]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
17 years agoMFC:
jinmei [Sun, 10 Jun 2007 17:17:55 +0000 (17:17 +0000)]
MFC:
fixed memory leak for IPv6 multicast membership information associated
with interface addresses.

Approved by:    gnn (mentor)

src/sys/netinet6/in6.c: 1.71
src/sys/netinet6/in6_ifattach.c: 1.36
src/sys/netinet6/in6_var.h: 1.31

17 years agoMFC rev. 1.103: make ipfw(8) show rules with mac/mac-type options
maxim [Sun, 10 Jun 2007 13:59:36 +0000 (13:59 +0000)]
MFC rev. 1.103: make ipfw(8) show rules with mac/mac-type options
correctly.
MFC rev. 1.105: teach get_mac_addr_mask() to not silently accept
incorrect MAC addresses.

17 years agoMFC remnant of rev. 1.7.
kib [Sun, 10 Jun 2007 11:09:39 +0000 (11:09 +0000)]
MFC remnant of rev. 1.7.

17 years agoMFC 1.30: Help ng_fec deal with multicast addresses.
dwmalone [Sun, 10 Jun 2007 07:41:25 +0000 (07:41 +0000)]
MFC 1.30: Help ng_fec deal with multicast addresses.

This isn't exactly the same patch as in -current as some of the
in-kernel multicast API is different.

17 years agoMFC 1.209: Don't verify UDP IPv4 checksum in-place.
dwmalone [Sun, 10 Jun 2007 07:29:08 +0000 (07:29 +0000)]
MFC 1.209: Don't verify UDP IPv4 checksum in-place.

17 years agoMFC: Fix some warnings by making things const, adding missing headers,
dwmalone [Sun, 10 Jun 2007 07:04:09 +0000 (07:04 +0000)]
MFC: Fix some warnings by making things const, adding missing headers,
removing some unused variables and making a variable unsigned.

17 years agoMFC:
matteo [Sat, 9 Jun 2007 09:28:30 +0000 (09:28 +0000)]
MFC:
rpcbind.c: rev. 1.15, 1.16, 1.17
rpcbind.8: rev. 1.10

1)Make it possible for rpcbind(8) to bind TCP listening socket to an IP
other than INADDR_ANY.

2) Add the -6 option to specify "IPv6 only".

PR: 84494, 1122566

17 years agoMFC: (revision 1.9)
ariff [Fri, 8 Jun 2007 17:33:38 +0000 (17:33 +0000)]
MFC: (revision 1.9)

Fix (enable) phone out for laptops with ALC655, specifically
for Amilo Pro V2055.

PR: kern/113101
Submitted by: konrad@egipt-medytacje.pl

17 years agoMFC 1.139: Fix a snafu in the changes in 1.138.
jhb [Fri, 8 Jun 2007 16:51:20 +0000 (16:51 +0000)]
MFC 1.139: Fix a snafu in the changes in 1.138.

PR: kern/113387
Submitted by: Andre Albsmeier

17 years agoMFC addition of bsd.port.options.mk
pav [Fri, 8 Jun 2007 16:05:32 +0000 (16:05 +0000)]
MFC addition of bsd.port.options.mk

17 years agoMFC:
bz [Thu, 7 Jun 2007 09:50:53 +0000 (09:50 +0000)]
MFC:
  1.200 sbin/ipfw/ipfw.8, 1.104 sbin/ipfw/ipfw2.c
  1.110 sys/netinet/ip_fw.h, 1.164 sys/netinet/ip_fw2.c

  Add support for filtering on Routing Header Type 0 and
  Mobile IPv6 Routing Header Type 2 in addition to filter
  on the non-differentiated presence of any Routing Header.

17 years agoMFC: Make gcore(1) 64 bit safe.
jhb [Wed, 6 Jun 2007 22:06:42 +0000 (22:06 +0000)]
MFC: Make gcore(1) 64 bit safe.

17 years agoMFC: Add support for kernel modules with a single PT_LOAD section.
jhb [Wed, 6 Jun 2007 21:43:35 +0000 (21:43 +0000)]
MFC: Add support for kernel modules with a single PT_LOAD section.

17 years agoMFC 1.176: Add vm map and object locking to each_writable_segment().
jhb [Wed, 6 Jun 2007 21:30:42 +0000 (21:30 +0000)]
MFC 1.176: Add vm map and object locking to each_writable_segment().

17 years agoMerge 1.47:
glebius [Wed, 6 Jun 2007 16:20:50 +0000 (16:20 +0000)]
Merge 1.47:
  Make it possible that carpdetach() unlocks on return. Then, in
  carp_clone_destroy() we are on a safe side, we don't need to
  unlock the cif, that can me already non-existent at this point.

  Reported by:    Anton Yuzhaninov <citrin rambler-co.ru>

17 years agoMFC all of the remaining changes. This brings over improved ioctl
ambrisko [Wed, 6 Jun 2007 15:59:29 +0000 (15:59 +0000)]
MFC all of the remaining changes.  This brings over improved ioctl
support and makes MegaCli work on amd64.  Scott supplied patches
to make the ioctl commands not block the controller.  MegaCli
still has some issues with our emulation and create some bogus
commands so flashing doesn't work.

This should loosely resolve 113232.  We were waiting on some other
fixes to go into -current before migrating this to -stable.

PR: 113232
Submitted by: scottl

17 years agoMFC r1.99
thompsa [Wed, 6 Jun 2007 00:59:20 +0000 (00:59 +0000)]
MFC r1.99
  Remove a KASSERT intended to help the developer, the condition is no longer
  valid since the span code was added.

17 years agoFix last-minute typo.
scottl [Tue, 5 Jun 2007 22:54:01 +0000 (22:54 +0000)]
Fix last-minute typo.

17 years agoMFC: MPI 1.5.16, SAS topology scan, SATA WC enable tunable
scottl [Tue, 5 Jun 2007 21:32:57 +0000 (21:32 +0000)]
MFC: MPI 1.5.16, SAS topology scan, SATA WC enable tunable

17 years agoAdd build glue for the mfip driver.
scottl [Tue, 5 Jun 2007 21:31:11 +0000 (21:31 +0000)]
Add build glue for the mfip driver.

17 years agoAdd the CAM passthrough driver for MFI
scottl [Tue, 5 Jun 2007 21:27:22 +0000 (21:27 +0000)]
Add the CAM passthrough driver for MFI

17 years agoMFC of a couple small fixes: 82571 LAA reset workaround,
jfv [Tue, 5 Jun 2007 17:44:01 +0000 (17:44 +0000)]
MFC of a couple small fixes: 82571 LAA reset workaround,
 and invalid eeprom problem on some PCI-E systems.

17 years agoMerged changes from HEAD, including:
davidch [Tue, 5 Jun 2007 01:56:08 +0000 (01:56 +0000)]
Merged changes from HEAD, including:

- Updated copyright date to 2007.

- Tested with BCM5706 A3.

- Added ID for BCM5708 B2.

- Removed unused driver version string.

- Modified BCE_PRINTF macro to automatically fill-in the sc pointer.

- Fixed a kernel panic when the driver was loaded as a module from the
command-line because the MII bus pointer was null (i.e. the MII bus
hadn't been enumerated yet).

- Added fix proposed by Vladimir Ivanov <wawa@yandex-team.ru> to prevent
driver state corruption when releasing the lock during the ISR in
bce_rx_intr() to send packets up the stack.

- Added new TX chain and register read sysctl interfaces for debugging.

- Cleaned up formatting for various other debug routines.

- Added a new statistic maintained by firmware which tracks the number
of received packets dropped because no receive buffers are available.

17 years agoUnbreak release notes build.
bmah [Mon, 4 Jun 2007 18:26:35 +0000 (18:26 +0000)]
Unbreak release notes build.

17 years agoMFC:
ariff [Mon, 4 Jun 2007 09:06:05 +0000 (09:06 +0000)]
MFC:

Simmilar, but different treatment:

Insert NULL pointer checking around devclass_get_maxunit(pcm_devclass, ..) .
Things can get ugly without it due to uninitialized class.

err.. perhaps we should teach devclass_get_maxunit() to return -1 ?

17 years agoThis commit was manufactured by cvs2svn to create branch 'RELENG_6'.
cvs2svn [Mon, 4 Jun 2007 02:50:31 +0000 (02:50 +0000)]
This commit was manufactured by cvs2svn to create branch 'RELENG_6'.

17 years agoMFC: Add support for Asus A3N laptops.
philip [Sun, 3 Jun 2007 22:44:23 +0000 (22:44 +0000)]
MFC: Add support for Asus A3N laptops.

Submitted by: Holger Jorra <holger_jorra -at- gmx.net>

17 years agoMFC:
brueffer [Sun, 3 Jun 2007 14:59:11 +0000 (14:59 +0000)]
MFC:

New release notes:
- touch -A
- xorg 7.2 update
- mxge(4)
- X11BASE -> LOCALBASE transition
- netcat update to OpenBSD 4.1 version
- fdisk -p

17 years agoMFC: Don't lie about saved reports.
philip [Sun, 3 Jun 2007 13:52:46 +0000 (13:52 +0000)]
MFC: Don't lie about saved reports.

PR: gnu/89777
Submitted by: edwin

17 years agoMFC: rev 1.10
simokawa [Sun, 3 Jun 2007 13:43:00 +0000 (13:43 +0000)]
MFC: rev 1.10
Discard backlog on GDB port when connected.

17 years agoMFC: moused.c rev 1.77 and rev 1.78
philip [Sun, 3 Jun 2007 13:36:04 +0000 (13:36 +0000)]
MFC: moused.c rev 1.77 and rev 1.78
     moused.8 rev 1.60 and rev 1.61
     Makefile rev 1.9

    Add dynamic acceleration to moused(8).  This introduces a '-A' flag to
    control the acceleration algorithm.  It can be used together with the '-a'
    flag for regular acceleration. [1]

    Convert macros to use C99's syntax for macros with a variable number of
    arguments.

    Markup fixes to moused(8).

Submitted by: Oliver Fromme <olli -at- lurza.secnetix.de> [1]
Behind on MFCs: philip

17 years agoMFC if_re.c, 1.84 to RELENG_6
yongari [Sun, 3 Jun 2007 07:47:40 +0000 (07:47 +0000)]
MFC if_re.c, 1.84 to RELENG_6
  The TCP checksum offload handling in the 8111B/8168B and 8101E PCIe can
  apparently be confused by short TCP segments that have been manually
  padded to the minimum ethernet frame size. The driver does short frame
  padding in software as a workaround for a bug in the 8169 PCI devices
  that causes short IP fragments to be corrupted due to an apparent
  conflict between the hardware autopadding and hardware IP checksumming.

  To fix this, we avoid software padding for short TCP segments, since
  the hardware seems to autopad and checksum these correctly (even the
  older 8169 NICs get these right). Short UDP packets appear to be
  handled correctly in all cases. This should work around the IP header
  checksum bug in the 8169 while not tripping the TCP checksum bug in
  the 8111B/8168B and 8101E.

PR: 112089

17 years agoMFC if_re.c, 1.81 to RELENG_6
yongari [Sun, 3 Jun 2007 07:43:15 +0000 (07:43 +0000)]
MFC if_re.c, 1.81 to RELENG_6
  Fix re_setmulti() so that it works correctly for PCIe chips where
  the multicast hash table are in reverse order compared to older
  devices.

PR: 112089

17 years agoMFC: remove extraneous "can't get _ADR" messages.
njl [Sat, 2 Jun 2007 19:15:58 +0000 (19:15 +0000)]
MFC: remove extraneous "can't get _ADR" messages.

17 years agoMFC: 1.78 + 1.85
brueffer [Sat, 2 Jun 2007 08:02:23 +0000 (08:02 +0000)]
MFC: 1.78 + 1.85

Enable automatic hardware notes generation for mxge(4).

17 years agoMFC: various improvements, including fileno allocation bug fix
des [Fri, 1 Jun 2007 20:56:18 +0000 (20:56 +0000)]
MFC: various improvements, including fileno allocation bug fix

17 years agoFix a breakage with "MODULES_WITH_WORLD=true make buildworld".
simokawa [Fri, 1 Jun 2007 00:50:59 +0000 (00:50 +0000)]
Fix a breakage with "MODULES_WITH_WORLD=true make buildworld".

PR: kern/113204

17 years agoFix comma placement error in previous commit.
cperciva [Thu, 31 May 2007 18:37:09 +0000 (18:37 +0000)]
Fix comma placement error in previous commit.

17 years agoMFC rev. 1.33 traceroute.c and rev. 1.18 traceroute6.c: make it
maxim [Thu, 31 May 2007 14:01:51 +0000 (14:01 +0000)]
MFC rev. 1.33 traceroute.c and rev. 1.18 traceroute6.c: make it
possible to use 1 sec wait time.

17 years agoMFC rev. 1.39 fdisk.8 and rev. 1.83 fdisk.c: add -p flag -- print
maxim [Thu, 31 May 2007 13:52:24 +0000 (13:52 +0000)]
MFC rev. 1.39 fdisk.8 and rev. 1.83 fdisk.c: add -p flag -- print
a slice table in fdisk configuration file format.

17 years agoMFC: netcat from OpenBSD 4.1-RELEASE.
delphij [Thu, 31 May 2007 07:23:11 +0000 (07:23 +0000)]
MFC: netcat from OpenBSD 4.1-RELEASE.

17 years agoMFC revision 1.137
delphij [Thu, 31 May 2007 02:42:37 +0000 (02:42 +0000)]
MFC revision 1.137
date: 2007/05/17 06:08:42;  author: delphij;  state: Exp;  lines: +8 -5
Tighten IP address check, prevent '..' from being passed
from the check.

Submitted by:   Ren Zhen <bg1tpt gmail com>

17 years agoMFC: rev. 1.7
simokawa [Wed, 30 May 2007 12:23:38 +0000 (12:23 +0000)]
MFC: rev. 1.7
Change default value of hw.firewire.dcons_crom.force_console to 0
for least astonishment.

17 years agoMFC rev. 1.13: Make compile without GDB option.
simokawa [Wed, 30 May 2007 12:22:08 +0000 (12:22 +0000)]
MFC rev. 1.13: Make compile without GDB option.

17 years agoMFC r1.13, 1.14
thompsa [Wed, 30 May 2007 08:47:15 +0000 (08:47 +0000)]
MFC r1.13, 1.14
- Fix a mbuf leak where sc_start fails or the protocol is none.
- packets on the input interface were counted twice
- Use IFQ_HANDOFF instead of rolling our own

17 years agoMFC 1.19:
yar [Wed, 30 May 2007 08:26:01 +0000 (08:26 +0000)]
MFC 1.19:

  Fix a formatting error: the list of control packets ends earlier.

17 years agoMFC the mxge 10GbE driver.
gallatin [Tue, 29 May 2007 16:37:16 +0000 (16:37 +0000)]
MFC the mxge 10GbE driver.

17 years agoMFC: Add a few sentences of caution for Microsoft Vista users who might
bmah [Tue, 29 May 2007 14:16:25 +0000 (14:16 +0000)]
MFC:  Add a few sentences of caution for Microsoft Vista users who might
want to try resizing partitions.

install.sgml 1.37 -> 1.39

17 years agoMFC: (1.78) handle temporary redirects (HTTP status code 307)
des [Tue, 29 May 2007 12:35:26 +0000 (12:35 +0000)]
MFC: (1.78) handle temporary redirects (HTTP status code 307)

17 years agoMFC rev. 1.309 of sys/kern/kern_descrip.c,
kib [Tue, 29 May 2007 10:09:43 +0000 (10:09 +0000)]
MFC rev. 1.309 of sys/kern/kern_descrip.c,
    rev. 1.438 of sys/kern/vfs_syscalls.c,
    rev. 1.77 of sys/sys/filedesc.h:
Mark the filedescriptor table entries with VOP_OPEN being performed for them
as UF_OPENING. Disable closing of that entries. This should fix the crashes
caused by devfs_open() (and fifo_open()) dereferencing struct file * by
index, while the filedescriptor is closed by parallel thread.

RELENG_6 testing by: Mark Kane <mark at mkproductions org>

17 years agoMFC: (1.143) break cyclic dependency
des [Tue, 29 May 2007 09:28:10 +0000 (09:28 +0000)]
MFC: (1.143) break cyclic dependency

17 years agoEnable 4 virtual playback channels by default, since the last stumbling
ariff [Mon, 28 May 2007 21:18:06 +0000 (21:18 +0000)]
Enable 4 virtual playback channels by default, since the last stumbling
block (snd_ich) has been dealt with.

17 years agoMFC (revision 1.75)
ariff [Mon, 28 May 2007 21:07:41 +0000 (21:07 +0000)]
MFC (revision 1.75)

* Fix possible integer overflow during speed calibration. Delay final
  initialization (pcm_setstatus) after calibration finished.

PR: kern/100169
Tested by: Kevin Overman <oberman@es.net>

* Fix (or rather workaround) Intel 440MX Errata #36
  - http://www.intel.com/design/chipsets/specupdt/245051.htm

AC97 Soft Audio and Soft Modem Master Abort Errata

Issue:
  Use of either soft audio or soft modem on an IntelĀ® 82443MX PCISet
  based platform running a 100 MHz Processor System Bus and an AC97 codec
  may result in failures. The system continues to function normally while
  the AC97 hardware may not resume and may require a cold-boot to
  recover. As a result of the failure, the Master Abort Status bit will
  be set in the audio or modem function PCI header space.

Workaround:
  Force uncacheable DMA on both BDL and pcm buffers.

Tested by: Emil Holmstr|m <emil@linux.se>

17 years agom_sanity isn't available on RELENG_6 - don't define with INVARIANTS
kmacy [Mon, 28 May 2007 07:37:47 +0000 (07:37 +0000)]
m_sanity isn't available on RELENG_6 - don't define with INVARIANTS

17 years agoadd new files to build
kmacy [Mon, 28 May 2007 07:08:44 +0000 (07:08 +0000)]
add new files to build

17 years ago- Update to 1.0.0.86 of vendor source
kmacy [Mon, 28 May 2007 06:26:07 +0000 (06:26 +0000)]
- Update to 1.0.0.86 of vendor source
- add some performance improvements for small packet workloads
  - Double the number of descriptors that a single call to send can use
  - Quadruple the number of descriptors that can be reclaimed per pass
  - only run reclaim twice per second
  - increase coalesce timer from 3.5us to 5us
- set OACTIVE so interface doesn't stall when the ring fills

Sponsored by: Chelsio Inc.

17 years agoRespect NO_INET6 knob (MFC in spirit of rev. 1.61).
delphij [Mon, 28 May 2007 06:16:40 +0000 (06:16 +0000)]
Respect NO_INET6 knob (MFC in spirit of rev. 1.61).

Submitted by: Alex Kozlov <spam rm-rf.kiev.ua>
PR: bin/96335

17 years agoMFC: Make sure fwsid is not NULL.
simokawa [Mon, 28 May 2007 04:36:37 +0000 (04:36 +0000)]
MFC: Make sure fwsid is not NULL.

17 years agoMFC: Simplify the bus probe routin using a kthread.
simokawa [Mon, 28 May 2007 04:33:43 +0000 (04:33 +0000)]
MFC: Simplify the bus probe routin using a kthread.

17 years agoMFC: Include stand.h for loader.
simokawa [Mon, 28 May 2007 04:31:23 +0000 (04:31 +0000)]
MFC: Include stand.h for loader.

17 years agoMFC pfsync_syncpeer option for rc.conf(5).
keramida [Mon, 28 May 2007 02:00:54 +0000 (02:00 +0000)]
MFC pfsync_syncpeer option for rc.conf(5).

: Add a pfsync_syncpeer option to /etc/defaults/rc.conf and rc.conf(5),
: which can be used to turn off multicast pfsync support, and enable
: the transmission of directed PFSYNC (IP protocol: 240) packets to
: a specific "sync peer" host.
:
: PR:             conf/111225
: Submitted by:   Bas van Beek <bas@tobin.nl>
: Approved by:    mtm, mlaier
: MFC after:      2 weeks
:
: Revision  Changes    Path
: 1.314     +1 -0      src/etc/defaults/rc.conf
: 1.5       +6 -1      src/etc/rc.d/pfsync
: 1.323     +20 -0     src/share/man/man5/rc.conf.5

17 years agoMFC:
le [Sun, 27 May 2007 10:15:18 +0000 (10:15 +0000)]
MFC:

   Fix handling of filenames containing whitespace.

   PR:      bin/112860

17 years agoThis commit was manufactured by cvs2svn to create branch 'RELENG_6'.
cvs2svn [Sun, 27 May 2007 04:39:30 +0000 (04:39 +0000)]
This commit was manufactured by cvs2svn to create branch 'RELENG_6'.

17 years agoMFC rev. 1.422: fix typo.
maxim [Sat, 26 May 2007 07:31:47 +0000 (07:31 +0000)]
MFC rev. 1.422: fix typo.

17 years agoo MFC revs. 1.33 - 1.35: make iostat -Ix work, style.
maxim [Sat, 26 May 2007 07:30:23 +0000 (07:30 +0000)]
o MFC revs. 1.33 - 1.35: make iostat -Ix work, style.

17 years agoMFC: add the debug.acpi.suspend_bounce sysctl.
njl [Sat, 26 May 2007 00:48:55 +0000 (00:48 +0000)]
MFC: add the debug.acpi.suspend_bounce sysctl.

17 years agoMFC rev 1.190 (avoid bogus geometry if transport error)
mjacob [Fri, 25 May 2007 20:14:02 +0000 (20:14 +0000)]
MFC rev 1.190 (avoid bogus geometry if transport error)
MFC rev 1.200-1.202 (EasyMP3 quirk)

17 years agoMFC 1.49: Get the correct softc pointer when checking to see if a PCI-PCI
jhb [Fri, 25 May 2007 17:07:32 +0000 (17:07 +0000)]
MFC 1.49: Get the correct softc pointer when checking to see if a PCI-PCI
bridge has MSI blacklisted.

17 years agoMFC: Support 64-bit scatter-gather DMA addresses in 64-bit capable AHCI and
jhb [Fri, 25 May 2007 17:06:46 +0000 (17:06 +0000)]
MFC: Support 64-bit scatter-gather DMA addresses in 64-bit capable AHCI and
Marvell controllers.

Approved by: sos

17 years agoMFC: Fix statistical accounting for bytes and packets during sack
jhb [Fri, 25 May 2007 16:36:39 +0000 (16:36 +0000)]
MFC: Fix statistical accounting for bytes and packets during sack
retransmits.

17 years agoAll right, pass me the pointy hat. Come to think of it, I'll take the
des [Fri, 25 May 2007 13:31:36 +0000 (13:31 +0000)]
All right, pass me the pointy hat.  Come to think of it, I'll take the
whole box...

17 years agoMFC:
jinmei [Fri, 25 May 2007 06:35:46 +0000 (06:35 +0000)]
MFC:
- Disabled responding to NI queries from a global address by default as
  specified in RFC4620.  A new flag for icmp6_nodeinfo was added to enable the
  feature.
- Also cleaned up the code so that the semantics of the icmp6_nodeinfo
  flags is clearer (i.e., defined specific macro names instead of using
  hard-coded values).

Approved by:    gnn (mentor)

src/sys/netinet/icmp6.h: 1.21
src/sys/netinet6/icmp6.c: 1.76
src/sys/netinet6/in6_proto.c: 1.43

17 years agoMFC 1.29:
pav [Thu, 24 May 2007 22:52:46 +0000 (22:52 +0000)]
MFC 1.29:
  Replace a fairly opaque sentence with a much clearer wording from NetBSD.

17 years agoRevert previous commit.
des [Thu, 24 May 2007 19:15:09 +0000 (19:15 +0000)]
Revert previous commit.

17 years agoMFC (forgotten in previous commit)
des [Thu, 24 May 2007 19:13:40 +0000 (19:13 +0000)]
MFC (forgotten in previous commit)

17 years agoFix assertion accidentally changed to HEAD version. This was breaking the
dwhite [Thu, 24 May 2007 18:55:28 +0000 (18:55 +0000)]
Fix assertion accidentally changed to HEAD version. This was breaking the
RELENG_6 build.

Pointy hat to: des

17 years agoMFC: fix infinite loop bug
des [Thu, 24 May 2007 16:18:50 +0000 (16:18 +0000)]
MFC: fix infinite loop bug

17 years agoMFC: simplify vnode reclamation, fix fileno allocation
des [Thu, 24 May 2007 16:17:44 +0000 (16:17 +0000)]
MFC: simplify vnode reclamation, fix fileno allocation

17 years agoMFC: remove T/TCP support
des [Thu, 24 May 2007 16:15:32 +0000 (16:15 +0000)]
MFC: remove T/TCP support

17 years agoMFC: add FILESYSTEMS
des [Thu, 24 May 2007 16:14:37 +0000 (16:14 +0000)]
MFC: add FILESYSTEMS

17 years agoMFC: apply "additional TCP options" earlier
des [Thu, 24 May 2007 16:10:52 +0000 (16:10 +0000)]
MFC: apply "additional TCP options" earlier

17 years agoMFC: expose vdropl()
des [Thu, 24 May 2007 16:09:38 +0000 (16:09 +0000)]
MFC: expose vdropl()

17 years agoMFC: (1.9) missing El
des [Thu, 24 May 2007 16:07:32 +0000 (16:07 +0000)]
MFC: (1.9) missing El

17 years agoMFC: (1.14) set default refresh interval to one second
des [Thu, 24 May 2007 16:06:54 +0000 (16:06 +0000)]
MFC: (1.14) set default refresh interval to one second

17 years agoMFC: (1.60, 1.61) errors in nocona / prescott support
des [Thu, 24 May 2007 16:05:25 +0000 (16:05 +0000)]
MFC: (1.60, 1.61) errors in nocona / prescott support

17 years agoMFC: add CPUTYPE support for "core" and "core2".
des [Thu, 24 May 2007 16:03:44 +0000 (16:03 +0000)]
MFC: add CPUTYPE support for "core" and "core2".

17 years agoMFC rev. 1.62: Complain about "tar -c /path/which/does/not/exist".
cperciva [Thu, 24 May 2007 09:27:50 +0000 (09:27 +0000)]
MFC rev. 1.62: Complain about "tar -c /path/which/does/not/exist".

17 years agoMFC: Use unsigned comparisons.
cperciva [Thu, 24 May 2007 09:24:18 +0000 (09:24 +0000)]
MFC: Use unsigned comparisons.

17 years agoMFC 1.308: Use kern_open() to open /dev/null in fdcheckstd().
jhb [Wed, 23 May 2007 20:36:24 +0000 (20:36 +0000)]
MFC 1.308: Use kern_open() to open /dev/null in fdcheckstd().

17 years agoRevert previous commit, was part of a different change.
jhb [Wed, 23 May 2007 20:29:41 +0000 (20:29 +0000)]
Revert previous commit, was part of a different change.

Reported by: kib

17 years agoMFC: rev. 1.47, 1.48 and relevant changes from 1.49
brueffer [Wed, 23 May 2007 19:45:49 +0000 (19:45 +0000)]
MFC: rev. 1.47, 1.48 and relevant changes from 1.49

- Use our standard section 4 SYNOPSIS.
- Add a bunch of missing Xrefs.
- Minor nits.

17 years agoMFC: Rework the support used by ABIs to override resource limits so that
jhb [Wed, 23 May 2007 18:24:43 +0000 (18:24 +0000)]
MFC: Rework the support used by ABIs to override resource limits so that
a 64-bit process exec'd by a 32-bit process doesn't end up with 32-bit
limits.

This doesn't break the ABI as neither of the 32-bit ABIs (COMPAT_LINUX32
and COMPAT_IA32) are buildable as modules on 6.x/amd64 and none of the
other ABIs use this hook.

17 years agoMFC 1.352: Fix a typo in a bootverbose printf.
jhb [Wed, 23 May 2007 18:12:00 +0000 (18:12 +0000)]
MFC 1.352: Fix a typo in a bootverbose printf.

17 years agoMFC 1.350: Use more specific local variable pointers to narrow some
jhb [Wed, 23 May 2007 18:11:30 +0000 (18:11 +0000)]
MFC 1.350: Use more specific local variable pointers to narrow some
expressions.

17 years agoFix buffer overflow in libmagic(3).
cperciva [Wed, 23 May 2007 16:13:20 +0000 (16:13 +0000)]
Fix buffer overflow in libmagic(3).

Security: FreeBSD-SA-07:04.file
Approved by: so (cperciva)

17 years agoMFC ver. 1.307 and 1.308:
mtm [Wed, 23 May 2007 00:56:58 +0000 (00:56 +0000)]
MFC ver. 1.307 and 1.308:
Comment out knobs that are supposed to be examples.

17 years agoMFC revision 1.6 and part of revision 1.4:
cperciva [Wed, 23 May 2007 00:31:15 +0000 (00:31 +0000)]
MFC revision 1.6 and part of revision 1.4:
 * Add missing '${BASEDIR}/'s, thereby fixing problems when the "-b basedir"
option is used.
 * Don't add new kernel modules into a non-existent kernel directory.
 * Remove files before the directory which contains them.

17 years agoMFC increase bitfield to be wide enough to take the default value that
mjacob [Wed, 23 May 2007 00:25:18 +0000 (00:25 +0000)]
MFC increase bitfield to be wide enough to take the default value that
gets set into it.

17 years agoMFC ver. 1.1:
mtm [Tue, 22 May 2007 22:17:44 +0000 (22:17 +0000)]
MFC ver. 1.1:
----------------------------
date: 2007/04/21 11:23:33;  author: mtm;  state: Exp;
Regression tests for recent changes to inet6_rth_* family of functions
regarding RFC3542 compliance.

17 years agoMFC:
mtm [Tue, 22 May 2007 22:07:49 +0000 (22:07 +0000)]
MFC:
sys/sys/socket.h     ver. 1.93
lib/libc/net/rthdr.c ver. 1.9
----------------------------
date: 2007/04/19 15:48:16;  author: mtm;  state: Exp;  lines: +4 -2
Make inet6_rth_* family of functions more compliant with RFC3542:
1. CMSG_NXTHDR(mhdr, cmsg) is supposed to dereference cmsg and return
   the next header in the chain. If cmsg is NULL it should return
   the first header, behaving essentially like CMSG_FIRSTHDR().
2. inet6_rth_(space|init|add) should do basic checking on their input
   to verify that the number of headers (segments) is
   between 0 and 127 inclusive.