MFC if_re.c rev. 1.91, if_rlreg.h rev. 1.66 to RELENG_6.
re(4) devices requires an external EEPROM. Depending on models it
would be 93C46(1Kbit) or 93C56(2Kbit). One of differences between them
is number of address lines required to access the EEPROM. For example,
93C56 EEPROM needs 8 address lines to read/write data. If 93C56
recevied premature end of required number of serial clock(CLK) to set
OP code/address of EEPROM, the result would be unexpected behavior.
Previously it tried to detect 93C46, which requires 6 address lines,
and then assumed it would be 93C56 if read data was not expected
value. However, this approach didn't work in some models/situations
as 93C56 requries 8 address lines to access its data. In order to fix
it, change EEPROM probing order such that 93C56 is detected reliably.
While I'm here change hard-coded address line numbers with defined
constant to enhance readability.
Move the IFT_IEEE8023ADLAG case above the copy to IFP2ENADDR(ifp). This chunk
of code has a slightly different layout to HEAD had was mismerged. This fixes
lagg(4) on RELENG_6.
Add a sysctl for the purge run interval so that it can
be tuned along with the rest of hostcache parameters.
The new sysctl name is `net.inet.tcp.hostcache.prune'.
Replace a constant with an already defined symbolic name for it.
Sync lagg to HEAD
- Implement the Marker Protocol for lacp
- Export the actor and partner peer info and show in ifconfig verbose
- Compare the partner system priority when choosing the aggregator.
- Cleanup whitespace, variable names and gc unused defines
Note: To avoid POLA violation, the merged module still lets root
in irrespective of login.conf settings. In HEAD, root has to have
an explicit "ignorenologin" capability to bypass nologin(5).
- Finally document the LOGIN_MECLASS hack in the manpage and comments.
- Fix login_getcaplist() prototype in the manpage.
- Improve the manpage markup.
- Add controller id for Intel 82801I (ICH9).
PR: kern/114399
Submitted by: Michael Fuckner <michael@fuckner.net>
- MSI support. Disable by default due to various issues with too many
broken hardwares. MSI can be enabled through device.hints(5) or
kenv(8) by setting "hint.pcm.%d.msi=1" (for RELENG_6, you need to
set/enable hw.pci.enable_msi=1).
Partially submitted by: kevlo
YAMAMOTO Taku <taku@tackymt.homeip.net>
Tested by: joel, kevlo, YAMAMOTO Taku
Add a new option for ppp.conf: rad_port_id. It allows to
change the way of what ppp submits to the RADIUS server
as NAS-Port-Id. Possible options are: the PID of the process
owning the corresponding interface, tun(4) interface number,
interface index (as it would get returned by if_nametoindex(3)),
or it's possible to keep the default behavior. Check the ppp(8)
manual page for details.
MFC ppp.8.m4 1.327:
Cleanup after previous commit.
MFC of the work committed on 2007-06-18 to improve the default
named.conf. This brings us in line with BCP, and makes our
installed configuration usable as a local resolver right out
of the box.
This includes:
1.359 of etc/Makefile
1.5 of etc/namedb/Makefile
1.1 of etc/namedb/master/Makefile
all for the new files in etc/namedb/master,
1.23 and 1.24 of etc/namedb/named.conf to bring in the new
configuration options,
The new files in etc/namedb/master to support those options,
And deletion of the old *localhost*.rev files.
See the commit message to HEAD and the comments in the new
named.conf file for more details about the changes and the
reasons for them.
New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page
mapping, due to recent changes in linux compatibility layer which
require it. All linux applications that using sound + mmap() (mostly games)
require this to be enabled. Disabled by default.
MFC (4 of X):
- don't limit number of syscalls to 255
- handle more socket options
- bug-/compatibility-fixes to linux
* file related (includes fixes which prevent creation of strange files
which can only be removed with a fsck)
* make ping work
* ...
- add devfs to the file system type handling/translation
Compile tested by: scf (i386, as part of a mega-MFC-patch)
Tested by: Arno J. Klaassen <arno@heho.snv.jussieu.fr> (amd64)
MFC (3 of X):
- In preparation of further linuxulator fixes MFC kern_descrip.c rev 1.296 and
syscallsubr.h rev 1.41 by jhb:
Add a kern_close() so that the ABIs can close a file descriptor w/o having
to populate a close_args struct and change some of the places that do.
Tested by: scf (i386, as part of a mega-MFC-patch),
Arno J. Klaassen <arno@heho.snv.jussieu.fr> (amd64)
MFC (2 of X):
- some linuxulator style(9) fixes
- add datatypes and definitions in preparation of further commits
- use one (l_sigval) of the new datatypes on amd64 and get rid of a now unused
include (this may or may not fix some signal handling issues on amd64) [1]
- convert some unconditional debug handling into the common linuxulator way
- use some more appropriate format string types in some debug printf's
- convert signal validation into a macro
Except for [1] on amd64 there is no visible change to a normal user.
Tested by: scf (i386), Arno J. Klaassen <arno@heho.snv.jussieu.fr> (amd64)
Add Viking Interworks 256MB as an ata device; this might give
some false positives but at this moment it is better to add
support then to dont have it at all (comment from Soren).
PR: kern/111516
Submitted by: Thomas Nystrom <thn at saeab dot se>
Approved by: re (kensmith)
Approved by: imp (mentor)
OK'ed by: sos (With the comment noted above about false
positives).
brian [Sat, 7 Jul 2007 00:54:46 +0000 (00:54 +0000)]
Fix a problem introduced in netinet/in.c 1.85.2.7 where
in_ifdetach() calls in_delmulti_ifp().
The code now *really* deletes the elements in in_multihead
for the ifp that's going away (rather than just decrementing
the reference count). Previously we were left with inm and
ifma structures containing bogus ifnet pointers after
destroying an interface that had more than one IP4 assignment
made to it in it's lifetime.
I've also added a if_delmulti_ent() to make deleting known
ifma structures possible rather than depending on
if_findmulti() to end up finding the same thing. It
will in fact always find the correct ifma *unless* the
passed sockaddr has a bogus sa_len of zero.
Finally, when adding a multicast address, we no longer
increment the refcount (well, we do, but then we decrement
it again). The refcount here is in fact bogus so hopefully
readers will see that now.
This code is going directly into -stable as it has been
rewritten in -current and those changes are deemed too
intrusive for -stable consumption right now.
MFC some comment changes and a reworking of one cast.
With this change, libarchive source in -CURRENT and 6-STABLE
is exactly identical. The only difference is the
version number (1.9 for 6-STABLE, 2.2 for -CURRENT) and
SHLIB_MAJOR (2 in 6-STABLE, 4 in -CURRENT). The
source has conditionals to compile the old libarchive 1
API/ABI or the newer libarchive 2 API/ABI depending on
the version number set at compile time.
Forced commit to note that the big MFC included
the recent fix to disable attempts to use lseek()
on non-regular files. In particular, this fixes
"tar -t" on tape drives.
This is identical to libarchive 2.2.3 from -CURRENT, with
all of the performance improvements, bug fixes, and new
formats, except that:
* archive_read_finish() and archive_write_finish() don't return errors
* archive_write_data() has a different return type
* user-provided skip callbacks have a slightly different signature
These differences maintain ABI compatibility with libarchive 1.3.1
(hence the "1" in "1.9") and allows libarchive 1.9 to have the
same SHLIB_MAJOR and otherwise serve as a straight replacement
for libarchive 1.3.1 (only better).
Thanks to: Many, many people who contributed error reports, bug fixes,
and suggestions over the last three years. ;-)
MFC rev XXXX:
Do not attempt to enable AHCI mode on ALi SATA controllers other
than the 5288.
It is not correctly implemented in earlier silicon, and the BIOS often
lies about AHCI capability on platforms where these chips are deployed.
With this change I am able to boot FreeBSD on the ASUS Vintage AH-1
barebones system.
- Fix input/microphone support for ASUS A8N-VMCSM series.
Submitted by: Simon Schubert <corecode@fs.ei.tum.de>
- Defer flushing unsolicited response into taskqueue thread rather
than handle it directly in interrupt handler, since few of its
operations (like measuring/calibrating jack impedance) are quite
expensive.
- Misc. debugging cleanups.
Note: The offending quirk should have been made model/codec specific,
but since there were no records / log which model requires it, the quirk
logic had to be inverted (blacklist instead of whitelist).
Fix a bug in sblock() that has existed since revision 1.1 from BSD:
correctly return an error if M_NOWAIT is passed to sblock() and the
operation might block. This remarkably subtle macro bug appears to
be responsible for quite a few undiagnosed socket buffer corruption
and mbuf-related kernel panics.
This bug has already been fixed in 7-CURRENT as part of the move to
using sx(9) locks to serialize simultaneous socket I/O, but is an
MFC candidate for all earlier FreeBSD -STABLE branches.
MFC after: 2 weeks
Found by: Isilon
Submitted by: jeff
Tested by: jhb, Yahoo!
jhb [Fri, 29 Jun 2007 21:04:22 +0000 (21:04 +0000)]
Properly initialize the address value in siginfo for Linux binaries on
amd64. amd64 has a separate tf_addr in the trapframe unlike i386 which
reuses tf_err to pass the address to signals.
ariff [Thu, 21 Jun 2007 20:58:44 +0000 (20:58 +0000)]
MFC: (revision 1.42)
- Fix headphone jack sensing support for Olivetti Olibook 610-430 XPSE [1].
- Drain all callout handlers during driver detach appropriately.
- M_NOWAIT -> M_WAITOK
kmacy [Thu, 21 Jun 2007 03:30:18 +0000 (03:30 +0000)]
MFC:
- switch adapter and port lock over to using sx so that resources
can be allocated atomically
- add debug macros for printing lock initialization / teardown
- add buffers to port_info and adapter to allow each lock to have a
unique name
- destroy mutexes initialized by cxgb_offload_init
- remove recursive calls to ADAPTER_LOCK
- move callout_drain calls so that they don't occur with the lock held
- ensure that only as many qsets as are needed are initialized and
destroyed
philip [Wed, 20 Jun 2007 08:29:18 +0000 (08:29 +0000)]
MFC: moused.c rev 1.79
Fix a (very) longstanding bug in moused(8) affecting high-resolution
rodents when linear acceleration (-a) was enabled with a <1 value to
slow them down.
Previously, rounding errors would eat small movements so the mouse
had to be moved a certain distance to get any movement at all. We
now calculate the rounding errors and take them into account when
reporting movement.
Submitted by: Oliver Fromme <olli -at- lurza.secnetix.de>
jhb [Mon, 18 Jun 2007 22:44:59 +0000 (22:44 +0000)]
MFC: Support for running 32-bit multithreaded binaries using libthr on
amd64 including:
- Add 32-bit wrappers for thr_new(), thr_suspend(), and the umtx system
calls.
- Add support to amd64 for constructing thread upcalls for 32-bit
processes.
- Leave %fs and %gs alone in the signal trampoline for 32-bit processes on
amd64.
- Add 'casuword32()' to amd64 and ia64.
gshapiro [Mon, 18 Jun 2007 04:35:36 +0000 (04:35 +0000)]
MFC: You can never have enough documentation.
Document the SENDMAIL_ALIASES, SENDMAIL_MAP_SRC, SENDMAIL_MAP_TYPE, and
SENDMAIL_START_SCRIPT
.Xr make.conf 5
variables.
These are used in
.Pa /etc/mail/Makefile .
kmacy [Sun, 17 Jun 2007 23:52:17 +0000 (23:52 +0000)]
MFC:
- switch over to standard method for initializing cdevs (contributed by scottl@)
- break out timer_reclaim_task to be per-port
- move msix teardown into separate function
- fix bus_setup_intr for msi-x for the multi-port case so that msi-x resources
are not corrupted on unload
- handle 10/100/1000 base-T media and auto negotiation
- bind qset to cpu even for singleq case
- white space cleanups
- remove recursive PORT_LOCK
- move mtu setting to separate function
- stop and re-init port when changing mtu
- replace all direct references to m_data with calls to mtod
- handle attach failure better by not trying to de-initialize
taskqueues when they have not been allocated
- no longer default to jumbo frames
- fix gcc-4.2 signedness compile errors
wilko [Sat, 16 Jun 2007 20:49:49 +0000 (20:49 +0000)]
Forced commit to clarify preceding unclear commit message:
- core was contacted by Joerg Peter Schley requesting 2 comment lines
be removed with his name & email address in it.
gshapiro [Fri, 15 Jun 2007 02:39:36 +0000 (02:39 +0000)]
MFC: Add a new rc.conf variable, sendmail_rebuild_aliases, which tells
/etc/rc.d/sendmail whether or not to run newaliases if the database
is missing or the aliases text file is newer than aliases.db.
In my opinion, the aliases file should never be automatically rebuilt.
The current text form could represent a work in progress. Therefore,
in FreeBSD 7.0, this new option will default to "NO". When this rc.d
change is MFC'ed, it will need to remain "YES" to maintain backward
compatibility.
PR: conf/86252
Approved by: re (kensmith)
MFC after: 3 days
davidch [Thu, 14 Jun 2007 21:28:14 +0000 (21:28 +0000)]
- Updated firmware to latest release (v3.4.8) to fix TSO + jumbo frame lockup
- Added MSI (hw.bce.msi_enable) sysctl
- Fixed kernel panic when MSI is used and module is unloaded
- Added several new debug routines
- Removed slack space for RX/TX chains since it only covers sloppy coding
- Fixed a potential problem when programming jumbo MTU size in hardware
- Various other comment changes
- Added 2.5G support for BCM5708S
davidch [Thu, 14 Jun 2007 20:02:41 +0000 (20:02 +0000)]
- Added a new Ethernet media type (2500BaseSX) to support BCM5708 controllers
which support a 2.5Gbps mode over fiber using next page extensions during
autonegotiation. Typically only found in blade systems which also include
a Broadcom 2.5Gbps capable switch.
bde [Thu, 14 Jun 2007 12:38:12 +0000 (12:38 +0000)]
MFC: all except 1.63:
1.59-1.62: already merged
1.63: removal of bridge(45) not merged
1.64: already merged
1.65: use uintmax_t
1.66: repair formatting to 79 columns
1.67: increase accuracy of timeouts and thus of per-interval counts
1.68: style fixes
Requested by: kuriyama (1.66)
1.66 changes the whitespace in output. Hopefully nothing depends on the
exact misformatting that it fixes.