In the lockmgr support:
- GIANT_RESTORE() is just called when the sleep finishes, so the current
code can ends up into a giant unlock problem. Fix it by appropriately
call GIANT_RESTORE() when needed. Note that this is not exactly ideal
because for any interation of the adaptive spinning we drop and restore
Giant, but the overhead should be not a factor.
- In the lock held in exclusive mode case, after the adaptive spinning is
brought to completition, we should just retry to acquire the lock
instead to fallthrough. Fix that.
- Fix a style nit
In the sx support:
- Call GIANT_SAVE() before than looping. This saves some overhead because
in the current code GIANT_SAVE() is called several times.
Tested by: Giovanni Trematerra <giovanni dot trematerra at gmail dot com>
a pointer-fetching specific operation check. Consequently, rename the
operation ASSERT_ATOMIC_LOAD_PTR().
* Fix the implementation of ASSERT_ATOMIC_LOAD_PTR() by checking
directly alignment on the word boundry, for all the given specific
architectures. That's a bit too strict for some common case, but it
assures safety.
* Add a comment explaining the scope of the macro
* Add a new stub in the lockmgr specific implementation
Tested by: marcel (initial version), marius
Reviewed by: rwatson, jhb (comment specific review)
Approved by: re (kib)
Actually, as it did receive few tuning, the support is disabled by
default, but it can opt-in with the option ADAPTIVE_LOCKMGRS.
Due to the nature of lockmgrs, adaptive spinning needs to be
selectively enabled for any interested lockmgr.
The support is bi-directional, or, in other ways, it will work in both
cases if the lock is held in read or write way. In particular, the
read path is passible of further tunning using the sysctls
debug.lockmgr.retries and debug.lockmgr.loops . Ideally, such sysctls
should be axed or compiled out before release.
Addictionally note that adaptive spinning doesn't cope well with
LK_SLEEPFAIL. The reason is that many (and probabilly all) consumers
of LK_SLEEPFAIL are mainly interested in knowing if the interlock was
dropped or not in order to reacquire it and re-test initial conditions.
This directly interacts with adaptive spinning because lockmgr needs
to drop the interlock while spinning in order to avoid a deadlock
(further details in the comments inside the patch).
Final note: finding someone willing to help on tuning this with
relevant workloads would be either very important and appreciated.
Tested by: jeff, pho
Requested by: many
adds probes for mutexes, reader/writer and shared/exclusive locks to
gather contention statistics and other locking information for
dtrace scripts, the lockstat(1M) command and other potential
consumers.
Reviewed by: attilio jhb jb
Approved by: gnn (mentor)
- Align the fifo output in fifo_print() with other vn_printf() output.
- Remove the leading space from lockmgr_printinfo() so its output lines up
in vn_printf().
- lockmgr_printinfo() now ends with a newline, so remove an extra newline
from vn_printf().
of spurious witness warnings since lockmgr grew witness support. Before
this, every time you passed an interlock to a lockmgr lock WITNESS treated
it as a LOR.
Reviewed by: attilio
routine wakes up proc0 so that proc0 can swap the thread back in.
Historically, this has been done by waking up proc0 directly from
setrunnable() itself via a wakeup(). When waking up a sleeping thread
that was swapped out (the usual case when waking proc0 since only sleeping
threads are eligible to be swapped out), this resulted in a bit of
recursion (e.g. wakeup() -> setrunnable() -> wakeup()).
With sleep queues having separate locks in 6.x and later, this caused a
spin lock LOR (sleepq lock -> sched_lock/thread lock -> sleepq lock).
An attempt was made to fix this in 7.0 by making the proc0 wakeup use
the ithread mechanism for doing the wakeup. However, this required
grabbing proc0's thread lock to perform the wakeup. If proc0 was asleep
elsewhere in the kernel (e.g. waiting for disk I/O), then this degenerated
into the same LOR since the thread lock would be some other sleepq lock.
Fix this by deferring the wakeup of the swapper until after the sleepq
lock held by the upper layer has been locked. The setrunnable() routine
now returns a boolean value to indicate whether or not proc0 needs to be
woken up. The end result is that consumers of the sleepq API such as
*sleep/wakeup, condition variables, sx locks, and lockmgr, have to wakeup
proc0 if they get a non-zero return value from sleepq_abort(),
sleepq_broadcast(), or sleepq_signal().
Discussed with: jeff
Glanced at by: sam
Tested by: Jurgen Weber jurgen - ish com au
MFC after: 2 weeks
the only one difference is that lockmgr*() functions now accept
LK_NOWITNESS flag which skips ordering for the instanced calling.
- Remove an unuseful stub in witness_checkorder() (because the above check
doesn't allow ever happening) and allow witness_upgrade() to accept
non-try operation too.
bit in order to allow per-bit checks on the options flag, in particular
in the consumers code [1]
- Re-enable the check against TDP_DEADLKTREAT as the anti-waiters
starvation patch allows exclusive waiters to override new shared
requests.
[1] Requested by: pjd, jeff
state transitioning flags and of msleep(9) callings.
Use, instead, an algorithm very similar to what sx(9) and rwlock(9)
alredy do and direct accesses to the sleepqueue(9) primitive.
In order to avoid writer starvation a mechanism very similar to what
rwlock(9) uses now is implemented, with the correspective per-thread
shared lockmgrs counter.
This patch also adds 2 new functions to lockmgr KPI: lockmgr_rw() and
lockmgr_args_rw(). These two are like the 2 "normal" versions, but they
both accept a rwlock as interlock. In order to realize this, the general
lockmgr manager function "__lockmgr_args()" has been implemented through
the generic lock layer. It supports all the blocking primitives, but
currently only these 2 mappers live.
The patch drops the support for WITNESS atm, but it will be probabilly
added soon. Also, there is a little race in the draining code which is
also present in the current CVS stock implementation: if some sharers,
once they wakeup, are in the runqueue they can contend the lock with
the exclusive drainer. This is hard to be fixed but the now committed
code mitigate this issue a lot better than the (past) CVS version.
In addition assertive KA_HELD and KA_UNHELD have been made mute
assertions because they are dangerous and they will be nomore supported
soon.
In order to avoid namespace pollution, stack.h is splitted into two
parts: one which includes only the "struct stack" definition (_stack.h)
and one defining the KPI. In this way, newly added _lockmgr.h can
just include _stack.h.
Kernel ABI results heavilly changed by this commit (the now committed
version of "struct lock" is a lot smaller than the previous one) and
KPI results broken by lockmgr_rw() / lockmgr_args_rw() introduction,
so manpages and __FreeBSD_version will be updated accordingly.
Tested by: kris, pho, jeff, danger
Reviewed by: jeff
Sponsored by: Google, Summer of Code program 2007
than rely on the lockmgr support [1]:
* bump the waiters only if the interlock is held
* let brelvp() return the waiters count
* rely on brelvp() instead than BUF_LOCKWAITERS() in order to check
for the waiters number
- Remove a namespace pollution introduced recently with lockmgr.h
including lock.h by including lock.h directly in the consumers and
making it mandatory for using lockmgr.
- Modify flags accepted by lockinit():
* introduce LK_NOPROFILE which disables lock profiling for the
specified lockmgr
* introduce LK_QUIET which disables ktr tracing for the specified
lockmgr [2]
* disallow LK_SLEEPFAIL and LK_NOWAIT to be passed there so that it
can only be used on a per-instance basis
- Remove BUF_LOCKWAITERS() and lockwaiters() as they are no longer
used
This patch breaks KPI so __FreBSD_version will be bumped and manpages
updated by further commits. Additively, 'struct buf' changes results in
a disturbed ABI also.
[2] Really, currently there is no ktr tracing in the lockmgr, but it
will be added soon.
[1] Submitted by: kib
Tested by: pho, Andrea Barberio <insomniac at slackware dot it>
always curthread.
As KPI gets broken by this patch, manpages and __FreeBSD_version will be
updated by further commits.
Tested by: Andrea Barberio <insomniac at slackware dot it>
the same operation of lockmgr() but accepting a custom wmesg, prio and
timo for the particular lock instance, overriding default values
lkp->lk_wmesg, lkp->lk_prio and lkp->lk_timo.
- Use lockmgr_args() in order to implement BUF_TIMELOCK()
- Cleanup BUF_LOCK()
- Remove LK_INTERNAL as it is nomore used in the lockmgr namespace
Tested by: Andrea Barberio <insomniac at slackware dot it>
A couple of notes for this:
* WITNESS support, when enabled, is only used for shared locks in order
to avoid problems with the "disowned" locks
* KA_HELD and KA_UNHELD only exists in the lockmgr namespace in order
to assert for a generic thread (not curthread) owning or not the
lock. Really, this kind of check is bogus but it seems very
widespread in the consumers code. So, for the moment, we cater this
untrusted behaviour, until the consumers are not fixed and the
options could be removed (hopefully during 8.0-CURRENT lifecycle)
* Implementing KA_HELD and KA_UNHELD (not surported natively by
WITNESS) made necessary the introduction of LA_MASKASSERT which
specifies the range for default lock assertion flags
* About other aspects, lockmgr_assert() follows exactly what other
locking primitives offer about this operation.
- Build real assertions for buffer cache locks on the top of
lockmgr_assert(). They can be used with the BUF_ASSERT_*(bp)
paradigm.
- Add checks at lock destruction time and use a cookie for verifying
lock integrity at any operation.
- Redefine BUF_LOCKFREE() in order to not use a direct assert but
let it rely on the aforementioned destruction time check.
KPI results evidently broken, so __FreeBSD_version bumping and
manpage update result necessary and will be committed soon.
Side note: lockmgr_assert() will be used soon in order to implement
real assertions in the vnode namespace replacing the legacy and still
bogus "VOP_ISLOCKED()" way.
Tested by: kris (earlier version)
Reviewed by: jhb
VOP_ISLOCKED(arg, curthread). Now, VOP_ISLOCKED() and lockstatus() should
only acquire curthread as argument; this will lead in axing the additional
argument from both functions, making the code cleaner.
Reviewed by: jeff, kib
This support tries to be as parallel as possible with other locking
primitives, but there are differences; more specifically:
- The base witness support is alredy equipped for allowing lock
duplication acquisition as lockmgr rely on this.
- In the case of lockmgr_disown() the lock result unlocked by witness
even if it is still held by the "kernel context"
- In the case of upgrading we can have 3 different situations:
* Total unlocking of the shared lock and nothing else
* Real witness upgrade if the owner is the first upgrader
* Shared unlocking and exclusive locking if the owner is not the first
upgrade but it is still allowed to upgrade
- LK_DRAIN is basically handled like an exclusive acquisition
Additively new options LK_NODUP and LK_NOWITNESS can now be used with
lockinit(): LK_NOWITNESS disables WITNESS for the specified lock while
LK_NODUP enable duplicated locks tracking. This will require manpages
update and a __FreeBSD_version bumping (addressed by further commits).
This patch also fixes a problem occurring if a lockmgr is held in
exclusive mode and the same owner try to acquire it in shared mode:
currently there is a spourious shared locking acquisition while what
we really want is a lock downgrade. Probabilly, this situation can be
better served with a EDEADLK failing errno return.
Side note: first testing on this patch alredy reveleated several LORs
reported, so please expect LORs cascades until resolved. NTFS also is
reported broken by WITNESS introduction. BTW, NTFS is exposing a lock
leak which needs to be fixed, and this patch can help it out if
rightly tweaked.
Tested by: kris, yar, Scot Hetzel <swhetzel at gmail dot com>
- Remove the "thread" argument from the lockmgr() function as it is
always curthread now
- Axe lockcount() function as it is no longer used
- Axe LOCKMGR_ASSERT() as it is bogus really and no currently used.
Hopefully this will be soonly replaced by something suitable for it.
- Remove the prototype for dumplockinfo() as the function is no longer
present
Addictionally:
- Introduce a KASSERT() in lockstatus() in order to let it accept only
curthread or NULL as they should only be passed
- Do a little bit of style(9) cleanup on lockmgr.h
KPI results heavilly broken by this change, so manpages and
FreeBSD_version will be modified accordingly by further commits.
Tested by: matteo
panic but it won't actually lock anything.
This can lead some paths to reach lockmgr_disown() with inconsistent
lock which will let trigger the relative assertions.
Fix those in order to recognize panic situation and to not trigger.
Reported by: pho
Submitted by: kib
Now, lockmgr() function can only be called passing curthread and the
KASSERT() is upgraded according with this.
In order to support on-the-fly owner switching, the new function
lockmgr_disown() has been introduced and gets used in BUF_KERNPROC().
KPI, so, results changed and FreeBSD version will be bumped soon.
Differently from previous code, we assume idle thread cannot try to
acquire the lockmgr as it cannot sleep, so loose the relative check[1]
in BUF_KERNPROC().
Tested by: kris
[1] kib asked for a KASSERT in the lockmgr_disown() about this
condition, but after thinking at it, as this is a well known general
rule, I found it not really necessary.
This option just adds complexity and the new implementation no longer
will support it, so axing it now that it is unused is probabilly the
better idea.
FreeBSD version is bumped in order to reflect the KPI breakage introduced
by this patch.
In the ports tree, kris found that only old OSKit code uses it, but as
it is thought to work only on 2.x kernels serie, version bumping will
solve any problem.
with the interlock), owner of the lock should be only curthread or at
least, for its limited usage, NULL which identifies LK_KERNPROC.
The thread "extra argument" for the lockmgr interface is going to be
removed in the near future, but for the moment, just let kernel run for
some days with this check on in order to find potential deadlocking
places around the kernel and fix them.
linker interfaces for looking up function names and offsets from
instruction pointers. Create two variants of each call: one that is
"DDB-safe" and avoids locking in the linker, and one that is safe for
use in live kernels, by virtue of observing locking, and in particular
safe when kernel modules are being loaded and unloaded simultaneous to
their use. This will allow them to be used outside of debugging
contexts.
Modify two of three current stack(9) consumers to use the DDB-safe
interfaces, as they run in low-level debugging contexts, such as inside
lockmgr(9) and the kernel memory allocator.
Update man page.
should never be moved by one lock to another.
As, luckily, nothing in our tree is using it, axe the function.
This breaks lockmgr KPI, so interested, third-party modules should update
their source code with appropriate replacement.
Ok'ed by: ups, rwatson
MFC after: 3 days
an unified way for all the lock primitives to express lock assertions.
Currenty, lockmgrs and rmlocks don't have assertions, so just panic in
that case.
This will be a base for more callout improvements.
Ok'ed by: jhb, jeff
opposed to what process. Since threads by default have teh name of the
process unless over-written with more useful information, just print the
thread name instead.
We can now use LOCK_CLASS() as a stronger check in lockmgr_chain() as a
result. This required putting back lk_flags as lockmgr's use of flags
conflicted with other flags in lo_flags otherwise.
- Tweak 'show lock' output for lockmgr to match sx, rw, and mtx.
These functions are intended to be used to drop a lock and then reacquire
it when doing an sleep such as msleep(9). Both functions accept a
'struct lock_object *' as their first parameter. The 'lc_unlock' function
returns an integer that is then passed as the second paramter to the
subsequent 'lc_lock' function. This can be used to communicate state.
For example, sx locks and rwlocks use this to indicate if the lock was
share/read locked vs exclusive/write locked.
Currently, spin mutexes and lockmgr locks do not provide working lc_lock
and lc_unlock functions.
- only collect timestamps when a lock is contested - this reduces the overhead
of collecting profiles from 20x to 5x
- remove unused function from subr_lock.c
- generalize cnt_hold and cnt_lock statistics to be kept for all locks
- NOTE: rwlock profiling generates invalid statistics (and most likely always has)
someone familiar with that should review
wait (time waited to acquire) and hold times for *all* kernel locks. If
the architecture has a system synchronized TSC, the profiling code will
use that - thereby minimizing profiling overhead. Large chunks of profiling
code have been moved out of line, the overhead measured on the T1 for when
it is compiled in but not enabled is < 1%.
Approved by: scottl (standing in for mentor rwatson)
Reviewed by: des and jhb
