Analysis of current linux futex code :
--------------------------------------
A central hash table futex_queues[] holds all contexts (futex_q) of waiting
threads.
Each futex_wait()/futex_wait() has to obtain a spinlock on a hash slot to
perform lookups or insert/deletion of a futex_q.
When a futex_wait() is done, calling thread has to :
1) - Obtain a read lock on mmap_sem to be able to validate the user pointer
(calling find_vma()). This validation tells us if the futex uses
an inode based store (mapped file), or mm based store (anonymous mem)
2) - compute a hash key
3) - Atomic increment of reference counter on an inode or a mm_struct
4) - lock part of futex_queues[] hash table
5) - perform the test on value of futex.
(rollback is value != expected_value, returns EWOULDBLOCK)
(various loops if test triggers mm faults)
6) queue the context into hash table, release the lock got in 4)
7) - release the read_lock on mmap_sem
<block>
8) Eventually unqueue the context (but rarely, as this part may be done
by the futex_wake())
Futexes were designed to improve scalability but current implementation has
various problems :
- Central hashtable :
This means scalability problems if many processes/threads want to use
futexes at the same time.
This means NUMA unbalance because this hashtable is located on one node.
- Using mmap_sem on every futex() syscall :
Even if mmap_sem is a rw_semaphore, up_read()/down_read() are doing atomic
ops on mmap_sem, dirtying cache line :
- lot of cache line ping pongs on SMP configurations.
mmap_sem is also extensively used by mm code (page faults, mmap()/munmap())
Highly threaded processes might suffer from mmap_sem contention.
mmap_sem is also used by oprofile code. Enabling oprofile hurts threaded
programs because of contention on the mmap_sem cache line.
- Using an atomic_inc()/atomic_dec() on inode ref counter or mm ref counter:
It's also a cache line ping pong on SMP. It also increases mmap_sem hold time
because of cache misses.
Most of these scalability problems come from the fact that futexes are in
one global namespace. As we use a central hash table, we must make sure
they are all using the same reference (given by the mm subsystem). We
chose to force all futexes be 'shared'. This has a cost.
But fact is POSIX defined PRIVATE and SHARED, allowing clear separation,
and optimal performance if carefuly implemented. Time has come for linux
to have better threading performance.
The goal is to permit new futex commands to avoid :
- Taking the mmap_sem semaphore, conflicting with other subsystems.
- Modifying a ref_count on mm or an inode, still conflicting with mm or fs.
This is possible because, for one process using PTHREAD_PROCESS_PRIVATE
futexes, we only need to distinguish futexes by their virtual address, no
matter the underlying mm storage is.
If glibc wants to exploit this new infrastructure, it should use new
_PRIVATE futex subcommands for PTHREAD_PROCESS_PRIVATE futexes. And be
prepared to fallback on old subcommands for old kernels. Using one global
variable with the FUTEX_PRIVATE_FLAG or 0 value should be OK.
PTHREAD_PROCESS_SHARED futexes should still use the old subcommands.
Compatibility with old applications is preserved, they still hit the
scalability problems, but new applications can fly :)
Note : the same SHARED futex (mapped on a file) can be used by old binaries
*and* new binaries, because both binaries will use the old subcommands.
Note : Vast majority of futexes should be using PROCESS_PRIVATE semantic,
as this is the default semantic. Almost all applications should benefit
of this changes (new kernel and updated libc)
Some bench results on a Pentium M 1.6 GHz (SMP kernel on a UP machine)
/* calling futex_wait(addr, value) with value != *addr */
433 cycles per futex(FUTEX_WAIT) call (mixing 2 futexes)
424 cycles per futex(FUTEX_WAIT) call (using one futex)
334 cycles per futex(FUTEX_WAIT_PRIVATE) call (mixing 2 futexes)
334 cycles per futex(FUTEX_WAIT_PRIVATE) call (using one futex)
For reference :
187 cycles per getppid() call
188 cycles per umask() call
181 cycles per ni_syscall() call
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Pierre Peiffer <pierre.peiffer@bull.net>
Cc: "Ulrich Drepper" <drepper@gmail.com>
Cc: "Nick Piggin" <nickpiggin@yahoo.com.au>
Cc: "Ingo Molnar" <mingo@elte.hu>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch provides the futex_requeue_pi functionality, which allows some
threads waiting on a normal futex to be requeued on the wait-queue of a
PI-futex.
This provides an optimization, already used for (normal) futexes, to be used
with the PI-futexes.
This optimization is currently used by the glibc in pthread_broadcast, when
using "normal" mutexes. With futex_requeue_pi, it can be used with
PRIO_INHERIT mutexes too.
Signed-off-by: Pierre Peiffer <pierre.peiffer@bull.net>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Ulrich Drepper <drepper@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch modifies futex_wait() to use an hrtimer + schedule() in place of
schedule_timeout().
schedule_timeout() is tick based, therefore the timeout granularity is the
tick (1 ms, 4 ms or 10 ms depending on HZ). By using a high resolution timer
for timeout wakeup, we can attain a much finer timeout granularity (in the
microsecond range). This parallels what is already done for futex_lock_pi().
The timeout passed to the syscall is no longer converted to jiffies and is
therefore passed to do_futex() and futex_wait() as an absolute ktime_t
therefore keeping nanosecond resolution.
Also this removes the need to pass the nanoseconds timeout part to
futex_lock_pi() in val2.
In futex_wait(), if there is no timeout then a regular schedule() is
performed. Otherwise, an hrtimer is fired before schedule() is called.
[akpm@linux-foundation.org: fix `make headers_check']
Signed-off-by: Sebastien Dugue <sebastien.dugue@bull.net>
Signed-off-by: Pierre Peiffer <pierre.peiffer@bull.net>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Ulrich Drepper <drepper@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Today, all threads waiting for a given futex are woken in FIFO order (first
waiter woken first) instead of priority order.
This patch makes use of plist (pirotity ordered lists) instead of simple list
in futex_hash_bucket.
All non-RT threads are stored with priority MAX_RT_PRIO, causing them to be
woken last, in FIFO order (RT-threads are woken first, in priority order).
Signed-off-by: Sebastien Dugue <sebastien.dugue@bull.net>
Signed-off-by: Pierre Peiffer <pierre.peiffer@bull.net>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Ulrich Drepper <drepper@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
LTP test sigaction_16_24 fails, because it expects sem_wait to be restarted
if SA_RESTART is set. sem_wait is implemented with futex_wait, that
currently doesn't support being restarted. Ulrich confirms that the call
should be restartable.
Implement a restart_block method to handle the relative timeout, and allow
restarts.
Signed-off-by: Nick Piggin <npiggin@suse.de>
Cc: Ulrich Drepper <drepper@redhat.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Roland McGrath <roland@redhat.com>
Cc: Oleg Nesterov <oleg@tv-sign.ru>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
lguest uses the convenient futex infrastructure for inter-domain I/O, so
expose get_futex_key, get_key_refs (renamed get_futex_key_refs) and
drop_key_refs (renamed drop_futex_key_refs). Also means we need to expose the
union that these use.
No code changes.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Testing of -rt by IBM uncovered a locking bug in wake_futex_pi(): the PI
state needs to be locked before we access it.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Chuck Ebbert <cebbert@redhat.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
- hrtimers did not use the hrtimer_restart enum and relied on the implict
int representation. Fix the prototypes and the functions using the enums.
- Use seperate name spaces for the enumerations
- Convert hrtimer_restart macro to inline function
- Add comments
No functional changes.
[akpm@osdl.org: fix input driver]
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: john stultz <johnstul@us.ibm.com>
Cc: Roman Zippel <zippel@linux-m68k.org>
Cc: Dmitry Torokhov <dtor@mail.ru>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Change all the uses of f_{dentry,vfsmnt} to f_path.{dentry,mnt} in
linux/kernel/.
Signed-off-by: Josef "Jeff" Sipek <jsipek@cs.sunysb.edu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
- move some file_operations structs into the .rodata section
- move static strings from policy_types[] array into the .rodata section
- fix generic seq_operations usages, so that those structs may be defined
as "const" as well
[akpm@osdl.org: couple of fixes]
Signed-off-by: Helge Deller <deller@gmx.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
When disassembling a kernel I found around over 90 sync Instructions from
mb, rmb and wmb calls in the kernel and only few of those make any sense to
me. So here's the first one - I think the wmb() in kernel/futex.c is not
needed on uniprocessors so should become an smb_wmb().
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Acked-by: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Introduce pagefault_{disable,enable}() and use these where previously we did
manual preempt increments/decrements to make the pagefault handler do the
atomic thing.
Currently they still rely on the increased preempt count, but do not rely on
the disabled preemption, this might go away in the future.
(NOTE: the extra barrier() in pagefault_disable might fix some holes on
machines which have too many registers for their own good)
[heiko.carstens@de.ibm.com: s390 fix]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Nick Piggin <npiggin@suse.de>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Apparently FUTEX_FD is unfixably racy and nothing uses it (or if it does, it
shouldn't).
Add a warning printk, give any remaining users six months to migrate off it.
Cc: Ulrich Drepper <drepper@redhat.com>
Cc: Ingo Molnar <mingo@elte.hu>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
File handles can be requested to send sigio and sigurg to processes. By
tracking the destination processes using struct pid instead of pid_t we make
the interface safe from all potential pid wrap around problems.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
It is ok to do find_task_by_pid() + get_task_struct() under
rcu_read_lock(), we cand drop tasklist_lock.
Note that testing of ->exit_state is racy with or without tasklist anyway.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Acked-by: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The current implementation of futex_lock_pi returns -ERESTART_RESTARTBLOCK
in case that the lock operation has been interrupted by a signal. This
results in a return of -EINTR to userspace in case there is an handler for
the signal. This is wrong, because userspace expects that the lock
function does not return in any case of signal delivery.
This was not caught by my insufficient test case, but triggered a nasty
userspace problem in an high load application scenario. Unfortunately also
glibc does not check for this invalid return value.
Using -ERSTARTNOINTR makes sure, that the interrupted syscall is restarted.
The restart block related code can be safely removed, as the possible
timeout argument is an absolute time value.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
futex_find_get_task:
if (p->state == EXIT_ZOMBIE || p->exit_state == EXIT_ZOMBIE)
return NULL;
I can't understand this. First, p->state can't be EXIT_ZOMBIE. The
->exit_state check looks strange too. Sub-threads or tasks whose ->parent
ignores SIGCHLD go directly to EXIT_DEAD state (I am ignoring a ptrace
case). Why EXIT_DEAD tasks should be ok? Yes, EXIT_ZOMBIE is more
important (a task may stay zombie for a long time), but this doesn't mean
we should explicitely ignore other EXIT_XXX states.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Acked-by: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
We found this issue last week w/ the -RT kernel, but it seems the same
issue is in mainline as well.
Basically it is possible for futex_unlock_pi to return without actually
freeing the lock. This is due to buggy logic in the use of
futex_handle_fault() and its attempt argument in a failure case.
Looking at futex.c the logic is as follows:
1) In futex_unlock_pi() we start w/ ret=0 and we go down to the first
futex_atomic_cmpxchg_inatomic(), where we find uval==-EFAULT. We then
jump to the pi_faulted label.
2) From pi_faulted: We increment attempt, unlock the sem and hit the
retry label.
3) From the retry label, with ret still zero, we again hit EFAULT on the
first futex_atomic_cmpxchg_inatomic(), and again goto the pi_faulted
label.
4) Again from pi_faulted: we increment attempt and enter the
conditional, where we call futex_handle_fault.
5) futex_handle_fault fails, and we goto the out_unlock_release_sem
label.
6) From out_unlock_release_sem we return, and since ret is still zero,
we return without error, while never actually unlocking the lock.
Issue #1: at the first futex_atomic_cmpxchg_inatomic() we should probably
be setting ret=-EFAULT before jumping to pi_faulted: However in our case
this doesn't really affect anything, as the glibc we're using ignores the
error value from futex_unlock_pi().
Issue #2: Look at futex_handle_fault(), its first conditional will return
-EFAULT if attempt is >= 2. However, from the "if(attempt++)
futex_handle_fault(attempt)" logic above, we'll *never* call
futex_handle_fault when attempt is less then two. So we never get a chance
to even try to fault the page in.
The following patch addresses these two issues by 1) Always setting ret to
-EFAULT if futex_handle_fault fails, and 2) Removing the = in
futex_handle_fault's (attempt >= 2) check.
I'm really not sure this is the right fix, but wanted to bring it up so
folks knew the issue is alive and well in the current -git tree. From
looking at the git logs the logic was first introduced (then later copied
to other places) in the following commit almost a year ago:
http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=4732efbeb997189d9f9b04708dc26bf8613ed721;hp=5b039e681b8c5f30aac9cc04385cc94be45d0823
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Ingo Molnar <mingo@elte.hu>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch adds a barrier() in futex unqueue_me to avoid aliasing of two
pointers.
On my s390x system I saw the following oops:
Unable to handle kernel pointer dereference at virtual kernel address
0000000000000000
Oops: 0004 [#1]
CPU: 0 Not tainted
Process mytool (pid: 13613, task: 000000003ecb6ac0, ksp: 00000000366bdbd8)
Krnl PSW : 0704d00180000000 00000000003c9ac2 (_spin_lock+0xe/0x30)
Krnl GPRS: 00000000ffffffff 000000003ecb6ac0 0000000000000000 0700000000000000
0000000000000000 0000000000000000 000001fe00002028 00000000000c091f
000001fe00002054 000001fe00002054 0000000000000000 00000000366bddc0
00000000005ef8c0 00000000003d00e8 0000000000144f91 00000000366bdcb8
Krnl Code: ba 4e 20 00 12 44 b9 16 00 3e a7 84 00 08 e3 e0 f0 88 00 04
Call Trace:
([<0000000000144f90>] unqueue_me+0x40/0xe4)
[<0000000000145a0c>] do_futex+0x33c/0xc40
[<000000000014643e>] sys_futex+0x12e/0x144
[<000000000010bb00>] sysc_noemu+0x10/0x16
[<000002000003741c>] 0x2000003741c
The code in question is:
static int unqueue_me(struct futex_q *q)
{
int ret = 0;
spinlock_t *lock_ptr;
/* In the common case we don't take the spinlock, which is nice. */
retry:
lock_ptr = q->lock_ptr;
if (lock_ptr != 0) {
spin_lock(lock_ptr);
/*
* q->lock_ptr can change between reading it and
* spin_lock(), causing us to take the wrong lock. This
* corrects the race condition.
[...]
and my compiler (gcc 4.1.0) makes the following out of it:
00000000000003c8 <unqueue_me>:
3c8: eb bf f0 70 00 24 stmg %r11,%r15,112(%r15)
3ce: c0 d0 00 00 00 00 larl %r13,3ce <unqueue_me+0x6>
3d0: R_390_PC32DBL .rodata+0x2a
3d4: a7 f1 1e 00 tml %r15,7680
3d8: a7 84 00 01 je 3da <unqueue_me+0x12>
3dc: b9 04 00 ef lgr %r14,%r15
3e0: a7 fb ff d0 aghi %r15,-48
3e4: b9 04 00 b2 lgr %r11,%r2
3e8: e3 e0 f0 98 00 24 stg %r14,152(%r15)
3ee: e3 c0 b0 28 00 04 lg %r12,40(%r11)
/* write q->lock_ptr in r12 */
3f4: b9 02 00 cc ltgr %r12,%r12
3f8: a7 84 00 4b je 48e <unqueue_me+0xc6>
/* if r12 is zero then jump over the code.... */
3fc: e3 20 b0 28 00 04 lg %r2,40(%r11)
/* write q->lock_ptr in r2 */
402: c0 e5 00 00 00 00 brasl %r14,402 <unqueue_me+0x3a>
404: R_390_PC32DBL _spin_lock+0x2
/* use r2 as parameter for spin_lock */
So the code becomes more or less:
if (q->lock_ptr != 0) spin_lock(q->lock_ptr)
instead of
if (lock_ptr != 0) spin_lock(lock_ptr)
Which caused the oops from above.
After adding a barrier gcc creates code without this problem:
[...] (the same)
3ee: e3 c0 b0 28 00 04 lg %r12,40(%r11)
3f4: b9 02 00 cc ltgr %r12,%r12
3f8: b9 04 00 2c lgr %r2,%r12
3fc: a7 84 00 48 je 48c <unqueue_me+0xc4>
400: c0 e5 00 00 00 00 brasl %r14,400 <unqueue_me+0x38>
402: R_390_PC32DBL _spin_lock+0x2
As a general note, this code of unqueue_me seems a bit fishy. The retry logic
of unqueue_me only works if we can guarantee, that the original value of
q->lock_ptr is always a spinlock (Otherwise we overwrite kernel memory). We
know that q->lock_ptr can change. I dont know what happens with the original
spinlock, as I am not an expert with the futex code.
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Ingo Molnar <mingo@redhat.com>
Cc: Thomas Gleixner <tglx@timesys.com>
Signed-off-by: Christian Borntraeger <borntrae@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Fix robust PI-futexes to be properly unlocked on unexpected exit.
For this to work the kernel has to know whether a futex is a PI or a
non-PI one, because the semantics are different. Since the space in
relevant glibc data structures is extremely scarce, the best solution is
to encode the 'PI' information in bit 0 of the robust list pointer.
Existing (non-PI) glibc robust futexes have this bit always zero, so the
ABI is kept. New glibc with PI-robust-futexes will set this bit.
Further fixes from Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Ulrich Drepper <drepper@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Fix pi_state->list handling bugs: list handling mishap, locking error.
Plus add more debug checks and fix a few style issues i noticed while
debugging this.
(reported by Ulrich Drepper and Jakub Jelinek.)
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Calling futex_lock_pi is called with a reference to a non PI futex and
waiters exist already, lookup_pi_state() oopses due to pi_state == NULL.
Check this condition and return -EINVAL to userspace.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jakub Jelinek <jakub@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Teach special (recursive) locking code to the lock validator. Introduces
double_lock_hb() to unify double- hash-bucket-lock taking.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Fix futex_wake() exit condition bug when handling the robust-list with PI
futexes on them.
(reported by Ulrich Drepper, debugged by the lock validator.)
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: Ulrich Drepper <drepper@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
lock_queue was getting called essentially twice in a row and was
continually incrementing the mm_count ref count, thus causing a memory
leak.
Dinakar Guniguntala provided a proper fix for the problem that simply grabs
the spinlock for the hash bucket queue rather than calling lock_queue.
The second time we do a queue_lock in futex_lock_pi, we really only need to
take the hash bucket lock.
Signed-off-by: Dinakar Guniguntala <dino@in.ibm.com>
Signed-off-by: Vernon Mauery <vernux@us.ibm.com>
Acked-by: Paul E. McKenney <paulmck@us.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
In futex_requeue(), when the 2 futexes keys hash to the same bucket, there
is no need to move the futex_q to the end of the bucket list.
Signed-off-by: Sebastien Dugue <sebastien.dugue@bull.net>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This adds the actual pi-futex implementation, based on rt-mutexes.
[dino@in.ibm.com: fix an oops-causing race]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Dinakar Guniguntala <dino@in.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
We are pleased to announce "lightweight userspace priority inheritance" (PI)
support for futexes. The following patchset and glibc patch implements it,
ontop of the robust-futexes patchset which is included in 2.6.16-mm1.
We are calling it lightweight for 3 reasons:
- in the user-space fastpath a PI-enabled futex involves no kernel work
(or any other PI complexity) at all. No registration, no extra kernel
calls - just pure fast atomic ops in userspace.
- in the slowpath (in the lock-contention case), the system call and
scheduling pattern is in fact better than that of normal futexes, due to
the 'integrated' nature of FUTEX_LOCK_PI. [more about that further down]
- the in-kernel PI implementation is streamlined around the mutex
abstraction, with strict rules that keep the implementation relatively
simple: only a single owner may own a lock (i.e. no read-write lock
support), only the owner may unlock a lock, no recursive locking, etc.
Priority Inheritance - why, oh why???
-------------------------------------
Many of you heard the horror stories about the evil PI code circling Linux for
years, which makes no real sense at all and is only used by buggy applications
and which has horrible overhead. Some of you have dreaded this very moment,
when someone actually submits working PI code ;-)
So why would we like to see PI support for futexes?
We'd like to see it done purely for technological reasons. We dont think it's
a buggy concept, we think it's useful functionality to offer to applications,
which functionality cannot be achieved in other ways. We also think it's the
right thing to do, and we think we've got the right arguments and the right
numbers to prove that. We also believe that we can address all the
counter-arguments as well. For these reasons (and the reasons outlined below)
we are submitting this patch-set for upstream kernel inclusion.
What are the benefits of PI?
The short reply:
----------------
User-space PI helps achieving/improving determinism for user-space
applications. In the best-case, it can help achieve determinism and
well-bound latencies. Even in the worst-case, PI will improve the statistical
distribution of locking related application delays.
The longer reply:
-----------------
Firstly, sharing locks between multiple tasks is a common programming
technique that often cannot be replaced with lockless algorithms. As we can
see it in the kernel [which is a quite complex program in itself], lockless
structures are rather the exception than the norm - the current ratio of
lockless vs. locky code for shared data structures is somewhere between 1:10
and 1:100. Lockless is hard, and the complexity of lockless algorithms often
endangers to ability to do robust reviews of said code. I.e. critical RT
apps often choose lock structures to protect critical data structures, instead
of lockless algorithms. Furthermore, there are cases (like shared hardware,
or other resource limits) where lockless access is mathematically impossible.
Media players (such as Jack) are an example of reasonable application design
with multiple tasks (with multiple priority levels) sharing short-held locks:
for example, a highprio audio playback thread is combined with medium-prio
construct-audio-data threads and low-prio display-colory-stuff threads. Add
video and decoding to the mix and we've got even more priority levels.
So once we accept that synchronization objects (locks) are an unavoidable fact
of life, and once we accept that multi-task userspace apps have a very fair
expectation of being able to use locks, we've got to think about how to offer
the option of a deterministic locking implementation to user-space.
Most of the technical counter-arguments against doing priority inheritance
only apply to kernel-space locks. But user-space locks are different, there
we cannot disable interrupts or make the task non-preemptible in a critical
section, so the 'use spinlocks' argument does not apply (user-space spinlocks
have the same priority inversion problems as other user-space locking
constructs). Fact is, pretty much the only technique that currently enables
good determinism for userspace locks (such as futex-based pthread mutexes) is
priority inheritance:
Currently (without PI), if a high-prio and a low-prio task shares a lock [this
is a quite common scenario for most non-trivial RT applications], even if all
critical sections are coded carefully to be deterministic (i.e. all critical
sections are short in duration and only execute a limited number of
instructions), the kernel cannot guarantee any deterministic execution of the
high-prio task: any medium-priority task could preempt the low-prio task while
it holds the shared lock and executes the critical section, and could delay it
indefinitely.
Implementation:
---------------
As mentioned before, the userspace fastpath of PI-enabled pthread mutexes
involves no kernel work at all - they behave quite similarly to normal
futex-based locks: a 0 value means unlocked, and a value==TID means locked.
(This is the same method as used by list-based robust futexes.) Userspace uses
atomic ops to lock/unlock these mutexes without entering the kernel.
To handle the slowpath, we have added two new futex ops:
FUTEX_LOCK_PI
FUTEX_UNLOCK_PI
If the lock-acquire fastpath fails, [i.e. an atomic transition from 0 to TID
fails], then FUTEX_LOCK_PI is called. The kernel does all the remaining work:
if there is no futex-queue attached to the futex address yet then the code
looks up the task that owns the futex [it has put its own TID into the futex
value], and attaches a 'PI state' structure to the futex-queue. The pi_state
includes an rt-mutex, which is a PI-aware, kernel-based synchronization
object. The 'other' task is made the owner of the rt-mutex, and the
FUTEX_WAITERS bit is atomically set in the futex value. Then this task tries
to lock the rt-mutex, on which it blocks. Once it returns, it has the mutex
acquired, and it sets the futex value to its own TID and returns. Userspace
has no other work to perform - it now owns the lock, and futex value contains
FUTEX_WAITERS|TID.
If the unlock side fastpath succeeds, [i.e. userspace manages to do a TID ->
0 atomic transition of the futex value], then no kernel work is triggered.
If the unlock fastpath fails (because the FUTEX_WAITERS bit is set), then
FUTEX_UNLOCK_PI is called, and the kernel unlocks the futex on the behalf of
userspace - and it also unlocks the attached pi_state->rt_mutex and thus wakes
up any potential waiters.
Note that under this approach, contrary to other PI-futex approaches, there is
no prior 'registration' of a PI-futex. [which is not quite possible anyway,
due to existing ABI properties of pthread mutexes.]
Also, under this scheme, 'robustness' and 'PI' are two orthogonal properties
of futexes, and all four combinations are possible: futex, robust-futex,
PI-futex, robust+PI-futex.
glibc support:
--------------
Ulrich Drepper and Jakub Jelinek have written glibc support for PI-futexes
(and robust futexes), enabling robust and PI (PTHREAD_PRIO_INHERIT) POSIX
mutexes. (PTHREAD_PRIO_PROTECT support will be added later on too, no
additional kernel changes are needed for that). [NOTE: The glibc patch is
obviously inofficial and unsupported without matching upstream kernel
functionality.]
the patch-queue and the glibc patch can also be downloaded from:
http://redhat.com/~mingo/PI-futex-patches/
Many thanks go to the people who helped us create this kernel feature: Steven
Rostedt, Esben Nielsen, Benedikt Spranger, Daniel Walker, John Cooper, Arjan
van de Ven, Oleg Nesterov and others. Credits for related prior projects goes
to Dirk Grambow, Inaky Perez-Gonzalez, Bill Huey and many others.
Clean up the futex code, before adding more features to it:
- use u32 as the futex field type - that's the ABI
- use __user and pointers to u32 instead of unsigned long
- code style / comment style cleanups
- rename hash-bucket name from 'bh' to 'hb'.
I checked the pre and post futex.o object files to make sure this
patch has no code effects.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Cc: Ulrich Drepper <drepper@redhat.com>
Cc: Jakub Jelinek <jakub@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Extend the get_sb() filesystem operation to take an extra argument that
permits the VFS to pass in the target vfsmount that defines the mountpoint.
The filesystem is then required to manually set the superblock and root dentry
pointers. For most filesystems, this should be done with simple_set_mnt()
which will set the superblock pointer and then set the root dentry to the
superblock's s_root (as per the old default behaviour).
The get_sb() op now returns an integer as there's now no need to return the
superblock pointer.
This patch permits a superblock to be implicitly shared amongst several mount
points, such as can be done with NFS to avoid potential inode aliasing. In
such a case, simple_set_mnt() would not be called, and instead the mnt_root
and mnt_sb would be set directly.
The patch also makes the following changes:
(*) the get_sb_*() convenience functions in the core kernel now take a vfsmount
pointer argument and return an integer, so most filesystems have to change
very little.
(*) If one of the convenience function is not used, then get_sb() should
normally call simple_set_mnt() to instantiate the vfsmount. This will
always return 0, and so can be tail-called from get_sb().
(*) generic_shutdown_super() now calls shrink_dcache_sb() to clean up the
dcache upon superblock destruction rather than shrink_dcache_anon().
This is required because the superblock may now have multiple trees that
aren't actually bound to s_root, but that still need to be cleaned up. The
currently called functions assume that the whole tree is rooted at s_root,
and that anonymous dentries are not the roots of trees which results in
dentries being left unculled.
However, with the way NFS superblock sharing are currently set to be
implemented, these assumptions are violated: the root of the filesystem is
simply a dummy dentry and inode (the real inode for '/' may well be
inaccessible), and all the vfsmounts are rooted on anonymous[*] dentries
with child trees.
[*] Anonymous until discovered from another tree.
(*) The documentation has been adjusted, including the additional bit of
changing ext2_* into foo_* in the documentation.
[akpm@osdl.org: convert ipath_fs, do other stuff]
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Nathan Scott <nathans@sgi.com>
Cc: Roland Dreier <rolandd@cisco.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The futex timeval is not checked for correctness. The change does not
break existing applications as the timeval is supplied by glibc (and glibc
always passes a correct value), but the glibc-internal tests for this
functionality fail.
Signed-off-by: Thomas Gleixner <tglx@tglx.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
- fix: initialize the robust list(s) to NULL in copy_process.
- doc update
- cleanup: rename _inuser to _inatomic
- __user cleanups and other small cleanups
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Arjan van de Ven <arjan@infradead.org>
Cc: Ulrich Drepper <drepper@redhat.com>
Cc: Andi Kleen <ak@muc.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Add the core infrastructure for robust futexes: structure definitions, the new
syscalls and the do_exit() based cleanup mechanism.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Arjan van de Ven <arjan@infradead.org>
Acked-by: Ulrich Drepper <drepper@redhat.com>
Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Make the futex code compilable and usable on NOMMU by making the attempt to
handle page faults conditional on CONFIG_MMU. If this is not enabled, then
we can assume that EFAULT returned from futex_atomic_op_inuser() is not
recoverable, and that the address lies outside of valid memory.
handle_mm_fault() is made to BUG if called on NOMMU without attempting to
invoke the actual handler (__handle_mm_fault).
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Fix a memory ordering problem that occurs on IA64. The "store" to q->lock_ptr
in wake_futex() can become visible before wake_up_all() clears the lock in the
futex_q.
Signed-off-by: Jack Steiner <steiner@sgi.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The follow_page changes in get_futex_key have left it with two almost
identical blocks, when handling the rare case of a futex in a nonlinear vma.
get_user_pages will itself do that follow_page, and its additional
find_extend_vma is hardly any overhead since the vma is already cached. Let's
just delete the follow_page block and let get_user_pages do it.
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The code for FUTEX_WAKE_OP calls an arch callback,
futex_atomic_op_inuser(). That callback can return an error code, but
currently the caller assumes any error is EFAULT, and will try various
things to resolve the fault before eventually giving up with EFAULT
(regardless of the original error code). This is not a theoretical case -
arch callbacks currently return -ENOSYS if the opcode they are given is
bogus.
This patch alters the code to detect non-EFAULT errors and return them
directly to the user.
Of course, whether -ENOSYS is the correct return value for the bogus opcode
case, or whether EINVAL would be more appropriate is another question.
Signed-off-by: David Gibson <dwg@au1.ibm.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jamie Lokier <jamie@shareable.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Final step in pushing down common core's page_table_lock. follow_page no
longer wants caller to hold page_table_lock, uses pte_offset_map_lock itself;
and so no page_table_lock is taken in get_user_pages itself.
But get_user_pages (and get_futex_key) do then need follow_page to pin the
page for them: take Daniel's suggestion of bitflags to follow_page.
Need one for WRITE, another for TOUCH (it was the accessed flag before:
vanished along with check_user_page_readable, but surely get_numa_maps is
wrong to mark every page it finds as accessed), another for GET.
And another, ANON to dispose of untouched_anonymous_page: it seems silly for
that to descend a second time, let follow_page observe if there was no page
table and return ZERO_PAGE if so. Fix minor bug in that: check VM_LOCKED -
make_pages_present ought to make readonly anonymous present.
Give get_numa_maps a cond_resched while we're there.
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch cleans up the error path of futex_fd() by removing duplicate
code.
Signed-off-by: Pekka Enberg <penberg@cs.helsinki.fi>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
ATM pthread_cond_signal is unnecessarily slow, because it wakes one waiter
(which at least on UP usually means an immediate context switch to one of
the waiter threads). This waiter wakes up and after a few instructions it
attempts to acquire the cv internal lock, but that lock is still held by
the thread calling pthread_cond_signal. So it goes to sleep and eventually
the signalling thread is scheduled in, unlocks the internal lock and wakes
the waiter again.
Now, before 2003-09-21 NPTL was using FUTEX_REQUEUE in pthread_cond_signal
to avoid this performance issue, but it was removed when locks were
redesigned to the 3 state scheme (unlocked, locked uncontended, locked
contended).
Following scenario shows why simply using FUTEX_REQUEUE in
pthread_cond_signal together with using lll_mutex_unlock_force in place of
lll_mutex_unlock is not enough and probably why it has been disabled at
that time:
The number is value in cv->__data.__lock.
thr1 thr2 thr3
0 pthread_cond_wait
1 lll_mutex_lock (cv->__data.__lock)
0 lll_mutex_unlock (cv->__data.__lock)
0 lll_futex_wait (&cv->__data.__futex, futexval)
0 pthread_cond_signal
1 lll_mutex_lock (cv->__data.__lock)
1 pthread_cond_signal
2 lll_mutex_lock (cv->__data.__lock)
2 lll_futex_wait (&cv->__data.__lock, 2)
2 lll_futex_requeue (&cv->__data.__futex, 0, 1, &cv->__data.__lock)
# FUTEX_REQUEUE, not FUTEX_CMP_REQUEUE
2 lll_mutex_unlock_force (cv->__data.__lock)
0 cv->__data.__lock = 0
0 lll_futex_wake (&cv->__data.__lock, 1)
1 lll_mutex_lock (cv->__data.__lock)
0 lll_mutex_unlock (cv->__data.__lock)
# Here, lll_mutex_unlock doesn't know there are threads waiting
# on the internal cv's lock
Now, I believe it is possible to use FUTEX_REQUEUE in pthread_cond_signal,
but it will cost us not one, but 2 extra syscalls and, what's worse, one of
these extra syscalls will be done for every single waiting loop in
pthread_cond_*wait.
We would need to use lll_mutex_unlock_force in pthread_cond_signal after
requeue and lll_mutex_cond_lock in pthread_cond_*wait after lll_futex_wait.
Another alternative is to do the unlocking pthread_cond_signal needs to do
(the lock can't be unlocked before lll_futex_wake, as that is racy) in the
kernel.
I have implemented both variants, futex-requeue-glibc.patch is the first
one and futex-wake_op{,-glibc}.patch is the unlocking inside of the kernel.
The kernel interface allows userland to specify how exactly an unlocking
operation should look like (some atomic arithmetic operation with optional
constant argument and comparison of the previous futex value with another
constant).
It has been implemented just for ppc*, x86_64 and i?86, for other
architectures I'm including just a stub header which can be used as a
starting point by maintainers to write support for their arches and ATM
will just return -ENOSYS for FUTEX_WAKE_OP. The requeue patch has been
(lightly) tested just on x86_64, the wake_op patch on ppc64 kernel running
32-bit and 64-bit NPTL and x86_64 kernel running 32-bit and 64-bit NPTL.
With the following benchmark on UP x86-64 I get:
for i in nptl-orig nptl-requeue nptl-wake_op; do echo time elf/ld.so --library-path .:$i /tmp/bench; \
for j in 1 2; do echo ( time elf/ld.so --library-path .:$i /tmp/bench ) 2>&1; done; done
time elf/ld.so --library-path .:nptl-orig /tmp/bench
real 0m0.655s user 0m0.253s sys 0m0.403s
real 0m0.657s user 0m0.269s sys 0m0.388s
time elf/ld.so --library-path .:nptl-requeue /tmp/bench
real 0m0.496s user 0m0.225s sys 0m0.271s
real 0m0.531s user 0m0.242s sys 0m0.288s
time elf/ld.so --library-path .:nptl-wake_op /tmp/bench
real 0m0.380s user 0m0.176s sys 0m0.204s
real 0m0.382s user 0m0.175s sys 0m0.207s
The benchmark is at:
http://sourceware.org/ml/libc-alpha/2005-03/txt00001.txt
Older futex-requeue-glibc.patch version is at:
http://sourceware.org/ml/libc-alpha/2005-03/txt00002.txt
Older futex-wake_op-glibc.patch version is at:
http://sourceware.org/ml/libc-alpha/2005-03/txt00003.txt
Will post a new version (just x86-64 fixes so that the patch
applies against pthread_cond_signal.S) to libc-hacker ml soon.
Attached is the kernel FUTEX_WAKE_OP patch as well as a simple-minded
testcase that will not test the atomicity of the operation, but at least
check if the threads that should have been woken up are woken up and
whether the arithmetic operation in the kernel gave the expected results.
Acked-by: Ingo Molnar <mingo@redhat.com>
Cc: Ulrich Drepper <drepper@redhat.com>
Cc: Jamie Lokier <jamie@shareable.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Yoichi Yuasa <yuasa@hh.iij4u.or.jp>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Convert most of the current code that uses _NSIG directly to instead use
valid_signal(). This avoids gcc -W warnings and off-by-one errors.
Signed-off-by: Jesper Juhl <juhl-lkml@dif.dk>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.
Let it rip!
