The ability to schedule multiple threads per process
(one one cpu) by making ALL system calls optionally asynchronous.
to come: ia64 and power-pc patches, patches for gdb, test program (in tools)
Reviewed by: Almost everyone who counts
(at various times, peter, jhb, matt, alfred, mini, bernd,
and a cast of thousands)
NOTE: this is still Beta code, and contains lots of debugging stuff.
expect slight instability in signals..
simple reads (and on IA32, a "pause" instruction for each interation of the
loop) to spin until either the mutex owner field changes, or the lock owner
stops executing.
Suggested by: tanimura
Tested on: i386
MI API with empty cpu_pause() functions on other arch's, but this
functionality is definitely unique to IA-32, so I decided to leave it
as i386-only and wrap it in #ifdef's. I should have dropped the cpu_
prefix when I made that decision.
Requested by: bde
Pentium 4's and newer IA32 processors. The "pause" instruction has been
verified by Intel to be a NOP on all currently existing IA32 processors
prior to the Pentium 4.
option is used (not on by default).
- In the case of trying to lock a mutex, if the MTX_CONTESTED flag is set,
then we can safely read the thread pointer from the mtx_lock member while
holding sched_lock. We then examine the thread to see if it is currently
executing on another CPU. If it is, then we keep looping instead of
blocking.
- In the case of trying to unlock a mutex, it is now possible for a mutex
to have MTX_CONTESTED set in mtx_lock but to not have any threads
actually blocked on it, so we need to handle that case. In that case,
we just release the lock as if MTX_CONTESTED was not set and return.
- We do not adaptively spin on Giant as Giant is held for long times and
it slows SMP systems down to a crawl (it was taking several minutes,
like 5-10 or so for my test alpha and sparc64 SMP boxes to boot up when
they adaptively spinned on Giant).
- We only compile in the code to do this for SMP kernels, it doesn't make
sense for UP kernels.
Tested on: i386, alpha, sparc64
the generic lock type for use with witness. If this argument is NULL then
the lock name is used as the lock type. Add a macro for a lock type name
for network driver locks.
even when the number of records approaches the size of the hash table.
Besides, the previous implementation (using linear probing) was broken :)
Also, use the newly introduced MTX_SYSINIT.
various machdep.c's to being declared in kern_mutex.c.
- Add a new function mutex_init() used to perform early initialization
needed for mutexes such as setting up thread0's contested lock list
and initializing MI mutexes. Change the various MD startup routines
to call this function instead of duplicating all the code themselves.
Tested on: alpha, i386
locks to be able to setup a SYSINIT call. This helps in places where
a lock is needed to protect some data, but the data is not truly
associated with a subsystem that can properly initialize it's lock.
The macros use the mtx_sysinit() and sx_sysinit() functions,
respectively, as the handler argument to SYSINIT().
Reviewed by: alfred, jhb, smp@
release times. Measurements are made and stored in nanoseconds but
presented in microseconds, which should be sufficient for the locks for
which we actually want this (those that are held long and / or often).
Also, rename some variables and structure members to unit-agnostic names.
following sysctl variables:
debug.mutex.prof.enable enable / disable profiling
debug.mutex.prof.acquisitions number of mutex acquisitions recorded
debug.mutex.prof.records number of acquisition points recorded
debug.mutex.prof.maxrecords max number of acquisition points
debug.mutex.prof.rejected number of rejections (due to full table)
debug.mutex.prof.hashsize hash size
debug.mutex.prof.collisions number of hash collisions
debug.mutex.prof.stats profiling statistics
The code records four numbers for each acquisition point (identified by
source file name and line number): longest time held, total time held,
number of non-recursive acquisitions, average time held. The measurements
are in clock cycles (as returned by get_cyclecount(9)); this may cause
measurements on some SMP systems to be unreliable. This can probably be
worked around by replacing get_cyclecount(9) by some incarnation of
nanotime(9).
This work was derived from initial patches by eivind.
with this flag. Remove the dup_list and dup_ok code from subr_witness. Now
we just check for the flag instead of doing string compares.
Also, switch the process lock, process group lock, and uma per cpu locks over
to this interface. The original mechanism did not work well for uma because
per cpu lock names are unique to each zone.
Approved by: jhb
seem to be too short for the 500 Mhz DS20 I'm testing on. The rather
arbitrary numbers are rather bogus anyways. We should probably have
variables for these limits that are calibrated in the MD startup code
somehow.
mutex releases to not require flags for the cases when preemption is
not allowed:
The purpose of the MTX_NOSWITCH and SWI_NOSWITCH flags is to prevent
switching to a higher priority thread on mutex releease and swi schedule,
respectively when that switch is not safe. Now that the critical section
API maintains a per-thread nesting count, the kernel can easily check
whether or not it should switch without relying on flags from the
programmer. This fixes a few bugs in that all current callers of
swi_sched() used SWI_NOSWITCH, when in fact, only the ones called from
fast interrupt handlers and the swi_sched of softclock needed this flag.
Note that to ensure that swi_sched()'s in clock and fast interrupt
handlers do not switch, these handlers have to be explicitly wrapped
in critical_enter/exit pairs. Presently, just wrapping the handlers is
sufficient, but in the future with the fully preemptive kernel, the
interrupt must be EOI'd before critical_exit() is called. (critical_exit()
can switch due to a deferred preemption in a fully preemptive kernel.)
I've tested the changes to the interrupt code on i386 and alpha. I have
not tested ia64, but the interrupt code is almost identical to the alpha
code, so I expect it will work fine. PowerPC and ARM do not yet have
interrupt code in the tree so they shouldn't be broken. Sparc64 is
broken, but that's been ok'd by jake and tmm who will be fixing the
interrupt code for sparc64 shortly.
Reviewed by: peter
Tested on: i386, alpha
- The MD functions critical_enter/exit are renamed to start with a cpu_
prefix.
- MI wrapper functions critical_enter/exit maintain a per-thread nesting
count and a per-thread critical section saved state set when entering
a critical section while at nesting level 0 and restored when exiting
to nesting level 0. This moves the saved state out of spin mutexes so
that interlocking spin mutexes works properly.
- Most low-level MD code that used critical_enter/exit now use
cpu_critical_enter/exit. MI code such as device drivers and spin
mutexes use the MI wrappers. Note that since the MI wrappers store
the state in the current thread, they do not have any return values or
arguments.
- mtx_intr_enable() is replaced with a constant CRITICAL_FORK which is
assigned to curthread->td_savecrit during fork_exit().
Tested on: i386, alpha
of Giant during the Giant unwinding phase, and start work on instrumenting
Giant for the file and proc mutexes.
These wrappers allow developers to turn on and off Giant around various
subsystems. DEVELOPERS SHOULD NEVER TURN OFF GIANT AROUND A SUBSYSTEM JUST
BECAUSE THE SYSCTL EXISTS! General developers should only considering
turning on Giant for a subsystem whos default is off (to help track down
bugs). Only developers working on particular subsystems who know what
they are doing should consider turning off Giant.
These wrappers will greatly improve our ability to unwind Giant and test
the kernel on a (mostly) subsystem by subsystem basis. They allow Giant
unwinding developers (GUDs) to emplace appropriate subsystem and structural
mutexes in the main tree and then request that the larger community test
the work by turning off Giant around the subsystem(s), without the larger
community having to mess around with patches. These wrappers also allow
GUDs to boot into a (more likely to be) working system in the midst of
their unwinding work and to test that work under more controlled
circumstances.
There is a master sysctl, kern.giant.all, which defaults to 0 (off). If
turned on it overrides *ALL* other kern.giant sysctls and forces Giant to
be turned on for all wrapped subsystems. If turned off then Giant around
individual subsystems are controlled by various other kern.giant.XXX sysctls.
Code which overlaps multiple subsystems must have all related subsystem Giant
sysctls turned off in order to run without Giant.
off to witness_init() making the check for double intializating a lock by
testing the LO_INITIALIZED flag moot. Workaround this by checking the
LO_INITIALIZED flag ourself before we bzero the lock structure.
all the debugging code into the function versions of the mutex operations
in kern_mutex.c. This reduced the __mtx_* macros to simply wrappers of
the _{get,rel}_lock_* macros, so the __mtx_* macros were also abolished in
favor of just calling the _{get,rel}_lock_* macros. The tangled hairy mass
of macros calling macros is at least a bit more sane now.
Note ALL MODULES MUST BE RECOMPILED
make the kernel aware that there are smaller units of scheduling than the
process. (but only allow one thread per process at this time).
This is functionally equivalent to teh previousl -current except
that there is a thread associated with each process.
Sorry john! (your next MFC will be a doosie!)
Reviewed by: peter@freebsd.org, dillon@freebsd.org
X-MFC after: ha ha ha ha
I'm at it also add a comment in mtx_validate() explaining the purpose
of the last change.
Basically, this fixes booting kernels compiled with MUTEX_DEBUG. What used
to happen is before we setidt from init386() [still using BTX idt], we
called mtx_init() on several mutex locks, notably Giant and some others.
This is a problem for MUTEX_DEBUG because it enables mtx_validate() which
calls kernacc(), some of which in turn requires Giant.
Fix by calling kernacc() from mtx_validate() only if (!cold).
switch. Count the context switch when preempting the current thread to let
a higher priority thread blocked on a mutex we just released run as an
involuntary context switch.
Reported by: bde
struct lock_instance that is stored in the per-process and per-CPU lock
lists. Previously, the lock lists just kept a pointer to each lock held.
That pointer is now replaced by a lock instance which contains a pointer
to the lock object, the file and line of the last acquisition of a lock,
and various flags about a lock including its recursion count.
- If we sleep while holding a sleepable lock, then mark that lock instance
as having slept and ignore any lock order violations that occur while
acquiring Giant when we wake up with slept locks. This is ok because of
Giant's special nature.
- Allow witness to differentiate between shared and exclusive locks and
unlocks of a lock. Witness will now detect the case when a lock is
acquired first in one mode and then in another. Mutexes are always
locked and unlocked exclusively. Witness will also now detect the case
where a process attempts to unlock a shared lock while holding an
exclusive lock and vice versa.
- Fix a bug in the lock list implementation where we used the wrong
constant to detect the case where a lock list entry was full.
other "system" header files.
Also help the deprecation of lockmgr.h by making it a sub-include of
sys/lock.h and removing sys/lockmgr.h form kernel .c files.
Sort sys/*.h includes where possible in affected files.
OK'ed by: bde (with reservations)
- Introduce lock classes and lock objects. Each lock class specifies a
name and set of flags (or properties) shared by all locks of a given
type. Currently there are three lock classes: spin mutexes, sleep
mutexes, and sx locks. A lock object specifies properties of an
additional lock along with a lock name and all of the extra stuff needed
to make witness work with a given lock. This abstract lock stuff is
defined in sys/lock.h. The lockmgr constants, types, and prototypes have
been moved to sys/lockmgr.h. For temporary backwards compatability,
sys/lock.h includes sys/lockmgr.h.
- Replace proc->p_spinlocks with a per-CPU list, PCPU(spinlocks), of spin
locks held. By making this per-cpu, we do not have to jump through
magic hoops to deal with sched_lock changing ownership during context
switches.
- Replace proc->p_heldmtx, formerly a list of held sleep mutexes, with
proc->p_sleeplocks, which is a list of held sleep locks including sleep
mutexes and sx locks.
- Add helper macros for logging lock events via the KTR_LOCK KTR logging
level so that the log messages are consistent.
- Add some new flags that can be passed to mtx_init():
- MTX_NOWITNESS - specifies that this lock should be ignored by witness.
This is used for the mutex that blocks a sx lock for example.
- MTX_QUIET - this is not new, but you can pass this to mtx_init() now
and no events will be logged for this lock, so that one doesn't have
to change all the individual mtx_lock/unlock() operations.
- All lock objects maintain an initialized flag. Use this flag to export
a mtx_initialized() macro that can be safely called from drivers. Also,
we on longer walk the all_mtx list if MUTEX_DEBUG is defined as witness
performs the corresponding checks using the initialized flag.
- The lock order reversal messages have been improved to output slightly
more accurate file and line numbers.
and change the u_int mtx_saveintr member of struct mtx to a critical_t
mtx_savecrit.
- On the alpha we no longer need a custom _get_spin_lock() macro to avoid
an extra PAL call, so remove it.
- Partially fix using mutexes with WITNESS in modules. Change all the
_mtx_{un,}lock_{spin,}_flags() macros to accept explicit file and line
parameters and rename them to use a prefix of two underscores. Inside
of kern_mutex.c, generate wrapper functions for
_mtx_{un,}lock_{spin,}_flags() (only using a prefix of one underscore)
that are called from modules. The macros mtx_{un,}lock_{spin,}_flags()
are mapped to the __mtx_* macros inside of the kernel to inline the
usual case of mutex operations and map to the internal _mtx_* functions
in the module case so that modules will use WITNESS and KTR logging if
the kernel is compiled with support for it.
if we hold a spin mutex, since we can trivially get into deadlocks if we
start switching out of processes that hold spinlocks. Checking to see if
interrupts were disabled was a sort of cheap way of doing this since most
of the time interrupts were only disabled when holding a spin lock. At
least on the i386. To fix this properly, use a per-process counter
p_spinlocks that counts the number of spin locks currently held, and
instead of checking to see if interrupts are disabled in the witness code,
check to see if we hold any spin locks. Since child processes always
start up with the sched lock magically held in fork_exit(), we initialize
p_spinlocks to 1 for child processes. Note that proc0 doesn't go through
fork_exit(), so it starts with no spin locks held.
Consulting from: cp
don't end up back at ourselves which would indicate deadlock.
- Add the proc lock to the witness dup_list as we may hold more than one
process lock at a time.
- Don't assert a mutex is owned in _mtx_unlock_sleep() as that is too late.
We do the checks in the macros instead.
update native priority, it is diffcult to get right and likely
to end up horribly wrong. Use an honestly wrong fixed value
that seems to work; PUSER for user threads, and the interrupt
priority for ithreads. Set it once when the process is created
and forget about it.
Suggested by: bde
Pointy hat: me
