]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
25 years agoWhen selecting the correct EEPROM offset to use for probing the station
wpaul [Tue, 26 Jan 1999 15:18:26 +0000 (15:18 +0000)]
When selecting the correct EEPROM offset to use for probing the station
address, account for cards which report the Texas Instruments PCI vendor ID
in addition to Compaq and Olicom. (I don't actually have a card that
reports the TI vendor/device ID, but it appears that some Racore adapters
work this way, and failing to account for it when we have the ID listed
in the supported devices list is a bug.)

25 years ago1. Adjust fs sizes to get floppies back under control.
jkh [Tue, 26 Jan 1999 15:14:11 +0000 (15:14 +0000)]
1. Adjust fs sizes to get floppies back under control.

2. Viciously slash all CD support out of boot.flp.  It's basically just
   a net boot floppy now.

25 years agoNO_LKM is deprecated
jkh [Tue, 26 Jan 1999 12:59:09 +0000 (12:59 +0000)]
NO_LKM is deprecated

25 years agos/wcd/acd/
jkh [Tue, 26 Jan 1999 10:18:52 +0000 (10:18 +0000)]
s/wcd/acd/

Noted by: des

25 years agoAllow /etc/rc.conf.site as well as /etc/rc.conf.local (and add rc_conf_files
jkh [Tue, 26 Jan 1999 10:15:57 +0000 (10:15 +0000)]
Allow /etc/rc.conf.site as well as /etc/rc.conf.local (and add rc_conf_files
variable for frobbing).

25 years agoAdded descriptions on new flags introduced in psm.c rev.1.57.
yokota [Tue, 26 Jan 1999 10:01:01 +0000 (10:01 +0000)]
Added descriptions on new flags introduced in psm.c rev.1.57.

25 years agoOops, one line was accidentally commented out in the previous commit.
yokota [Tue, 26 Jan 1999 10:00:02 +0000 (10:00 +0000)]
Oops, one line was accidentally commented out in the previous commit.

25 years agoPull down the splash screen when someone is about to read from the
yokota [Tue, 26 Jan 1999 09:58:37 +0000 (09:58 +0000)]
Pull down the splash screen when someone is about to read from the
keyboard.  Do this in scread(), rather than in scopen().

25 years agoFrob the upgrade target to be a bit more inclusive. This appears to
jkh [Tue, 26 Jan 1999 09:12:21 +0000 (09:12 +0000)]
Frob the upgrade target to be a bit more inclusive.  This appears to
be a good bit more successful at doing automated upgrades than the
previous aout-to-elf target, anyway.  I'll MF3 in a couple of days
depending on how this does.

25 years ago Ok, people didn't like kern.conf_dir. Poof, backed out.
dillon [Tue, 26 Jan 1999 07:37:11 +0000 (07:37 +0000)]
Ok, people didn't like kern.conf_dir.  Poof, backed out.

25 years ago Remove use of kern.conf_dir sysctl. conf_dir is left as a local
dillon [Tue, 26 Jan 1999 07:36:20 +0000 (07:36 +0000)]
Remove use of kern.conf_dir sysctl.  conf_dir is left as a local
    variable and propogates back to /etc/rc where it will be used to
    locate the rc.local file.  The local variable will also be used by
    /etc/rc.conf.  Note that /etc/rc.conf reverts to its prior operation
    of accessing /etc/rc.conf.local if run standalone.

25 years ago Remove sysctl's from rc.conf, there seems to be a concensus that
dillon [Tue, 26 Jan 1999 07:32:59 +0000 (07:32 +0000)]
Remove sysctl's from rc.conf, there seems to be a concensus that
    rc.conf should only contain simple ops.  We still keep the conf_dir
    override, however, and this will be used when rc.conf is run from
    /etc/rc in a diskless configuration.

25 years agoRipped out EDITOR=ee with extreme prejudice.
dg [Tue, 26 Jan 1999 07:28:31 +0000 (07:28 +0000)]
Ripped out EDITOR=ee with extreme prejudice.

25 years agoMove reading of rc.conf sooner as requested by Greg. I'm a tad nervous
peter [Tue, 26 Jan 1999 05:14:36 +0000 (05:14 +0000)]
Move reading of rc.conf sooner as requested by Greg.  I'm a tad nervous
about this becase that makes it get run *before* the filesystems are
mounted.  If people have added stuff to their rc.conf or rc.conf.local
that uses stuff outside of /bin and /sbin, this will break.

25 years agoThe vinum setup tool automatically loads the vinum module if it's needed,
peter [Tue, 26 Jan 1999 04:59:43 +0000 (04:59 +0000)]
The vinum setup tool automatically loads the vinum module if it's needed,
there is no need to duplicate the logic here.

XXX rc.conf needs to be read sooner..

25 years agoCheck if the intpm controller is configured first before stopping
peter [Tue, 26 Jan 1999 04:54:38 +0000 (04:54 +0000)]
Check if the intpm controller is configured first before stopping
recognition of the 82371AB device.

25 years agoAutomatically load the vn module if it isn't already in the kernel.
peter [Tue, 26 Jan 1999 04:53:09 +0000 (04:53 +0000)]
Automatically load the vn module if it isn't already in the kernel.

25 years ago1. Properly chflags libraries before moving (otherwise they don't).
jkh [Tue, 26 Jan 1999 04:13:03 +0000 (04:13 +0000)]
1. Properly chflags libraries before moving (otherwise they don't).
2. Add my helper script for upgrades; not enabled yet.

25 years agoAdd support for the USR3031 PNP modem. Also fix a minor typo since I was
steve [Tue, 26 Jan 1999 04:02:54 +0000 (04:02 +0000)]
Add support for the USR3031 PNP modem.  Also fix a minor typo since I was
here.

25 years agoMove bsd.port.*mk to under ${PORTSDIR}/Mk (already repository copied).
asami [Tue, 26 Jan 1999 03:40:12 +0000 (03:40 +0000)]
Move bsd.port.*mk to under ${PORTSDIR}/Mk (already repository copied).
There are only skeletons left here; they merely serve as a backup to
include the real versions under ${PORTSDIR}/Mk while we update the ports
tree to include them directly.

25 years ago o Add info about Julian's Linux Threads checkin (one of these
imp [Tue, 26 Jan 1999 03:13:22 +0000 (03:13 +0000)]
o Add info about Julian's Linux Threads checkin (one of these
  days I'm going to write a make kvm top level target).
o Add warning about the
  short lived boot block breakage and briefly what to do to
  fix it.

25 years agoMostly remove the VM_STACK OPTION.
julian [Tue, 26 Jan 1999 02:49:52 +0000 (02:49 +0000)]
Mostly remove the VM_STACK OPTION.
This changes the definitions of a few items so that structures are the
same whether or not the option itself is enabled. This allows
people to enable and disable the option without recompilng the world.

As the author says:

|I ran into a problem pulling out the VM_STACK option.  I was aware of this
|when I first did the work, but then forgot about it.  The VM_STACK stuff
|has some code changes in the i386 branch.  There need to be corresponding
|changes in the alpha branch before it can come out completely.

what is done:
|
|1) Pull the VM_STACK option out of the header files it appears in.  This
|really shouldn't affect anything that executes with or without the rest
|of the VM_STACK patches.  The vm_map_entry will then always have one
|extra element (avail_ssize).  It just won't be used if the VM_STACK
|option is not turned on.
|
|I've also pulled the option out of vm_map.c.  This shouldn't harm anything,
|since the routines that are enabled as a result are not called unless
|the VM_STACK option is enabled elsewhere.
|
|2) Add what appears to be appropriate code the the alpha branch, still
|protected behind the VM_STACK switch.  I don't have an alpha machine,
|so we would need to get some testers with alpha machines to try it out.
|
|Once there is some testing, we can consider making the change permanent
|for both i386 and alpha.
|
[..]
|
|Once the alpha code is adequately tested, we can pull VM_STACK out
|everywhere.
|

Submitted by: "Richard Seaman, Jr." <dick@tar.com>

25 years agoEnable Linux threads support by default.
julian [Tue, 26 Jan 1999 02:38:12 +0000 (02:38 +0000)]
Enable Linux threads support by default.
This takes the conditionals out of the code that has been tested by
various people for a while.
ps and friends (libkvm) will need a recompile as some proc structure
changes are made.

Submitted by: "Richard Seaman, Jr." <dick@tar.com>

25 years agoFix a couple of further bugs: missing argument to sprintf() and
rnordier [Mon, 25 Jan 1999 23:07:02 +0000 (23:07 +0000)]
Fix a couple of further bugs: missing argument to sprintf() and
"==" used for "=".

25 years agoUpdate pkg_add's remote package feature to reflect the new structure of
billf [Mon, 25 Jan 1999 21:08:13 +0000 (21:08 +0000)]
Update pkg_add's remote package feature to reflect the new structure of
wcarchive. This needs to be updated when:

(a) The directory structure changes on wcarchive
(b) getosreldate() has a new significance
(ie. get ELF packages/get a.out packages)
(c) Branches are changed around

As we stand right now, 3.0-postELFday and 4.0-x all point to the same
directory, but when 4.0 starts to have its own packages, this file will
need to be changed.

25 years agoTerminate commit for the Intel PIIX4 SMBus support. Already committed files
nsouch [Mon, 25 Jan 1999 19:34:27 +0000 (19:34 +0000)]
Terminate commit for the Intel PIIX4 SMBus support. Already committed files
are sys/pci/intpm*

Submitted by: Takanori Watanabe <takawata@shidahara1.planet.sci.kobe-u.ac.jp>

25 years agoIntel PIIX4 Power Management Unit for smbus(4).
nsouch [Mon, 25 Jan 1999 19:31:48 +0000 (19:31 +0000)]
Intel PIIX4 Power Management Unit for smbus(4).

Submitted by: Takanori Watanabe <takawata@shidahara1.planet.sci.kobe-u.ac.jp>

25 years ago Add kern.conf_dir sysctl. This is a R+W string used to specify the
dillon [Mon, 25 Jan 1999 18:26:09 +0000 (18:26 +0000)]
Add kern.conf_dir sysctl.  This is a R+W string used to specify the
    directory containing rc.conf.local and rc.local, and possibly other
    things in the future.

    This sysctl is used by the diskless startup code and new rc.conf.  If
    it cannot be found or is empty, the system should revert to using /etc.

25 years ago Commit first rc.diskless startup plus modifications to rc.conf and Makefile
dillon [Mon, 25 Jan 1999 18:24:46 +0000 (18:24 +0000)]
Commit first rc.diskless startup plus modifications to rc.conf and Makefile
    to make it work.  NOTE!  Uses new kern.conf_dir sysctl.

25 years ago Finish up /etc/rc adjustments to handle diskless read-only-root booting.
dillon [Mon, 25 Jan 1999 18:07:25 +0000 (18:07 +0000)]
Finish up /etc/rc adjustments to handle diskless read-only-root booting.
    This will be followed up by rc.diskless and documentation.

25 years agoPort NetBSD's 19990120-accept bug fix. This works around the race condition
fenner [Mon, 25 Jan 1999 16:58:56 +0000 (16:58 +0000)]
Port NetBSD's 19990120-accept bug fix.  This works around the race condition
where select(2) can return that a listening socket has a connected socket
queued, the connection is broken, and the user calls accept(2), which then
blocks because there are no connections queued.

Reviewed by: wollman
Obtained from: NetBSD
(ftp://ftp.NetBSD.ORG/pub/NetBSD/misc/security/patches/19990120-accept)

25 years agoDon't free the socket address if soaccept() / pru_accept() doesn't
fenner [Mon, 25 Jan 1999 16:53:53 +0000 (16:53 +0000)]
Don't free the socket address if soaccept() / pru_accept() doesn't
 return one.

25 years agoIdentify the TI1250 PCMCIA/CardBus bridge. It seems that it's compatible
torstenb [Mon, 25 Jan 1999 12:59:31 +0000 (12:59 +0000)]
Identify  the TI1250 PCMCIA/CardBus bridge. It seems that it's compatible
to the TI1131. At least it works in my Compaq Armada 7800.

25 years agoSync with sys/i386/conf/Makefile.i386 revision 1.137.
kato [Mon, 25 Jan 1999 12:12:01 +0000 (12:12 +0000)]
Sync with sys/i386/conf/Makefile.i386 revision 1.137.

25 years agoUse __XSTRING() from cdefs.h instead of rolling my own.
eivind [Mon, 25 Jan 1999 11:25:24 +0000 (11:25 +0000)]
Use __XSTRING() from cdefs.h instead of rolling my own.

25 years agoCorrectly record the end of the a.out symbol table. In practice, a
rnordier [Mon, 25 Jan 1999 10:34:35 +0000 (10:34 +0000)]
Correctly record the end of the a.out symbol table.  In practice, a
bootinfo structure where bi_esymtab < bi_symtab was being passed
to the kernel.  In the case of older 2.x kernels, this was causing
garbage to be printed to the video console, followed by an exception.

This should resolve a problem reported on -current by Peter Jeremy
<peter.jeremy@auss2.alcatel.com.au>.

25 years agoDon't SEGV when ``set proctitle'' is used in the default
brian [Mon, 25 Jan 1999 10:19:03 +0000 (10:19 +0000)]
Don't SEGV when ``set proctitle'' is used in the default
section.
Submitted by: Dan Lukes <dan@obluda.cz>
PR: 9669

25 years agoBack out a couple of i386 conditionals which aren't needed with the new
dfr [Mon, 25 Jan 1999 08:48:49 +0000 (08:48 +0000)]
Back out a couple of i386 conditionals which aren't needed with the new
syscons and fix a few printfs.

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

25 years agoDon't try to call SYSUNINIT functions if there was a link error.
dfr [Mon, 25 Jan 1999 08:42:24 +0000 (08:42 +0000)]
Don't try to call SYSUNINIT functions if there was a link error.

Reviewed by: Peter Wemm <peter@netplex.com.au>

25 years agoLink everything against libcrypt. ELF builds complain without it.
markm [Mon, 25 Jan 1999 06:49:43 +0000 (06:49 +0000)]
Link everything against libcrypt. ELF builds complain without it.
Clean up the master makefile a bit and add a "dekerberise" target
for those who have shot themselves in the foot.

Thanks to: Randy Bush

25 years agoCorrect a typo and an unparseable sentence.
jkoshy [Mon, 25 Jan 1999 06:22:31 +0000 (06:22 +0000)]
Correct a typo and an unparseable sentence.

PR: docs/9659
Submitted by: Kazuo Horikawa <horikawa@jp.freebsd.org>

25 years agoPlay with MFS size a little.
jkh [Mon, 25 Jan 1999 04:57:07 +0000 (04:57 +0000)]
Play with MFS size a little.

25 years ago Introduce rc script for BOOTP 'diskless' boot. Well, not quite diskless
dillon [Mon, 25 Jan 1999 04:40:53 +0000 (04:40 +0000)]
Introduce rc script for BOOTP 'diskless' boot.  Well, not quite diskless
    since the kernel must be booted from something ( like a floppy ).  This
    script must occur near the beginning of the rc file in order to support
    read-only NFS mounts, which in turn allows all the BOOTP machines to use
    the same / and /usr.

    The companion rc.diskless script is forthcoming.

25 years agoForce the order of the setdefs* so that make -jN doesn't build the
peter [Mon, 25 Jan 1999 04:08:28 +0000 (04:08 +0000)]
Force the order of the setdefs* so that make -jN doesn't build the
setdefs* multiple times at the same time that it's compiling setdefs[0].c.
I think this was leading to unterminated linker sets etc.

25 years ago Fix swap radix tree dump formatting ( pstat -ss ), it was printing the
dillon [Mon, 25 Jan 1999 04:07:07 +0000 (04:07 +0000)]
Fix swap radix tree dump formatting ( pstat -ss ), it was printing the
    wrong radix for recursive subnodes.

25 years agoNO_LKM is no longer an option. LKM support is an option itself.
peter [Mon, 25 Jan 1999 03:51:51 +0000 (03:51 +0000)]
NO_LKM is no longer an option.  LKM support is an option itself.

25 years agogethostbyname2() was broken for lookups via NIS on FreeBSD/alpha
gallatin [Mon, 25 Jan 1999 03:26:45 +0000 (03:26 +0000)]
gethostbyname2() was broken for lookups via NIS on FreeBSD/alpha
due to _gethostbynis() setting h.h_length to sizeof(u_long), which
works out to 8 on alphas.  And 8!= NS_INADDRSZ.

25 years agoFix an aout-to-elf upgrade failure. Don't let the kernel Makefile
peter [Mon, 25 Jan 1999 01:47:49 +0000 (01:47 +0000)]
Fix an aout-to-elf upgrade failure.  Don't let the kernel Makefile
think it's in ELF mode and upgrading /kernel.

25 years agoSMBus support for the Intel PIIX4 power management unit. See smbus(4),
nsouch [Sun, 24 Jan 1999 18:13:31 +0000 (18:13 +0000)]
SMBus support for the Intel PIIX4 power management unit. See smbus(4),
iicbus(4) and smb(4).

User programs are available to retrieve SDRAM and sensor info, contact
the author.

Submitted by: Takanori Watanabe <takawata@shidahara1.planet.sci.kobe-u.ac.jp>
Reviewed by: Mike Smith <msmith@freebsd.org>

25 years ago'Nother place where libcrypt needs to be linked.
markm [Sun, 24 Jan 1999 15:28:41 +0000 (15:28 +0000)]
'Nother place where libcrypt needs to be linked.

Thanks to: Randy Bush

25 years agoReplace a bunch of "ln foo bar"'s with "ln -f foo bar".
markm [Sun, 24 Jan 1999 13:51:26 +0000 (13:51 +0000)]
Replace a bunch of "ln foo bar"'s with "ln -f foo bar".
This means that if(when) you go "sh MAKEDEV all" in /dev
the devices get remade; you don't get errors.
A lot of the changes are for info only; they are commented out.

Not exactly shot to pieces by: bde

25 years agoMore libcrypt backout.
markm [Sun, 24 Jan 1999 13:46:33 +0000 (13:46 +0000)]
More libcrypt backout.

25 years agoFix logic surrounding the noconn option.
obrien [Sun, 24 Jan 1999 10:10:20 +0000 (10:10 +0000)]
Fix logic surrounding the noconn option.
(this should also shutup the warning message everybody sees)

Submitted by: David Wolfskill <dhw@whistle.com>

25 years agoUnder ELF, this must be linked against libcrypt.
markm [Sun, 24 Jan 1999 08:12:56 +0000 (08:12 +0000)]
Under ELF, this must be linked against libcrypt.
Found by: Randy Bush

25 years agoSupport 'O MaxHeaderLines=' to override the default header count and line
peter [Sun, 24 Jan 1999 07:54:30 +0000 (07:54 +0000)]
Support 'O MaxHeaderLines=' to override the default header count and line
length limits.  The configuration keyword is: confMAX_HEADER_LINES

25 years agoFix symlinking. Without the -f "force" option, the wrong version
markm [Sun, 24 Jan 1999 07:51:33 +0000 (07:51 +0000)]
Fix symlinking. Without the -f "force" option, the wrong version
can be found.
Submitted by:   Bruce

25 years agoMerge changes from vendor branch into our version
peter [Sun, 24 Jan 1999 07:46:43 +0000 (07:46 +0000)]
Merge changes from vendor branch into our version

25 years agoThis commit was generated by cvs2svn to compensate for changes in r43148,
peter [Sun, 24 Jan 1999 07:40:56 +0000 (07:40 +0000)]
This commit was generated by cvs2svn to compensate for changes in r43148,
which included commits to RCS files with non-trunk default branches.

25 years agoCheck the patch obtained from sendmail.org for the header denial-of-service
peter [Sun, 24 Jan 1999 07:40:56 +0000 (07:40 +0000)]
Check the patch obtained from sendmail.org for the header denial-of-service
attack into the vendor branch.  It is a little unusual doing it this way
but it will eliminate (or minimize anyway) conflicts when 8.9.3 comes out.

Obtained from: sendmail.org (as posted on bugtraq, but without broken tabs)

25 years agoDrop outdated FreeBSD version number from test script.
rnordier [Sun, 24 Jan 1999 07:10:34 +0000 (07:10 +0000)]
Drop outdated FreeBSD version number from test script.
Submitted by : Sue Blake <sue@vedanix.welearn.com.au>
PR : 9650

25 years ago Undo last commit - not a bug, just duplicate code. PG_MAPPED and
dillon [Sun, 24 Jan 1999 07:06:52 +0000 (07:06 +0000)]
Undo last commit - not a bug, just duplicate code.  PG_MAPPED and
    PG_WRITEABLE are already cleared by vm_page_protect().

25 years agoBacked out most of previous commit to go with backing out support for
bde [Sun, 24 Jan 1999 06:43:30 +0000 (06:43 +0000)]
Backed out most of previous commit to go with backing out support for
revoke() on non-device files.

25 years agoGo back to only supporting revoke() for bdevs and cdevs. It is very
bde [Sun, 24 Jan 1999 06:28:37 +0000 (06:28 +0000)]
Go back to only supporting revoke() for bdevs and cdevs.  It is very
buggy for fifos, and no one seems to have investigated its behaviour
on other types of files.  It has been broken since the Lite2 merge
in rev.1.54.

Nagged about by: Brian Feldman (green@unixhelp.org)

25 years agoFrom the PR:
msmith [Sun, 24 Jan 1999 06:06:21 +0000 (06:06 +0000)]
From the PR:

I added a FICL_TRACE-conditioned trace facility based on "see".
It is ugly because words' functions are almost all static, and ficlExec,
where the trace has to be located, can't get their pointers. So, #ifdef
this staticization, and add most of see's body into ficlExec. Duplication
of code, uglyness, etc. But it is cleanly #ifdef'ed, and works like a
charm.

It does not provide "step" facility, though, just trace. It is
tunable at run-time through "trace!". If anyone (most likely me :) ever
wants a step facility, I'll add it. Should be easy.

PR: bin/9652
Submitted by: "Daniel C. Sobral" <dcs@newsguy.com>

25 years ago Change all manual settings of vm_page_t->dirty = VM_PAGE_BITS_ALL
dillon [Sun, 24 Jan 1999 06:04:52 +0000 (06:04 +0000)]
Change all manual settings of vm_page_t->dirty = VM_PAGE_BITS_ALL
    to use the vm_page_dirty() inline.

    The inline can thus do sanity checks ( or not ) over all cases.

25 years agoBleh. Missing 'unit =' doesn't help legacy num_ide_disks code.
msmith [Sun, 24 Jan 1999 06:03:44 +0000 (06:03 +0000)]
Bleh.  Missing 'unit =' doesn't help legacy num_ide_disks code.

PR: bin/9651
Submitted by: "Daniel C. Sobral" <dcs@newsguy.com>

25 years ago vm_map_split() used to dirty the page manually after calling
dillon [Sun, 24 Jan 1999 06:00:31 +0000 (06:00 +0000)]
vm_map_split() used to dirty the page manually after calling
    vm_page_rename(), but never pulled the page off PQ_CACHE if it was on
    PQ_CACHE.  Dirty pages in PQ_CACHE are not allowed and a KASSERT was
    added in -4.x to test for this... and got hit.

    In -4.x, vm_page_rename() automatically dirties the page.  This commit
    also has it deal with the PQ_CACHE case, deactivating the page in that
    case.

25 years agoFrom the PR:
msmith [Sun, 24 Jan 1999 05:58:18 +0000 (05:58 +0000)]
From the PR:

FICL's TYPE copies the counted string to HERE, as abial has
remarked. Answering to abial's question, this is NOT garanteed to have
enough space.
...
We have dynamic memory. Even before memory-alloc got in, we
already had dynamic memory. Use it, then! (ficlMalloc is sysdep, so I
suppose that's why it was not used for TYPE; ficl is probably designed
to work without a working ficlFree).

Submitted by: "Daniel C. Sobral" <dcs@newsguy.com>

25 years ago Add vm_page_dirty() inline with PQ_CACHE sanity check
dillon [Sun, 24 Jan 1999 05:57:50 +0000 (05:57 +0000)]
Add vm_page_dirty() inline with PQ_CACHE sanity check

25 years agoCorrect #include in vinumioctl.c to compile correctly
grog [Sun, 24 Jan 1999 05:14:39 +0000 (05:14 +0000)]
Correct #include in vinumioctl.c to compile correctly
  without VINUMDEBUG

Submitted-by: peter
25 years ago Addendum: The original code that the last commit 'fixed' actually did
dillon [Sun, 24 Jan 1999 03:49:58 +0000 (03:49 +0000)]
Addendum:  The original code that the last commit 'fixed' actually did
    not have a bug in it, but the last commit did make it more readable so
    we are keeping it.

25 years ago vm_pager_put_pages() is passed an rcval array to hold per-page return
dillon [Sun, 24 Jan 1999 02:32:15 +0000 (02:32 +0000)]
vm_pager_put_pages() is passed an rcval array to hold per-page return
    values.  The 'int' return value for the procedure was never used and
    not well defined in any case when there are mixed errors on pages, so
    it has been removed.  vm_pager_put_pages() and associated vm_pager
    functions now return void.

25 years ago Clear PG_MAPPED as well as PG_WRITEABLE when a page is moved to the
dillon [Sun, 24 Jan 1999 02:29:26 +0000 (02:29 +0000)]
Clear PG_MAPPED as well as PG_WRITEABLE when a page is moved to the
    cache.

25 years ago Added warning printf ( needs INVARIANTS ) when busy cache page is found
dillon [Sun, 24 Jan 1999 01:33:22 +0000 (01:33 +0000)]
Added warning printf ( needs INVARIANTS ) when busy cache page is found
    while trying to free memory.

25 years agoAdd comment indicating that the vinum pseudo-device is experimental,
grog [Sun, 24 Jan 1999 01:22:37 +0000 (01:22 +0000)]
Add comment indicating that the vinum pseudo-device is experimental,
and that vinum should be started as a kld.

Tripped-over-by: many people
25 years ago There was a situation where sendfile() might attempt to initiate I/O
dillon [Sun, 24 Jan 1999 01:15:58 +0000 (01:15 +0000)]
 There was a situation where sendfile() might attempt to initiate I/O
     on a PG_BUSY page, due to a bug in its sequencing of a conditional.

25 years ago It is possible for a page in the cache to be busy. vm_pageout.c was not
dillon [Sun, 24 Jan 1999 01:06:31 +0000 (01:06 +0000)]
It is possible for a page in the cache to be busy.  vm_pageout.c was not
    checking for this condition while it tried to free cache pages.  Fixed.

25 years ago Add invariants to vm_page_busy() and vm_page_wakeup() to check for
dillon [Sun, 24 Jan 1999 01:05:15 +0000 (01:05 +0000)]
Add invariants to vm_page_busy() and vm_page_wakeup() to check for
    PG_BUSY stupidity.

25 years ago Clear PG_WRITEABLE in vm_page_cache(). This may or may not be a bug,
dillon [Sun, 24 Jan 1999 01:04:04 +0000 (01:04 +0000)]
Clear PG_WRITEABLE in vm_page_cache().  This may or may not be a bug,
    but the bit should definitely be cleared.

25 years ago Depreciate vm_object_pmap_copy() - nobody uses it. Everyone uses
dillon [Sun, 24 Jan 1999 01:01:38 +0000 (01:01 +0000)]
Depreciate vm_object_pmap_copy() - nobody uses it.  Everyone uses
    vm_object_pmap_copt_1() now, apparently.

25 years ago Get rid of unused old_m in vm_fault. Add INVARIANTS to test whether
dillon [Sun, 24 Jan 1999 00:55:04 +0000 (00:55 +0000)]
Get rid of unused old_m in vm_fault.  Add INVARIANTS to test whether
    page is still busy after all the hell vm_fault goes through.. it is
    supposed to be, and printf() if it isn't.  don't panic, though.

25 years ago Don't try to calculate B_CACHE for an NFS related bp that has a
dillon [Sun, 24 Jan 1999 00:51:11 +0000 (00:51 +0000)]
Don't try to calculate B_CACHE for an NFS related bp that has a
    > 0 b_validend.  This will screw up small-writes, causing lots of
    little writes out the network.

    We will assume that NFS handles B_CACHE properly.

25 years agoDocument root_disk_unit
msmith [Sun, 24 Jan 1999 00:12:17 +0000 (00:12 +0000)]
Document root_disk_unit

25 years agobiosdisk.c
msmith [Sun, 24 Jan 1999 00:12:04 +0000 (00:12 +0000)]
biosdisk.c
Implement a new variable 'root_disk_unit' which supersedes
'num_ide_disks' and makes it possible to explicitly set the
root device unit number regardless of type considerations.

bootinfo.c
If we can't calculate a dev_t for the root disk, complain and
don't proceed to boot with an invalid boot device.

25 years agoFill in the bi_bios_dev field in the bootinfo struct; the loader doesn't
msmith [Sun, 24 Jan 1999 00:10:10 +0000 (00:10 +0000)]
Fill in the bi_bios_dev field in the bootinfo struct; the loader doesn't
guess this value well in anything other than the simplest of situations.

25 years agoMove kernel-only declarations to within #ifdef KERNEL
archie [Sat, 23 Jan 1999 23:59:50 +0000 (23:59 +0000)]
Move kernel-only declarations to within #ifdef KERNEL
Prompted by: gcc warnings when compiling /sbin/ipfw

25 years agoAdd the texinfo'ed docs to the build/install.
markm [Sat, 23 Jan 1999 21:24:49 +0000 (21:24 +0000)]
Add the texinfo'ed docs to the build/install.

25 years agoReplace DEBUG by PLIP_DEBUG option
nsouch [Sat, 23 Jan 1999 17:07:49 +0000 (17:07 +0000)]
Replace DEBUG by PLIP_DEBUG option

25 years agoAdd various documented ppbus options
nsouch [Sat, 23 Jan 1999 17:06:01 +0000 (17:06 +0000)]
Add various documented ppbus options

25 years agoFix wrong exit code.
rnordier [Sat, 23 Jan 1999 17:05:54 +0000 (17:05 +0000)]
Fix wrong exit code.
PR: 9321
Submitted by: Bob Van Valzah <Bob@whitebarn.com>

25 years agoRemove a workaround for the alpha port using an outdated version of syscons.
dfr [Sat, 23 Jan 1999 17:04:13 +0000 (17:04 +0000)]
Remove a workaround for the alpha port using an outdated version of syscons.

25 years agoUpdate the alpha port to use the new syscons.
dfr [Sat, 23 Jan 1999 16:53:30 +0000 (16:53 +0000)]
Update the alpha port to use the new syscons.

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

25 years agoI may have forgotten to upgrade this value, but that will never happen
eivind [Sat, 23 Jan 1999 15:07:10 +0000 (15:07 +0000)]
I may have forgotten to upgrade this value, but that will never happen
again.  (Fully clone the value of __FreeBSD__ from the compiler.)

25 years agoFix a dot too many in path.
rnordier [Sat, 23 Jan 1999 14:50:28 +0000 (14:50 +0000)]
Fix a dot too many in path.
PR: 9445
Noticed by: Ying-Chieh Liao <ijliao@Terry.Dorm10.NCTU.edu.tw>

25 years agoRemove imm.4, merged with vpo.4
nsouch [Sat, 23 Jan 1999 14:01:55 +0000 (14:01 +0000)]
Remove imm.4, merged with vpo.4

Fix vpo.4 manpage generation in Makefile
Fix bootflags bits description in ppc.4

25 years agoAdded detection of Hauppauge IR remote control.
roger [Sat, 23 Jan 1999 11:32:06 +0000 (11:32 +0000)]
Added detection of Hauppauge IR remote control.
and MSP34xx Audio chip. Fixed i2c read error.
Hauppauge supplied details of new Tuner Types.
Added tuner type 0x1a.
Danny Braniss <danny@cs.huji.ac.il> submitted Bt878
AverMedia detection with PCI subsystem vendor id.

25 years agoSubmitted by: Roger Hardiman <roger@cs.strath.ac.uk>
roger [Sat, 23 Jan 1999 11:28:16 +0000 (11:28 +0000)]
Submitted by: Roger Hardiman <roger@cs.strath.ac.uk>
Updated to support Hauppauge IR Remote Control

25 years ago__FreeBSD__ is also used here.
jhay [Sat, 23 Jan 1999 10:59:29 +0000 (10:59 +0000)]
__FreeBSD__ is also used here.

Forgotten by: Lots of people.
Pointed out by: make world.

25 years agoSuppress warnings when compiling LINT. No functional change.
yokota [Sat, 23 Jan 1999 10:55:16 +0000 (10:55 +0000)]
Suppress warnings when compiling LINT.  No functional change.

25 years ago- Remove reference to obsolete options.
yokota [Sat, 23 Jan 1999 10:51:58 +0000 (10:51 +0000)]
- Remove reference to obsolete options.
- Describe options for the vga driver.
Reviewed by: bde