -H)
Michael Haro [Wed, 15 Dec 1999 03:44:09 +0000 (03:44 +0000)]
add human readable output (-h and -H)

Obtained from: parts of human readable code from OpenBSD
Reviewed by: obrien

add POSIX, byte and megabyte block size ouput flags

PR: 13579 (POSIX flag)
Submitted by: Mike Meyer <mwm@phone.net>

24 years agoPut up some more helpful dialog boxes.
Jordan K. Hubbard [Wed, 15 Dec 1999 01:30:52 +0000 (01:30 +0000)]
Put up some more helpful dialog boxes.
Adjust some text to make more sense.

24 years agoapic_irq() returns -1 when there is no match for (IOAPIC, intpin) pair.
Tor Egge [Wed, 15 Dec 1999 01:14:56 +0000 (01:14 +0000)]
apic_irq() returns -1 when there is no match for (IOAPIC, intpin) pair.

Adjust some comments to better match the code.

24 years agoReplace spaces with a tab to fix build breakage
Chris Timmons [Wed, 15 Dec 1999 01:07:28 +0000 (01:07 +0000)]
Replace spaces with a tab to fix build breakage

24 years agoStep #2 of fixing the fixit floppy.
Joerg Wunsch [Tue, 14 Dec 1999 22:41:09 +0000 (22:41 +0000)]
Step #2 of fixing the fixit floppy.

Kill duplicates for programs that have been in the boot crunched image
as well as on the fixit floppy (pwd, newfs, hostname, test).  Our
space is really too valuable to have them around there twice.  I doubt
pwd needs to be there at all since it's a builtin into sh(1) anyway
(oh, and the same applies to test(1) IIRC), but heck, leave them by
now.

Use the new `fixit' target in MAKEDEV to create the /dev nodes on
the floppy, instead of including the kitchensink...

Finally, tune the values used for creating the floppy.  I currently
end up with

Filesystem  1K-blocks     Used    Avail Capacity iused   ifree  %iused
/dev/vnn0c       1363     1301      -47   104%     368      14    96%

...which is not quite ideal yet, but at least a working configuration
again.

24 years agoAdd NetBSD's explanation of why it would be hard to emulate the Mach
Andrew Gallatin [Tue, 14 Dec 1999 22:39:27 +0000 (22:39 +0000)]
Add NetBSD's explanation of why it would be hard to emulate the Mach
system calls.

obtained from: NetBSD

24 years agoAdd generated files
Andrew Gallatin [Tue, 14 Dec 1999 22:37:09 +0000 (22:37 +0000)]
Add generated files

24 years agoThe target `all' evoluted into a kitchen-sink these days... It's no
Joerg Wunsch [Tue, 14 Dec 1999 22:36:03 +0000 (22:36 +0000)]
The target `all' evoluted into a kitchen-sink these days...  It's no
longer really suitable as a default to create the various /dev nodes
to be contained on the fixit floppy, since all our proud new devide
nodes finally made the fixit floppy run over...

So instead create a new target titled `fixit' which creates just the
dev nodes for a single unit per each default driver; whoever needs
more of them is free to create whatever he needs, perhaps after
killing unused nodes before.  There were more than 700 /dev nodes on
the floppy before that action, and it's still around 350 now.  I doubt
all the various /dev/ugen* entries are really useful on such a tool,
so people, please check, and if you feel like more could be eliminated
from that floppy, kill'em.

While i was at it, removed traces of ft(8) that still survived even
though the driver has long since been dead.

That's step #1.  #2 will follow...

24 years agoFinally add the Alpha OSF/1 compat code. I will add it to the
Andrew Gallatin [Tue, 14 Dec 1999 22:35:36 +0000 (22:35 +0000)]
Finally add the Alpha OSF/1 compat code.  I will add it to the
sys/modules Makefile after completing a buildworld.

History:

The bulk of this code was obtained from NetBSD approximately one year
ago (I have taken care to preserve the original NetBSD copyrights and
I thank the authors for their work.) At that time, the OSF/1 code was
what was left over from their initial bootstrapping off of OSF/1 and
did not provide support for executing shared binaries.

I have independently added support for shared libraries, and support
for some of the more obscure system calls.  This code has been
available for testing and comment since January of 1999 and running on
production machines here at Duke since April.

Known working applications include:

- Netscape (all versions I've tried)
- Mathematica 3.0.2
- Splus 3.4
- ArcInfo 7.1
- Matlab (version unknown)
- SimOS
- Atom instrumented binaries (built on a real OSF/1 system)

Applications which are known not to work:

- All applications linking to libmach
- Adobe Acrobat  (uses libmach)

This has been tested with applications running against shared
libraries from OSF/1 (aka Tru64) 4.0D and 4.0F.

Reviewed by: marcel, obrien
BDE-lint by: obrien
Agreed in principal to by: msmith

24 years agoupdate to isdn4bsd beta release 0.90
Hellmuth Michaelis [Tue, 14 Dec 1999 21:14:28 +0000 (21:14 +0000)]
update to isdn4bsd beta release 0.90

24 years agoupdate to isdn4bsd beta release 0.90
Hellmuth Michaelis [Tue, 14 Dec 1999 20:54:38 +0000 (20:54 +0000)]
update to isdn4bsd beta release 0.90

24 years agoupdate to isdn4bsd beta release 0.90
Hellmuth Michaelis [Tue, 14 Dec 1999 20:52:41 +0000 (20:52 +0000)]
update to isdn4bsd beta release 0.90
drivers which are likely to be ported to newbus are commented out for now

24 years agoupdate to isdn4bsd beta release 0.90
Hellmuth Michaelis [Tue, 14 Dec 1999 20:48:35 +0000 (20:48 +0000)]
update to isdn4bsd beta release 0.90

24 years agoMake setjmp, longjmp, sigsetjmp, and siglongjmp weak aliases for
Jason Evans [Tue, 14 Dec 1999 20:17:52 +0000 (20:17 +0000)]
Make setjmp, longjmp, sigsetjmp, and siglongjmp weak aliases for
__setjmp, __longjmp, __sigsetjmp, and __siglongjmp, respectively.
This supports cancellation in the linuxthreads port.  In the long run,
a much more comprehensive solution will necessitate more dramatic changes
to libc symbol naming, and these aliases will probably need modification
at that time.

24 years ago Fix two problems: First, fix the append seek position race that can
Matthew Dillon [Tue, 14 Dec 1999 19:07:54 +0000 (19:07 +0000)]
Fix two problems:  First, fix the append seek position race that can
    occur due to np->n_size potentially changing if nfs_getcacheblk()
    blocks in nfs_write().

    Second, under -current we must supply the proper bufsize when obtaining
    buffers that straddle the EOF, but due to the fact that np->n_size can
    change out from under us it is possible that we may specify the wrong
    buffer size and wind up truncating dirty data written by another
    process.

    Both problems are solved by implementing nfs_rslock(), which allows us
    to lock around sensitive buffer cache operations such as those that
    occur when appending to a file.

    It is believed that this race is responsible for causing dirtyoff/dirtyend
    and (in stable) validoff/validend to exceed the buffer size.  Therefore
    we have now added a warning printf for the dirtyoff/end case in current.

    However, we have introduced a new problem which we need to fix at some
    point, and that is that soft or intr NFS mounts may become
    uninterruptable from the point of view of process A which is stuck waiting
    on rslock while process B is stuck doing the rpc.  To unstick process A,
    process B would have to be interrupted first.

Reviewed by: Alfred Perlstein <bright@wintelcom.net>

24 years agoStop calling everthing an XP1000.
Andrew Gallatin [Tue, 14 Dec 1999 17:50:00 +0000 (17:50 +0000)]
Stop calling everthing an XP1000.

24 years agoImprove the mapping between the hardware PCI bus numbering on multi-hose
Andrew Gallatin [Tue, 14 Dec 1999 17:35:08 +0000 (17:35 +0000)]
Improve the mapping between the hardware PCI bus numbering on multi-hose
tsunami systems and the PCI bus-numbering system of FreeBSD.  Eg, the former
allows for 2 PCI bus 2's (one each on hoses 0 and 1) while the latter
needs to give each PCI bus a unique monotonically increasing number.

It has been fairly well tested and correctly maps machines with a ppb on
hose 1 as well as machines with ppbs on both hoses.

DS10s remain untested, as I do not have a pci card with a ppb which will
pass POST in a tsunami.

This is a house of cards.

24 years agoCommented out
Nik Clayton [Tue, 14 Dec 1999 16:56:46 +0000 (16:56 +0000)]
Commented out

    MAN8+= rstat_svc.8

The file it talks about doesn't exist on FreeBSD, so there's no point in
installing the manual page.  There was already a comment to this effect in
this file, but the entry hadn't been commented out.

rstat.1 and rstat_svc.8 can probably actually be removed.

PR:             docs/13767
Submitted by:   Seth <seth@freebie.dp.ny.frb.org>

24 years agoUse SEQ_* macros for comparing sequence space numbers.
Jonathan Lemon [Tue, 14 Dec 1999 15:43:56 +0000 (15:43 +0000)]
Use SEQ_* macros for comparing sequence space numbers.

Reviewed by: truckman

24 years agoRemove x-ref to itself.
Bill Fumerola [Tue, 14 Dec 1999 11:34:47 +0000 (11:34 +0000)]
Remove x-ref to itself.

Reviewed by: mpp

24 years agoCorrect "standard compilance" notes
Alexey Zelkin [Tue, 14 Dec 1999 10:35:09 +0000 (10:35 +0000)]
Correct "standard compilance" notes

Reminded by: bde

24 years agoHave another go at the VIA support, this time use the PCI ID form the
Søren Schmidt [Tue, 14 Dec 1999 10:25:28 +0000 (10:25 +0000)]
Have another go at the VIA support, this time use the PCI ID form the
main component in the southbridge chip to determine which VIA chip
we are dealing with.

Try to enable DMA on generic controllers that say they has the
capability, instead of relying on the BIOS to have set it up.

24 years agoUse the traditional bmake magic.
Bill Fumerola [Tue, 14 Dec 1999 08:47:59 +0000 (08:47 +0000)]
Use the traditional bmake magic.

Reviewed/approved by: sos

24 years ago-Wall.
Bill Fumerola [Tue, 14 Dec 1999 08:47:42 +0000 (08:47 +0000)]
-Wall.

Reviewed/approved by: sos

24 years agoFix the command to launch DMA for 16bit samples.
Seigo Tanimura [Tue, 14 Dec 1999 06:34:48 +0000 (06:34 +0000)]
Fix the command to launch DMA for 16bit samples.

Tested by: Steve Reid <sreid@sea-to-sky.net>

24 years agoDocument `chown's move.
David E. O'Brien [Tue, 14 Dec 1999 04:54:55 +0000 (04:54 +0000)]
Document `chown's move.

24 years agoInstall `chown' into /sbin rather than /usr/sbin as `chown' is needed by
David E. O'Brien [Tue, 14 Dec 1999 04:48:17 +0000 (04:48 +0000)]
Install `chown' into /sbin rather than /usr/sbin as `chown' is needed by
`MAKEDEV'.  And one might need to `MAKEDEV' inorder to get a device node
in order to mount /usr from.

24 years agoCompletely rip-out and redesign sysinstall's refresh model as well
Jordan K. Hubbard [Tue, 14 Dec 1999 04:25:29 +0000 (04:25 +0000)]
Completely rip-out and redesign sysinstall's refresh model as well
as redoing all the menus to have proper, or at least non-hallucinogenic,
keyboard accelerators.

This requires my recent update to libdialog to work properly and will
probably also exhibit some other "interesting" behavior while the last
few missing screen clears are found (which is why I'm not going to MFC
immediately).  At least now, however, sysinstall does not gratuitously
redraw random screens at the drop of a hat and drive serial console
installers out of their minds.

24 years agoPrevent spaces from being interpreted as accelerators.
Jordan K. Hubbard [Tue, 14 Dec 1999 04:18:22 +0000 (04:18 +0000)]
Prevent spaces from being interpreted as accelerators.

24 years agoBug fix:
Brian Feldman [Tue, 14 Dec 1999 02:23:14 +0000 (02:23 +0000)]
Bug fix:
   The variables "m_mclalloc_wid" and "m_mballoc_wid" were not in the
proper place.  They should have been in uipc_mbuf.c and have been global,
not in mbuf.h and local per each file that uses mbuf.h.

Sorta bug fix:
   In mbuf.h, the definitions of various things for KERNEL and not
KERNEL cases were very screwy.   This fixes all of that which I could
find.

24 years agoMake kerberos4 conditional.
Jordan K. Hubbard [Tue, 14 Dec 1999 00:47:23 +0000 (00:47 +0000)]
Make kerberos4 conditional.

24 years agoConditionalize kerberos4 to keep sources in sync.
Jordan K. Hubbard [Tue, 14 Dec 1999 00:44:05 +0000 (00:44 +0000)]
Conditionalize kerberos4 to keep sources in sync.

24 years agoFix grammar error. (missing verb)
Bill Fumerola [Mon, 13 Dec 1999 23:37:56 +0000 (23:37 +0000)]
Fix grammar error. (missing verb)

24 years agoFix some problems reported by Mike Pritchard:
Bill Paul [Mon, 13 Dec 1999 21:45:13 +0000 (21:45 +0000)]
Fix some problems reported by Mike Pritchard:

- Add a flag DC_TX_INTR_ALWAYS which causes the transmit code to
  request a TX done interrupt for every packet. The PNIC seems to need
  this to insure that the sent TX buffers get reaped in a timely fashion.

- Try to unreset the SIA as soon as possible after resetting the whole
  chip.

- Change dcphy to support either 10/100 or 10Mbps only NICs. The
  built-in 21143 ethernet in Compaq Presario machines is 10Mbps only
  and it doesn't work right if we try to advertise 100Mbps modes during
  autoneg. When restricted to only 10mbps modes, it works fine.

  Note that for now, I detect this condition by checking the PCI
  subsystem ID on this NIC (which has a Compaq vendor/device ID).
  Yes, I know that's what the SROM is supposed to be for. I'm deliberately
  ignoring the SROM wherever possible. Sue me.

The latter two fixes allow if_dc to work correctly with the built-in
ethernet on certain Compaq Presario boxes. There are liable to be quite
a few people using these as their home systems who might want to try
FreeBSD; may as well be nice to them.

Now if anybody out there has an Alpha miata with 10Mbps ethernet and
can show me the output from pciconf -l on their system, I'd be grateful.

24 years agoAdd a `build-tools' target for make_hash and make_keys.
Marcel Moolenaar [Mon, 13 Dec 1999 21:25:08 +0000 (21:25 +0000)]
Add a `build-tools' target for make_hash and make_keys.

24 years agoAdd new option, -L that will listen the various listen queue lengths.
Guido van Rooij [Mon, 13 Dec 1999 20:08:44 +0000 (20:08 +0000)]
Add new option, -L that will listen the various listen queue lengths.

Reviewed by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu>,
Clive Lin <clive@GnatS.CirX.ORG>

24 years agoAdd the NMEA and the ONCORE driver.
Ollivier Robert [Mon, 13 Dec 1999 19:44:07 +0000 (19:44 +0000)]
Add the NMEA and the ONCORE driver.

Submitted by: wollman, phk

24 years agominor comment update
Matt Jacob [Mon, 13 Dec 1999 19:40:00 +0000 (19:40 +0000)]
minor comment update

24 years agoFix bug with internal functions when locale is active
Andrey A. Chernov [Mon, 13 Dec 1999 19:11:53 +0000 (19:11 +0000)]
Fix bug with internal functions when locale is active

PR: 15370
Reviewed by: no responce from maintainer
Obtained from: perl development sources

24 years agoHandle PPPoPTY correctly.
Brian Somers [Mon, 13 Dec 1999 18:45:21 +0000 (18:45 +0000)]
Handle PPPoPTY correctly.

24 years agoAdd notes from recent HEADS UP messages:
Warner Losh [Mon, 13 Dec 1999 17:55:23 +0000 (17:55 +0000)]
Add notes from recent HEADS UP messages:
o soren updated the ata driver.  Please use 1213 or newer if
  you have problems with the old stuff.
o Enshrine the night of bared axes against wd.  Strongly push
  users to move to ata driver.
o Alpha users need to recompile klds after 11/29.
o Mention the sound driver bridge code commit of 11/22.
o change examples in 1205 entry from wd to ad.
o Talk about new dc driver replacing all of the drivers that
  Bill Paul has ever written.[*]

Corrections to these entries welcome.

[*] Well, at least those based on tulip clones...

24 years ago Add a readahead heuristic to the NFS server side code. While the server
Matthew Dillon [Mon, 13 Dec 1999 17:34:45 +0000 (17:34 +0000)]
Add a readahead heuristic to the NFS server side code.  While the server
    cannot unilaterally pass data to a client it can reduce the physical
    disk transaction overhead by reading larger blocks.  This results in
    better pipelining of requests/responses over the network and an almost
    100% increase in cpu efficiency on the server.  On a 100BaseTX network
    NFS read performance increases from 8.5 MBytes/sec to 10 MB/sec (maxed
    out), and cpu efficiency increases from 72% idle to 80% idle on the server.

Reviewed by: Alfred Perlstein <bright@wintelcom.net>

24 years ago Forgot a log message in the last commit. Fix an mbuf leak in
Matthew Dillon [Mon, 13 Dec 1999 17:09:34 +0000 (17:09 +0000)]
Forgot a log message in the last commit.  Fix an mbuf leak in
    nfs_srvcache.c related duplicate requests on a heavily loaded
    server.  If a duplicate request is received while the previous
    request is still being processed from the cache, the previous
    mbuf chain in the cache will be replaced without first being freed.

PR: kern/15222

24 years agoPR: kern/15222
Matthew Dillon [Mon, 13 Dec 1999 17:07:03 +0000 (17:07 +0000)]
PR: kern/15222
Submitted by: Ian Dowse <iedowse@maths.tcd.ie>

24 years agoAdd simple manual page for amd driver
Alexey Zelkin [Mon, 13 Dec 1999 17:00:59 +0000 (17:00 +0000)]
Add simple manual page for amd driver

Requested by: folks on fido7.ru.unix.bsd

24 years agoAdd module for if_ef driver and make it compile.
Boris Popov [Mon, 13 Dec 1999 16:42:17 +0000 (16:42 +0000)]
Add module for if_ef driver and make it compile.

24 years agoAdd a warning about the perils of using an older version of sysinstall
Sheldon Hearn [Mon, 13 Dec 1999 16:30:55 +0000 (16:30 +0000)]
Add a warning about the perils of using an older version of sysinstall
to upgrade to a newer version of FreeBSD.

PR: 12777
Reported by: wayne@staff.msen.com (Michael R. Wayne)

24 years agoBring up an if_ef driver which allows support for four ethernet
Boris Popov [Mon, 13 Dec 1999 16:24:22 +0000 (16:24 +0000)]
Bring up an if_ef driver which allows support for four ethernet
frame types. Currently it supports only IPX protocol and doesn't
affect existing functionality when not loaded.

Reviewed by: Ollivier Robert <roberto@keltia.freenix.fr>

24 years agoAllow ifunit() routine to understand names like ed0f2. Also
Boris Popov [Mon, 13 Dec 1999 15:57:11 +0000 (15:57 +0000)]
Allow ifunit() routine to understand names like ed0f2. Also
fix a bug caused by using bcmp() instead of strcmp().

Reviewed by: Garrett Wollman <wollman@khavrinen.lcs.mit.edu>

24 years agoOops. HAVE_INET6_STRUCTS was supposed to be HAS_INET6_STRUCTS. Luckily
Peter Wemm [Mon, 13 Dec 1999 15:43:32 +0000 (15:43 +0000)]
Oops.  HAVE_INET6_STRUCTS was supposed to be HAS_INET6_STRUCTS.  Luckily
it was implicitly on because we had AF_INET6 defined.

24 years agoMerge from sys/i386/conf/files.i386 rev 1.289.
KATO Takenori [Mon, 13 Dec 1999 14:57:51 +0000 (14:57 +0000)]
Merge from sys/i386/conf/files.i386 rev 1.289.

Submitted by: yokota

24 years agoPlease welcome the brand new ntp4 daemon & utilities in the FreeBSD tree.
Ollivier Robert [Mon, 13 Dec 1999 14:09:56 +0000 (14:09 +0000)]
Please welcome the brand new ntp4 daemon & utilities in the FreeBSD tree.
This is the second part of the commit (the third -- link in usr.sbin/Makefile)
will be done after a more complete review by phk & obrien.

NOTE: the number of drivers included in the default configuration is very
minimal, mainly local clocks and the one I use RAWDCF. Anyone wanting to
have a more complete version will find recompilation very easy.

It builds and runs on both alpha & i386. It also does survive "make world".

Reviewed by: phk, obrien (partly)

24 years agoMerge from sys/dev/kbd/kbd.c rev 1.13.
KATO Takenori [Mon, 13 Dec 1999 13:19:38 +0000 (13:19 +0000)]
Merge from sys/dev/kbd/kbd.c rev 1.13.

Submitted by: yokota

24 years agoMerge from sys/dev/syscons/syscons.c rev 1.330.
KATO Takenori [Mon, 13 Dec 1999 13:17:24 +0000 (13:17 +0000)]
Merge from sys/dev/syscons/syscons.c rev 1.330.

24 years agoMerge from sys/i386/conf/Makefile.i386 rev 1.166.
KATO Takenori [Mon, 13 Dec 1999 13:15:32 +0000 (13:15 +0000)]
Merge from sys/i386/conf/Makefile.i386 rev 1.166.

24 years ago- Pull in kbd.c when sc or vt is included in the kernel, even if
Kazutaka YOKOTA [Mon, 13 Dec 1999 13:01:00 +0000 (13:01 +0000)]
- Pull in kbd.c when sc or vt is included in the kernel, even if
  no keyboard driver is defined in the kernel config file.

24 years agoDocument VGLGetXY and VGLSetXY functions.
Andrzej Bialecki [Mon, 13 Dec 1999 10:48:22 +0000 (10:48 +0000)]
Document VGLGetXY and VGLSetXY functions.

Reviewed by: sos

24 years ago- Add a module event function to the ukbd driver and make the ukbd KLD
Kazutaka YOKOTA [Mon, 13 Dec 1999 10:36:36 +0000 (10:36 +0000)]
- Add a module event function to the ukbd driver and make the ukbd KLD
  module work.
- Delete unnecessary #include.

24 years agoAdd DMA support for the SiS 5591 and old PIIX chipsets.
Søren Schmidt [Mon, 13 Dec 1999 10:19:39 +0000 (10:19 +0000)]
Add DMA support for the SiS 5591 and old PIIX chipsets.

Add a missing DELAY(1) in ata_wait.

Change the info from ad_version, so the ATA version from the disk can
be used to quantify the DAM modes valid for this drive, ie be more
selective with turning DMA on on older disks that should not support it..

Fix the probe for BIOS enabled DMA in the generic case, master/slave
was reversed in the test.

Check the return for ata_command in all cases, and print warnings if
it fails.

Call ata_dmainit with all dmamodes off when falling back to PIO mode,
that should take care of both the Promise & HPT366 controllers not
being able to handle the fallback...

Cleanup the printf's in the drivers, use the prober device name (if
possible) instead of ataN-master/slave.

24 years ago- Remember the keyboard repeat delay and rate.
Kazutaka YOKOTA [Mon, 13 Dec 1999 09:31:43 +0000 (09:31 +0000)]
- Remember the keyboard repeat delay and rate.
- Add a new ioctl, KDGETREPEAT, to retrieve the keyboard repeat rate.
- Delete unnecessary #include.

24 years agoAdd a run of Linux ldconfig.
Martin Cracauer [Mon, 13 Dec 1999 08:38:22 +0000 (08:38 +0000)]
Add a run of Linux ldconfig.

24 years agoExtra capital letter in COnsole
Chris Piazza [Mon, 13 Dec 1999 06:08:24 +0000 (06:08 +0000)]
Extra capital letter in COnsole

24 years ago Fix a timeout deadlock that can occur when the process holding the
Matthew Dillon [Mon, 13 Dec 1999 04:24:55 +0000 (04:24 +0000)]
Fix a timeout deadlock that can occur when the process holding the
    receive lock hasn't yet managed to send its own request.

PR: kern/15055
Submitted by: Ian Dowse iedowse@maths.tcd.ie

24 years agomore dma underrun fixes, using a per-channel 2nd buffer
Cameron Grant [Mon, 13 Dec 1999 03:29:09 +0000 (03:29 +0000)]
more dma underrun fixes, using a per-channel 2nd buffer

Submitted by: tanimura

24 years agoFix two problems with pipe_write():
Tor Egge [Mon, 13 Dec 1999 02:55:47 +0000 (02:55 +0000)]
Fix two problems with pipe_write():

 1. Data written beyond end of pipe buffer, causing kernel memory corruption.

    - Check that space is still valid after obtaining the pipe lock.

    - Defer the calculation of transfer size until the pipe
      lock has been obtained.

    - Update the pipe buffer pointers while holding the pipe lock.

 2. Writes of size <= PIPE_BUF not always atomic.

    - Allow an internal write to span two contiguous segments,
      so writes of size <= PIPE_BUF can be kept atomic
      when wrapping around from the end to the start of the
      pipe buffer.

PR: 15235
Reviewed by: Matt Dillon <dillon@FreeBSD.org>

24 years agoexplain that ioflags can be used to give read-ahead hints to the underlying
Alfred Perlstein [Mon, 13 Dec 1999 02:28:53 +0000 (02:28 +0000)]
explain that ioflags can be used to give read-ahead hints to the underlying
filesystem.

24 years agoThe current code incorrectly assumes that all vlans
Jordan K. Hubbard [Mon, 13 Dec 1999 01:57:00 +0000 (01:57 +0000)]
The current code incorrectly assumes that all vlans
are configured, and/or associated with a parent device.  If you
receive a frame for a VLAN that's not in the list, you walk off
the end of the list.  Boom.

Submitted by: C. Stephen Gunn <csg@waterspout.com>
PR: 15291

24 years agosys/net/if_vlan.c fails to maintain the IFF_RUNNING flag on the
Jordan K. Hubbard [Mon, 13 Dec 1999 01:38:14 +0000 (01:38 +0000)]
sys/net/if_vlan.c fails to maintain the IFF_RUNNING flag on the
vlan interfaces it manages.  This prevents the interface from
actually sending or receiving data.

Submitted by: C. Stephen Gunn <csg@waterspout.com>
PR: 15290

24 years agoKeep fortunes short, so motd won't scroll outside the screen.
Daniel C. Sobral [Mon, 13 Dec 1999 01:05:15 +0000 (01:05 +0000)]
Keep fortunes short, so motd won't scroll outside the screen.

24 years agoExercise sufficient editorial license and further shorten motd.
Daniel C. Sobral [Mon, 13 Dec 1999 00:58:53 +0000 (00:58 +0000)]
Exercise sufficient editorial license and further shorten motd.

Stolen from: jkh

24 years agoAlways set INP_IPV4 flag for IPv4 pcb entries, because netstat needs it
Yoshinobu Inoue [Mon, 13 Dec 1999 00:39:20 +0000 (00:39 +0000)]
Always set INP_IPV4 flag for IPv4 pcb entries, because netstat needs it
to print out protocol specific pcb info.

A patch submitted by guido@gvr.org, and asmodai@wxs.nl also reported
the problem.
Thanks and sorry for your troubles.

Submitted by: guido@gvr.org
Reviewed by: shin

24 years agoAdd colldef to the list of tools. It's needed on older systems.
Marcel Moolenaar [Sun, 12 Dec 1999 22:24:08 +0000 (22:24 +0000)]
Add colldef to the list of tools. It's needed on older systems.

24 years agoArgh, fix a stupid typo. :-(
Peter Wemm [Sun, 12 Dec 1999 21:36:19 +0000 (21:36 +0000)]
Argh, fix a stupid typo. :-(

24 years agoo Add gensetdefs to the list of tools.
Marcel Moolenaar [Sun, 12 Dec 1999 21:24:07 +0000 (21:24 +0000)]
o  Add gensetdefs to the list of tools.
o  Don't set CFLAGS in the bootstrap env. It is very likely to be
   overridden my any CFLAGS setting in /etc/make.conf. Setting it
   here is almost useless. So far, it doesn't seem necessary.

24 years agoMove mlx, ncp and nwfs to the common area, they build on the Alpha and
Peter Wemm [Sun, 12 Dec 1999 21:17:08 +0000 (21:17 +0000)]
Move mlx, ncp and nwfs to the common area, they build on the Alpha and
are marked cross-platform in conf/files..

24 years agoAllow it to be compiled as cross-tool.
Marcel Moolenaar [Sun, 12 Dec 1999 21:16:46 +0000 (21:16 +0000)]
Allow it to be compiled as cross-tool.

24 years agoZap unused CFLAGS += -DNCP
Peter Wemm [Sun, 12 Dec 1999 21:15:51 +0000 (21:15 +0000)]
Zap unused CFLAGS += -DNCP

24 years agoFix pointer problem for the Alpha
Peter Wemm [Sun, 12 Dec 1999 21:10:53 +0000 (21:10 +0000)]
Fix pointer problem for the Alpha

24 years agoNote that Tekram controllers are supported again.
Jordan K. Hubbard [Sun, 12 Dec 1999 21:06:44 +0000 (21:06 +0000)]
Note that Tekram controllers are supported again.

PR: 15090
Reported by: Ilmar S. Habibulin <ilmar@ints.ru>

24 years agoMove mlx from x86-only to generic. It builds on the Alpha and is in the
Peter Wemm [Sun, 12 Dec 1999 21:00:42 +0000 (21:00 +0000)]
Move mlx from x86-only to generic.  It builds on the Alpha and is in the
common conf/files for the main kernel..

24 years agotime_t is an int on the Alpha, not long.
Peter Wemm [Sun, 12 Dec 1999 20:59:29 +0000 (20:59 +0000)]
time_t is an int on the Alpha, not long.

24 years agoMove amr from x86-specific to generic. (it's in the generic conf/files
Peter Wemm [Sun, 12 Dec 1999 20:55:49 +0000 (20:55 +0000)]
Move amr from x86-specific to generic.  (it's in the generic conf/files
in the main kernel and builds fine on the alpha here...)

24 years agoDon't waste time creating amr.h since it (and NAMR) aren't used.
Peter Wemm [Sun, 12 Dec 1999 20:54:53 +0000 (20:54 +0000)]
Don't waste time creating amr.h since it (and NAMR) aren't used.

24 years agoFix a printf format problem on the Alpha, where size_t == unsigned long,
Peter Wemm [Sun, 12 Dec 1999 20:53:31 +0000 (20:53 +0000)]
Fix a printf format problem on the Alpha, where size_t == unsigned long,
not unsigned int as on the x86.

24 years agoFix joy and put it back in the MI section. (yes, it works on Alpha)
Peter Wemm [Sun, 12 Dec 1999 20:42:58 +0000 (20:42 +0000)]
Fix joy and put it back in the MI section.  (yes, it works on Alpha)

24 years agoMake this kld'able (#include "joy.h" no longer required as there are no
Peter Wemm [Sun, 12 Dec 1999 20:39:35 +0000 (20:39 +0000)]
Make this kld'able  (#include "joy.h" no longer required as there are no
references to NJOY any more after newbusification)

24 years agoChange "atapi" (the old one) from "present but broken" to "not present".
Peter Wemm [Sun, 12 Dec 1999 20:35:47 +0000 (20:35 +0000)]
Change "atapi" (the old one) from "present but broken" to "not present".

24 years agoRemove a whole bunch of "CFLAGS+= -DFSNAME" cruft. It hasn't been
Peter Wemm [Sun, 12 Dec 1999 20:34:11 +0000 (20:34 +0000)]
Remove a whole bunch of "CFLAGS+= -DFSNAME" cruft.  It hasn't been
needed for ages, but keeps getting cut/pasted into new Makefiles.
(Once apon a time it was used to activate mount arguments in
 <sys/mount.h>, but that was killed with extreme prejudice long ago)

24 years agosd->da, wd->ad
Chris Piazza [Sun, 12 Dec 1999 19:12:45 +0000 (19:12 +0000)]
sd->da, wd->ad

24 years agoLook for disk devices rather than bdevs.
Poul-Henning Kamp [Sun, 12 Dec 1999 17:43:12 +0000 (17:43 +0000)]
Look for disk devices rather than bdevs.

Approved by: grog

24 years agoBring these more into line with other modules that have .h files generated
Peter Wemm [Sun, 12 Dec 1999 16:47:33 +0000 (16:47 +0000)]
Bring these more into line with other modules that have .h files generated
on the fly.

24 years agoUse the -c and -h args of vnode_if.sh to generate only the .h file that
Peter Wemm [Sun, 12 Dec 1999 16:45:58 +0000 (16:45 +0000)]
Use the -c and -h args of vnode_if.sh to generate only the .h file that
we use.  The .c half is statically compiled into the kernel.  It's kinda
silly to generate a .h file on the fly that has inlines to call the
.c stuff when the .c code is fixed.

Also, zap the special treatment for VFS_KLD modules.  This treatment
applies to lots of things, not just VFS's.

24 years ago Make the cvs-supfile cover the entire repository - add the doc hierarchy.
Matthew Dillon [Sun, 12 Dec 1999 16:44:26 +0000 (16:44 +0000)]
Make the cvs-supfile cover the entire repository - add the doc hierarchy.

Approved by: jkh

24 years agoUse a seperate -c and -h mode. The vnode_if.c file is compiled only into
Peter Wemm [Sun, 12 Dec 1999 16:43:05 +0000 (16:43 +0000)]
Use a seperate -c and -h mode.  The vnode_if.c file is compiled only into
the kernel while the vnode_if.h header is a bunch of inlines to call the
code that is in the kernel. Generating the .h file on the fly is kinda
bogus because it has to match the one compiled into the kernel.

IMHO we should have kern/vnode_if.c and sys/vnode_if.h committed in the
tree but that's another battle.

24 years agoPut on asbestos suit and put a splcam() around the 'Mounting root from..'
Peter Wemm [Sun, 12 Dec 1999 16:34:43 +0000 (16:34 +0000)]
Put on asbestos suit and put a splcam() around the 'Mounting root from..'
message to stop it splitting.  Every single scsi machine I've seen seems
to reliably collide with this and it's rather annoying.