]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
20 years agoAdd fgetwln(), a wide character version of fgetln().
Tim J. Robbins [Fri, 16 Jul 2004 06:06:09 +0000 (06:06 +0000)]
Add fgetwln(), a wide character version of fgetln().

20 years agoRename slbexpand() to __slbexpand() and make it available outside
Tim J. Robbins [Fri, 16 Jul 2004 05:52:51 +0000 (05:52 +0000)]
Rename slbexpand() to __slbexpand() and make it available outside
of fgetln.c (non-static).

20 years agoAvoid passing negative values to <ctype.h> functions on machines with
Tim J. Robbins [Fri, 16 Jul 2004 05:10:46 +0000 (05:10 +0000)]
Avoid passing negative values to <ctype.h> functions on machines with
signed chars.

20 years agoDocument incorrect handling of multibyte characters.
Tim J. Robbins [Fri, 16 Jul 2004 05:08:16 +0000 (05:08 +0000)]
Document incorrect handling of multibyte characters.

20 years agoUpdate config.h to account for the prgregset_t and psaddr_t types
Marcel Moolenaar [Fri, 16 Jul 2004 05:03:42 +0000 (05:03 +0000)]
Update config.h to account for the prgregset_t and psaddr_t types
that have been added to <sys/procfs.h>. This change has no effect
because the source file that would be affected is not compiled on
FreeBSD. Hence, this is for completeness only.

20 years agoAdd the change of the sound drivers.
Seigo Tanimura [Fri, 16 Jul 2004 04:04:29 +0000 (04:04 +0000)]
Add the change of the sound drivers.

20 years agoRename the sound device drivers:
Seigo Tanimura [Fri, 16 Jul 2004 04:00:08 +0000 (04:00 +0000)]
Rename the sound device drivers:

- `sound'
  The generic sound driver, always required.

- `snd_*'
  Device-dependent drivers, named after the sound module names.
  Configure accordingly to your hardware.

In addition, rename the `snd_pcm' module to `sound' in order to sync
with the driver names.

Suggested by: cg

20 years agoRemove unused fields from the pmap.
Alan Cox [Fri, 16 Jul 2004 03:42:45 +0000 (03:42 +0000)]
Remove unused fields from the pmap.

20 years agoAs per discussions on current@, protect unsuspecting users from
Marcel Moolenaar [Fri, 16 Jul 2004 02:51:28 +0000 (02:51 +0000)]
As per discussions on current@, protect unsuspecting users from
trying to upgrade their system with make world instead of following
the preferred and suggested sequence of commands. The fact remains
that make world does not upgrade the kernel.
Allow make world when DESTDIR has been specified, including when
DESTDIR specifies the root file system. Otherwise, print a useful
warning and fail.

Reviewed, tested and scrutinized by: gad@

20 years agoUpdate Biba and MLS man pages to take into account recent renaming of
Robert Watson [Fri, 16 Jul 2004 02:04:41 +0000 (02:04 +0000)]
Update Biba and MLS man pages to take into account recent renaming of
the 'single' label element to 'effective.

20 years agoRename Biba and MLS _single label elements to _effective, which more
Robert Watson [Fri, 16 Jul 2004 02:03:50 +0000 (02:03 +0000)]
Rename Biba and MLS _single label elements to _effective, which more
accurately represents the intention of the 'single' label element in
Biba and MLS labels.  It also approximates the use of 'effective' in
traditional UNIX credentials, and avoids confusion with 'singlelabel'
in the context of file systems.

Inspired by: trhodes

20 years agoClarify getfsstat(2) usage.
Alfred Perlstein [Fri, 16 Jul 2004 01:18:13 +0000 (01:18 +0000)]
Clarify getfsstat(2) usage.

The getfsstat(2) function expects a buffer and a count, and returns a count.

The confusing part is that the count it takes is a byte count, while the
return value is a count of the number of structures it has filled out.

Spell this out.

20 years agoWhen entering soclose(), assert that SS_NOFDREF is not already set.
Robert Watson [Fri, 16 Jul 2004 00:37:34 +0000 (00:37 +0000)]
When entering soclose(), assert that SS_NOFDREF is not already set.

20 years agoUse qsort_r() instead of qsort() when sorting links by latency
Gleb Smirnoff [Fri, 16 Jul 2004 00:07:44 +0000 (00:07 +0000)]
Use qsort_r() instead of qsort() when sorting links by latency
This helps us to remove a global variable and a mutex protecting it.

Reviewed by: rwatson
Approved by: julian (mentor)

20 years agoCopy qsort_r(3) from libc to libkern.
Gleb Smirnoff [Thu, 15 Jul 2004 23:58:23 +0000 (23:58 +0000)]
Copy qsort_r(3) from libc to libkern.

Reviewed by: phk
Approved by: julian (mentor)

20 years agofix array index out of bounds in rpc->rc_srtt[], rpc->rc_sdrtt[]
Jim Rees [Thu, 15 Jul 2004 22:21:25 +0000 (22:21 +0000)]
fix array index out of bounds in rpc->rc_srtt[], rpc->rc_sdrtt[]

Noticed by: tedu
Approved by: alfred

20 years agoPreparation commit for the tty cleanups that will follow in the near
Poul-Henning Kamp [Thu, 15 Jul 2004 20:47:41 +0000 (20:47 +0000)]
Preparation commit for the tty cleanups that will follow in the near
future:

rename ttyopen() -> tty_open() and ttyclose() -> tty_close().

We need the ttyopen() and ttyclose() for the new generic cdevsw
functions for tty devices in order to have consistent naming.

20 years agoMFNetBSD.
Lukas Ertl [Thu, 15 Jul 2004 19:25:06 +0000 (19:25 +0000)]
MFNetBSD.

rev. 1.67, author: mycroft
   Fix a byte order error.

rev. 1.68, author: mycroft
   Adjust some silliness that was causing us to do extra work for
   "frame list rollover" interrupts, which we pretty much ignore.

Obtained from:   NetBSD

20 years agoAdd security/freebsd-update back for non-ia64 releases. (The port is
Colin Percival [Thu, 15 Jul 2004 18:22:21 +0000 (18:22 +0000)]
Add security/freebsd-update back for non-ia64 releases.  (The port is
currently marked ONLY_FOR_ARCHS= i386 amd64 alpha sparc64.)

Approved by: scottl

20 years agoPush down the acquisition and release of the page queues lock into
Alan Cox [Thu, 15 Jul 2004 18:00:43 +0000 (18:00 +0000)]
Push down the acquisition and release of the page queues lock into
pmap_protect() and pmap_remove().  In general, they require the lock in
order to modify a page's pv list or flags.  In some cases, however,
pmap_protect() can avoid acquiring the lock.

20 years agoAdd a few comments about ACPI probing.
Nate Lawson [Thu, 15 Jul 2004 16:47:31 +0000 (16:47 +0000)]
Add a few comments about ACPI probing.

20 years agoAdd fdc_acpi to module build, bump WARNS to 2.
Nate Lawson [Thu, 15 Jul 2004 16:44:14 +0000 (16:44 +0000)]
Add fdc_acpi to module build, bump WARNS to 2.

20 years agoHook up fdc_acpi for the kernel build.
Nate Lawson [Thu, 15 Jul 2004 16:43:52 +0000 (16:43 +0000)]
Hook up fdc_acpi for the kernel build.

20 years agoClean up resources properly if attach fails. Always reset ISA drives on
Nate Lawson [Thu, 15 Jul 2004 16:41:07 +0000 (16:41 +0000)]
Clean up resources properly if attach fails.  Always reset ISA drives on
probe.

20 years agoRe-work for fdc_acpi. Expose fdc_add_child() and move the static
Nate Lawson [Thu, 15 Jul 2004 16:39:40 +0000 (16:39 +0000)]
Re-work for fdc_acpi.  Expose fdc_add_child() and move the static
hints-based probe to fdc_hints_probe().

Also:
* Fix some resource leaks when attach fails.
* Remove the FDC_ATTACHED flag.  It was supposed to prevent multiple
  unloads but this is not necessary.

20 years agoAdd an ACPI floppy drive attachment that probes via the _FDE and _FDI
Nate Lawson [Thu, 15 Jul 2004 16:38:07 +0000 (16:38 +0000)]
Add an ACPI floppy drive attachment that probes via the _FDE and _FDI
methods.  It also now handles ordinary floppy drive probing for drives
attached to ACPI.

Reviewed by: imp

20 years agoFix a typo in a comment.
John Baldwin [Thu, 15 Jul 2004 16:37:48 +0000 (16:37 +0000)]
Fix a typo in a comment.

20 years agoUpdate the interface for child drivers. Add acpi_scan_children, which
Nate Lawson [Thu, 15 Jul 2004 16:29:08 +0000 (16:29 +0000)]
Update the interface for child drivers.  Add acpi_scan_children, which
allows a bus to re-enumerate its child handles and optionally replace
them with new children, arranged to the bus's liking.  (The current device
space is flat with all devices immediately under acpi0).  Add comments
for each interface.

20 years agoMove references to the recently removed worm directory.
Scott Long [Thu, 15 Jul 2004 15:37:56 +0000 (15:37 +0000)]
Move references to the recently removed worm directory.

20 years agoMove the fdc_alloc_resources function into the bus front end.
Yoshihiro Takahashi [Thu, 15 Jul 2004 15:00:02 +0000 (15:00 +0000)]
Move the fdc_alloc_resources function into the bus front end.

20 years agoReenable nullfs for local links. The change from 'struct vfsconf'
Martin Blapp [Thu, 15 Jul 2004 13:24:44 +0000 (13:24 +0000)]
Reenable nullfs for local links. The change from 'struct vfsconf'
to 'struct xvfsconf' broke auto configuration. Patches have been
submitted to the vendor.

20 years agoDocument incorrect handling of multibyte characters.
Tim J. Robbins [Thu, 15 Jul 2004 11:10:09 +0000 (11:10 +0000)]
Document incorrect handling of multibyte characters.

20 years agoAvoid passing negative values to isdigit() on machines with signed chars.
Tim J. Robbins [Thu, 15 Jul 2004 10:26:38 +0000 (10:26 +0000)]
Avoid passing negative values to isdigit() on machines with signed chars.

20 years agoFix ``set ifaddr''. The code was actually using an uninitialised variable,
Brian Somers [Thu, 15 Jul 2004 09:42:16 +0000 (09:42 +0000)]
Fix ``set ifaddr''.  The code was actually using an uninitialised variable,
but conveniently, because ncpaddr.ncpaddr_family != AF_INET, the call to
ncpaddr_getip4addr() became a no-op leaving the local address as it was
(defaulting to whatever my hostname resolves to).

PR: 62050
Submitted by: Peter Jeremy <peter.jeremy@alcatel.com.au>
MFC after: 3 days

20 years agoAdd a publisher variable with the URL for the FreeBSD Project, and put
Murray Stokely [Thu, 15 Jul 2004 09:28:03 +0000 (09:28 +0000)]
Add a publisher variable with the URL for the FreeBSD Project, and put
this into the ISO headers by specifying the -P option to mkisofs.

Obtained from: share/examples/worm/makecdfs.sh

20 years agoAvoid passing negative values to isspace() on machines with signed chars.
Tim J. Robbins [Thu, 15 Jul 2004 09:23:04 +0000 (09:23 +0000)]
Avoid passing negative values to isspace() on machines with signed chars.

20 years agoRespect locale settings from the environment.
Tim J. Robbins [Thu, 15 Jul 2004 08:54:41 +0000 (08:54 +0000)]
Respect locale settings from the environment.

20 years agoAvoid passing negative values to isspace() on systems with signed chars.
Tim J. Robbins [Thu, 15 Jul 2004 08:27:04 +0000 (08:27 +0000)]
Avoid passing negative values to isspace() on systems with signed chars.

20 years agoDo a pass over all modules in the kernel and make them return EOPNOTSUPP
Poul-Henning Kamp [Thu, 15 Jul 2004 08:26:07 +0000 (08:26 +0000)]
Do a pass over all modules in the kernel and make them return EOPNOTSUPP
for unknown events.

A number of modules return EINVAL in this instance, and I have left
those alone for now and instead taught MOD_QUIESCE to accept this
as "didn't do anything".

20 years agoRespect locale settings from the environment.
Tim J. Robbins [Thu, 15 Jul 2004 08:13:56 +0000 (08:13 +0000)]
Respect locale settings from the environment.

20 years agoCleanup shutdown output.
Alfred Perlstein [Thu, 15 Jul 2004 08:01:00 +0000 (08:01 +0000)]
Cleanup shutdown output.

20 years agoA couple of grammar fixes in the bktr options section.
Peter Pentchev [Thu, 15 Jul 2004 07:52:27 +0000 (07:52 +0000)]
A couple of grammar fixes in the bktr options section.

PR: 66828 (mostly)
Submitted by: Michel Lavondes <fox@vader.aacc.cc.md.us>
MFC after: 2 weeks

20 years agoFix 'camcontrol rescan' syntax and use the proper indentation for
Peter Pentchev [Thu, 15 Jul 2004 07:34:14 +0000 (07:34 +0000)]
Fix 'camcontrol rescan' syntax and use the proper indentation for
the FILES list.

PR: 66827
Submitted by: Michel Lavondes <fox@vader.aacc.cc.md.us>
MFC after: 2 weeks

20 years agoDon't pass negative values into <ctype.h> functions on machines
Tim J. Robbins [Thu, 15 Jul 2004 07:33:56 +0000 (07:33 +0000)]
Don't pass negative values into <ctype.h> functions on machines
with signed chars.

20 years agoPrint size_t with %zu, not %d.
Tim J. Robbins [Thu, 15 Jul 2004 07:30:15 +0000 (07:30 +0000)]
Print size_t with %zu, not %d.

20 years agoRespect locale settings from the environment.
Tim J. Robbins [Thu, 15 Jul 2004 07:26:20 +0000 (07:26 +0000)]
Respect locale settings from the environment.

20 years agoAdd incorrect multibyte character handling to the already long list
Tim J. Robbins [Thu, 15 Jul 2004 06:59:34 +0000 (06:59 +0000)]
Add incorrect multibyte character handling to the already long list
of bugs.

20 years agoUse freopen() instead of a nasty hack.
Tim J. Robbins [Thu, 15 Jul 2004 06:57:36 +0000 (06:57 +0000)]
Use freopen() instead of a nasty hack.

20 years agoDocument incorrect handling of multibyte characters.
Tim J. Robbins [Thu, 15 Jul 2004 06:43:52 +0000 (06:43 +0000)]
Document incorrect handling of multibyte characters.

20 years agoEnsure that suffix matches occur on character boundaries.
Tim J. Robbins [Thu, 15 Jul 2004 06:15:10 +0000 (06:15 +0000)]
Ensure that suffix matches occur on character boundaries.

20 years agoDocument line length and multibyte character limitations.
Tim J. Robbins [Thu, 15 Jul 2004 05:12:04 +0000 (05:12 +0000)]
Document line length and multibyte character limitations.

20 years agoRemove Walnut Creek CDROM script that Jordan wrote in the 90s.
Murray Stokely [Thu, 15 Jul 2004 05:04:39 +0000 (05:04 +0000)]
Remove Walnut Creek CDROM script that Jordan wrote in the 90s.
This functionality is largely in src/release/${ARCH}/mkisofs.sh now.

PR: docs/43569

20 years agoUse err() instead of perror(), exit().
Tim J. Robbins [Thu, 15 Jul 2004 04:56:41 +0000 (04:56 +0000)]
Use err() instead of perror(), exit().

20 years agoRemove some serious foot-shooting potential from the release Makefile. For
John Baldwin [Thu, 15 Jul 2004 04:52:57 +0000 (04:52 +0000)]
Remove some serious foot-shooting potential from the release Makefile.  For
some unknown reason, when LOCAL_PATCHES and LOCAL_SCRIPT were originally
added, they were silently ignored if the actual file did not exist.  As a
result, if one mistyped the pathname to a patch or script, then the release
silently succeeded.  However, it was not built with the desired changes and
no warning was given to inform the builder either.  This commit explicitly
checks to see that all of the defined patches and scripts exist up front
and bails if any of them do not exist.  I lost several hours of valuable
sleeping time this evening due to this "feature" so I've finally gone and
ripped out.  I've tripped over this in the past several other times as
well.

Glanced at by: scottl

20 years agoUse warn() instead of perror().
Tim J. Robbins [Thu, 15 Jul 2004 04:51:21 +0000 (04:51 +0000)]
Use warn() instead of perror().

20 years agoAdd libthread_db assisted debugging support module.
David Xu [Thu, 15 Jul 2004 04:51:04 +0000 (04:51 +0000)]
Add libthread_db assisted debugging support module.

20 years agoUse warn() instead of perror().
Tim J. Robbins [Thu, 15 Jul 2004 04:45:24 +0000 (04:45 +0000)]
Use warn() instead of perror().

20 years agoTidy up system shutdown.
Alfred Perlstein [Thu, 15 Jul 2004 04:29:48 +0000 (04:29 +0000)]
Tidy up system shutdown.

20 years agotype prgregset_t really should be an array. this is odd, however, other
David Xu [Thu, 15 Jul 2004 03:52:17 +0000 (03:52 +0000)]
type prgregset_t really should be an array. this is odd, however, other
systems defined interfaces in thread_db.h use prgregset_t but not
prgregset_t * to be a output parameter, this is the only way to maintain
source code compatible with them.

20 years agoDisable SIGIO for now, leave a comment as to why it's busted and hard
Alfred Perlstein [Thu, 15 Jul 2004 03:49:52 +0000 (03:49 +0000)]
Disable SIGIO for now, leave a comment as to why it's busted and hard
to fix.

20 years agoAdd proc_service.h, the common file both debugger and libthread_db will
David Xu [Thu, 15 Jul 2004 03:43:18 +0000 (03:43 +0000)]
Add proc_service.h, the common file both debugger and libthread_db will
use, program wants to load libthread_db.so should provid proc service
interface.

20 years agoAdd my initial work of libthread_db. The library is used by gdb to debug
David Xu [Thu, 15 Jul 2004 03:36:35 +0000 (03:36 +0000)]
Add my initial work of libthread_db. The library is used by gdb to debug
threaded process. Current, only libpthread is supported, but macrel will
work on it to support libthr and libc_r.

20 years agoClean up the output on reboot by keeping completion messages on the same
Nate Lawson [Thu, 15 Jul 2004 03:20:08 +0000 (03:20 +0000)]
Clean up the output on reboot by keeping completion messages on the same
line as the announcement.  Someone should probably update the "buffers
remaining" message since we now no longer should have any buffers remaining
at that point.

20 years agoA loop in pmap_remove() should use TAILQ_FOREACH_SAFE(), not
Alan Cox [Thu, 15 Jul 2004 03:20:00 +0000 (03:20 +0000)]
A loop in pmap_remove() should use TAILQ_FOREACH_SAFE(), not
TAILQ_FOREACH(), because the loop deletes elements from the list.

Reviewed by: marcel@

20 years agoMake the day/month ordering dependent on the current locale by
Tim Kientzle [Thu, 15 Jul 2004 03:14:46 +0000 (03:14 +0000)]
Make the day/month ordering dependent on the current locale by
testing the locale at program startup and setting a flag, then
using that flag to determine appropriate strftime() arguments.

20 years agoStyle: rename 'mkdirpath' so it's clearer exactly what it does.
Tim Kientzle [Thu, 15 Jul 2004 03:13:27 +0000 (03:13 +0000)]
Style: rename 'mkdirpath' so it's clearer exactly what it does.
(To be precise, it creates the parent dir of the provided path.)

20 years agoA module with no modevent function gets modevent_nop() as default.
Poul-Henning Kamp [Wed, 14 Jul 2004 22:37:36 +0000 (22:37 +0000)]
A module with no modevent function gets modevent_nop() as default.

Until now the function has just returned zero for any event, but
that is downright wrong for MOD_UNLOAD and not very useful for any
future events we add where it may be crucial to be able to tell
if the event was unhandled or successful.

Change the function to return as follows:

MOD_LOAD -> 0
MOD_UNLOAD -> EBUSY
anything else -> EOPNOTSUPP

20 years agoAdd a comment separator.
Nate Lawson [Wed, 14 Jul 2004 22:09:32 +0000 (22:09 +0000)]
Add a comment separator.

20 years agoAdd a HARDWARE section which lists supported devices. The actual
Simon L. B. Nielsen [Wed, 14 Jul 2004 21:58:17 +0000 (21:58 +0000)]
Add a HARDWARE section which lists supported devices.  The actual
device listings has been moved (and in some cases more or less
rewritten) from the DESCRIPTION section.

20 years agoAdd a note indicating that the eh_prototype field used to construct
Robert Watson [Wed, 14 Jul 2004 20:31:37 +0000 (20:31 +0000)]
Add a note indicating that the eh_prototype field used to construct
ethernet headers is unsynchronized.

20 years agoAdd a mutex ng_tty_mtx to protect the global variable ngt_unit. Note
Robert Watson [Wed, 14 Jul 2004 20:31:05 +0000 (20:31 +0000)]
Add a mutex ng_tty_mtx to protect the global variable ngt_unit.  Note
that the locking of globals here isn't complete, and there's also a
locking issue relating to calling into and out of the tty code.

20 years agoAdd ng_ppp_latencies_mtx, a global mutex to protect the latency list.
Robert Watson [Wed, 14 Jul 2004 20:29:54 +0000 (20:29 +0000)]
Add ng_ppp_latencies_mtx, a global mutex to protect the latency list.
Note that the table is a hack, and so is this mutex.

Reviewed by: glebius

20 years agoIntroduce a new mutex, ng_fec_mtx, to protect the global unit list to
Robert Watson [Wed, 14 Jul 2004 20:27:33 +0000 (20:27 +0000)]
Introduce a new mutex, ng_fec_mtx, to protect the global unit list to
synchronization allocation of FEC unit numbers.

Reviewed by: glebius

20 years agoIntroduce a new mutex, ng_eiface_mtx, to protect the global unit list
Robert Watson [Wed, 14 Jul 2004 20:26:29 +0000 (20:26 +0000)]
Introduce a new mutex, ng_eiface_mtx, to protect the global unit list
lock used to synchronize allocation of unit numbers for new netgraph
ethernet interfaces.

Reviewed by: glebius

20 years agoIntroduce a new mutex, ng_iface_mtx, to protect the global unit list
Robert Watson [Wed, 14 Jul 2004 20:24:21 +0000 (20:24 +0000)]
Introduce a new mutex, ng_iface_mtx, to protect the global unit list
lock used to synchronize allocation of unit numbers for new netgraph
interfaces.

Reviewed by: glebius
Tested by: glebius

20 years agoSome laptops report the "design-capacity" instead of the "real-capacity"
Mark Santcroos [Wed, 14 Jul 2004 19:31:31 +0000 (19:31 +0000)]
Some laptops report the "design-capacity" instead of the "real-capacity"
when the battery is fully charged. That breaks some of the arithmetic in
calculating the remaining capacity (ends up with more than 100%).
This commit makes sure the max is 100.

Approved by: njl

20 years agoIn addition to the real user ID check, do an explicit jail
Christian S.J. Peron [Wed, 14 Jul 2004 19:04:31 +0000 (19:04 +0000)]
In addition to the real user ID check, do an explicit jail
check to ensure that the caller is not prison root.

The intention is to fix file descriptor creation so that
prison root can not use the last remaining file descriptors.
This privilege should be reserved for non-jailed root users.

Approved by: bmilekic (mentor)

20 years agoCorrect bounds check in lapic_create().
John Baldwin [Wed, 14 Jul 2004 18:12:15 +0000 (18:12 +0000)]
Correct bounds check in lapic_create().

Submitted by: "Ted Unangst" tedu at coverity.com

20 years agoUnbreak LINT: device card no longer takes a count.
Dag-Erling Smørgrav [Wed, 14 Jul 2004 17:50:08 +0000 (17:50 +0000)]
Unbreak LINT: device card no longer takes a count.

20 years agoMake sure to update the mnt_stats before UFS1 extattr tried to
Poul-Henning Kamp [Wed, 14 Jul 2004 14:19:32 +0000 (14:19 +0000)]
Make sure to update the mnt_stats before UFS1 extattr tried to
do I/O on the device.  Otherwise the blocksize is undefined in the
buffer cache.

20 years agoAdd a reference to glabel(8).
Pawel Jakub Dawidek [Wed, 14 Jul 2004 13:32:00 +0000 (13:32 +0000)]
Add a reference to glabel(8).

20 years agoFix a copy-and-paste-o in IFQ_DRV_PREPEND - all pointyhats to me.
Max Laier [Wed, 14 Jul 2004 13:31:41 +0000 (13:31 +0000)]
Fix a copy-and-paste-o in IFQ_DRV_PREPEND - all pointyhats to me.
While here also fix a (not less stupid) braino in IFQ_DRV_PURGE.

Reported-by: clement
Tested-by: clement (_PREPEND in sis(4))

20 years agoRemove a stray backslash
Brian Somers [Wed, 14 Jul 2004 13:31:17 +0000 (13:31 +0000)]
Remove a stray backslash

20 years agoDocument the -E and -a options as being extensions.
Tim J. Robbins [Wed, 14 Jul 2004 10:37:21 +0000 (10:37 +0000)]
Document the -E and -a options as being extensions.

20 years agoUpdate BUGS section to reflect current state of multibyte character support.
Tim J. Robbins [Wed, 14 Jul 2004 10:33:29 +0000 (10:33 +0000)]
Update BUGS section to reflect current state of multibyte character support.

20 years agoMake the 'y' (translate) command aware of multibyte characters.
Tim J. Robbins [Wed, 14 Jul 2004 10:06:22 +0000 (10:06 +0000)]
Make the 'y' (translate) command aware of multibyte characters.

20 years agoFix description of cmap_lookup_hard().
Tim J. Robbins [Wed, 14 Jul 2004 08:36:09 +0000 (08:36 +0000)]
Fix description of cmap_lookup_hard().

20 years agoRemove unused member of struct csclass: csc_value.
Tim J. Robbins [Wed, 14 Jul 2004 08:35:11 +0000 (08:35 +0000)]
Remove unused member of struct csclass: csc_value.

20 years agoSplay the left and right subtrees on min - 1 and max + 1, respectively,
Tim J. Robbins [Wed, 14 Jul 2004 08:33:14 +0000 (08:33 +0000)]
Splay the left and right subtrees on min - 1 and max + 1, respectively,
before trying to coalesce. Forgetting to splay caused us to miss many
opportunities for coalescing.

20 years agoLike on i386, eliminate pv_ptem (which was suggested by alc). This
Peter Wemm [Wed, 14 Jul 2004 07:13:35 +0000 (07:13 +0000)]
Like on i386, eliminate pv_ptem (which was suggested by alc).  This
reduces the size of the pv_entry structure a small but significant amount.

This is implemented a little differently because it isn't so cheap to get
the physical address of the page tabke page on amd64.. instead of it
being directly accessible from the top level page directory, it is now
two additional tree levels down.  However.. In almost all cases, we
recently had the physical address if the page table page a short while
before we needed it, but it slipped through our fingers.  This patch
saves it for when we do need it.  Also, for the one case where we do not
have the ptp paddr, we are always running in curproc context and so we
can do a vtopte-like trick.  I've implemented vtopde() for this purpose.

There is still a CYA entry in pmap_unuse_pt() that needs to be removed.  I
think it can be removed now but I forgot to test with it gone.

20 years agoRemove duplicate entry.
Murray Stokely [Wed, 14 Jul 2004 07:10:15 +0000 (07:10 +0000)]
Remove duplicate entry.

Submitted by: mlaier@

20 years agoRemove fdc_alloc_resources, which should have happened in last commit.
Warner Losh [Wed, 14 Jul 2004 07:04:17 +0000 (07:04 +0000)]
Remove fdc_alloc_resources, which should have happened in last commit.

20 years agoMake FIOASYNC, FIOSETOWN and FIOGETOWN work on kqueues.
Alfred Perlstein [Wed, 14 Jul 2004 07:02:03 +0000 (07:02 +0000)]
Make FIOASYNC, FIOSETOWN and FIOGETOWN work on kqueues.

20 years agoFix the pccard attachment to have a chance of working.
Warner Losh [Wed, 14 Jul 2004 06:59:58 +0000 (06:59 +0000)]
Fix the pccard attachment to have a chance of working.
Move the resource allocation into the bus front ends.

20 years agoFollow PnP location string change in acpi.c.
Takanori Watanabe [Wed, 14 Jul 2004 06:58:39 +0000 (06:58 +0000)]
Follow PnP location string change in acpi.c.

20 years agodo { } while(0) KNOTE macro, whitespace
Alfred Perlstein [Wed, 14 Jul 2004 06:47:32 +0000 (06:47 +0000)]
do { } while(0) KNOTE macro, whitespace

20 years agoAdditional pmap locking
Alan Cox [Wed, 14 Jul 2004 05:49:44 +0000 (05:49 +0000)]
Additional pmap locking

20 years agoSwitch snoop device to using C99 initialization for struct linesw.
Robert Watson [Wed, 14 Jul 2004 05:32:55 +0000 (05:32 +0000)]
Switch snoop device to using C99 initialization for struct linesw.

20 years agoMore technical terms.
Murray Stokely [Wed, 14 Jul 2004 05:23:16 +0000 (05:23 +0000)]
More technical terms.

MFC After: 3 days

20 years agoConvert SLIP to using C99 structure initialization for its struct
Robert Watson [Wed, 14 Jul 2004 05:01:40 +0000 (05:01 +0000)]
Convert SLIP to using C99 structure initialization for its struct
linesw.