Allow rw_tryenter(&lock, RW_READER) to recurse, for vfs_busy().

lockdebug: try to detect recursive acquirements of read-write locks.

Remove clause 3 and 4 from TNF licenses

lockdebug_barrier: disable preemption using the interrupt priority level,
not crit_enter/crit_exit. Since this is called from mi_switch(), crit_exit
could recurse and skew statistics.

Replace use of CACHE_LINE_SIZE in some obvious places.

Make it compile if !LOCKDEBUG.

Try to prevent more than one CPUs panicing when an error is detected.

Need sys/lock.h for SPINLOCK_SPIN_HOOK.

Start detangling lock.h from intr.h. This is likely to cause short term
breakage, but the mess of dependencies has been regularly breaking the
build recently anyhow.

Merge more changes from vmlocking2, mainly:

- Locking improvements.
- Use pool_cache for more items.

Merge changes from vmlocking2 branch.

Use membar_*().

lockdebug_mem_check: turn a redundant check into a panic.

fix lockdebug_mem_check.

lockdebug_alloc: detect double initialization explicitly.

re-enable lockdebug_mem_check. ad@ said disabling it was a mistake
in the first place.
the implementation has been changed to use rb-tree rather than hash
so that it can check regions larger than pa

re-enable lockdebug_mem_check. ad@ said disabling it was a mistake
in the first place.
the implementation has been changed to use rb-tree rather than hash
so that it can check regions larger than page size as well.

show more ...

remove #if 0'ed code.

make kmutex_t and krwlock_t smaller by killing lock id.
ok'ed by Andrew Doran.

Don't static init simplelock (back out those changes).

use __SIMPLELOCK_UNLOCKED, not SIMPLELOCK_INITIALIZER

Change some initialization of static queues to compile time.
(xxx_INIT to xxx_HEAD_INITIALIZER). Drop code which inits
non-auto (global or static) variables to 0 since that's
already implied by bein

Change some initialization of static queues to compile time.
(xxx_INIT to xxx_HEAD_INITIALIZER). Drop code which inits
non-auto (global or static) variables to 0 since that's
already implied by being non-auto. Init some static/global
cpu_simple_locks at compile time.

show more ...

Merge scheduler changes from the vmlocking branch. All discussed on
tech-kern:

- Invert priority space so that zero is the lowest priority. Rearrange
number and type of priority levels into bands.

Merge scheduler changes from the vmlocking branch. All discussed on
tech-kern:

- Invert priority space so that zero is the lowest priority. Rearrange
number and type of priority levels into bands. Add new bands like
'kernel real time'.
- Ignore the priority level passed to tsleep. Compute priority for
sleep dynamically.
- For SCHED_4BSD, make priority adjustment per-LWP, not per-process.

show more ...

Fix spelling.

Merge from vmlocking:

- G/C spinlockmgr() and simple_lock debugging.
- Always include the kernel_lock functions, for LKMs.
- Slightly improved subr_lockdebug code.
- Keep sizeof(struct lock) the sam

Merge from vmlocking:

- G/C spinlockmgr() and simple_lock debugging.
- Always include the kernel_lock functions, for LKMs.
- Slightly improved subr_lockdebug code.
- Keep sizeof(struct lock) the same if LOCKDEBUG.

show more ...

Include sys/cdefs.h first.