]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
20 years agoRework the logic for `-t <tty>', such that it accepts "ttyp0" and "console",
gad [Thu, 24 Jun 2004 01:57:59 +0000 (01:57 +0000)]
Rework the logic for `-t <tty>', such that it accepts "ttyp0" and "console",
in addition to "/dev/ttyp0" or "p0"  and  "/dev/console" or "co".

20 years agoAttach the correct handle, not the one that was just deleted. Also,
njl [Thu, 24 Jun 2004 01:57:31 +0000 (01:57 +0000)]
Attach the correct handle, not the one that was just deleted.  Also,
remove some duplicated code.

20 years agoClean up use of ng_hdhasb_mtx a little:
rwatson [Thu, 24 Jun 2004 01:47:31 +0000 (01:47 +0000)]
Clean up use of ng_hdhasb_mtx a little:

- Assert the mutex in NG_IDHASH_FIND() since the mutex is required to
  safely walk the node lists in the ng_ID_hash table.

- Acquire the ng_nodelist_mtx when walking ng_allnodes or ng_allhooks
  to generate state dump output from the netgraph sysctls.

20 years agoAcquire socket lock in the "waiting for connection" loop in
rwatson [Thu, 24 Jun 2004 01:43:23 +0000 (01:43 +0000)]
Acquire socket lock in the "waiting for connection" loop in
kern_connect(), replacing tsleep() with msleep() with the socket
mutex.

20 years agoIntroduce sbreserve_locked(), which asserts the socket buffer lock on
rwatson [Thu, 24 Jun 2004 01:37:04 +0000 (01:37 +0000)]
Introduce sbreserve_locked(), which asserts the socket buffer lock on
the socket buffer having its limits adjusted.  sbreserve() now acquires
the lock before calling sbreserve_locked().  In soreserve(), acquire
socket buffer locks across read-modify-writes of socket buffer fields,
and calls into sbreserve/sbrelease; make sure to acquire in keeping
with the socket buffer lock order.  In tcp_mss(), acquire the socket
buffer lock in the calling context so that we have atomic read-modify
-write on buffer sizes.

20 years agoSlide socket buffer lock earlier in sopoll() to cover the call into
rwatson [Thu, 24 Jun 2004 00:54:26 +0000 (00:54 +0000)]
Slide socket buffer lock earlier in sopoll() to cover the call into
selrecord(), setting up select and flagging the socker buffers as SB_SEL
and setting up select under the lock.

20 years agoRun the power off code directly instead of using indirection through
njl [Thu, 24 Jun 2004 00:48:45 +0000 (00:48 +0000)]
Run the power off code directly instead of using indirection through
smp_rendezvous() to ensure we run on the BSP.  This reverts rev 1.128.
Add a comment indicating that MI code should be the one that runs all
shutdown functions on the BSP with the APs halted.  This should work
around problems in power off while waiting for the MI code to be improved.

20 years agoRemove spls from portal_open(). Acquire socket lock while sleeping
rwatson [Thu, 24 Jun 2004 00:47:23 +0000 (00:47 +0000)]
Remove spls from portal_open().  Acquire socket lock while sleeping
waiting for the socket to connect and use msleep() on the socket
mute rather than tsleep().  Acquire socket buffer mutexes around
read-modify-write of socket buffer flags.

20 years agoUse uintmax_t for CPU statistics and add a cast to prevent truncation of
njl [Thu, 24 Jun 2004 00:38:51 +0000 (00:38 +0000)]
Use uintmax_t for CPU statistics and add a cast to prevent truncation of
the statistics in a multiply.

Pointed out by: YONETANI Tomokazu

20 years agos/ARCH/ARCH_SUBDIR/g -- This reduces the chance of possible conflicts
marcel [Thu, 24 Jun 2004 00:02:32 +0000 (00:02 +0000)]
s/ARCH/ARCH_SUBDIR/g -- This reduces the chance of possible conflicts
with the user's environment.

Wondered why his cross-builds kept failing: marcel

20 years agoAdd a function to clean up RAID5 packets and use it when I/O has
le [Wed, 23 Jun 2004 23:52:55 +0000 (23:52 +0000)]
Add a function to clean up RAID5 packets and use it when I/O has
finished or when building the complete packet fails.

20 years agoChange "struct varent" to use the standard queue(8) macros, instead of
gad [Wed, 23 Jun 2004 23:48:09 +0000 (23:48 +0000)]
Change "struct varent" to use the standard queue(8) macros, instead of
using it's own version of the same basic algorithm.

Submitted by: part by Cyrille Lefevre, part of it done by me

20 years agos/udi_peed/udi_speed/
markus [Wed, 23 Jun 2004 23:25:58 +0000 (23:25 +0000)]
s/udi_peed/udi_speed/

Approved by: brueffer
MFC after: 3 days

20 years agoRemove two debugging printfs that are currently rather disturbing
le [Wed, 23 Jun 2004 22:32:01 +0000 (22:32 +0000)]
Remove two debugging printfs that are currently rather disturbing
than helpful.

20 years agoFix a test of bit-flag "P_SA" by adding parenthesis around the expression.
gad [Wed, 23 Jun 2004 21:59:56 +0000 (21:59 +0000)]
Fix a test of bit-flag "P_SA" by adding parenthesis around the expression.

Submitted by: Cyrille Lefevre

20 years agoMake uma_mtx MTX_RECURSE. Here's why:
bmilekic [Wed, 23 Jun 2004 21:59:03 +0000 (21:59 +0000)]
Make uma_mtx MTX_RECURSE.  Here's why:

The general UMA lock is a recursion-allowed lock because
there is a code path where, while we're still configured
to use startup_alloc() for backend page allocations, we
may end up in uma_reclaim() which calls zone_foreach(zone_drain),
which grabs uma_mtx, only to later call into startup_alloc()
because while freeing we needed to allocate a bucket.  Since
startup_alloc() also takes uma_mtx, we need to be able to
recurse on it.

This exact explanation also added as comment above mtx_init().

Trace showing recursion reported by: Peter Holm <peter-at-holm.cc>

20 years agoMake the udf_vnops side endian clean.
scottl [Wed, 23 Jun 2004 21:49:03 +0000 (21:49 +0000)]
Make the udf_vnops side endian clean.

20 years agoMove the sack sysctl's under net.inet.tcp.sack
ps [Wed, 23 Jun 2004 21:34:07 +0000 (21:34 +0000)]
Move the sack sysctl's under net.inet.tcp.sack

net.inet.tcp.do_sack -> net.inet.tcp.sack.enable
net.inet.tcp.sackhole_limit -> net.inet.tcp.sack.sackhole_limit

Requested by: wollman

20 years agoReplace a call to strncpy() with a call to strlcpy()
gad [Wed, 23 Jun 2004 21:31:43 +0000 (21:31 +0000)]
Replace a call to strncpy() with a call to strlcpy()

Submitted by: Cyrille Lefevre

20 years agoAdd a check for defunct processes in saveuser(), so the output for "args"
gad [Wed, 23 Jun 2004 21:17:25 +0000 (21:17 +0000)]
Add a check for defunct processes in saveuser(), so the output for "args"
(aka "command") will display "<defunct>", as does the output from "comm"
for those processes.  Also do better checking for malloc() failures.

Submitted by: Cyrille Lefevre

20 years agoAccept "sd len 0" and auto-size the subdisk correctly.
le [Wed, 23 Jun 2004 21:15:55 +0000 (21:15 +0000)]
Accept "sd len 0" and auto-size the subdisk correctly.

Spotted by: csjp

20 years agoAdd support for TCP Selective Acknowledgements. The work for this
ps [Wed, 23 Jun 2004 21:04:37 +0000 (21:04 +0000)]
Add support for TCP Selective Acknowledgements.  The work for this
originated on RELENG_4 and was ported to -CURRENT.

The scoreboarding code was obtained from OpenBSD, and many
of the remaining changes were inspired by OpenBSD, but not
taken directly from there.

You can enable/disable sack using net.inet.tcp.do_sack. You can
also limit the number of sack holes that all senders can have in
the scoreboard with net.inet.tcp.sackhole_limit.

Reviewed by: gnn
Obtained from: Yahoo! (Mohan Srinivasan, Jayanth Vijayaraghavan)

20 years agoFirst half of making UDF be endian-clean. This addresses the vfsops side.
scottl [Wed, 23 Jun 2004 19:36:09 +0000 (19:36 +0000)]
First half of making UDF be endian-clean.  This addresses the vfsops side.

20 years agoUse bus_dma* instead of contigmalloc()+vtophys() for RELENG_5.
rik [Wed, 23 Jun 2004 18:13:10 +0000 (18:13 +0000)]
Use bus_dma* instead of contigmalloc()+vtophys() for RELENG_5.

20 years agoFinally implement bus_config_intr() support for I/O APIC interrupt sources.
jhb [Wed, 23 Jun 2004 18:11:33 +0000 (18:11 +0000)]
Finally implement bus_config_intr() support for I/O APIC interrupt sources.
This should fix problems with older SMP systems that only have ISA/EISA
IRQs when routing virgin PCI interrupts as well as on other boxes whose
MADT does not have any interrupt override entries for ISA IRQs that are
used to route PCI interrupts even in APIC mode.

20 years agoFetch the actual acpi0 device_t and use device_is_attached() to see if
jhb [Wed, 23 Jun 2004 17:59:01 +0000 (17:59 +0000)]
Fetch the actual acpi0 device_t and use device_is_attached() to see if
it's alive rather than trying to fetch its softc pointer via its devclass.

Glanced at by: imp, njl

20 years agoel(4) stopped needing to me a count device in December 2000.
brooks [Wed, 23 Jun 2004 17:33:25 +0000 (17:33 +0000)]
el(4) stopped needing to me a count device in December 2000.

20 years ago- Defer BUS_CONFIG_INTR() on ACPI IRQ resources until the resources are
jhb [Wed, 23 Jun 2004 17:21:02 +0000 (17:21 +0000)]
- Defer BUS_CONFIG_INTR() on ACPI IRQ resources until the resources are
  actually used.  For most ACPI devices this means deferring the call
  until bus_alloc_resource().
  - Add a function acpi_config_intr() to call BUS_CONFIG_INTR() for an
    ACPI IRQ resource using the trigger mode and polarity information
    stored in the ACPI resource object.
  - Add a function acpi_lookup_irq_resource() to lookup the ACPI IRQ
    resource that corresponds to a specified rid and new-bus resource.
  - Have the ACPI PCI bridge driver call BUS_CONFIG_INTR() on interrupts
    that it routes through link devices.
- Remove needactivate variable from acpi_alloc_resource() by changing the
  function not modify the flags variable but just mask off RF_ACTIVE when
  calling rman_reserve_resource().

Reviewed by: njl (1, an earlier version)

20 years agoVarious cleanups in support of a future ioapic_config_intr() function:
jhb [Wed, 23 Jun 2004 15:29:20 +0000 (15:29 +0000)]
Various cleanups in support of a future ioapic_config_intr() function:
- Allow ioapic_set_{nmi,smi,extint}() to be called multiple times on the
  same pin so long as the pin's mode is the same as the mode being
  requested.
- Add a notion of bus type for the interrupt associated with interrupt pin.
  This is needed so that we can force all EISA interrupts to be active high
  in the forthcoming ioapic_config_intr().
- Fix a bug for EISA systems that didn't remap IRQs.  This would have broken
  EISA systems that tried to disable mixed mode for IRQ 0.

20 years agoIn swap_pager_getpages(), bp->b_dev can be NULL, particularly for the
bms [Wed, 23 Jun 2004 15:15:07 +0000 (15:15 +0000)]
In swap_pager_getpages(), bp->b_dev can be NULL, particularly for the
case of NFS mounted swap, so do not try to dereference it.

While we're here, brucify the printf() call which happens when we
time out on acquisition of vm_page_queue_mtx.

PR: kern/67898
Submitted by: bde (style)

20 years agoNow that we associate a device_t with ACPI device handles, lookup the
jhb [Wed, 23 Jun 2004 15:08:40 +0000 (15:08 +0000)]
Now that we associate a device_t with ACPI device handles, lookup the
device associated with any PCI devices that are enumerated in the ACPI
tree when adding children to an ACPI PCI bus and remove the duplicate
ACPI-only device_t and replace the device_t associated with the handle with
the ACPI and PCI aware device_t.

20 years agoClean CLEANFILES.
rik [Wed, 23 Jun 2004 14:22:14 +0000 (14:22 +0000)]
Clean CLEANFILES.

20 years agoImprove the kludge assignment of sub-bus numbers a little by fixing an
imp [Wed, 23 Jun 2004 13:49:46 +0000 (13:49 +0000)]
Improve the kludge assignment of sub-bus numbers a little by fixing an
off by one error.  subbus must be > than pribus, so the right test is
<= not just <.

Reported by: Peter Losher

20 years agoAvoid padding the value of "ucomm" when it is the last column in the line.
gad [Wed, 23 Jun 2004 12:18:36 +0000 (12:18 +0000)]
Avoid padding the value of "ucomm" when it is the last column in the line.

Submitted by: Cyrille Lefevre

20 years agoMake sure the value of "upr" (scheduling priority on return from system call)
gad [Wed, 23 Jun 2004 11:56:57 +0000 (11:56 +0000)]
Make sure the value of "upr" (scheduling priority on return from system call)
is scaled in the same way that "pri" (scheduling priority) is scaled.

Submitted by: Cyrille Lefevre

20 years agoInclude the `-c' option in the usage() message.
gad [Wed, 23 Jun 2004 11:40:56 +0000 (11:40 +0000)]
Include the `-c' option in the usage() message.

Submitted by: Cyrille Lefevre

20 years agoIn the sorting routine, sort by thread-id if two processes have the same PID.
gad [Wed, 23 Jun 2004 11:28:17 +0000 (11:28 +0000)]
In the sorting routine, sort by thread-id if two processes have the same PID.

Submitted by: Cyrille Lefevre

20 years agoMake code more clean: backout support for 3.x branch.
rik [Wed, 23 Jun 2004 11:23:54 +0000 (11:23 +0000)]
Make code more clean: backout support for 3.x branch.

20 years agoMFNetBSD 1.60, author: augustss
le [Wed, 23 Jun 2004 11:16:12 +0000 (11:16 +0000)]
MFNetBSD 1.60, author: augustss
   Several changes:
    * Implement read for ulpt.
    * If the device is not opened for reading, occasionally drain any
      data the printer might have (but don't hammer the printer with reads).
    * Lower the buffer size to one page.
   The driver seems to work with more printers now.

Obtained from: NetBSD

20 years agosave a few redundant lines by moving the retry loop further backwards.
phk [Wed, 23 Jun 2004 10:28:09 +0000 (10:28 +0000)]
save a few redundant lines by moving the retry loop further backwards.

20 years agoIt's /usr/sbin/nologin not /sbin/nologin
mlaier [Wed, 23 Jun 2004 09:42:19 +0000 (09:42 +0000)]
It's /usr/sbin/nologin not /sbin/nologin

Found-by: brueffer
Pointy-hat-to: mlaier

20 years agoFix some style bugs I introduced pointed out by bde. Also add a blank
dwmalone [Wed, 23 Jun 2004 08:26:31 +0000 (08:26 +0000)]
Fix some style bugs I introduced pointed out by bde. Also add a blank
line after the empty variable declarations.

Reviewed by: md5

20 years agoSince we grew struct xtty, t_dev doesn't have to be schizofrenic anymore.
phk [Wed, 23 Jun 2004 08:25:23 +0000 (08:25 +0000)]
Since we grew struct xtty, t_dev doesn't have to be schizofrenic anymore.

20 years ago - Remove unused definitions.
alc [Wed, 23 Jun 2004 08:06:52 +0000 (08:06 +0000)]
 - Remove unused definitions.
 - Move a definition inside the scope of a #ifdef _KERNEL.

20 years agoUpdate copyright years.
den [Wed, 23 Jun 2004 07:40:52 +0000 (07:40 +0000)]
Update copyright years.

20 years agoMFamd64: Document the machdep.hlt_cpus sysctl MIB variable.
bms [Wed, 23 Jun 2004 07:12:34 +0000 (07:12 +0000)]
MFamd64: Document the machdep.hlt_cpus sysctl MIB variable.

PR: i386/65729
Submitted by: Craig Rodrigues

20 years agoNow that <runetype.h> no longer brings in namespace pollution,
tjr [Wed, 23 Jun 2004 07:11:39 +0000 (07:11 +0000)]
Now that <runetype.h> no longer brings in namespace pollution,
bring back the inline functions for the !__BSD_VISIBLE case.

20 years agoPrefix the names of members of _RuneLocale and its sub-structures
tjr [Wed, 23 Jun 2004 07:01:44 +0000 (07:01 +0000)]
Prefix the names of members of _RuneLocale and its sub-structures
with ``__'' to avoid polluting the namespace. This doesn't change the
documented rune interface at all, but breaks applications that accessed
_RuneLocale directly.

20 years agoInclude <sys/mutex.h> and its prerequisite <sys/lock.h> instead of
bde [Wed, 23 Jun 2004 06:47:49 +0000 (06:47 +0000)]
Include <sys/mutex.h> and its prerequisite <sys/lock.h> instead of
depending on namespace pollution in <sys/vnode.h> for the definition
of mutex interfaces used in SOCKBUF_*LOCK().

Sorted includes.

Removed unused includes.

20 years agoInclude <sys/mutex.h> and its prerequisite <sys/lock.h> instead of
bde [Wed, 23 Jun 2004 06:35:43 +0000 (06:35 +0000)]
Include <sys/mutex.h> and its prerequisite <sys/lock.h> instead of
depending on namespace pollution in <sys/vnode.h> for the definition
of GIANT_REQUIRED.

Sorted includes.

20 years agoImplement the protection check required by the pmap_extract_and_hold()
alc [Wed, 23 Jun 2004 04:37:14 +0000 (04:37 +0000)]
Implement the protection check required by the pmap_extract_and_hold()
specification.  This enables the elimination of Giant from that function.

20 years agoTurn back on the building of GDB info files.
obrien [Wed, 23 Jun 2004 04:36:04 +0000 (04:36 +0000)]
Turn back on the building of GDB info files.

20 years agoAvoid calling bpf_filter() with len == 0, which causes a change in semantics
archie [Wed, 23 Jun 2004 02:37:10 +0000 (02:37 +0000)]
Avoid calling bpf_filter() with len == 0, which causes a change in semantics
(it treats the buffer pointer as an mbuf pointer) and subsequent panic.

MFC after: 3 days
Reported by: Tony Hariman <tony@cbn.net.id>

20 years agoAdd required includes for post-sorwakeup() change to fix FAST_IPSEC
bms [Wed, 23 Jun 2004 01:58:22 +0000 (01:58 +0000)]
Add required includes for post-sorwakeup() change to fix FAST_IPSEC
compilation.

20 years agoSwap order of ruleset load and enabling pf to work around a problem on altq
mlaier [Wed, 23 Jun 2004 01:42:06 +0000 (01:42 +0000)]
Swap order of ruleset load and enabling pf to work around a problem on altq
startup. Moreover, this is the "more logic" order.

20 years agoAdd "privsep" user/group _pflogd:_pflogd (64:64) to make pflogd(8) work
mlaier [Wed, 23 Jun 2004 01:32:28 +0000 (01:32 +0000)]
Add "privsep" user/group _pflogd:_pflogd (64:64) to make pflogd(8) work
again. This user/group is not required for install* targets, hence do not
add them to CHECK_UIDS/CHECK_GIDS in Makefile.inc1 (no need to annoy
people).

Discussed-on: -current

20 years agoRemove unlocked read annotation for sbspace(); the read is locked.
rwatson [Wed, 23 Jun 2004 00:35:50 +0000 (00:35 +0000)]
Remove unlocked read annotation for sbspace(); the read is locked.

20 years agoDocument the new passthrough interface.
scottl [Wed, 23 Jun 2004 00:27:56 +0000 (00:27 +0000)]
Document the new passthrough interface.

20 years agoBe specific about which socket properties are inherited from the original
bms [Wed, 23 Jun 2004 00:20:46 +0000 (00:20 +0000)]
Be specific about which socket properties are inherited from the original
socket upon accept(2).

PR: docs/54995, kern/45733

20 years agoFix an inconsistency in socket option propagation on accept(). Propagate
bms [Tue, 22 Jun 2004 23:58:09 +0000 (23:58 +0000)]
Fix an inconsistency in socket option propagation on accept(). Propagate
the SS_NBIO flag from the parent socket to the child socket during an
accept() operation.

The file descriptor O_NONBLOCK flag would have been propagated already
by the fflag assignment, and therefore would have been inconsistent
with the underlying socket's so_state member.

This makes accept() more closely adhere to the API contract we effectively
outline in the manual page. Note also that Linux continues to differ here;
O_NONBLOCK is not propagated. The other BSDs do propagate the flag, as
does Solaris. The Single UNIX Specification does not offer specific
advice on this issue.

PR: kern/45733
Requested by: Jayanth Vijayaraghavan
Reviewed by: rwatson

20 years agoEnable 'vinum' on AMD64.
obrien [Tue, 22 Jun 2004 23:15:01 +0000 (23:15 +0000)]
Enable 'vinum' on AMD64.

20 years agoReset the update flag when scanning for new devices.
ps [Tue, 22 Jun 2004 23:05:06 +0000 (23:05 +0000)]
Reset the update flag when scanning for new devices.

20 years agoThrow out pfsync and pflog, the modules were removed
brueffer [Tue, 22 Jun 2004 22:57:13 +0000 (22:57 +0000)]
Throw out pfsync and pflog, the modules were removed

20 years agoCorrect a misleading comment regarding the IPSEC_FILTERGIF option.
bms [Tue, 22 Jun 2004 22:02:57 +0000 (22:02 +0000)]
Correct a misleading comment regarding the IPSEC_FILTERGIF option.

PR: 57125
Requested by: Adrian Steinmann

20 years agoUpdate the copyright to use a 3 clause BSD license rather than one that
jhb [Tue, 22 Jun 2004 21:55:22 +0000 (21:55 +0000)]
Update the copyright to use a 3 clause BSD license rather than one that
is less clear about allowing redistribution of modified copies.

Requested by: Tom Cosgrove tom dot cosgrove at arches-consulting dot com

20 years agoRemove a bogus increment of %di when scanning the list of loader paths.
jhb [Tue, 22 Jun 2004 21:52:20 +0000 (21:52 +0000)]
Remove a bogus increment of %di when scanning the list of loader paths.
%di will already point to the character after the nul char when the
'repnz scasb' terminates.

Submitted by: Tom Cosgrove tom dot cosgrove at arches-consulting dot com

20 years agoImprove source-code compatibility with Linux applications using the
bms [Tue, 22 Jun 2004 21:46:49 +0000 (21:46 +0000)]
Improve source-code compatibility with Linux applications using the
IPX stack.

PR: kern/65217
Submitted by: Radim Kolar

20 years agoFix whitespace errors in previous commit.
brooks [Tue, 22 Jun 2004 21:07:54 +0000 (21:07 +0000)]
Fix whitespace errors in previous commit.

Reported by: simon

20 years agoRemove the TIOCDCDTIMESTAMP option.
phk [Tue, 22 Jun 2004 20:32:17 +0000 (20:32 +0000)]
Remove the TIOCDCDTIMESTAMP option.

The RFC-2783 PPS-API (<sys/timepps.h>) provides better and more
configurable service.

20 years agoAdd entry for network interface cloning changes.
brooks [Tue, 22 Jun 2004 20:28:03 +0000 (20:28 +0000)]
Add entry for network interface cloning changes.

20 years agoFix a few spelling mistakes in comments and clean them up a bit.
le [Tue, 22 Jun 2004 20:22:24 +0000 (20:22 +0000)]
Fix a few spelling mistakes in comments and clean them up a bit.

20 years agoCatch up with the dev_t -> struct cdev change.
le [Tue, 22 Jun 2004 20:18:15 +0000 (20:18 +0000)]
Catch up with the dev_t -> struct cdev change.

20 years agoBump FreeBSD_version for if_clone overhaul.
brooks [Tue, 22 Jun 2004 20:15:42 +0000 (20:15 +0000)]
Bump FreeBSD_version for if_clone overhaul.

20 years agoMajor overhaul of pseudo-interface cloning. Highlights include:
brooks [Tue, 22 Jun 2004 20:13:25 +0000 (20:13 +0000)]
Major overhaul of pseudo-interface cloning.  Highlights include:

 - Split the code out into if_clone.[ch].
 - Locked struct if_clone. [1]
 - Add a per-cloner match function rather then simply matching names of
   the form <name><unit> and <name>.
 - Use the match function to allow creation of <interface>.<tag>
   vlan interfaces.  The old way is preserved unchanged!
 - Also the match function to allow creation of stf(4) interfaces named
   stf0, stf, or 6to4.  This is the only major user visible change in
   that "ifconfig stf" creates the interface stf rather then stf0 and
   does not print "stf0" to stdout.
 - Allow destroy functions to fail so they can refuse to delete
   interfaces.  Currently, we forbid the deletion of interfaces which
   were created in the init function, particularly lo0, pflog0, and
   pfsync0.  In the case of lo0 this was a panic implementation so it
   does not count as a user visiable change. :-)
 - Since most interfaces do not need the new functionality, an family of
   wrapper functions, ifc_simple_*(), were created to wrap old style
   cloner functions.
 - The IF_CLONE_INITIALIZER macro is replaced with a new incompatible
   IFC_CLONE_INITIALIZER and ifc_simple consumers use IFC_SIMPLE_DECLARE
   instead.

Submitted by:   Maurycy Pawlowski-Wieronski <maurycy at fouk.org> [1]
Reviewed by:    andre, mlaier
Discussed on: net

20 years agoImplement the protection check required by the pmap_extract_and_hold()
alc [Tue, 22 Jun 2004 19:35:43 +0000 (19:35 +0000)]
Implement the protection check required by the pmap_extract_and_hold()
specification.  This enables the elimination of Giant from that function.

Reviewed by: tegge@

20 years agoFix a problem with multiple-delivery (i.e. broadcast) in ng_bridge.
green [Tue, 22 Jun 2004 18:56:47 +0000 (18:56 +0000)]
Fix a problem with multiple-delivery (i.e. broadcast) in ng_bridge.
Only the first link0..link$NLINKS hooks would be utilized, whereas
the link hooks may be connected sparsely.

Add a counter variable so that the link hook array is only traversed
while there is still work to do, but that it continues up to the end
if it has to.

20 years agoNo need to free the softc, because it wasn't allocated.
le [Tue, 22 Jun 2004 18:13:43 +0000 (18:13 +0000)]
No need to free the softc, because it wasn't allocated.

20 years agoUse the maximum physical bus when rescanning those targets. This
ps [Tue, 22 Jun 2004 17:39:57 +0000 (17:39 +0000)]
Use the maximum physical bus when rescanning those targets.  This
fixes a panic on cards which do not have any drives attached.

20 years agoFix a paste-o in key_cmpspidx_withmask().
bms [Tue, 22 Jun 2004 17:17:44 +0000 (17:17 +0000)]
Fix a paste-o in key_cmpspidx_withmask().

PR: misc/67013
Submitted by: Zhenmin <zli4@cs.uiuc.edu>

20 years agoAdjust the system endian and a.out headers to be more MI and cross-building
obrien [Tue, 22 Jun 2004 17:05:39 +0000 (17:05 +0000)]
Adjust the system endian and a.out headers to be more MI and cross-building
friendly.  Use the systems headers rather than local versions.

Reviewed by: ru

20 years agoUse nanoseconds and then lexicographic ordering when the seconds of
dwmalone [Tue, 22 Jun 2004 16:02:29 +0000 (16:02 +0000)]
Use nanoseconds and then lexicographic ordering when the seconds of
the [acm]time are the same. I was going to use Scott's patch, but I
couldn't get the style quite right, so I used a patch of my own.

Submitted by: Scott Mitchell <scott+freebsd at fishballoon.org>
MFC after: 3 weeks

20 years agoDon't sleep in the g_down path. More error checks to come.
le [Tue, 22 Jun 2004 14:54:31 +0000 (14:54 +0000)]
Don't sleep in the g_down path.  More error checks to come.

20 years agoMerged from sys/dev/sio/sio.c revisions 1.439 and 1.440.
nyan [Tue, 22 Jun 2004 12:12:10 +0000 (12:12 +0000)]
Merged from sys/dev/sio/sio.c revisions 1.439 and 1.440.

20 years agoMFi386: revision 1.592.
nyan [Tue, 22 Jun 2004 12:11:20 +0000 (12:11 +0000)]
MFi386: revision 1.592.

20 years agoUse the right ordering of args on mtx_init(). No functional changes
sos [Tue, 22 Jun 2004 11:18:25 +0000 (11:18 +0000)]
Use the right ordering of args on mtx_init(). No functional changes
since the args in question was all zero's.

Found by: Jimmy Olgeni <olgeni@FreeBSD.org>

20 years agoWhitespace cleanup.
des [Tue, 22 Jun 2004 10:33:58 +0000 (10:33 +0000)]
Whitespace cleanup.

20 years agoUpdate reference to FAQ (3.16 -> 5.3, still hardcoded)
den [Tue, 22 Jun 2004 10:11:26 +0000 (10:11 +0000)]
Update reference to FAQ (3.16 -> 5.3, still hardcoded)

20 years agong_fec(4) is not anymore incompatible with ng_ether(4).
ru [Tue, 22 Jun 2004 09:23:31 +0000 (09:23 +0000)]
ng_fec(4) is not anymore incompatible with ng_ether(4).

Submitted by: Gleb Smirnoff

20 years agofix handling of an unknown progeam name
eik [Tue, 22 Jun 2004 09:18:50 +0000 (09:18 +0000)]
fix handling of an unknown progeam name

Noticed by: Joshua Goodall <joshua@roughtrade.net>
Approved by: ru

20 years agoFixed copyright owner.
ru [Tue, 22 Jun 2004 09:11:01 +0000 (09:11 +0000)]
Fixed copyright owner.

20 years agoDocument the "fec" Netgraph node type.
ru [Tue, 22 Jun 2004 09:09:18 +0000 (09:09 +0000)]
Document the "fec" Netgraph node type.

Submitted by: Gleb Smirnoff
OK'ed by: julian, wpaul

20 years agoSet up logging before trying to lock the sandbox; otherwise tbmaster(1)
des [Tue, 22 Jun 2004 08:34:46 +0000 (08:34 +0000)]
Set up logging before trying to lock the sandbox; otherwise tbmaster(1)
will just log a mysterious "tinderbox returned exit code 35".

20 years agoFix typos and add spaces before `(' in some comments
brueffer [Tue, 22 Jun 2004 08:23:26 +0000 (08:23 +0000)]
Fix typos and add spaces before `(' in some comments

Submitted by: markus
Approved by: njl
MFC after: 3 days

20 years agoCatch up with g_dev_open() changes and pass over the current thread
le [Tue, 22 Jun 2004 06:38:01 +0000 (06:38 +0000)]
Catch up with g_dev_open() changes and pass over the current thread
pointer.

Submitted by:    cperciva

20 years agorevert section of code that calls netisr_queue() to match v1.33 of this file
darrenr [Tue, 22 Jun 2004 05:20:30 +0000 (05:20 +0000)]
revert section of code that calls netisr_queue() to match v1.33 of this file

20 years ago#ifdef's for FreeBSD are wrong, causing too many variable declaractions to
darrenr [Tue, 22 Jun 2004 05:16:29 +0000 (05:16 +0000)]
#ifdef's for FreeBSD are wrong, causing too many variable declaractions to
disappear.

20 years agoRegenerate after updating syscalls.master.
rwatson [Tue, 22 Jun 2004 04:36:25 +0000 (04:36 +0000)]
Regenerate after updating syscalls.master.

20 years agoMark unlink() as MPSAFE as we now acquire Giant in the unlink()
rwatson [Tue, 22 Jun 2004 04:34:55 +0000 (04:34 +0000)]
Mark unlink() as MPSAFE as we now acquire Giant in the unlink()
system call.

20 years agoAcquire Giant in link() so that the system call can be marked
rwatson [Tue, 22 Jun 2004 04:34:05 +0000 (04:34 +0000)]
Acquire Giant in link() so that the system call can be marked
MPSAFE.  Don't want to acquire Giant in kern_link() sync linux
compat code performs actions requiring Giant prior to calling
kern_link().

20 years agoRebuild following marking link() as MPSAFE.
rwatson [Tue, 22 Jun 2004 04:29:59 +0000 (04:29 +0000)]
Rebuild following marking link() as MPSAFE.