bms [Fri, 25 Aug 2006 13:01:33 +0000 (13:01 +0000)]
MFC: Add identifier for the Epson CX3650 all-in-one scanner function.
This enables the scanner function on these devices to be detected
and probed by uscanner(4), but only when ulpt is not loaded.
bms [Fri, 25 Aug 2006 12:55:50 +0000 (12:55 +0000)]
MFC: Block a variety of signals which may affect reboot(8), before killing
init(8), to avoid losing a race to them and dying before being able
to call reboot(2).
rik [Thu, 24 Aug 2006 11:17:48 +0000 (11:17 +0000)]
MFC: -j 1.10 -j 1.13
----------------------------
revision 1.13
date: 2006/08/23 23:51:29; author: rik; state: Exp; lines: +1 -1
Fix path for source file in files.FOO.
----------------------------
revision 1.12
date: 2006/08/09 21:23:42; author: rik; state: Exp; lines: +122 -22
- Added ability to provide (optional) path to the kernel sources
- Added check of dirs used by example generator
- Fixed path for ${s}/conf/files.FOO and include it via files
directive from config file
- Changed kernel configuration example with the driver: it is
not produced by copying Generic but by including it
- KDB is added to config (for DDB)
- Added module building instead and fixed kernel building
rik [Thu, 24 Aug 2006 11:13:30 +0000 (11:13 +0000)]
MFC: -j 1.21 -j 1.24
----------------------------
revision 1.24
date: 2006/08/09 10:53:26; author: rik; state: Exp; lines: +24 -3
- Added 'files "files.FOO"' to config for inclusion of user specific files.
- Added KDB option to config file (for DDB).
- Added ability to compile custom kernel.
Discussed with: julian@
----------------------------
revision 1.23
date: 2006/08/06 11:06:35; author: rik; state: Exp; lines: +32 -19
- Added ability to provide (optional) path to the kernel sources
- Fixed path for ${s}/conf/files.FOO and a note that it should be merged into
corresponding file to be able to compile the kernel
- Changed kernel configuration example with the driver: it is not produced by
copying Generic but by including it
- Changed from automatic module building to asking an user if it whants to
yar [Thu, 24 Aug 2006 10:32:40 +0000 (10:32 +0000)]
MFC rev 1.65:
Allow for setting negative priority (niceness) when $foo_user is non-root.
The order in _doit must be "nice su", not "su nice", for that.
In addition, don't ignore the exit status from "cd $foo_chdir".
brooks [Tue, 22 Aug 2006 19:41:49 +0000 (19:41 +0000)]
MFC: sync /etc/network.subr and /etc/pccard_ether and their defaults
with HEAD. Key changes include:
- Support a limited set of punctuation in interface names ".-/+" by
converting them to '_' before looking up interface variables.
- Don't flush routes by default when an interface is removed.
- Don't generate unnecessicary noise on startup when interfaces are
renamed.
kib [Tue, 22 Aug 2006 11:04:01 +0000 (11:04 +0000)]
MFC rev. 1.97:
Lock the vnode around the call to VOP_GETATTR. Move the locked code
and vn_fullpath (that call malloc(..., M_WAITOK)) from under the
vm object lock, since sleep is not allowed while holding the mutex.
Being there, wrap VOP_GETATTR call with conditional Giant aquire.
Currently this is (almost) noop because pseudofs is Giant-locked.
dwmalone [Mon, 21 Aug 2006 15:59:48 +0000 (15:59 +0000)]
Sync with -current:
- allow matching on subject: ranges of uid, ranges of gid, jail id
and object: ranges of uid, ranges of gid, filesystem, object is
suid, object is sgid, object matches subject uid/gid, object type.
This involves an ABI change between the kernel module and libugidfw,
but no change between applications and ugidfw.
mjacob [Mon, 21 Aug 2006 00:48:37 +0000 (00:48 +0000)]
MFC 1.120: Fix RELENG_4 code version- isp_roles wasn't getting
initialized so it ended up defaulting to ISP_ROLE_NONE. My testing
hadn't caught it because I was deliberatly setting role via ioctl.
dwmalone [Sun, 20 Aug 2006 19:28:43 +0000 (19:28 +0000)]
MFC: Make net.inet.ip.portrange.reservedhigh and
net.inet.ip.portrange.reservedlow apply to IPv6 aswell as IPv4. Update
a cut'n'paste comment so that it is a bit more up to date.
imp [Sun, 20 Aug 2006 05:23:07 +0000 (05:23 +0000)]
MFC:
Radio Shack USB Cable
Prolific Pharos
Willcom W-SIM
HAMLET exagerate XUR232
No functional changes to the driver, so at worst, I've merged support
for devices that didn't work before and though i think they should
work now, they don't. Please let me know if there are problems.
rwatson [Fri, 18 Aug 2006 14:05:17 +0000 (14:05 +0000)]
Merge if_tun.c:1.158 from HEAD to RELENG_6:
Add kqueue support to if_tun. Loosely based on if_tap changes.
Two almost identical patches based on the if_tap work were submitted
via GNATS; I started out with the patch in 100796 from David Gilbert,
but could have easily started with the patch from Vilmos Nebehaj which
I found only later.
yar [Fri, 18 Aug 2006 10:48:11 +0000 (10:48 +0000)]
MFC:
Optionally pad outgoing frames to the minimum of 60 bytes (excl. FCS)
before tagging them. This can help to work around brain-damage in some
switches that fail to pad a frame after untagging it if its length drops
below the minimum. This option is blessed by IEEE Std 802.1Q (2003 Ed.),
paragraph C.4.4.3.b. It's controlled by sysctl net.link.vlan.soft_pad.
jhb [Thu, 17 Aug 2006 19:53:06 +0000 (19:53 +0000)]
MFC several DDB improvements including the simpler 'ps', 'show proc',
'show thread' (currently doesn't work due to a ddb bug that in 6.x
that causes 'show thread' to be treated as 'show threads'), 'show lockmgr',
'show sleepchain', 'show lockchain', 'show allchains', 'show locktree',
'show sleepq', and 'show turnstile'.
sobomax [Thu, 17 Aug 2006 01:20:36 +0000 (01:20 +0000)]
MFC:
Remove mention of the `W' flag, which has been turned to no-op by the
neworder change. Keep the option in a config file parser, to not
violate POLA.
yongari [Thu, 17 Aug 2006 00:13:07 +0000 (00:13 +0000)]
MFC if_xl.c, 1.205 to RELENG_6:
Don't reset Tx threshold value whenever xl_init_locked() is called.
Instead the threshould is initialized in device attach. Later the
threshold could be increased in Tx underrun error and the new
threshold should be used in xl_init_locked().
yongari [Thu, 17 Aug 2006 00:08:26 +0000 (00:08 +0000)]
MFC rgephy.c, 1.11 to RELENG_6:
Due to the poor PHY documentation from RealTek I can't sure but I
think the RealTek PHY needs driver to set RGEPHY_BMCR_AUTOEN bit of
RGEPHY_MII_BMCR register and proper ANAR register setting for manual
media type selection.
This fixes long standing manual media type selection bug in rgephy(4).
simon [Wed, 16 Aug 2006 11:20:49 +0000 (11:20 +0000)]
MFC 1.17-1.20:
- In addition to -b (suppress builds) add -k (suppress kernel build) and
-w (suppress world).
- Create the diskless magic files in /etc during the install_etc()
step instead of setup_nanobsd().
- When creating the NanoBSD disk image, install boot blocks from the
newly-built world, in preference to those from the host system.
- 'make distribution' must be run from src, not src/etc.
yar [Tue, 15 Aug 2006 19:43:32 +0000 (19:43 +0000)]
MFC rev. 1.102:
Track interface department events and detach vlans from
departing trunk so that we don't get into trouble later
by dereferencing a stale pointer to dead trunk's things.
alc [Tue, 15 Aug 2006 17:51:02 +0000 (17:51 +0000)]
MFC revision 1.360
Ensure that vm_object_deallocate() doesn't dereference a stale object
pointer: When vm_object_deallocate() sleeps because of a non-zero
paging in progress count on either object or object's shadow,
vm_object_deallocate() must ensure that object is still the shadow's
backing object when it reawakens. In fact, object may have been
deallocated while vm_object_deallocate() slept. If so, reacquiring
the lock on object can lead to a deadlock.
yar [Tue, 15 Aug 2006 15:59:25 +0000 (15:59 +0000)]
MFC rev. 1.71, 1.72:
Guess the number of arguments to a function somewhat better.
Now GCC likes to stick a "mov %eax, %FOO" instruction before
"addl $BAR, %esp" if the function just called returns an int,
which is a very common case in the kernel.
Return -1 from db_numargs() if number of args couldn't be guessed.
Use this later to indicate in backtrace output that args shown are
uncertain.