]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
24 years agoThe wrong version of the file was committed previously which explains the
Kris Kennaway [Sun, 16 Jan 2000 21:00:06 +0000 (21:00 +0000)]
The wrong version of the file was committed previously which explains the
problems seen here.

24 years agoAdd the apm -e argument, which will toggle the "enabled" status of APM.
Brian Feldman [Sun, 16 Jan 2000 20:25:57 +0000 (20:25 +0000)]
Add the apm -e argument, which will toggle the "enabled" status of APM.

24 years agoget pd_type from inquiry data itself
Matt Jacob [Sun, 16 Jan 2000 20:15:43 +0000 (20:15 +0000)]
get pd_type from inquiry data itself

24 years agouse XPT_GDEV_STATS- XPT_GDEV_TYPE was deprecated a while back
Matt Jacob [Sun, 16 Jan 2000 20:03:34 +0000 (20:03 +0000)]
use XPT_GDEV_STATS- XPT_GDEV_TYPE was deprecated a while back

24 years agoFix breakage from last commit- macro was broken. I appreciate the
Matt Jacob [Sun, 16 Jan 2000 19:02:29 +0000 (19:02 +0000)]
Fix breakage from last commit- macro was broken. I appreciate the
committer's fixing things for alpha, but I sure wish they could
have compiled on one!

24 years agoRemoved outdated comment on experimental nature of ata. Added comment to esp
Wilko Bulte [Sun, 16 Jan 2000 18:46:21 +0000 (18:46 +0000)]
Removed outdated comment on experimental nature of ata. Added comment to esp

24 years agoAdd and document support for handling comments and blank lines
Chris Costello [Sun, 16 Jan 2000 18:12:42 +0000 (18:12 +0000)]
Add and document support for handling comments and blank lines
in /etc/sysctl.conf.

Reviewed by: imp, jkh

24 years agofix kernel panic at rtfree() in INET6 enabled envrionment.
Yoshinobu Inoue [Sun, 16 Jan 2000 18:00:06 +0000 (18:00 +0000)]
fix kernel panic at rtfree() in INET6 enabled envrionment.
  This is probably due to twice rtfree() in in6_pcbdetach(),
  one for inp->in6p_route.ro_rt, and another one for inp->inp_route.ro_rt.
  But these 2 are actually shared in inpcb, so 2nd rtfree() is not necessary.

24 years agoFix !crypto releases once more.
Poul-Henning Kamp [Sun, 16 Jan 2000 17:21:28 +0000 (17:21 +0000)]
Fix !crypto releases once more.

24 years agoImplement setres[ug]id() and getres[ug]id(). This has been sitting in
Peter Wemm [Sun, 16 Jan 2000 16:34:26 +0000 (16:34 +0000)]
Implement setres[ug]id() and getres[ug]id().  This has been sitting in
my tree for ages (~2 years) waiting for an excuse to commit it.  Now Linux
has implemented it and it seems that Staroffice (when using the
linux_base6.1 port's libc) calls this in the linux emulator and dies in
setup.  The Linux emulator can call these now.

24 years agoAdd missing share/examples/ses directory.
Poul-Henning Kamp [Sun, 16 Jan 2000 14:02:51 +0000 (14:02 +0000)]
Add missing share/examples/ses directory.

24 years agoALPHAHW.TXT described the supported machine types for FreeBSD/alpha
Wilko Bulte [Sun, 16 Jan 2000 13:18:08 +0000 (13:18 +0000)]
ALPHAHW.TXT described the supported machine types for FreeBSD/alpha
along with their quirks along with generic info on things like SRM,
bootable adapters etc.

24 years agoThere's not /etc/pccard_ether_remove on -current.
Tatsumi Hosokawa [Sun, 16 Jan 2000 13:13:12 +0000 (13:13 +0000)]
There's not /etc/pccard_ether_remove on -current.

24 years agoI've forgotten one hardwired device name in insert/remove line.
Tatsumi Hosokawa [Sun, 16 Jan 2000 13:11:23 +0000 (13:11 +0000)]
I've forgotten one hardwired device name in insert/remove line.

24 years agofix missing \n in sndstat output
Cameron Grant [Sun, 16 Jan 2000 12:52:22 +0000 (12:52 +0000)]
fix missing \n in sndstat output

24 years agoCorrect comments / point to right LINT file (./LINT does not exist for Alpha)
Wilko Bulte [Sun, 16 Jan 2000 12:39:24 +0000 (12:39 +0000)]
Correct comments / point to right LINT file (./LINT does not exist for Alpha)

24 years agooops, best play format was set to unsigned 16 bit instead of signed 16 bit
Cameron Grant [Sun, 16 Jan 2000 12:05:07 +0000 (12:05 +0000)]
oops, best play format was set to unsigned 16 bit instead of signed 16 bit
stereo.  remenant from testing.

24 years agoUpdate this for the recient import.
David E. O'Brien [Sun, 16 Jan 2000 10:22:31 +0000 (10:22 +0000)]
Update this for the recient import.

24 years agoUpdate for 1.0.5a.
David E. O'Brien [Sun, 16 Jan 2000 10:21:28 +0000 (10:21 +0000)]
Update for 1.0.5a.

24 years agoMerge rev 1.2 into BC 1.0.5a, which fixes a spelling error.
David E. O'Brien [Sun, 16 Jan 2000 10:14:50 +0000 (10:14 +0000)]
Merge rev 1.2 into BC 1.0.5a, which fixes a spelling error.

24 years agoGNU has figured out they should fix the bug we fixed in rev 1.2 themselves.
David E. O'Brien [Sun, 16 Jan 2000 10:10:15 +0000 (10:10 +0000)]
GNU has figured out they should fix the bug we fixed in rev 1.2 themselves.
So return this file back to the stock version.

24 years agoThis commit was generated by cvs2svn to compensate for changes in r56101,
David E. O'Brien [Sun, 16 Jan 2000 10:01:07 +0000 (10:01 +0000)]
This commit was generated by cvs2svn to compensate for changes in r56101,
which included commits to RCS files with non-trunk default branches.

24 years agoImport GNU bc 1.05a.
David E. O'Brien [Sun, 16 Jan 2000 10:01:07 +0000 (10:01 +0000)]
Import GNU bc 1.05a.

24 years agoConvert last examples of `controller' to `device'.
Jeroen Ruigrok van der Werven [Sun, 16 Jan 2000 09:37:18 +0000 (09:37 +0000)]
Convert last examples of `controller' to `device'.

24 years agoCleanup some more remaining bdev fluff.
Poul-Henning Kamp [Sun, 16 Jan 2000 09:25:34 +0000 (09:25 +0000)]
Cleanup some more remaining bdev fluff.

24 years agoCleanup some remaining bdev fluff.
Poul-Henning Kamp [Sun, 16 Jan 2000 09:25:10 +0000 (09:25 +0000)]
Cleanup some remaining bdev fluff.

24 years agoCosmetic change: space instead of tab in front of $FreeBSD$.
Sheldon Hearn [Sun, 16 Jan 2000 08:25:40 +0000 (08:25 +0000)]
Cosmetic change: space instead of tab in front of $FreeBSD$.

Submitted by: bde

24 years agoThe kernel side of per-process unaligned access control (printing, fixing &
Andrew Gallatin [Sun, 16 Jan 2000 07:07:33 +0000 (07:07 +0000)]
The kernel side of per-process unaligned access control (printing, fixing &
delivering SIGBUS).  This will allow a non-superuser to control unaligned
access behaviour on a per-process basis once a userland control program
(uac) is written.

Reviewed by: obrien
Tested by:   obrien

24 years agoThis fixes a bug that /etc/pccard_ether did not work without DHCP.
Tatsumi Hosokawa [Sun, 16 Jan 2000 06:44:48 +0000 (06:44 +0000)]
This fixes a bug that /etc/pccard_ether did not work without DHCP.
For example, when /etc/pccard.conf had ed0 in config line, but kernel
refused this name and said

devclass_alloc_unit: ed0 already exists, using next availale unit
number

Kernel used ed1 as device name and it did not match with config and
insert/remove lines.  Fortunately, dhclient was called without args,
and it works, but if we wanted to use static IP address for PC-card,
it did not work.

This modification makes pccardd to execute insert/remove lines with
the true device name that returns from kernel.  (Last change to
etc/pccard.conf.sample eliminated all hardwired device name from
insert/remove lines in /etc/pccard.conf)

24 years agoFix an instance of rman_get_start() than should be rman_get_bushandle().
Bill Paul [Sun, 16 Jan 2000 06:41:49 +0000 (06:41 +0000)]
Fix an instance of rman_get_start() than should be rman_get_bushandle().

24 years agoFixed pedantic syntax errors caused by a trailing semicolon in the
Bruce Evans [Sun, 16 Jan 2000 06:11:33 +0000 (06:11 +0000)]
Fixed pedantic syntax errors caused by a trailing semicolon in the
definition of EVENTHANDLER_DECLARE().

Removed all other trailing semicolons in macro definitions.  The
ones after `do ... while (0)' wrappers defeated the point of the
wrappers but were harmless.

Enforce semicolons after invocation of declaration-like macros.

24 years agoRemove reference to floppy tape drive.
Kris Kennaway [Sun, 16 Jan 2000 05:55:34 +0000 (05:55 +0000)]
Remove reference to floppy tape drive.

24 years agoMention openssl and auditing fixes.
Kris Kennaway [Sun, 16 Jan 2000 05:52:55 +0000 (05:52 +0000)]
Mention openssl and auditing fixes.

24 years agoTurn back on openssl building.
Kris Kennaway [Sun, 16 Jan 2000 05:25:26 +0000 (05:25 +0000)]
Turn back on openssl building.

24 years agoTurn back on libcrypto and libssl building.
Kris Kennaway [Sun, 16 Jan 2000 05:24:47 +0000 (05:24 +0000)]
Turn back on libcrypto and libssl building.

24 years ago(Forced commit)
Kris Kennaway [Sun, 16 Jan 2000 05:23:21 +0000 (05:23 +0000)]
(Forced commit)

Support RSA via RSAref. This autodetects the RSAref package, and if it
is not found, compiles without RSA.

24 years ago*** empty log message ***
Kris Kennaway [Sun, 16 Jan 2000 05:19:27 +0000 (05:19 +0000)]
*** empty log message ***

24 years agoReplace a conjunction "you're" with "you are". This makes
Chris Costello [Sun, 16 Jan 2000 05:15:48 +0000 (05:15 +0000)]
Replace a conjunction "you're" with "you are".  This makes
reading by non-native English speakers a little bit easier, and
follows some basic doc guidelines.

24 years agoThis commit was generated by cvs2svn to compensate for changes in r56083,
Kris Kennaway [Sun, 16 Jan 2000 05:14:57 +0000 (05:14 +0000)]
This commit was generated by cvs2svn to compensate for changes in r56083,
which included commits to RCS files with non-trunk default branches.

24 years agoImport the RSA support code. There shouldn't be any actual RSA
Kris Kennaway [Sun, 16 Jan 2000 05:14:57 +0000 (05:14 +0000)]
Import the RSA support code. There shouldn't be any actual RSA
cryptography here.

24 years agoFix for missing symbol in -DRSAref case.
Kris Kennaway [Sun, 16 Jan 2000 04:45:18 +0000 (04:45 +0000)]
Fix for missing symbol in -DRSAref case.

24 years agoFixed stray backslash in previous commit.
Bruce Evans [Sun, 16 Jan 2000 04:43:11 +0000 (04:43 +0000)]
Fixed stray backslash in previous commit.

Fixed bitrot in comments about library dependencies.  The list has been
maintained better in ../Makefile.inc1, except it has been uninverted
there so it is hard to use manually.

24 years agoAdd MAINTAINER tag so people don't feel the need to randomly frob with this.
Kris Kennaway [Sun, 16 Jan 2000 02:20:03 +0000 (02:20 +0000)]
Add MAINTAINER tag so people don't feel the need to randomly frob with this.

24 years ago Add SPLASSERT() macro. SPLASSERT() compiles to a no-op
Jonathan M. Bresler [Sun, 16 Jan 2000 02:10:21 +0000 (02:10 +0000)]
    Add SPLASSERT() macro.  SPLASSERT() compiles to a no-op
        unless both "option INVARIANTS" and "options INVARIANT_SUPPORT"
        are defined in the kernel's config(8) file.

        SPLASSERT(expression, msg) used KASSERT to check that the
        expression is true, panic()ing the kernel otherwise.

Approved by: jkh
Reviewed by: jdp, dfr, phk, eivind and green

24 years ago Add SPLASSERT.9 man page. SPLASSERT() commit to follow immediately.
Jonathan M. Bresler [Sun, 16 Jan 2000 02:08:37 +0000 (02:08 +0000)]
Add SPLASSERT.9 man page.  SPLASSERT() commit to follow immediately.

Approved by: jkh

24 years agoSet ifq_maxlen to default (IFQ_MAXLEN).
Matthew N. Dodd [Sun, 16 Jan 2000 00:48:33 +0000 (00:48 +0000)]
Set ifq_maxlen to default (IFQ_MAXLEN).

24 years agoRemove Bison from the build. This is a prelude to possibly removing it
David E. O'Brien [Sun, 16 Jan 2000 00:11:34 +0000 (00:11 +0000)]
Remove Bison from the build.  This is a prelude to possibly removing it
from 4.0-R.

24 years agoOops, didn't commit the Makefile for libposix1e--this should fix build
Robert Watson [Sat, 15 Jan 2000 23:33:49 +0000 (23:33 +0000)]
Oops, didn't commit the Makefile for libposix1e--this should fix build
problems.

Reviewed by: eivind

24 years agoK&Rify
Brian Somers [Sat, 15 Jan 2000 23:08:30 +0000 (23:08 +0000)]
K&Rify

24 years agobuglet fix
Matt Jacob [Sat, 15 Jan 2000 23:05:29 +0000 (23:05 +0000)]
buglet fix

24 years agoAdd a minimal example set of how to use the SES ioctls to monitor
Matt Jacob [Sat, 15 Jan 2000 22:47:17 +0000 (22:47 +0000)]
Add a minimal example set of how to use the SES ioctls to monitor
SES/SAF-TE devices. It's usable enough (in chpmon) to be alerted as you'd
like to be if a powersupply goes south.

24 years agoOnly log when there is a real change to the configuration state.
David E. O'Brien [Sat, 15 Jan 2000 22:46:40 +0000 (22:46 +0000)]
Only log when there is a real change to the configuration state.

Requested and Tested by: unfurl

24 years agoDo a wad of cleanup and put into place at least *one* functional
Matt Jacob [Sat, 15 Jan 2000 22:41:27 +0000 (22:41 +0000)]
Do a wad of cleanup and put into place at least *one* functional
portion of SES retrieval (native SES itself) and implement all
the appropriate ioctls. Seems to work...

24 years agoMove the kernel specific stuff into scsi_ses.c. Redo the ioctls
Matt Jacob [Sat, 15 Jan 2000 22:40:04 +0000 (22:40 +0000)]
Move the kernel specific stuff into scsi_ses.c. Redo the ioctls
to be more platform independent. Add a ses_hlptxt structure definition
and retrieval ioctl for when we are able to retrieve object help text

SES Objects can have up to 64 KBytes of associated 'help' text- the
Sun A5000 uses this, for example, to give physical location information
(e.g., 'left power supply').

24 years agoUserland build of Kerberos5 (AKA Heimdal). More to come.
Mark Murray [Sat, 15 Jan 2000 21:38:08 +0000 (21:38 +0000)]
Userland build of Kerberos5 (AKA Heimdal). More to come.

This is not ready for primetime yet! Please hold off on the bug reports.

24 years agoUse "$device" instead of hardwired device names in insert/remove lines
Tatsumi Hosokawa [Sat, 15 Jan 2000 21:29:18 +0000 (21:29 +0000)]
Use "$device" instead of hardwired device names in insert/remove lines
of /etc/pccard.conf.sample

24 years agoadd SES driver
Matt Jacob [Sat, 15 Jan 2000 21:04:13 +0000 (21:04 +0000)]
add SES driver

24 years agoEnsure that DMA mappings are freed in error situations.
Mike Spengler [Sat, 15 Jan 2000 21:01:04 +0000 (21:01 +0000)]
Ensure that DMA mappings are freed in error situations.

24 years agoNotice when an interface's ATM address changes when running a local arpserver.
Mike Spengler [Sat, 15 Jan 2000 20:46:07 +0000 (20:46 +0000)]
Notice when an interface's ATM address changes when running a local arpserver.

24 years agoProcess SPANS ARP requests even if source IP is 0.0.0.0 (or broadcast). This
Mike Spengler [Sat, 15 Jan 2000 20:34:55 +0000 (20:34 +0000)]
Process SPANS ARP requests even if source IP is 0.0.0.0 (or broadcast).  This
can occur when talking to a Fore host which is using "bonded" interfaces.

24 years agoAdded sn driver to network interface menu.
Tatsumi Hosokawa [Sat, 15 Jan 2000 20:30:22 +0000 (20:30 +0000)]
Added sn driver to network interface menu.

24 years agoFix the tl driver so it doesn't run off the end of the RX descriptor list
Bill Paul [Sat, 15 Jan 2000 20:14:49 +0000 (20:14 +0000)]
Fix the tl driver so it doesn't run off the end of the RX descriptor list
and panic in out of mbufs condition.

24 years agoHandle GetResponse errors correctly.
Mike Spengler [Sat, 15 Jan 2000 19:51:44 +0000 (19:51 +0000)]
Handle GetResponse errors correctly.

24 years agoEnable the building of libposix1e
Robert Watson [Sat, 15 Jan 2000 19:47:01 +0000 (19:47 +0000)]
Enable the building of libposix1e

Reviewed by: eivind

24 years ago|The hard limit for the BPF buffer size is 32KB, which appears too low
Poul-Henning Kamp [Sat, 15 Jan 2000 19:46:12 +0000 (19:46 +0000)]
|The hard limit for the BPF buffer size is 32KB, which appears too low
|for high speed networks (even at 100Mbit/s this corresponds to 1/300th
|of a second). The default buffer size is 4KB, but libpcap and ipfilter
|both override this (using the BIOCSBLEN ioctl) and allocate 32KB.
|
|The following patch adds an sysctl for bpf_maxbufsize, similar to the
|one for bpf_bufsize that you added back in December 1995. I choose to
|make the default for this limit 512KB (the value suggested by NFR).

Submitted by: se
Reviewed by: phk

24 years ago|Since I had suffered from partially applied CTM deltas, some time ago,
Poul-Henning Kamp [Sat, 15 Jan 2000 19:45:18 +0000 (19:45 +0000)]
|Since I had suffered from partially applied CTM deltas, some time ago,
|I made ctm ignore deltas for files that match the "after edit" MD5.
|(In one case, I had the compiler fill all temporary space while CTM
|was editing files.)

Submitted by: se
Reviewed by: phk

24 years agolibposix1e provides userland library calls for the POSIX.1e security
Robert Watson [Sat, 15 Jan 2000 19:44:27 +0000 (19:44 +0000)]
libposix1e provides userland library calls for the POSIX.1e security
interface.  This commit introduces the library, as well as a modest
subset of the ACL calls, with some modifications to support multiple
ACL semantics.

Reviewed by: eivind

24 years agoDon't clobber the sources when plain "make" is run. Just print a message
Bruce Evans [Sat, 15 Jan 2000 19:40:08 +0000 (19:40 +0000)]
Don't clobber the sources when plain "make" is run.  Just print a message
as in all other Makefiles that create syscall files.

Fixed some style bugs.

24 years agoadd SEND/RECEIVE diagnostic opcodes, SEND is a Mandatory command
Matt Jacob [Sat, 15 Jan 2000 19:05:29 +0000 (19:05 +0000)]
add SEND/RECEIVE diagnostic opcodes, SEND is a Mandatory command

24 years agoMinor enhancement: set the 'dual link LED' bit in the auxmode register
Bill Paul [Sat, 15 Jan 2000 18:43:07 +0000 (18:43 +0000)]
Minor enhancement: set the 'dual link LED' bit in the auxmode register
of the Broadcom BCM5201 PHY on the LinkSys USB100TX adapter so that the
link LED correctly (lights up amber for 10mbps link, green for 100mbps
link).

Note that the sticker on the bottom of the adapter says amber for 10
and green for 100, but the appendix in the manual that comes with
the adapter says green for 10 and amber for 100. Given that there doesn't
seem to be any way to make the hardware produce the latter combination,
I think it's safe to say the sticker is right and the manual is wrong.
I'm just shocked, shocked I tell you.

24 years agoActually read the PCI device ID when testing the device ID value against
Bill Paul [Sat, 15 Jan 2000 18:04:28 +0000 (18:04 +0000)]
Actually read the PCI device ID when testing the device ID value against
the supported devices list.

24 years agoWe cannot have libcrypto, and therefore OpenSSL at all, without RSA.
Brian Feldman [Sat, 15 Jan 2000 18:02:10 +0000 (18:02 +0000)]
We cannot have libcrypto, and therefore OpenSSL at all, without RSA.
If you need examples of breakage, I'm ready to provide more than a
few.

24 years agoDon't mix up KNF and style(9). This isn't KNF, it's a derivative of it.
Peter Wemm [Sat, 15 Jan 2000 16:34:07 +0000 (16:34 +0000)]
Don't mix up KNF and style(9).  This isn't KNF, it's a derivative of it.

24 years agoAdded "sn" and "pcic" entries to UserConfig.
Tatsumi Hosokawa [Sat, 15 Jan 2000 16:33:49 +0000 (16:33 +0000)]
Added "sn" and "pcic" entries to UserConfig.

24 years agoEnable the 't' option, and document it in the manpage.
Søren Schmidt [Sat, 15 Jan 2000 15:51:47 +0000 (15:51 +0000)]
Enable the 't' option, and document it in the manpage.

Submitted by: green@freebsd.org <Brian Fundakowski Feldman>

24 years agoFix handling of svr4_sigsets, which are implemented in SysVR4 as a sequence
Mark Newton [Sat, 15 Jan 2000 15:38:17 +0000 (15:38 +0000)]
Fix handling of svr4_sigsets, which are implemented in SysVR4 as a sequence
of 4 longs used as a bitmask.  sv4r4_sigfillset has been broken for a
while, probably since rev 1.5.

This patch fixes SVR4_NSIG (i.e.: sets it to the actual number of signals,
instead of the number of bits in the mask) because some SysVR4 clients
honestly seem to care about whether bits in the signal mask are set for
non-existant signals.

Additionally, the svr4_sigfillset macro has been replaced by a
fully fledged function, because the macro didn't actually work
(it returned an all-ones mask, but we don't want that:  we want 0's
set where FreeBSD doesn't actually have a signal which is the same
as an SysVR4 signal, for example).

SysVR4 clients can now successfully ignore signals, although catching
them remains problematic (see commit log message for rev1.13 of
sys/i386/svr4/svr4_machdep.c for more info).

24 years agoRemove some all-too-wordy debugging prints
Mark Newton [Sat, 15 Jan 2000 15:30:44 +0000 (15:30 +0000)]
Remove some all-too-wordy debugging prints

24 years agoPre-freeze cleanups: Conditionalize debugging writes on DEBUG_SVR4,
Mark Newton [Sat, 15 Jan 2000 15:29:37 +0000 (15:29 +0000)]
Pre-freeze cleanups:  Conditionalize debugging writes on DEBUG_SVR4,
eliminate warnings, etc.

Note that svr4_setcontext() and svr4_getcontext() currently don't work
correctly, which makes returning from signal handlers somewhat problematic
(for reference:  the SysVR4 setcontext() and getcontext() syscalls behave
like a low-level version of setjmp() and longjmp(), in that they save and
restore process context.  SysVR4 uses this to implement its signal handler
trampoline:  The context which is saved before a signal handler is called
is restored by an implicit call to the setcontext() syscall when the signal
handler returns.  That currently doesn't work right in this emulator,
although it used to).

I'm committing this anyway, with a caveat that it's buggy, so that I can
(a) note the bug for anyone who is wondering about it, and (b) get the
stuff I've done to improve (but not fix) the situation in the tree before
4.0 is released.  If I have time to fix it totally RSN I'll talk to Jordan
to see if I can bend him into letting me commit a bugfix :-)

Note that the situation now is somewhat better than it was yesterday
anyway, because I've fixed the handling of svr4_sigsets which previously
was causing signal handlers to not get called at all.  Stay tuned for
an upcoming commit to svr4_signal.c...

24 years agoThis is the 3rd stage of syscons code reorganization.
Kazutaka YOKOTA [Sat, 15 Jan 2000 15:25:43 +0000 (15:25 +0000)]
This is the 3rd stage of syscons code reorganization.

- Split terminal emulation code from the main part of the driver so
that we can have alternative terminal emulator modules if we like in
the future.  (We are not quite there yet, though.)

- Put sysmouse related code in a separate file, thus, simplifying the
main part of the driver.

As some files are added to the source tree, you need to run config(8)
before you compile a new kernel next time.

You shouldn't see any functional change by this commit; this is only
internal code reorganization.

24 years agoInclude #warning about unconditionalizing VM86, which isn't supposed to work
Mark Newton [Sat, 15 Jan 2000 15:23:14 +0000 (15:23 +0000)]
Include #warning about unconditionalizing VM86, which isn't supposed to work
in the svr4 emulator yet.

24 years agoFixed the problem that IPsec connection hangs when bigger data is sent.
Yoshinobu Inoue [Sat, 15 Jan 2000 14:56:38 +0000 (14:56 +0000)]
Fixed the problem that IPsec connection hangs when bigger data is sent.
  -opt_ipsec.h was missing on some tcp files (sorry for basic mistake)
  -made buildable as above fix
  -also added some missing IPv4 mapped IPv6 addr consideration into
   ipsec4_getpolicybysock

24 years agogroff now is 1.15, not 1.11.
Alexey Zelkin [Sat, 15 Jan 2000 14:42:06 +0000 (14:42 +0000)]
groff now is 1.15, not 1.11.

24 years agoAdded missing 'else' for 'if (isipv6)' at IPv6 length setting in tcp_respond().
Yoshinobu Inoue [Sat, 15 Jan 2000 14:34:56 +0000 (14:34 +0000)]
Added missing 'else' for 'if (isipv6)' at IPv6 length setting in tcp_respond().
By this bug, IPv6 reset was not sent.
(I checked around same kind of bug, but no other found.)

24 years agoThis is another in Martin Blapp's N-series of mount-related cleanups :)
Brian Feldman [Sat, 15 Jan 2000 14:28:14 +0000 (14:28 +0000)]
This is another in Martin Blapp's N-series of mount-related cleanups :)
Changes are:
 - rpc.umntall is called at the right places now in /etc/rc*
 - rpc.umntall timeout has been lowered from two days (too high) to one
 - verbose messages in rpc.umntall have been clarified
 - kill double entries in /var/db/mounttab when rpc.umntall is invoked
 - ${early_nfs_mounts} has been removed from /etc/rc
 - patched mount(8) -p to print different pass/dump values for ufs filesystems.
   (last patch recieved from dan <bugg@bugg.strangled.net>)

Submitted by: Martin Blapp <mbr@imp.ch>, dan <bugg@bugg.strangled.net>

24 years agoFixed a bug in PCIC polling mode.
Tatsumi Hosokawa [Sat, 15 Jan 2000 12:02:03 +0000 (12:02 +0000)]
Fixed a bug in PCIC polling mode.
(it does not work without this fix)

24 years agoMore style fixes, both to code and man page.
Boris Popov [Sat, 15 Jan 2000 09:06:37 +0000 (09:06 +0000)]
More style fixes, both to code and man page.

Submitted by: Philippe Charnier <charnier@xp11.frmug.org>

24 years agoRevert previous change, which is reported to have broken world. I
Sheldon Hearn [Sat, 15 Jan 2000 09:00:03 +0000 (09:00 +0000)]
Revert previous change, which is reported to have broken world.  I
did test this through a ``make world'', but of course I already
had a working lint binary (one that does not call cpp -undef)
installed.

Reported by: "Pierre Y. Dampure" <Pierre.Dampure@barclayscapital.com>

24 years agoCheck if module was compiled without SMP support and running on
Boris Popov [Sat, 15 Jan 2000 08:35:48 +0000 (08:35 +0000)]
Check if module was compiled without SMP support and running on
an SMP system.

24 years agoAdd VT_NWFS tag.
Boris Popov [Sat, 15 Jan 2000 08:28:03 +0000 (08:28 +0000)]
Add VT_NWFS tag.

24 years agoAdd reference to SES device (it won't be in GENERIC unless folks clamor
Matt Jacob [Sat, 15 Jan 2000 07:46:39 +0000 (07:46 +0000)]
Add reference to SES device (it won't be in GENERIC unless folks clamor
for it). Document it's only option.

24 years agoRemove inclusion of dkbad.h- seems like a bug, but I had a stale
Matt Jacob [Sat, 15 Jan 2000 07:39:05 +0000 (07:39 +0000)]
Remove inclusion of dkbad.h- seems like a bug, but I had a stale
dkbad.h in /usr/include/sys- the local build process shouldn't have
been happy with that, but not finding dkbad.h in ../../sys. Hmm.

24 years agoClear ro->ro_rt just after RTFREE().
Yoshinobu Inoue [Sat, 15 Jan 2000 07:27:12 +0000 (07:27 +0000)]
Clear ro->ro_rt just after RTFREE().
  Pleases let me make sure that no one touch the invalid ro_rt pointer,
  after splx(s) and before next ro_rt initialization.
  Though usually this seems to be already called at splnet,
  I still sometime experience kernel crash at rtfree() in my
  INET6 enabled environment where IPv6 connection is frequently used.
  (Off-course, it might be just due to another bug.)

24 years agoadd options collateral for SES/SAF-TE driver
Matt Jacob [Sat, 15 Jan 2000 07:12:09 +0000 (07:12 +0000)]
add options collateral for SES/SAF-TE driver

24 years agoadd file collateral for SES/SAF-TE driver
Matt Jacob [Sat, 15 Jan 2000 07:11:36 +0000 (07:11 +0000)]
add file collateral for SES/SAF-TE driver

24 years agoadd functional but stub SES/SAF-TE driver
Matt Jacob [Sat, 15 Jan 2000 07:09:12 +0000 (07:09 +0000)]
add functional but stub SES/SAF-TE driver

24 years agoRemove compile warning not seen when compiling with target mode enabled.
Matt Jacob [Sat, 15 Jan 2000 07:06:44 +0000 (07:06 +0000)]
Remove compile warning not seen when compiling with target mode enabled.

24 years agoA processor with the CPUID of 0x?8? is Pentium III.
Seigo Tanimura [Sat, 15 Jan 2000 06:29:03 +0000 (06:29 +0000)]
A processor with the CPUID of 0x?8? is Pentium III.
(aka Coppermine)

Noticed by: Satoshi Sawada <k-sawata@gnoc2.comminet.or.jp>
Reviewd by: Takuma Yamada <fuzzy2@st.rim.or.jp>

24 years ago- Introduce a new section 'Multimedia' for sound/
Seigo Tanimura [Sat, 15 Jan 2000 06:25:41 +0000 (06:25 +0000)]
- Introduce a new section 'Multimedia' for sound/
  video capture cards.
- Add the description of newly supported sound cards.

24 years ago- Add the man pages for sound card bridge drivers.
Seigo Tanimura [Sat, 15 Jan 2000 06:25:09 +0000 (06:25 +0000)]
- Add the man pages for sound card bridge drivers.
  (sbc(4), gusc(4), csa(4))
- Update pcm(4) for newpcm.

24 years agoFix the 3Com Fast Etherlink 3C574TX's ID string as we now get the
David E. O'Brien [Sat, 15 Jan 2000 05:45:38 +0000 (05:45 +0000)]
Fix the 3Com Fast Etherlink 3C574TX's ID string as we now get the
complete description from the PCCARD system.