]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
25 years agoChange parameters of unlockrange to allow calling from revive_block,
Greg Lehey [Tue, 24 Aug 1999 02:23:21 +0000 (02:23 +0000)]
Change parameters of unlockrange to allow calling from revive_block,
which doesn't use request groups.

25 years agoAdd declaration for vinum_setstate_by_force.
Greg Lehey [Tue, 24 Aug 1999 02:21:46 +0000 (02:21 +0000)]
Add declaration for vinum_setstate_by_force.

25 years agogive_sd_to_drive: If we don't have space, deallocate the subdisk.
Greg Lehey [Tue, 24 Aug 1999 02:20:30 +0000 (02:20 +0000)]
give_sd_to_drive: If we don't have space, deallocate the subdisk.
                  Previously we could end up with a valid subdisk
                  entry pointing to offset -1 on the drive.

25 years agovinumclose: close raw plexes and subdisks.
Greg Lehey [Tue, 24 Aug 1999 02:18:55 +0000 (02:18 +0000)]
vinumclose: close raw plexes and subdisks.

25 years agoAdd support for logging subdisk I/O.
Greg Lehey [Tue, 24 Aug 1999 02:17:42 +0000 (02:17 +0000)]
Add support for logging subdisk I/O.

Change parameters of unlockrange to allow calling from revive_block,
which doesn't use request groups.

25 years agoThis commit was generated by cvs2svn to compensate for changes in r50276,
Peter Wemm [Tue, 24 Aug 1999 01:06:48 +0000 (01:06 +0000)]
This commit was generated by cvs2svn to compensate for changes in r50276,
which included commits to RCS files with non-trunk default branches.

25 years agoImport unmodified (but trimmed) ncurses 5.0 prerelease 990821.
Peter Wemm [Tue, 24 Aug 1999 01:06:48 +0000 (01:06 +0000)]
Import unmodified (but trimmed) ncurses 5.0 prerelease 990821.
This contains the full eti (panel, form, menu) extensions.
bmake glue to follow.

Obtained from: ftp://ftp.clark.net/pub/dickey/ncurses

25 years agoCast pointers to uintptr_t instead of casting them to u_long, and/or vice
Bruce Evans [Tue, 24 Aug 1999 00:56:50 +0000 (00:56 +0000)]
Cast pointers to uintptr_t instead of casting them to u_long, and/or vice
versa.  Cosmetic.

25 years agoCast pointers to [u]intptr_t instead of casting them to [u_]long. Don't
Bruce Evans [Tue, 24 Aug 1999 00:48:19 +0000 (00:48 +0000)]
Cast pointers to [u]intptr_t instead of casting them to [u_]long.  Don't
depend on gcc's feature of casting lvalues, especially for direct
assignment where it doesn't even simplify the syntax.  Cosmetic.

25 years agoCast pointers to uintptr_t instead of casting them to u_long. They
Bruce Evans [Tue, 24 Aug 1999 00:38:52 +0000 (00:38 +0000)]
Cast pointers to uintptr_t instead of casting them to u_long.  They
are still converted to u_long by assignment of the uintptr_t, and
address calculations are still done using u_long.  This is OK for
currently supported machines, but addresses should be represented
by vm_offset_t or uintptr_t in case pointers are longer than longs.

"Fixed" size of linker_path[].  MAXPATHLEN + 1 was 1 too large for
search paths with only one file path in them, but much too small
for search paths with several long file paths in them.

25 years agoFixed a misplaced cast to uintptr_t. Cosmetic.
Bruce Evans [Tue, 24 Aug 1999 00:22:10 +0000 (00:22 +0000)]
Fixed a misplaced cast to uintptr_t.  Cosmetic.

Use device_get_nameunit() instead of rolling our own.

25 years agoCast pointers to uintptr_t instead of casting them to u_long. Don't
Bruce Evans [Tue, 24 Aug 1999 00:04:45 +0000 (00:04 +0000)]
Cast pointers to uintptr_t instead of casting them to u_long.  Don't
depend on gcc's feature of casting lvalues.  Cosmetic.

25 years agoUse devtoname to print dev_t's instead of casting them to u_long for
Bruce Evans [Mon, 23 Aug 1999 23:55:03 +0000 (23:55 +0000)]
Use devtoname to print dev_t's instead of casting them to u_long for
misprinting with %lx.

Cast pointers to intptr_t instead of casting them to long.  Cosmetic.

25 years ago`bootdev' is an ordinary u_long, so don't cast it to a pointer to print it.
Bruce Evans [Mon, 23 Aug 1999 23:42:00 +0000 (23:42 +0000)]
`bootdev' is an ordinary u_long, so don't cast it to a pointer to print it.
gcc warns about the cast on i386's with 64-bit longs.

Print `bootdev' in all cases when we bail out because it is unreasonable.

25 years agoCosmetic: bring closer to RELENG_3
Brian Somers [Mon, 23 Aug 1999 23:13:59 +0000 (23:13 +0000)]
Cosmetic:  bring closer to RELENG_3

25 years agoImplement a version of s_lock_try that doesn't cause the next s_lock
Alan Cox [Mon, 23 Aug 1999 22:25:11 +0000 (22:25 +0000)]
Implement a version of s_lock_try that doesn't cause the next s_lock
call to panic when SL_DEBUG is set.  (SL_DEBUG is currently set
by default.)

25 years agoOops, the previous commit was missing a new include.
Bruce Evans [Mon, 23 Aug 1999 22:05:49 +0000 (22:05 +0000)]
Oops, the previous commit was missing a new include.

25 years agoComplain if this file is included.
Poul-Henning Kamp [Mon, 23 Aug 1999 21:51:21 +0000 (21:51 +0000)]
Complain if this file is included.

25 years agoNow that we can bind cdevsw to the individual dev_t, divorce the PERFMON
Poul-Henning Kamp [Mon, 23 Aug 1999 21:34:23 +0000 (21:34 +0000)]
Now that we can bind cdevsw to the individual dev_t, divorce the PERFMON
stuff from mem.c.  If PERFMON is there, it will "steal" a minor from
mem.c, but mem.c doesn't need to know about this.

Fixed type of cmd argument in perfmon_ioctl().

25 years agoInitialise fsids with (user) device numbers again. Bitrot when dev_t's
Bruce Evans [Mon, 23 Aug 1999 21:07:13 +0000 (21:07 +0000)]
Initialise fsids with (user) device numbers again.  Bitrot when dev_t's
were changed to pointers was obscured by casting dev_t's to longs.
fsids haven't even been comprised of longs since the Lite2 merge.

25 years agoReset the UHCI controller when the device comes back from suspend.
Nick Hibma [Mon, 23 Aug 1999 21:00:08 +0000 (21:00 +0000)]
Reset the UHCI controller when the device comes back from suspend.

This should be replaced by proper support for suspend one day (global
suspend).

Submitted-by: Christopher Masto <chris@netmonger.net>
25 years agoConvert DEVFS hooks in (most) drivers to make_dev().
Poul-Henning Kamp [Mon, 23 Aug 1999 20:59:21 +0000 (20:59 +0000)]
Convert DEVFS hooks in (most) drivers to make_dev().

Diskslice/label code not yet handled.

Vinum, i4b, alpha, pc98 not dealt with (left to respective Maintainers)

Add the correct hook for devfs to kern_conf.c

The net result of this excercise is that a lot less files depends on DEVFS,
and devtoname() gets more sensible output in many cases.

A few drivers had minor additional cleanups performed relating to cdevsw
registration.

A few drivers don't register a cdevsw{} anymore, but only use make_dev().

25 years agoUse devtoname() to print dev_t's instead of casting them to long or u_long
Bruce Evans [Mon, 23 Aug 1999 20:35:21 +0000 (20:35 +0000)]
Use devtoname() to print dev_t's instead of casting them to long or u_long
for misprinting in %lx format.

25 years agoThe nexus_attach() code works a lot better if it's actually connected to
Peter Wemm [Mon, 23 Aug 1999 19:23:33 +0000 (19:23 +0000)]
The nexus_attach() code works a lot better if it's actually connected to
the device methods... Also, don't fail to add eisa/isa because a previous
device failed to attach.

25 years agoModify the macros IMASK_UNLOCK, CPL_UNLOCK, and REL_FAST_INTR_LOCK
Alan Cox [Mon, 23 Aug 1999 19:14:18 +0000 (19:14 +0000)]
Modify the macros IMASK_UNLOCK, CPL_UNLOCK, and REL_FAST_INTR_LOCK
to perform the s_unlock inline.

25 years agoAdd 'options MFS_ROOT' back in to kernel configs -- not going to get far
Doug White [Mon, 23 Aug 1999 18:49:52 +0000 (18:49 +0000)]
Add 'options MFS_ROOT' back in to kernel configs -- not going to get far
without that!

25 years agoCorrect the inconsistent formatting in struct vm_map.
Alan Cox [Mon, 23 Aug 1999 18:16:05 +0000 (18:16 +0000)]
Correct the inconsistent formatting in struct vm_map.

Addendum to rev 1.47: submitted by dillon.

25 years agostruct vm_map:
Alan Cox [Mon, 23 Aug 1999 18:08:34 +0000 (18:08 +0000)]
struct vm_map:
The lock structure cannot be the first element of the vm_map
because this can result in livelock between two or more system
processes trying to kmem_alloc_wait.

25 years agoRecognise the GPS receiver from Trimble
Nick Hibma [Mon, 23 Aug 1999 14:49:22 +0000 (14:49 +0000)]
Recognise the GPS receiver from Trimble
in NMEA mode (without date).

        Trimble OEM Ace-II receiver. Low cost PCB with single
        voltage input, external active antenna and two serial
        ports with either NMEA and ITAPs output. Programmable
        to be tuned for 'time' accuracy in fixed station config.

Author of tool notified.

Submitted by: Dirk van Gulik <dirk.vangulik@jrc.it>

25 years ago- Make binary size smaller.
KATO Takenori [Mon, 23 Aug 1999 13:54:35 +0000 (13:54 +0000)]
- Make binary size smaller.
- Fix typo in TDK/LANX stuff.

Submitted by: chi@bd.mbn.or.jp (Chiharu Shibata)

25 years agoFix a mistake in my last SA_SIGINFO commit. Processes could block
Martin Cracauer [Mon, 23 Aug 1999 13:53:25 +0000 (13:53 +0000)]
Fix a mistake in my last SA_SIGINFO commit. Processes could block
SIGKILL and SIGSTOP.

PR: kern/13293
Submitted by: dwmalone@maths.tcd.ie
Obtained from: PR had correct fix

25 years agoUpdated to new keyboard driver.
KATO Takenori [Mon, 23 Aug 1999 13:50:08 +0000 (13:50 +0000)]
Updated to new keyboard driver.

Submitted by: yokota & nyan

25 years agoMerge from sys/dev/syscons/syscons.c revision 1.319.
KATO Takenori [Mon, 23 Aug 1999 13:45:28 +0000 (13:45 +0000)]
Merge from sys/dev/syscons/syscons.c revision 1.319.

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

25 years agoMerge from sys/i386/isa/npx.c revision 1.76.
KATO Takenori [Mon, 23 Aug 1999 13:42:22 +0000 (13:42 +0000)]
Merge from sys/i386/isa/npx.c revision 1.76.

25 years agoMerge from sys/i386/i386/userconfig.c revision 1.153.
KATO Takenori [Mon, 23 Aug 1999 13:41:00 +0000 (13:41 +0000)]
Merge from sys/i386/i386/userconfig.c revision 1.153.

25 years agoMerge from sys/i386/conf/options.i386 revision .1.20.
KATO Takenori [Mon, 23 Aug 1999 13:39:32 +0000 (13:39 +0000)]
Merge from sys/i386/conf/options.i386 revision .1.20.

25 years agoFix a few problems on the Alpha, and fix -Wall while here.
Peter Wemm [Mon, 23 Aug 1999 12:59:03 +0000 (12:59 +0000)]
Fix a few problems on the Alpha, and fix -Wall while here.
Before:
isab0@pci0:8:0: class=0x060100 card=0x00000000 chip=0xffffffffc6931080 ...
After:
isab0@pci0:8:0: class=0x060100 card=0x00000000 chip=0xc6931080 ...

25 years agoCorrect xref: setlocal -> setlocale(3).
Sheldon Hearn [Mon, 23 Aug 1999 12:55:19 +0000 (12:55 +0000)]
Correct xref: setlocal -> setlocale(3).

PR: 13321
Submitted by: Alexey Zelkin <phantom@cris.net>

25 years agoDisable some apparently stray debug printfs:
Peter Wemm [Mon, 23 Aug 1999 12:08:45 +0000 (12:08 +0000)]
Disable some apparently stray debug printfs:
ahc0: <Adaptec 274X SCSI host adapter> at 0x1c00-0x1cff, irq 11 (edge)
ahc0: on eisa0 slot 1
ahc0: aic7770 >= Rev E, SBLKCTL = 0x8
SSTAT0 = 0x0
SFUNCT = 0x0
Twin Channel, A SCSI Id=7, B SCSI Id=7, primary A, 4/255 SCBs

Not objected to by: gibbs

25 years agoDocument ENOSYS error.
Chris Costello [Mon, 23 Aug 1999 11:07:29 +0000 (11:07 +0000)]
Document ENOSYS error.

PR: docs/13290

25 years agoAdd new option to fdisk: -e
Poul-Henning Kamp [Mon, 23 Aug 1999 11:06:19 +0000 (11:06 +0000)]
Add new option to fdisk: -e

This wipes the MBR and creates slice 1 as a FreeBSD slice covering the
disk starting from the second track to the cylinder aligned end of the disk.

This is the most compatibly layout we have as far as I know.

25 years agoAdd missing files from previous commits.
Luigi Rizzo [Mon, 23 Aug 1999 11:04:23 +0000 (11:04 +0000)]
Add missing files from previous commits.

25 years agoAdd missing file from previous commit
Luigi Rizzo [Mon, 23 Aug 1999 11:02:08 +0000 (11:02 +0000)]
Add missing file from previous commit

25 years agoRemove some files probably forgotten from previous commit.
Luigi Rizzo [Mon, 23 Aug 1999 10:23:32 +0000 (10:23 +0000)]
Remove some files probably forgotten from previous commit.

25 years agoAdd missing files probably forgotten in previous commit
Luigi Rizzo [Mon, 23 Aug 1999 10:19:15 +0000 (10:19 +0000)]
Add missing files probably forgotten in previous commit

25 years agoAdd missing files probably forgotten in previous commit
Luigi Rizzo [Mon, 23 Aug 1999 10:13:06 +0000 (10:13 +0000)]
Add missing files probably forgotten in previous commit

25 years agoAdd missing file probably forgotten from previous commit
Luigi Rizzo [Mon, 23 Aug 1999 10:05:20 +0000 (10:05 +0000)]
Add missing file probably forgotten from previous commit

25 years agoFix the breakage caused by hanging softc off dev_t. This is only
Poul-Henning Kamp [Mon, 23 Aug 1999 09:35:12 +0000 (09:35 +0000)]
Fix the breakage caused by hanging softc off dev_t.  This is only
a workaround.

25 years agoMore cleanup: move device entries in Picobsd makefiles in a more
Luigi Rizzo [Mon, 23 Aug 1999 08:52:05 +0000 (08:52 +0000)]
More cleanup: move device entries in Picobsd makefiles in a more
accessible place.

25 years agoAdditional cleanup of the picobsd scripts.
Luigi Rizzo [Mon, 23 Aug 1999 08:33:08 +0000 (08:33 +0000)]
Additional cleanup of the picobsd scripts.
Also bump Version to 0.445

25 years agoBacked out my -n change to imply -l by request of sheldonh.
Chris Costello [Mon, 23 Aug 1999 05:39:40 +0000 (05:39 +0000)]
Backed out my -n change to imply -l by request of sheldonh.

25 years agoFix script in README to actually work, empty lines produce as
Rodney W. Grimes [Mon, 23 Aug 1999 05:12:43 +0000 (05:12 +0000)]
Fix script in README to actually work, empty lines produce as
much a diff as lines with the wrong stuff on it.

Add references in mtree data files to the README.

25 years agoMake -n flag compliant to the Single Unix Specification.
Chris Costello [Mon, 23 Aug 1999 01:17:58 +0000 (01:17 +0000)]
Make -n flag compliant to the Single Unix Specification.

To quote their ls(1) specification:

   -n
          The same as -l, except that the owner's UID and GID numbers are
          written, rather than the associated character strings.

Reviewed by: green

25 years agoAn explination of how to keep these files in good working order.
Rodney W. Grimes [Mon, 23 Aug 1999 01:16:26 +0000 (01:16 +0000)]
An explination of how to keep these files in good working order.

Reviewed by: bde

25 years agoRemove cross-reference to removed man page od.4
Chris Costello [Sun, 22 Aug 1999 23:52:01 +0000 (23:52 +0000)]
Remove cross-reference to removed man page od.4

sd(4) reference was also removed in the sentence:
  "In general the interfaces are similar to
   those described by wd(4) sd(4) and od(4)."

25 years agoThe previous fix didn't do anything if you didn't have pnp. The ICU
Peter Wemm [Sun, 22 Aug 1999 23:49:00 +0000 (23:49 +0000)]
The previous fix didn't do anything if you didn't have pnp.  The ICU
macros are only called in the !APIC_IO case, include icu.h there.

25 years agoFinish unbreaking autoconf.c includes (for non-SMP.)
Brian Feldman [Sun, 22 Aug 1999 23:44:33 +0000 (23:44 +0000)]
Finish unbreaking autoconf.c includes (for non-SMP.)

25 years agoRemove od.4 -- Kenneth Merry writes:
Chris Costello [Sun, 22 Aug 1999 23:40:41 +0000 (23:40 +0000)]
Remove od.4 -- Kenneth Merry writes:

 It's not supported any more.  It was never ported to CAM, and that
 functionality has been taken over by the da driver.  So the man page can be
 removed.

Reviewed by: ken

25 years agoAallow ppp to work with Nortel Networks Extranet Switch
Brian Somers [Sun, 22 Aug 1999 23:32:01 +0000 (23:32 +0000)]
Aallow ppp to work with Nortel Networks Extranet Switch
product and Windows NT tunneling.

Submitted by: Chain Lee <chain@nortelnetworks.com>

25 years agoppp_alias -> ppp_nat
Brian Somers [Sun, 22 Aug 1999 23:26:05 +0000 (23:26 +0000)]
ppp_alias -> ppp_nat

Submitted by: Josef L. Karthauser <joe@FreeBSD.org.uk>

25 years agoDecrease the isdn recovery time to 1 second and mention that it
Brian Somers [Sun, 22 Aug 1999 23:18:42 +0000 (23:18 +0000)]
Decrease the isdn recovery time to 1 second and mention that it
should always be less than ppps redial timeout.

25 years agoNOALIAS -> NONAT
Brian Somers [Sun, 22 Aug 1999 23:17:57 +0000 (23:17 +0000)]
NOALIAS -> NONAT
Reminded by: jkh

25 years agoMake a comment I added a bit nicer.
Brian Feldman [Sun, 22 Aug 1999 22:32:41 +0000 (22:32 +0000)]
Make a comment I added a bit nicer.

25 years agoFix a small typo (missing comma) to make the alpha kernel build again
Andrew Gallatin [Sun, 22 Aug 1999 21:47:32 +0000 (21:47 +0000)]
Fix a small typo (missing comma) to make the alpha kernel build again

25 years agoMajor PicoBSD update by luigi with some debugging help from myself.
Doug White [Sun, 22 Aug 1999 21:45:24 +0000 (21:45 +0000)]
Major PicoBSD update by luigi with some debugging help from myself.

. Now builds on -STABLE (-CURRENT is broken due to bugs)
. etc directory contents centralized instead of in each
  type directory (can exclude & override as desired)
. Removed extraneous language files (lang files for rc really necessary?)
. dialog-based build tool with support for custom floppy builds
. MFS image loads as a mfs_root module instead of compiled into kernel

THIS IS BROKEN ON CURRENT.  I'll MFC to -STABLE immediately following.
luigi tells me I have an OK from jkh on the MFC.

Submitted by: luigi

25 years agoOops, that wasn't so clever after all. struct isa_device is still a
Peter Wemm [Sun, 22 Aug 1999 21:44:35 +0000 (21:44 +0000)]
Oops, that wasn't so clever after all.  struct isa_device is still a
prerequisite for this old pnp.h.

25 years agoZap a heap of unused cruft now. We don't need the ISA/EISA/PCI hooks
Peter Wemm [Sun, 22 Aug 1999 19:59:29 +0000 (19:59 +0000)]
Zap a heap of unused cruft now.  We don't need the ISA/EISA/PCI hooks
here any more as they are self identifying.  Only PNP remains but that
will be replaced any day now.
Also reword a comment that had been XXX'ed to death to make it clear[er]
why we don't enable interrupts before probing.
PCIBIOS interrupt routing controls may make this possible to fix one day.

25 years agoTake advantage of the apm/npx code and let them identify themselves rather
Peter Wemm [Sun, 22 Aug 1999 19:56:55 +0000 (19:56 +0000)]
Take advantage of the apm/npx code and let them identify themselves rather
than having explicit hooks here.
Treat the eisa/isa attach a little differently so that we defer the
decision about to attach eisa/isa to the motherboard directly only if
the PCI probe (if it exists) fails to turn up a PCI->EISA/ISA bridge.
This restores the original device geometry where ISA and/or EISA attach
to their bridge rather than bypassing and going to the root.

25 years agoMake the identify routine add itself with priority 100 to make sure it
Peter Wemm [Sun, 22 Aug 1999 19:54:11 +0000 (19:54 +0000)]
Make the identify routine add itself with priority 100 to make sure it
goes after the npx/apm devices and any other motherboard devices that
may get added down the track.

25 years agoAdd an identify method to allow npx to arrange itself to be attached to
Peter Wemm [Sun, 22 Aug 1999 19:52:51 +0000 (19:52 +0000)]
Add an identify method to allow npx to arrange itself to be attached to
the nexus without explicit code in the nexus to do so.

25 years agoAdd an identify method to allow apm to attach itself to the nexus
Peter Wemm [Sun, 22 Aug 1999 19:51:33 +0000 (19:51 +0000)]
Add an identify method to allow apm to attach itself to the nexus
without hooks in there.

25 years agomake pnp.h self contained, for what good it'll do since it's days are
Peter Wemm [Sun, 22 Aug 1999 19:46:05 +0000 (19:46 +0000)]
make pnp.h self contained, for what good it'll do since it's days are
numbered (new pnp-aware isa code in the pipeline).

25 years agoGrammar and formatting touch-ups to comments.
Tim Vanderhoek [Sun, 22 Aug 1999 19:42:40 +0000 (19:42 +0000)]
Grammar and formatting touch-ups to comments.

25 years agoTypo: 102 => 192 (PR: docs/13310 - Maxim Sobolev <sobomax@altavista.net>)
Tim Vanderhoek [Sun, 22 Aug 1999 19:23:33 +0000 (19:23 +0000)]
Typo: 102 => 192  (PR: docs/13310 - Maxim Sobolev <sobomax@altavista.net>)

25 years agoSlightly sanitize the isa/eisa bridge detection. The pci->eisa bridge
Peter Wemm [Sun, 22 Aug 1999 19:10:39 +0000 (19:10 +0000)]
Slightly sanitize the isa/eisa bridge detection.  The pci->eisa bridge
logic (I use the term loosely) was a bit whacky.

25 years agoThe vty monster has feasted and now provides 8 vtys by default.
Tim Vanderhoek [Sun, 22 Aug 1999 19:09:19 +0000 (19:09 +0000)]
The vty monster has feasted and now provides 8 vtys by default.

PR: docs/13310 (Maxim Soboleva <sobomax@altavista.ne>)

25 years agoCatch an at_shutdown I didn't think of in the last batch.
Mike Smith [Sun, 22 Aug 1999 18:50:43 +0000 (18:50 +0000)]
Catch an at_shutdown I didn't think of in the last batch.

Note to self: port to AS2100 so as to have a test system.

Submitted by: jkh

25 years agoUse the final version of the patch for the overflow, not the next to
Warner Losh [Sun, 22 Aug 1999 17:24:25 +0000 (17:24 +0000)]
Use the final version of the patch for the overflow, not the next to
final.

Noticed by: eivind

25 years agoGo back to using microtime() to get the timestamps for {f,l,}utimes(path,
John Polstra [Sun, 22 Aug 1999 16:50:30 +0000 (16:50 +0000)]
Go back to using microtime() to get the timestamps for {f,l,}utimes(path,
NULL) for now.  Bruce says I jumped the gun with my change in
revision 1.131, or maybe it should use nanotime(), or maybe it
shouldn't be decided in the VFS layer at all.  I'm leaving it with
the old behavior until the Trans-Pacific Internet Vulcan Mind Meld
yields fuller understanding.

25 years agoMention that the 3Com 3cSOHO100-TX adapter (cheap version of the 3c905B) is
Bill Paul [Sun, 22 Aug 1999 16:34:52 +0000 (16:34 +0000)]
Mention that the 3Com 3cSOHO100-TX adapter (cheap version of the 3c905B) is
supported, as well as the 3c980B server adapter (newer rev of the 3c980,
same PCI ID) and the 3c905C-TX.

25 years ago1) Add HEADER section. Similar to CODE.
Nick Hibma [Sun, 22 Aug 1999 16:21:29 +0000 (16:21 +0000)]
1) Add HEADER section. Similar to CODE.
2) deindent the two sections (CODE and HEADER), based on the indent found
   on the first line of the section; to make the output more readable.

25 years agoPut full month names in another form ("genitive case"), the only form
Dmitrij Tejblum [Sun, 22 Aug 1999 16:12:13 +0000 (16:12 +0000)]
Put full month names in another form ("genitive case"), the only form
suitable for printing dates (like in strftime(... "%e %B %Y" ...)).

25 years agoModify yp_mkdb so that it only checks for bogus '+' and '-' characters
Bill Paul [Sun, 22 Aug 1999 15:34:47 +0000 (15:34 +0000)]
Modify yp_mkdb so that it only checks for bogus '+' and '-' characters
in source input if the -f flag is used, and modify Makefile.yp to only
use -f for the passwd, master.passwd and group maps. These should be
the only ones for which the + and - characters have special meaning
that make it important for us to avoid letting them into any of the map
databases. In some cases (namely the automounter maps) we have to allow
at least the - character through in order to create the map properly.

This closes PR #8699.

25 years agoAdd a missing underscore that made EVENTHANDLER_FAST_INVOKE broken.
Brian Feldman [Sun, 22 Aug 1999 15:21:52 +0000 (15:21 +0000)]
Add a missing underscore that made EVENTHANDLER_FAST_INVOKE broken.

25 years agoFix `key release event prevent suspend' problem. We don't need
Mitsuru IWASAKI [Sun, 22 Aug 1999 14:48:00 +0000 (14:48 +0000)]
Fix `key release event prevent suspend' problem.  We don't need
`sleep 1; zzz' trick now.

- APM BIOS Call for suspend/standby now should be issued with delay.
- Delay for suspend/standby can be adjusted by using sysctl(8) interface
  (eg. sysctl -w machdep.apm_suspend_delay=3).

25 years ago- Remove cdevsw entry points in individual keyboard drivers;
Kazutaka YOKOTA [Sun, 22 Aug 1999 09:52:33 +0000 (09:52 +0000)]
- Remove cdevsw entry points in individual keyboard drivers;
  instead, use generic entry points for all drivers.
- Eliminate bogus makedev().
- Eliminate softc in the lower drivers, as it is no longer necessary.

Submitted (95%) by: phk

25 years agoRevise description on moused options:
Kazutaka YOKOTA [Sun, 22 Aug 1999 06:12:58 +0000 (06:12 +0000)]
Revise description on moused options:
- Clearly distinguish PS/2, bus, and serial protocols.
- Explicitly state that serial mouse protocols don't work with
  the PS/2 and bus mice.

25 years agoImprove the detection code for GlidePoint. This is still a kludge,
Kazutaka YOKOTA [Sun, 22 Aug 1999 06:11:52 +0000 (06:11 +0000)]
Improve the detection code for GlidePoint.  This is still a kludge,
but better than before...

PR: kern/13215

25 years agoMake "s_unlock" an inline function. (Inlining this function takes
Alan Cox [Sun, 22 Aug 1999 05:37:18 +0000 (05:37 +0000)]
Make "s_unlock" an inline function.  (Inlining this function takes
less space than calling it.  A callable version still exists for
use by some assembly code.)

25 years agoFix a temp file race occurring only when -j is used.
Tim Vanderhoek [Sun, 22 Aug 1999 05:28:13 +0000 (05:28 +0000)]
Fix a temp file race occurring only when -j is used.

Noticed by: ru
Obtained from: OpenBSD

25 years agoFix assorted style problems and bring up to date.
Jordan K. Hubbard [Sun, 22 Aug 1999 05:14:23 +0000 (05:14 +0000)]
Fix assorted style problems and bring up to date.

Submitted by: Alexey M. Zelkin <phantom@cris.net>
PR: 13271

25 years agoFix potential buffer overflow when using gtags.
Tim Vanderhoek [Sun, 22 Aug 1999 03:48:31 +0000 (03:48 +0000)]
Fix potential buffer overflow when using gtags.

Submitted by: Shigio Yamaguchi [3]shigio@wafu.netgate.net (gtags author)
PR: bin/7607

25 years agoAssorted bug fixes.
Kazutaka YOKOTA [Sun, 22 Aug 1999 03:31:13 +0000 (03:31 +0000)]
Assorted bug fixes.

keyboard.c
- Call tcsetattr() in VGLKeyboardEnd() to restore tty, only when
  tty attributes have been previously saved.
  PR: misc/9524
  Submitted by: Katusyuki 'kei' Maeda (kei@nanet.co.jp)
- Set up the tty raw mode correctly.

main.c
- Restore VESA_800x600 raster text mode correctly in VGLEnd().
  Submitted by: des

text.c
- Allocate the correct size of a font buffer in VGLSetFontFile().
  I forgot the submitter ;-(

simple.c, bitmap.c
- Fix address calculation for the VGA mode X in VGLGetXY() and
  VGLBitmapCopy().
- Fix typo (dsty -> dstx) in __VGLBitmapCopy().

Reviewed by: sos

25 years agoUse the new vfs_timestamp() function to create the timestamps used
John Polstra [Sun, 22 Aug 1999 01:46:57 +0000 (01:46 +0000)]
Use the new vfs_timestamp() function to create the timestamps used
by utimes(path, NULL).  This gives them the same precision as the
timestamps produced by write operations.  Do likewise for lutimes()
and futimes().

Suggested by bde.

25 years agoAllow authkey to be changed independently of the current phase.
Brian Somers [Sun, 22 Aug 1999 01:29:53 +0000 (01:29 +0000)]
Allow authkey to be changed independently of the current phase.

25 years agoCrap, I knew I was going to forget something: add missing miibus method
Bill Paul [Sun, 22 Aug 1999 00:56:39 +0000 (00:56 +0000)]
Crap, I knew I was going to forget something: add missing miibus method
description file which slipped through the cracks.

Pointed out by: Doug <Doug@gorean.org>

25 years agoSupport full-precision file timestamps. Until now, only the seconds
John Polstra [Sun, 22 Aug 1999 00:15:16 +0000 (00:15 +0000)]
Support full-precision file timestamps.  Until now, only the seconds
have been maintained, and that is still the default.  A new sysctl
variable "vfs.timestamp_precision" can be used to enable higher
levels of precision:

      0 = seconds only; nanoseconds zeroed (default).
      1 = seconds and nanoseconds, accurate within 1/HZ.
      2 = seconds and nanoseconds, truncated to microseconds.
    >=3 = seconds and nanoseconds, maximum precision.

Level 1 uses getnanotime(), which is fast but can be wrong by up
to 1/HZ.  Level 2 uses microtime().  It might be desirable for
consistency with utimes() and friends, which take timeval structures
rather than timespecs.  Level 3 uses nanotime() for the higest
precision.

I benchmarked levels 0, 1, and 3 by copying a 550 MB tree with
"cpio -pdu".  There was almost negligible difference in the system
times -- much less than 1%, and less than the variation among
multiple runs at the same level.  Bruce Evans dreamed up a torture
test involving 1-byte reads with intervening fstat() calls, but
the cpio test seems more realistic to me.

This feature is currently implemented only for the UFS (FFS and
MFS) filesystems.  But I think it should be easy to support it in
the others as well.

An earlier version of this was reviewed by Bruce.  He's not to
blame for any breakage I've introduced since then.

Reviewed by: bde (an earlier version of the code)

25 years agoRemove two unused variable declarations.
Alan Cox [Sun, 22 Aug 1999 00:01:46 +0000 (00:01 +0000)]
Remove two unused variable declarations.

25 years agoClean up after removing sys/eventhandler.h from sys/systm.h at the last
Mike Smith [Sat, 21 Aug 1999 22:10:49 +0000 (22:10 +0000)]
Clean up after removing sys/eventhandler.h from sys/systm.h at the last
minute.  This should cover all of the missed cases (and should let LINT
build again).

25 years agoDon't print a "," after the last superblock.
Bill Fumerola [Sat, 21 Aug 1999 22:07:27 +0000 (22:07 +0000)]
Don't print a "," after the last superblock.

Submitted by: adrian