]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
20 years agoLook up ${CUSTOMIZE} in ${.CURDIR} first, and then as a path name by itself,
thomas [Fri, 28 May 2004 21:16:14 +0000 (21:16 +0000)]
Look up ${CUSTOMIZE} in ${.CURDIR} first, and then as a path name by itself,
assuming in the latter case that it is an absolute path name.

20 years agoAdd the integrated NICs on the Dell PowerEdge 1750 and 2650 to the list
simon [Fri, 28 May 2004 21:16:13 +0000 (21:16 +0000)]
Add the integrated NICs on the Dell PowerEdge 1750 and 2650 to the list
of supported devices.

20 years agoRemove a broken micro-optimization from pmap_enter(). The ill effect
alc [Fri, 28 May 2004 19:42:02 +0000 (19:42 +0000)]
Remove a broken micro-optimization from pmap_enter().  The ill effect
of this micro-optimization occurs when we call pmap_enter() to wire an
already mapped page.  Because of the micro-optimization, we fail to
mark the PTE as wired.  Later, on teardown of the address space,
pmap_remove_pages() destroys the PTE before vm_fault_unwire() has
unwired the page.  (pmap_remove_pages() is not supposed to destroy
wired PTEs.  They are destroyed by a later call to pmap_remove().)
Thus, the page becomes lost.

Note: The page is not lost if the application called munlock(2), only
if it relies on teardown of the address space to unwire its pages.

For the historically inclined, this bug was introduced by a
megacommit, revision 1.182, roughly six years ago.

Leak observed by: green@ and dillon independently
Patch submitted by: dillon at backplane dot com
Reviewed by: tegge@
MFC after: 1 week

20 years agoReenable ithread preemption for interrupts that occur while executing in
jhb [Fri, 28 May 2004 17:50:07 +0000 (17:50 +0000)]
Reenable ithread preemption for interrupts that occur while executing in
the kernel.  I accidentally broke this with the new interrupt code that
came in prior to 5.2.

Submitted by: bde

20 years agoInclude <net/ethernet.h> to have a prototype for ether_ntoa().
stefanf [Fri, 28 May 2004 17:44:11 +0000 (17:44 +0000)]
Include <net/ethernet.h> to have a prototype for ether_ntoa().

Approved by: das (mentor)

20 years agoInclude <sys/uio.h> for the complete type of struct iovec.
stefanf [Fri, 28 May 2004 17:41:36 +0000 (17:41 +0000)]
Include <sys/uio.h> for the complete type of struct iovec.

Approved by: das (mentor)

20 years agoDon't assume that the current setting (_CRS) of a PCI link device is
jhb [Fri, 28 May 2004 17:31:32 +0000 (17:31 +0000)]
Don't assume that the current setting (_CRS) of a PCI link device is
correct.  Instead, check it against the possible settings (_PRS) when
the link is probed.  This is important when using APIC mode but link
devices still have PIC mode settings.  This is also what Linux does.

Additional prodding by: Len Brown len dot brown at intel dot com

20 years ago- introduce a new primary `-depth n', which tests whether
eik [Fri, 28 May 2004 17:17:15 +0000 (17:17 +0000)]
- introduce a new primary `-depth n', which tests whether
  the depth of the current file relative to the starting
  point of the traversal is n. The usual +/- modifiers
  to the argument apply.

- while I'm here, fix -maxdepth in the case of a depth-first
  traversal

Print the top ten maintainers of python module ports
(works with p5-* too):

find /usr/ports -depth 2 \! -name 'py-*' -prune -o \
  -depth 3 -name Makefile -execdir make -VMAINTAINER \; \
  | sort | uniq -c | sort -nr | head

PR: 66667
Reviewed by: ru, joerg
Approved by: joerg
MFC after: 2 weeks

20 years agoStyle fixes.
njl [Fri, 28 May 2004 16:38:37 +0000 (16:38 +0000)]
Style fixes.

20 years agoStyle.
maxim [Fri, 28 May 2004 14:20:06 +0000 (14:20 +0000)]
Style.

Submitted by: bde

20 years agoUnify the start/size parameters for the RSDP search area. Don't bother
njl [Fri, 28 May 2004 07:25:23 +0000 (07:25 +0000)]
Unify the start/size parameters for the RSDP search area.  Don't bother
trying to exclude the top end of the range since it should hurt to overlap
by 4 bytes in the off-chance the RSDP signature appears incorrectly at the
very top of our search space.

20 years agoNow that we properly disable GPEs before entering a sleep state, including
njl [Fri, 28 May 2004 07:16:52 +0000 (07:16 +0000)]
Now that we properly disable GPEs before entering a sleep state, including
S5 (soft off), we don't need to disable ACPI when powering off.  This may
fix some systems that don't power off correctly.

20 years agoStyle cleanups. "extern" is unneeded for function prototypes.
njl [Fri, 28 May 2004 07:15:55 +0000 (07:15 +0000)]
Style cleanups.  "extern" is unneeded for function prototypes.

20 years agoUpdate the new suspend/resume GPE methods to properly limit the GPE
njl [Fri, 28 May 2004 07:04:09 +0000 (07:04 +0000)]
Update the new suspend/resume GPE methods to properly limit the GPE
based on the destination sleep state.  Add a method to restore the old
state on resume.  This is needed for the case of suspending to a very low
state disabling a GPE (i.e. S4), resuming, and then suspending to a higher
state (i.e. S3).  This case should now keep the proper GPEs enabled.

20 years agoPass a pointer to the sleep state instead of casting gymnastics to pass
njl [Fri, 28 May 2004 06:32:16 +0000 (06:32 +0000)]
Pass a pointer to the sleep state instead of casting gymnastics to pass
the value itself in the pointer.

20 years agoFix paste-o.
njl [Fri, 28 May 2004 06:29:30 +0000 (06:29 +0000)]
Fix paste-o.

20 years agoAttach per-device sysctls to allow users to set whether or not a given
njl [Fri, 28 May 2004 06:28:55 +0000 (06:28 +0000)]
Attach per-device sysctls to allow users to set whether or not a given
device can wake the system.  For example:

    dev.root0.nexus0.acpi0.acpi_lid0.wake: 1
    dev.root0.nexus0.acpi0.acpi_button0.wake: 1
    dev.root0.nexus0.acpi0.pcib0.wake: 0
    dev.root0.nexus0.acpi0.sio0.wake: 0

20 years agoRemove a redundant include directive
scottl [Fri, 28 May 2004 04:42:10 +0000 (04:42 +0000)]
Remove a redundant include directive

20 years agoFix LP64 environments: cast a pointer type to intptr_t before casting
marcel [Fri, 28 May 2004 02:50:43 +0000 (02:50 +0000)]
Fix LP64 environments: cast a pointer type to intptr_t before casting
to int and vice versa.

20 years agoDon't nest includes here. There's no need and the subset included is lame.
imp [Fri, 28 May 2004 02:24:07 +0000 (02:24 +0000)]
Don't nest includes here.  There's no need and the subset included is lame.

Submitted by: bde

20 years agoInclude required machine/bus.h
imp [Fri, 28 May 2004 02:23:10 +0000 (02:23 +0000)]
Include required machine/bus.h

20 years agoDefine another HOTPLUG event.
ps [Fri, 28 May 2004 00:49:57 +0000 (00:49 +0000)]
Define another HOTPLUG event.

20 years agoAdding ixgb(4) for Intel 10GbE Adapters
tackerman [Fri, 28 May 2004 00:27:31 +0000 (00:27 +0000)]
Adding ixgb(4) for Intel 10GbE Adapters

20 years agoFirst release of ixgb driver for the Intel(R) PRO/10GbE Family of Adapters. This...
tackerman [Fri, 28 May 2004 00:23:00 +0000 (00:23 +0000)]
First release of ixgb driver for the Intel(R) PRO/10GbE Family of Adapters.  This driver has
been developed for use with FreeBSD, version 4.8 and later.

Submitted by: Hema Joyce
Reviewed by:  Prafulla Deuskar
Approved by:  Prafulla Deuskar
MFC after: 1 week

20 years agoAdd FreeBSD 4.10-RELEASE.
hrs [Fri, 28 May 2004 00:12:38 +0000 (00:12 +0000)]
Add FreeBSD 4.10-RELEASE.

MFC after: 3 days

20 years agogive out a little more information in case of a missing dependency
eik [Fri, 28 May 2004 00:05:28 +0000 (00:05 +0000)]
give out a little more information in case of a missing dependency

PR: 56549
Submitted by: edwin
Reviewed by: joerg, ru
Approved by: joerg
MFC after: 2 weeks

20 years agoConnect libarchive decompress support to the build.
kientzle [Thu, 27 May 2004 23:57:45 +0000 (23:57 +0000)]
Connect libarchive decompress support to the build.
Also, add it to archive_read_support_compression_all()
so that typical clients get it pulled in by default.

20 years agoFreeBSD-ify the manpage. Our inetd does not support bind-address:port syntax
mlaier [Thu, 27 May 2004 23:51:05 +0000 (23:51 +0000)]
FreeBSD-ify the manpage. Our inetd does not support bind-address:port syntax
Christian will follow up with some additional words about how to protect
this from the outside world.

Submitted-by: brueffer
Approved-by: bms(mentor)

20 years agoAdd ftp-proxy 8021/tcp in order to make the inetd.conf entry actually work.
mlaier [Thu, 27 May 2004 23:42:26 +0000 (23:42 +0000)]
Add ftp-proxy 8021/tcp in order to make the inetd.conf entry actually work.

Submitted-by: brueffer
Approved-by: bms(mentor)

20 years agoUse rman_get_start in preference to reaching into the rman structure.
imp [Thu, 27 May 2004 22:43:10 +0000 (22:43 +0000)]
Use rman_get_start in preference to reaching into the rman structure.

20 years ago'gnutar' is now handled by the 'tar' reader, so
kientzle [Thu, 27 May 2004 21:27:42 +0000 (21:27 +0000)]
'gnutar' is now handled by the 'tar' reader, so
there's no need to enable support for it separately
from 'tar.'  (The call to enable gnutar support is
now just an alias for the tar support, left in to
avoid API breakage.)

20 years agoCall nfsm_clget_nolock() instead of nfsm_clget() when holding the NFS
rwatson [Thu, 27 May 2004 20:34:04 +0000 (20:34 +0000)]
Call nfsm_clget_nolock() instead of nfsm_clget() when holding the NFS
subsystem lock to avoid tripping over an assertion regarding whether
the lock is held or not.  This is likely to be the cause of a panic
tripped over by Andrea Campi.

20 years agoCheck for >= 255 since sign extension from byte to u_int sometimes makes
njl [Thu, 27 May 2004 19:23:27 +0000 (19:23 +0000)]
Check for >= 255 since sign extension from byte to u_int sometimes makes
the value for "unknown" 0xffffffff.  The underlying kernel drivers should
be updated to only return 255 but the ABI is used by too many userland
utilities.

Also, make this WARNS 6 compatible.

20 years agoRestructure the wake GPE API. Now there are three functions:
njl [Thu, 27 May 2004 18:38:45 +0000 (18:38 +0000)]
Restructure the wake GPE API.  Now there are three functions:

acpi_wake_init:
    Evaluate _PRW and set the GPE type
acpi_wake_set_enable:
    Enable or disable a device's GPE.
acpi_wake_sleep_prep:
    Perform any last-minute changes to the device to prepare it for
    entering the given sleep state.

Also, walk the entire namespace when transitioning to a sleep state,
disabling any GPEs which aren't appropriate for the given state.  Transition
acpi_lid and acpi_button to the new API.

This clears the way for non-ACPI-aware devices to wake the system (i.e.
modems) and fixes a problem where systems power up after shutdown when a
GPE is triggered.

20 years agoFix stupid patch(1) tricks. Apparently patch thinks all files match the
brooks [Thu, 27 May 2004 17:23:32 +0000 (17:23 +0000)]
Fix stupid patch(1) tricks.  Apparently patch thinks all files match the
empty file so if you accidently apply a patch created with diff -N
twice, you get files with duplicate contents.

Reported by: Antoine Brodin <antoine.brodin at laposte.net>

20 years agoFixing disorder is the hardest thing in the world: Learn to sort :-)
imp [Thu, 27 May 2004 15:21:51 +0000 (15:21 +0000)]
Fixing disorder is the hardest thing in the world: Learn to sort :-)

Submitted by: bde

20 years agoPOLA dictates that VLAN_MTU be enabled by default.
yar [Thu, 27 May 2004 14:36:58 +0000 (14:36 +0000)]
POLA dictates that VLAN_MTU be enabled by default.
In particular, disabling it was likely to break configurations
involving ng_vlan(4) since the latter couldn't control
the parent's VLAN_MTU in the way vlan(4) did.

Pointed out by: ru

20 years agoFix an off-by-one error in the range check for the maximal -i or -o
joerg [Thu, 27 May 2004 13:31:16 +0000 (13:31 +0000)]
Fix an off-by-one error in the range check for the maximal -i or -o
block size.

20 years agoMake "envctrl" a known master driver for iicbus.
joerg [Thu, 27 May 2004 13:29:09 +0000 (13:29 +0000)]
Make "envctrl" a known master driver for iicbus.

20 years agoCorrect some types in the yp structures; this fixes a number of problems
tmm [Thu, 27 May 2004 11:34:21 +0000 (11:34 +0000)]
Correct some types in the yp structures; this fixes a number of problems
on sparc64. Obtained from and cross-checked with the NetBSD version
of this file and the rpcgen-generated code.

20 years agoTouch .Dd because the content of the manpage has been updated.
yar [Thu, 27 May 2004 11:12:07 +0000 (11:12 +0000)]
Touch .Dd because the content of the manpage has been updated.

Pointed out by: ru

20 years agoBuffer partial wide characters more efficiently: instead of storing the
tjr [Thu, 27 May 2004 10:54:34 +0000 (10:54 +0000)]
Buffer partial wide characters more efficiently: instead of storing the
multibyte representation in conversion state objects, store the
accumulated wide character, set number and number of bytes remaining
to avoid having to derive them every time mbrtowc() is called.

20 years agoBring back the macro versions of getwc(), getwchar(), putwc() and
tjr [Thu, 27 May 2004 10:08:44 +0000 (10:08 +0000)]
Bring back the macro versions of getwc(), getwchar(), putwc() and
putwchar(), but this time avoid redundantly declaring __stdinp and
__stdoutp when source files include both <stdio.h> and <wchar.h>.

20 years agoPreviously, restoring an archive with hardlinked files that had
kientzle [Thu, 27 May 2004 05:02:35 +0000 (05:02 +0000)]
Previously, restoring an archive with hardlinked files that had
certain flags set (e.g., schg or uappend) would fail because the flags
were restored before the hardlink was created.

To address this, I've generalized the existing machinery for deferring
directory timestamp/mode restoration and used it to defer the
restoration of highly-restrictive flags to the end of the extraction,
after any links have been created.

Pointed out by: Pawel Jakub Dawidek (pjd@)

20 years agoMinor corrections to error handling and user notifications:
kientzle [Thu, 27 May 2004 04:35:36 +0000 (04:35 +0000)]
Minor corrections to error handling and user notifications:
* Move format/compression reporting to end of output, since
  we don't always know the input format until then.
* Set bsdtar exit value to 1 if any file could not be restored.
* Generate gtar-style warning when stripping leading '/' characters.
* Warn when removing symlinks.

20 years agoProvide framework for exiting with non-zero value on non-critical errors.
kientzle [Thu, 27 May 2004 04:30:59 +0000 (04:30 +0000)]
Provide framework for exiting with non-zero value on non-critical errors.

20 years agoDocument support for reading .Z compressed archives.
kientzle [Thu, 27 May 2004 04:21:52 +0000 (04:21 +0000)]
Document support for reading .Z compressed archives.
Correct a few other minor nits.

20 years agoGC some no-longer-used constants.
kientzle [Thu, 27 May 2004 04:01:58 +0000 (04:01 +0000)]
GC some no-longer-used constants.

20 years agoAdd prototypes for .Z compression support.
kientzle [Thu, 27 May 2004 04:00:25 +0000 (04:00 +0000)]
Add prototypes for .Z compression support.

20 years agoAdd read-only support for .Z compressed archives.
kientzle [Thu, 27 May 2004 03:58:55 +0000 (03:58 +0000)]
Add read-only support for .Z compressed archives.

20 years agoFix disordering of pccarddevs.h noticed by bde. Also remove a few
imp [Thu, 27 May 2004 03:49:45 +0000 (03:49 +0000)]
Fix disordering of pccarddevs.h noticed by bde.  Also remove a few
redundant includes and fix some of the include disordering.

Submitted by: bde

20 years agoThese are now not needed, after repo copy and conversion.
imp [Thu, 27 May 2004 01:31:11 +0000 (01:31 +0000)]
These are now not needed, after repo copy and conversion.

20 years agoSort includes, remove unnecssary ones.
imp [Thu, 27 May 2004 01:27:04 +0000 (01:27 +0000)]
Sort includes, remove unnecssary ones.

Submitted by: bde

20 years agoThose sysctls shouldn't be writtable from inside a jail.
pjd [Wed, 26 May 2004 23:03:27 +0000 (23:03 +0000)]
Those sysctls shouldn't be writtable from inside a jail.

20 years agoUse new eui64(3) functions to print EUI-64s and to allow access to nodes
brooks [Wed, 26 May 2004 22:59:55 +0000 (22:59 +0000)]
Use new eui64(3) functions to print EUI-64s and to allow access to nodes
by EUI-64 and name.

Reviewed by: simokawa

20 years agoAdd support for an /etc/eui64 file modeled on /etc/ethers. The API is
brooks [Wed, 26 May 2004 22:58:06 +0000 (22:58 +0000)]
Add support for an /etc/eui64 file modeled on /etc/ethers.  The API is
modeled on ethers(3) except that all functions are thread-safe.

Reviewed by: simokawa

20 years agoAdd PCI ID for via 8237.
matk [Wed, 26 May 2004 22:05:50 +0000 (22:05 +0000)]
Add PCI ID for via 8237.

Submitted by: Josh Elsasser <jre@vineyard.net>
Approved by: tanimura (mentor)
PR: kern/61730

20 years agoIt seems that clearing the MCR_IE bit in the modem control register
tmm [Wed, 26 May 2004 21:59:01 +0000 (21:59 +0000)]
It seems that clearing the MCR_IE bit in the modem control register
does not reliably prevent the triggering of interrupts for all supported
configurations. Thus, the FIFO size probe could cause an interrupt,
which could lead to an interrupt storm in the shared interrupt case.

To prevent this, change ns8250_bus_probe() to use the overflow bit in
the line status register instead of the RX ready bit in the interrupt
identification register to detect whether the FIFO has filled up.
This allows us to clear all bits in the interrupt enable register during
the probe, which should prevent interrupts reliably.
Additionally, the detected FIFO size may be a bit more accurate, because
the overflow bit is only set when the FIFO did actually fill up, while
interrupts would trigger a bit early.

Reviewed and tested on a lot of hardware by: marcel

20 years agoSilence some constness and printf type warnings. Most of the
dwmalone [Wed, 26 May 2004 21:14:13 +0000 (21:14 +0000)]
Silence some constness and printf type warnings. Most of the
const fixes are ugly 'cos the types in an iovec aren't quite
right for a writev.

20 years agoUnbreak build if
josef [Wed, 26 May 2004 20:18:58 +0000 (20:18 +0000)]
Unbreak build if
        options BKTR_USE_FREEBSD_SMBUS
is defined in kernel configuration.

Submitted by:   Tom Convery <tpc@tomfoo.com>
PR:             kern/60458
Approved by:    imp

20 years agoFix order.
pjd [Wed, 26 May 2004 17:51:53 +0000 (17:51 +0000)]
Fix order.

20 years agoSysctl hw.bus.devctl_disable shouldn't be writtable from inside a jail.
pjd [Wed, 26 May 2004 16:36:32 +0000 (16:36 +0000)]
Sysctl hw.bus.devctl_disable shouldn't be writtable from inside a jail.

Approved by: imp

20 years agoAdd pccarddevs.h and usbdevs.h as depends, ala miidevs.h, in the right
imp [Wed, 26 May 2004 16:31:41 +0000 (16:31 +0000)]
Add pccarddevs.h and usbdevs.h as depends, ala miidevs.h, in the right
places.  This should have been committed last night with the rest of
my changes, but wasn't.

Pointy hat to: imp

20 years agoMFi386: revision 1.493.
nyan [Wed, 26 May 2004 13:10:37 +0000 (13:10 +0000)]
MFi386: revision 1.493.

20 years agoRetire cpu_sched_exit(); it is not used any more.
tmm [Wed, 26 May 2004 12:09:39 +0000 (12:09 +0000)]
Retire cpu_sched_exit(); it is not used any more.

20 years agoMove the per-CPU vmspace pointer fixup that is required before a
tmm [Wed, 26 May 2004 12:06:52 +0000 (12:06 +0000)]
Move the per-CPU vmspace pointer fixup that is required before a
struct vmspace is freed from cpu_sched_exit() to pmap_release().

This has the advantage of being able to rely on MI code to decide
when a free should occur, instead of having to inspect the reference
count ourselves.

At the same time, turn the per-CPU vmspace pointer into a pmap pointer,
so that pmap_release() can deal with pmaps exclusively.

Reviewed (and embrassing bug spotted) by: jake

20 years agoif_printf() won't emit a newline unless told to.
yar [Wed, 26 May 2004 11:41:26 +0000 (11:41 +0000)]
if_printf() won't emit a newline unless told to.

20 years agoDump some more informations:
pjd [Wed, 26 May 2004 11:36:27 +0000 (11:36 +0000)]
Dump some more informations:
- device state
- list of used providers
- total number of disks
- number of disks online

Prodded by: Alex Deiter <tiamat@komi.mts.ru>

20 years agoUpdate the status of VLAN support in interface drivers.
yar [Wed, 26 May 2004 11:34:51 +0000 (11:34 +0000)]
Update the status of VLAN support in interface drivers.

20 years agoPrint provider's size in human-readable form as well.
pjd [Wed, 26 May 2004 11:08:35 +0000 (11:08 +0000)]
Print provider's size in human-readable form as well.

20 years agoQuick fix for overflow when tsc_freq >= 2^31. "int profrate" in struct
bde [Wed, 26 May 2004 09:43:38 +0000 (09:43 +0000)]
Quick fix for overflow when tsc_freq >= 2^31.  "int profrate" in struct
gmon and struct gmonhdr was originally just to represent the kernel
(profiling) clock frequency and it remains poorly suited to representing
the frequencies of fast counters like the TSC.  It broke a year or two
ago.  This quick fix keeps it working for another year or month or two
until TSC frequencies can exceed 2^32, by dividing the frequency by 2.
Dividing the frequency by 4 would work for a little longer but would
lose a little too much precision.

20 years agoMFamd64:
bde [Wed, 26 May 2004 07:43:41 +0000 (07:43 +0000)]
MFamd64:

Fixed profiling of trap, syscall and interrupt handlers and some
ordinary functions, essentially by backing out half of rev.1.106 of
i386/exception.s.  The handlers must be between certain labels for
the purposes of profiling, and this was broken by scattering them in
separately compiled .s files, especially for ordinary functions that
ended up between the labels.  Merge the files by #including them as
before, except with different pathnames and better comments and
organization.  Changes to the scattered files are minimal -- just
move the labels to the file that does the #includes.

This also partly fixes profiling of IPIs -- all IPI handlers are now
correctly classified as interrupt handlers, but many are still missing
mcount calls.

vm86bios.s is included as before, but it is now between the labels for
interrupt handlers again, which seems to be wrong since half of it is
for a non-interrupt handler.

20 years agoMove to generating pccarddevs.h on the fly, both for the kernel and
imp [Wed, 26 May 2004 00:53:10 +0000 (00:53 +0000)]
Move to generating pccarddevs.h on the fly, both for the kernel and
the modules.

Also generate usbdevs.h automatically now, but a non-kernel file is
stopping that at the moment.

20 years agoFix disorder introduce in 1.862 by sorting emu10k before miidevs.h,
imp [Wed, 26 May 2004 00:38:52 +0000 (00:38 +0000)]
Fix disorder introduce in 1.862 by sorting emu10k before miidevs.h,
rather than after.

bde inspired words: disorder

20 years agodevlist2h.awk is too generic a name for what it does. It really
imp [Wed, 26 May 2004 00:19:39 +0000 (00:19 +0000)]
devlist2h.awk is too generic a name for what it does.  It really
converts miidevs to a .h file, so rename to reflect that.

The usb and pccard versions have also been renamed and will be hooked
into the build system shortly (I've made the conversion in my p4
tree).

20 years agoKeepalive timer should be added if we does not have any sppp consumers before
rik [Tue, 25 May 2004 21:54:07 +0000 (21:54 +0000)]
Keepalive timer should be added if we does not have any sppp consumers before
and should be deleted if we do not have any anymore.

20 years agoHumanize_number(3) is a part of libutil.
pjd [Tue, 25 May 2004 20:11:50 +0000 (20:11 +0000)]
Humanize_number(3) is a part of libutil.

20 years ago'ld' lists drives, not volumes.
le [Tue, 25 May 2004 20:08:50 +0000 (20:08 +0000)]
'ld' lists drives, not volumes.

20 years agoDefinition of recursion.
wollman [Tue, 25 May 2004 19:15:02 +0000 (19:15 +0000)]
Definition of recursion.

20 years agoYou want to include libutil.h, not util.h.
trhodes [Tue, 25 May 2004 18:53:54 +0000 (18:53 +0000)]
You want to include libutil.h, not util.h.
Some minor sentence tweaking.

20 years agoMFS: vm_map.c rev 1.187.2.27 through 1.187.2.29, fix MS_INVALIDATE
des [Tue, 25 May 2004 18:40:53 +0000 (18:40 +0000)]
MFS: vm_map.c rev 1.187.2.27 through 1.187.2.29, fix MS_INVALIDATE
semantics but provide a sysctl knob for reverting to old ones.

20 years agoBack out previous commit; it went to the wrong file.
des [Tue, 25 May 2004 18:28:52 +0000 (18:28 +0000)]
Back out previous commit; it went to the wrong file.

20 years agoFix spelling
brueffer [Tue, 25 May 2004 17:02:20 +0000 (17:02 +0000)]
Fix spelling

20 years agoMFS: rev 1.187.2.27 through 1.187.2.29, fix MS_INVALIDATE semantics but
des [Tue, 25 May 2004 16:31:49 +0000 (16:31 +0000)]
MFS: rev 1.187.2.27 through 1.187.2.29, fix MS_INVALIDATE semantics but
provide a sysctl knob for reverting to old ones.

20 years agoEnable GPE at runtime rather than suspend time. This is to match the
njl [Tue, 25 May 2004 16:17:39 +0000 (16:17 +0000)]
Enable GPE at runtime rather than suspend time.  This is to match the
new behavior in ACPI-CA that defers GPE configuration.  This is a temporary
measure while reworking the GPE interface.

20 years agoAdd a super-user check to ipfw_ctl() to make sure that the calling
csjp [Tue, 25 May 2004 15:02:12 +0000 (15:02 +0000)]
Add a super-user check to ipfw_ctl() to make sure that the calling
process is a non-prison root. The security.jail.allow_raw_sockets
sysctl variable is disabled by default, however if the user enables
raw sockets in prisons, prison-root should not be able to interact
with firewall rule sets.

Approved by: rwatson, bmilekic (mentor)

20 years agoUse humanize_number(3).
pjd [Tue, 25 May 2004 14:53:47 +0000 (14:53 +0000)]
Use humanize_number(3).

Reminded by: jhb

20 years agoTeach fxp(4) to control VLAN_MTU in the hardware.
yar [Tue, 25 May 2004 14:49:46 +0000 (14:49 +0000)]
Teach fxp(4) to control VLAN_MTU in the hardware.
Now reception of extended frames can be toggled
through ioctl(SIOCSIFCAP).

The card will also receive extended frames when
in promiscuous mode.

20 years agoAfter all the relevant drivers have been fixed, fix vlan(4) itself
yar [Tue, 25 May 2004 14:30:12 +0000 (14:30 +0000)]
After all the relevant drivers have been fixed, fix vlan(4) itself
WRT manipulating capabilities of the parent interface:

- use ioctl(SIOCSIFCAP) to toggle VLAN_MTU (the way that was done
  before was just wrong);

- use the right order of conditional clauses to set the MTU fudge
  (that is logically independent from toggling VLAN_MTU.)

20 years agoAdd pgrep(1) and pkill(1) to the cross-reference section of ps(1).
gad [Tue, 25 May 2004 14:05:25 +0000 (14:05 +0000)]
Add pgrep(1) and pkill(1) to the cross-reference section of ps(1).

20 years agoRevert to rev. 1.36 until issues with -Wredundant-decls are sorted out.
tjr [Tue, 25 May 2004 12:41:02 +0000 (12:41 +0000)]
Revert to rev. 1.36 until issues with -Wredundant-decls are sorted out.

20 years agoActually negative size is possible for file system, but not for disk.
pjd [Tue, 25 May 2004 12:11:13 +0000 (12:11 +0000)]
Actually negative size is possible for file system, but not for disk.

20 years agoAs previously threatened, give each device its own sysctl context and
des [Tue, 25 May 2004 12:06:26 +0000 (12:06 +0000)]
As previously threatened, give each device its own sysctl context and
subtree (under the new dev top-level node).  This should greatly simplify
drivers which need per-device sysctl variables (such as ndis).

20 years agoFix typo in previous: getwc() should call fgetwc(), not the function
tjr [Tue, 25 May 2004 12:02:49 +0000 (12:02 +0000)]
Fix typo in previous: getwc() should call fgetwc(), not the function
version of itself.

Noticed by: stefanf

20 years agoSet the IFCAP_VLAN_MTU bit in capenable as well.
des [Tue, 25 May 2004 11:33:19 +0000 (11:33 +0000)]
Set the IFCAP_VLAN_MTU bit in capenable as well.

Reminded by: ru

20 years agoSet baudrate to 100 Mbps, and advertise our ability to handle extended
des [Tue, 25 May 2004 11:04:01 +0000 (11:04 +0000)]
Set baudrate to 100 Mbps, and advertise our ability to handle extended
frames (802.1q).

Submitted by: Steinar Haug <sthaug@nethelp.no>

20 years agoScan the source string for invalid wide characters in wcsrtombs()
tjr [Tue, 25 May 2004 10:45:24 +0000 (10:45 +0000)]
Scan the source string for invalid wide characters in wcsrtombs()
in the dst == NULL case.

20 years agoo Fix typo: s/bslim/bhlim/.
maxim [Tue, 25 May 2004 10:43:46 +0000 (10:43 +0000)]
o Fix typo: s/bslim/bhlim/.

PR: docs/67170
Submitted by: Anatoly Zherdev
MFC after: 3 days

20 years agoProvide trivial macro implementations of getwc(), getwchar(), putwc() and
tjr [Tue, 25 May 2004 10:42:52 +0000 (10:42 +0000)]
Provide trivial macro implementations of getwc(), getwchar(), putwc() and
putwchar() to reduce function call overhead.

20 years agoRound #1 of improving pcf(4).
joerg [Tue, 25 May 2004 07:42:45 +0000 (07:42 +0000)]
Round #1 of improving pcf(4).

This splits the driver into a bus-independant backend, plus bus-specific
frontends.  The old pcf(4) (i386/ISA) frontend is now in pcf_isa.c, the
frontend in envctrl.c is for sparc64/Ebus2 (Sun device name: SUNW,envctrl
from Sun E450 machines).  More frontends are expected to appear in future.

This is not yet ready for public consumption, but it basically works.
Nicolas will bring over his ISA-specific fixes soon.

Reviewed by: nsouch