Joel Dahl [Sun, 28 Jan 2007 20:38:07 +0000 (20:38 +0000)]
Clean up the BSD license to match the preferred license in
/usr/share/examples/etc/bsd-style-copyright. I've fixed a
few minor wording and formatting differences.
Approved by: matk, Hannu Savolainen <hannu@opensound.com>
Reviewed by: imp
It is possible that GEOM taste provider before SMP is started.
We can't bind to a CPU which is not yet on-line, so add code that wait for
CPUs to go on-line before binding to them.
Reported by: Alin-Adrian Anton <aanton@spintech.ro>
MFC after: 2 weeks
Sam Leffler [Sun, 28 Jan 2007 18:35:46 +0000 (18:35 +0000)]
ath and ath_rate_sample now have a compile-time dependency on the hal
so we need to build them only on architectures where there's a released
hal; this hack can be eliminated when an ia64 hal build is present
Sam Leffler [Sun, 28 Jan 2007 04:36:05 +0000 (04:36 +0000)]
for newer hal's we need opt_ah.h as it specifies how the hal has been
configured and that in turn controls the descriptor layout; the rate
control module has no business peeking inside the descriptor but until
we can change the api so the driver records the tx rates and passes
them deal with it
Ariff Abdullah [Sun, 28 Jan 2007 03:16:54 +0000 (03:16 +0000)]
Add speaker control for HP xw4300. This hardware doesn't respond to
unsolicited pin sense event and need manual control to turn off speaker
volume while attaching headphone.
Max Laier [Sat, 27 Jan 2007 19:52:08 +0000 (19:52 +0000)]
In case we are supplied with an imagename that matches a module, but not a
firmware in that module (eventhough this is a programming error) - drop the
reference to the module again.
Joseph Koshy [Sat, 27 Jan 2007 18:13:24 +0000 (18:13 +0000)]
Use a known good stack at the time of servicing an NMI --- reuse
the space allocated for the double fault handler since this space
is otherwise unused till the time a double fault occurs.
This change should have been committed alongside r1.127 of
"exception.S", but I somehow missed doing so.
Robert Watson [Sat, 27 Jan 2007 18:12:50 +0000 (18:12 +0000)]
Remove BSD < 199103 compatibility entries in the bpf_d structure: they are
not used in any of our code. Also remove explicit padding variable that
kept the bpf_d structure the same size before and after the change in
select implementation, since binary compatibility is not required for this
data structure on 7-CURRENT.
Bruce A. Mah [Fri, 26 Jan 2007 23:22:58 +0000 (23:22 +0000)]
Revert nd6.c revs. 1.67, 1.68, 1.69, 1.70 in an attempt to unbreak
IPv6 over point-to-point gif(4) tunnels.
These revisions caused a host route to the destination of a
point-to-point gif(4) interface to not get installed when the interface
and destination addresses were assigned. This caused
"no route to host" errors when trying to send traffic over the
interface. The first packet arriving inbound over the tunnel,
however, would cause the correct route to get installed, allowing
subsequent outbound traffic to be routed correctly.
gif(4) interfaces with prefix lengths of less than 128 bits
(i.e. no explicit destination address assigned) were not affected
by this bug.
This bug fix is a possible candidate for a 6.2-RELEASE errata note.
Doug White [Fri, 26 Jan 2007 17:05:24 +0000 (17:05 +0000)]
Collapse 5706C and 5708C PHYs into one entry. ID 0x15 is actually used for
the SERDES PHY on these chips and we want gentbi to pick this up, not brgphy.
Xin LI [Fri, 26 Jan 2007 14:57:17 +0000 (14:57 +0000)]
While we do not expect any change before and after GNU gzip
is replaced with BSD gzip, let's make it possible to
distinguish between the two with a __FreeBSDversion bump,
just in case some developers want it.
Xin LI [Fri, 26 Jan 2007 10:19:08 +0000 (10:19 +0000)]
Replace the GNU gzip with a slightly modified NetBSD gzip. The
NetBSD version is a feature-to-feature re-implementation of GNU
gzip using the freely-redistributable zlib and this version is
expected to be mostly bug-to-bug compatible with the GNU
implementation.
- Because this is a piece of mature code and we want to make
changes so it is added directly rather than importing to
src/contrib.
- Connect newly added code to src/usr.bin/ and rescue/rescue
build.
- Disconnect the GNU gzip code from build for now, they will
be eventually removed completely.
- Provide two new src.conf(5) knobs, WITHOUT_BZIP2_SUPPORT and
WITHOUT_BZIP2.
Tested by: kris (full exp-7 pointyhat build)
Approved by: core (importing a 4-clause BSD licensed file)
Approved by: re (adding new utility during -HEAD code slush)
Marcel Moolenaar [Fri, 26 Jan 2007 06:22:34 +0000 (06:22 +0000)]
o Remove screen, created by the unnecessary ofw_console(4)
o Remove ttyyX, created by the obsolete zs(4)
o Replace ttydX by ttyuX, which is created by uart(4)
o Enable ttyu0 as a console to better support the xserve.
Jeff Roberson [Thu, 25 Jan 2007 23:51:59 +0000 (23:51 +0000)]
- Implement much more intelligent ipi sending. This algorithm tries to
minimize IPIs and rescheduling when scheduling like tasks while keeping
latency low for important threads.
1) An idle thread is running.
2) The current thread is worse than realtime and the new thread is
better than realtime. Realtime to realtime doesn't preempt.
3) The new thread's priority is less than the threshold.
Peter Wemm [Thu, 25 Jan 2007 23:12:19 +0000 (23:12 +0000)]
Bah. Kris says the default-to-a.out knowledge has migrated into the
official gnu configure scripts and a couple of other places.
Add an example noisy, loud and annoying placeholder for /usr/bin/objformat
if it turns out to too much trouble to be gone. It is not connected to
the build yet.
Gleb Smirnoff [Thu, 25 Jan 2007 21:16:50 +0000 (21:16 +0000)]
- Create ng_ppp_bypass() function, that prepares a packet
with bypass header, to send it out to userland.
- Use ng_ppp_bypass() in ng_ppp_proto_recv().
- Use ng_ppp_bypass() in ng_ppp_comp_recv() and in
ng_ppp_crypt_recv() if compression or encryption is
disabled, respectively.
- Any LCP packet goes directly to ng_ppp_bypass(), instead
of passing through PPP stack.
- Any non-LCP packet on disabled link is discarded. This
is behavior defined in RFC.
MFp4: - When new object is created, it's group ID can be set to process'
effective group ID or to group ID of its parent directory.
- Add some comments from POSIX.
- Verify that after successful O_TRUNC open, size is equal to 0.
MFp4: When user is not a member of the group which owns a file, even if
he is the file's owner, he can't set set-gid bit.
POSIX requires to return 0 and clear the bit, but FreeBSD returns
EPERM for UFS in such case. For now do the same in ZFS.
Jeff Roberson [Thu, 25 Jan 2007 19:14:11 +0000 (19:14 +0000)]
- Get rid of the unused DIDRUN flag. This was really only present to
support sched_4bsd.
- Rename the KTR level for non schedgraph parsed events. They take event
space from things we'd like to graph.
- Reset our slice value after we sleep. The slice is simply there to
prevent starvation among equal priorities. A thread which had almost
exhausted it's slice and then slept doesn't need to be rescheduled a
tick after it wakes up.
- Set the maximum slice value to a more conservative 100ms now that it is
more accurately enforced.
Gleb Smirnoff [Thu, 25 Jan 2007 18:03:40 +0000 (18:03 +0000)]
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>
Bill Paul [Thu, 25 Jan 2007 17:30:30 +0000 (17:30 +0000)]
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.
Bruce Evans [Thu, 25 Jan 2007 14:33:13 +0000 (14:33 +0000)]
Rename some functions and variables from nfs_* to nfs4_* to avoid
collisions with nfsclient's names. Even static names should have a
unique prefix so that they can be debugged easily.
Hide the unused colliding variable nfsv3_commit_on_close in "#if 0"
together with other unused sysctl variables. Duplicating the nfs sysctl
under nfs4 is probably just a bug.
Bruce Evans [Thu, 25 Jan 2007 14:18:40 +0000 (14:18 +0000)]
Rename some functions and variables (mainly vfsops entry points) from
nfs_* to nfs4_* to avoid collisions with nfsclient's names. Even
static names should have a unique prefix so that they can be debugged
easily.
Most of the renamed functions can probably be shared. nfs4_cmount()
and nfs4_sync() are identical to the nfs_* versions, and all the others
except nfs4_vfsops() seem to be idendentical except for style bugs,
missing support for mountroot, and bugs.
Mike Pritchard [Thu, 25 Jan 2007 12:42:18 +0000 (12:42 +0000)]
Add a BUGS section that shows that ids that appear to be
negative are now ignored by the quota system and that extremely
large ids may make quotacheck run for a very long time.
Also mention that "options QUOTA" is required for the kernel
to provide quota support.
When the following conditions are meet:
- First configured key is based only on keyfile (no passphrase).
- Device is attached.
- User changes first key (setkey) from keyfile to passphrase and doesn't
specify number of iterations (with -i option).
...geli(8) won't store calculated number of iterations in metadata.
This result in device beeing unaccesable after detach.
One can recover from this situation by guessing number of iterations
generated, storing it in metadata and trying to attach device.
Recovery procedure isn't nice, but one's data is not lost.
Reported by: Thomas Nickl <T.Nickl@gmx.net>
MFC after: 1 week
Mohan Srinivasan [Thu, 25 Jan 2007 01:05:23 +0000 (01:05 +0000)]
Fix for problems that occur when all mbuf clusters migrate to the mbuf packet
zone. Cluster allocations fail when this happens. Also processes that may have
blocked on cluster allocations will never be woken up. Thanks to rwatson for
an overview of the issue and pointers to the mbuma paper and his tool to dump
out UMA zones.
Mohan Srinivasan [Wed, 24 Jan 2007 22:49:11 +0000 (22:49 +0000)]
Fix for a bug where only one process (of multiple) blocked on
maxpages on a zone is woken up, with the rest never being woken up as
a result of the ZFLAG_FULL flag being cleared. Wakeup all such blocked
procsses instead. This change introduces a thundering herd, but since
this should be relatively infrequent, optimizing this (by introducing
a count of blocked processes, for example) may be premature.
Jeff Roberson [Wed, 24 Jan 2007 21:19:56 +0000 (21:19 +0000)]
- Add a horrible bit of code to detect tsc differences between processors.
This only works if there is no significant drift and all processors are
running at the same frequency. Fortunately, schedgraph traces on MP
machines tend to cover less than a second so drift shouldn't be an issue.
- KTRFile::synchstamp() iterates once over the whole list to determine the
lowest tsc value and syncs adjusts all other values to match. We assume
that the first tick recorded on all cpus happened at the same instant to
start with.
- KTRFile::monostamp() iterates again over the whole file and checks for
a cpu agnostic monotonically increasing clock. If the time ever goes
backwards the cpu responsible is adjusted further to fit. This will
make the possible incorrect delta between cpus as small as the shortest
time between two events. This time can be fairly large due to sched_lock
essentially protecting all events.
- KTRFile::checkstamp() now returns an adjusted timestamp.
- StateEvent::draw() detects states that occur out of order in time and
draws them as 0 pixels after printing a warning.
Jeff Roberson [Wed, 24 Jan 2007 18:18:43 +0000 (18:18 +0000)]
- With a sleep time over 2097 seconds hzticks and slptime could end up
negative. Use unsigned integers for sleep and run time so this doesn't
disturb sched_interact_score(). This should fix the invalid interactive
priority panics reported by several users.
Randall Stewart [Wed, 24 Jan 2007 12:59:56 +0000 (12:59 +0000)]
Fixes the MSG_PEEK for sctp_generic_recvmsg() the msg_flags
were not being copied in properly so PEEK and any other
msg_flags input operation were not being performed right.
Approved by: gnn
Jeff Roberson [Tue, 23 Jan 2007 22:19:27 +0000 (22:19 +0000)]
- Print clock information so we know if something is not reported correctly
from the tsc.
- Set skipnext = 1 for yielding and preempted events so we don't show the
event that adds us back to the run queue. It used to be 2 so we would
skip the ksegrp run queue addition and the system run queue addition
but the ksegrp run queue has gone away.
- Don't display down to nanosecond resolution for scheduling events right
now. This can sometimes cause a division by zero.
Bruno Ducrot [Tue, 23 Jan 2007 19:20:30 +0000 (19:20 +0000)]
o introduce a flags 'errata' for HW bugs onto the softc.
o remove errata_a0 and introduce the corresponding flags into 'errata'.
o introduce a new errata for K8, namely some platform might set the
PENDING_BIT but aren't able to unset it, also don't loop forever
waiting PENDING_BIT being cleared.
o try to introduce a workaround for the PENDING_BIT stuck problem,
o support now half multipliers for K8.
Cylinder group bitmaps and blocks containing inode for a snapshot
file are after snaplock, while other ffs device buffers are before
snaplock in global lock order. By itself, this could cause deadlock
when bdwrite() tries to flush dirty buffers on snapshotted ffs. If,
during the flush, COW activity for snapshot needs to allocate block
and ffs_alloccg() selects the cylinder group that is being written
by bdwrite(), then kernel would panic due to recursive buffer lock
acquision.
Avoid dealing with buffers in bdwrite() that are from other side of
snaplock divisor in the lock order then the buffer being written. Add
new BOP, bop_bdwrite(), to do dirty buffer flushing for same vnode in
the bdwrite(). Default implementation, bufbdflush(), refactors the code
from bdwrite(). For ffs device buffers, specialized implementation is
used.
Reviewed by: tegge, jeff, Russell Cattelan (cattelan xfs org, xfs changes)
Tested by: Peter Holm
X-MFC after: 3 weeks (if ever: it changes ABI)
Jeff Roberson [Tue, 23 Jan 2007 08:50:34 +0000 (08:50 +0000)]
- Catch up to setrunqueue/choosethread/etc. api changes.
- Define our own maybe_preempt() as sched_preempt(). We want to be able
to preempt idlethread in all cases.
- Define our idlethread to require preemption to exit.
- Get the cpu estimation tick from sched_tick() so we don't have to worry
about errors from a sampling interval that differs from the time
domain. This was the source of sched_priority prints/panics and
inaccurate pctcpu display in top.
Bruce Evans [Tue, 23 Jan 2007 08:48:26 +0000 (08:48 +0000)]
Oops, pc98 is independent of i386 for clock.c and machdep.c but not
for clock.h, so changing th i386 clock.h broke it. MFi386 (not tested):
Cleaned up declaration and initialization of clock_lock. It is only
used by clock code, so don't export it to the world for machdep.c to
initialize. There is a minor problem initializing it before it is
used, since although clock initialization is split up so that parts
of it can be done early, the first part was never done early enough
to actually work. Split it up a bit more and do the first part as
late as possible to document the necessary order. The functions that
implement the split are still bogusly exported.
Cleaned up initialization of the i8254 clock hardware using the new
split. Actually initialize it early enough, and don't work around it
not being initialized in DELAY() when DELAY() is called early for
initialization of some console drivers.
This unfortunately moves a little more code before the early debugger
breakpoint so that it is harder to debug. The ordering of console and
related initialization is delicate because we want to do as little as
possible before the breakpoint, but must initialize a console.
Jeff Roberson [Tue, 23 Jan 2007 08:46:51 +0000 (08:46 +0000)]
- Remove setrunqueue and replace it with direct calls to sched_add().
setrunqueue() was mostly empty. The few asserts and thread state
setting were moved to the individual schedulers. sched_add() was
chosen to displace it for naming consistency reasons.
- Remove adjustrunqueue, it was 4 lines of code that was ifdef'd to be
different on all three schedulers where it was only called in one place
each.
- Remove the long ifdef'd out remrunqueue code.
- Remove the now redundant ts_state. Inspect the thread state directly.
- Don't set TSF_* flags from kern_switch.c, we were only doing this to
support a feature in one scheduler.
- Change sched_choose() to return a thread rather than a td_sched. Also,
rely on the schedulers to return the idlethread. This simplifies the
logic in choosethread(). Aside from the run queue links kern_switch.c
mostly does not care about the contents of td_sched.
Discussed with: julian
- Move the idle thread loop into the per scheduler area. ULE wants to
do something different from the other schedulers.
Bruce Evans [Tue, 23 Jan 2007 08:01:20 +0000 (08:01 +0000)]
Cleaned up declaration and initialization of clock_lock. It is only
used by clock code, so don't export it to the world for machdep.c to
initialize. There is a minor problem initializing it before it is
used, since although clock initialization is split up so that parts
of it can be done early, the first part was never done early enough
to actually work. Split it up a bit more and do the first part as
late as possible to document the necessary order. The functions that
implement the split are still bogusly exported.
Cleaned up initialization of the i8254 clock hardware using the new
split. Actually initialize it early enough, and don't work around it
not being initialized in DELAY() when DELAY() is called early for
initialization of some console drivers.
This unfortunately moves a little more code before the early debugger
breakpoint so that it is harder to debug. The ordering of console and
related initialization is delicate because we want to do as little as
possible before the breakpoint, but must initialize a console.
Craig Rodrigues [Tue, 23 Jan 2007 07:17:10 +0000 (07:17 +0000)]
Merge mount_nfs4.c and mount_nfs.c into one program.
If argv[0] == "mount_nfs4", then default to mounting NFSv4,
otherwise if argv[0] == "mount_nfs", default to the old mount_nfs behavior.
- Add a -4 option.
- Add the University of Michigan copyright from mount_nfs4.c, for the
code merged from mount_nfs4.c.