* Copyright (C) 2006 Red Hat, Inc., Ingo Molnar <mingo@redhat.com>
* Copyright (C) 2006 Timesys Corp., Thomas Gleixner <tglx@timesys.com>
*
+ * PRIVATE futexes by Eric Dumazet
+ * Copyright (C) 2007 Eric Dumazet <dada1@cosmosbay.com>
+ *
* Thanks to Ben LaHaise for yelling "hashed waitqueues" loudly
* enough at me, Linus for the original (flawed) idea, Matthew
* Kirkwood for proof-of-concept implementation.
#include "rtmutex_common.h"
-#ifdef CONFIG_DEBUG_RT_MUTEXES
-# include "rtmutex-debug.h"
-#else
-# include "rtmutex.h"
-#endif
-
#define FUTEX_HASHBITS (CONFIG_BASE_SMALL ? 4 : 8)
/*
/* Optional priority inheritance state: */
struct futex_pi_state *pi_state;
struct task_struct *task;
-
- /*
- * This waiter is used in case of requeue from a
- * normal futex to a PI-futex
- */
- struct rt_mutex_waiter waiter;
};
/*
/* Futex-fs vfsmount entry: */
static struct vfsmount *futex_mnt;
+/*
+ * Take mm->mmap_sem, when futex is shared
+ */
+static inline void futex_lock_mm(struct rw_semaphore *fshared)
+{
+ if (fshared)
+ down_read(fshared);
+}
+
+/*
+ * Release mm->mmap_sem, when the futex is shared
+ */
+static inline void futex_unlock_mm(struct rw_semaphore *fshared)
+{
+ if (fshared)
+ up_read(fshared);
+}
+
/*
* We hash on the keys returned from get_futex_key (see below).
*/
&& key1->both.offset == key2->both.offset);
}
-/*
- * Get parameters which are the keys for a futex.
+/**
+ * get_futex_key - Get parameters which are the keys for a futex.
+ * @uaddr: virtual address of the futex
+ * @shared: NULL for a PROCESS_PRIVATE futex,
+ * ¤t->mm->mmap_sem for a PROCESS_SHARED futex
+ * @key: address where result is stored.
+ *
+ * Returns a negative error code or 0
+ * The key words are stored in *key on success.
*
* For shared mappings, it's (page->index, vma->vm_file->f_path.dentry->d_inode,
* offset_within_page). For private mappings, it's (uaddr, current->mm).
* We can usually work out the index without swapping in the page.
*
- * Returns: 0, or negative error code.
- * The key words are stored in *key on success.
- *
- * Should be called with ¤t->mm->mmap_sem but NOT any spinlocks.
+ * fshared is NULL for PROCESS_PRIVATE futexes
+ * For other futexes, it points to ¤t->mm->mmap_sem and
+ * caller must have taken the reader lock. but NOT any spinlocks.
*/
-int get_futex_key(u32 __user *uaddr, union futex_key *key)
+int get_futex_key(u32 __user *uaddr, struct rw_semaphore *fshared,
+ union futex_key *key)
{
unsigned long address = (unsigned long)uaddr;
struct mm_struct *mm = current->mm;
* The futex address must be "naturally" aligned.
*/
key->both.offset = address % PAGE_SIZE;
- if (unlikely((key->both.offset % sizeof(u32)) != 0))
+ if (unlikely((address % sizeof(u32)) != 0))
return -EINVAL;
address -= key->both.offset;
+ /*
+ * PROCESS_PRIVATE futexes are fast.
+ * As the mm cannot disappear under us and the 'key' only needs
+ * virtual address, we dont even have to find the underlying vma.
+ * Note : We do have to check 'uaddr' is a valid user address,
+ * but access_ok() should be faster than find_vma()
+ */
+ if (!fshared) {
+ if (unlikely(!access_ok(VERIFY_WRITE, uaddr, sizeof(u32))))
+ return -EFAULT;
+ key->private.mm = mm;
+ key->private.address = address;
+ return 0;
+ }
/*
* The futex is hashed differently depending on whether
* it's in a shared or private mapping. So check vma first.
if (unlikely((vma->vm_flags & (VM_IO|VM_READ)) != VM_READ))
return (vma->vm_flags & VM_IO) ? -EPERM : -EACCES;
- /* Save the user address in the ley */
- key->uaddr = uaddr;
-
/*
* Private mappings are handled in a simple way.
*
* mappings of _writable_ handles.
*/
if (likely(!(vma->vm_flags & VM_MAYSHARE))) {
+ key->both.offset |= FUT_OFF_MMSHARED; /* reference taken on mm */
key->private.mm = mm;
key->private.address = address;
return 0;
* Linear file mappings are also simple.
*/
key->shared.inode = vma->vm_file->f_path.dentry->d_inode;
- key->both.offset++; /* Bit 0 of offset indicates inode-based key. */
+ key->both.offset |= FUT_OFF_INODE; /* inode-based key. */
if (likely(!(vma->vm_flags & VM_NONLINEAR))) {
key->shared.pgoff = (((address - vma->vm_start) >> PAGE_SHIFT)
+ vma->vm_pgoff);
* Take a reference to the resource addressed by a key.
* Can be called while holding spinlocks.
*
- * NOTE: mmap_sem MUST be held between get_futex_key() and calling this
- * function, if it is called at all. mmap_sem keeps key->shared.inode valid.
*/
inline void get_futex_key_refs(union futex_key *key)
{
- if (key->both.ptr != 0) {
- if (key->both.offset & 1)
+ if (key->both.ptr == 0)
+ return;
+ switch (key->both.offset & (FUT_OFF_INODE|FUT_OFF_MMSHARED)) {
+ case FUT_OFF_INODE:
atomic_inc(&key->shared.inode->i_count);
- else
+ break;
+ case FUT_OFF_MMSHARED:
atomic_inc(&key->private.mm->mm_count);
+ break;
}
}
EXPORT_SYMBOL_GPL(get_futex_key_refs);
*/
void drop_futex_key_refs(union futex_key *key)
{
- if (key->both.ptr != 0) {
- if (key->both.offset & 1)
+ if (key->both.ptr == 0)
+ return;
+ switch (key->both.offset & (FUT_OFF_INODE|FUT_OFF_MMSHARED)) {
+ case FUT_OFF_INODE:
iput(key->shared.inode);
- else
+ break;
+ case FUT_OFF_MMSHARED:
mmdrop(key->private.mm);
+ break;
}
}
EXPORT_SYMBOL_GPL(drop_futex_key_refs);
-static inline int get_futex_value_locked(u32 *dest, u32 __user *from)
+static u32 cmpxchg_futex_value_locked(u32 __user *uaddr, u32 uval, u32 newval)
+{
+ u32 curval;
+
+ pagefault_disable();
+ curval = futex_atomic_cmpxchg_inatomic(uaddr, uval, newval);
+ pagefault_enable();
+
+ return curval;
+}
+
+static int get_futex_value_locked(u32 *dest, u32 __user *from)
{
int ret;
}
/*
- * Fault handling. Called with current->mm->mmap_sem held.
+ * Fault handling.
+ * if fshared is non NULL, current->mm->mmap_sem is already held
*/
-static int futex_handle_fault(unsigned long address, int attempt)
+static int futex_handle_fault(unsigned long address,
+ struct rw_semaphore *fshared, int attempt)
{
struct vm_area_struct * vma;
struct mm_struct *mm = current->mm;
+ int ret = -EFAULT;
- if (attempt > 2 || !(vma = find_vma(mm, address)) ||
- vma->vm_start > address || !(vma->vm_flags & VM_WRITE))
- return -EFAULT;
+ if (attempt > 2)
+ return ret;
- switch (handle_mm_fault(mm, vma, address, 1)) {
- case VM_FAULT_MINOR:
- current->min_flt++;
- break;
- case VM_FAULT_MAJOR:
- current->maj_flt++;
- break;
- default:
- return -EFAULT;
+ if (!fshared)
+ down_read(&mm->mmap_sem);
+ vma = find_vma(mm, address);
+ if (vma && address >= vma->vm_start &&
+ (vma->vm_flags & VM_WRITE)) {
+ int fault;
+ fault = handle_mm_fault(mm, vma, address, 1);
+ if (unlikely((fault & VM_FAULT_ERROR))) {
+#if 0
+ /* XXX: let's do this when we verify it is OK */
+ if (ret & VM_FAULT_OOM)
+ ret = -ENOMEM;
+#endif
+ } else {
+ ret = 0;
+ if (fault & VM_FAULT_MAJOR)
+ current->maj_flt++;
+ else
+ current->min_flt++;
+ }
}
- return 0;
+ if (!fshared)
+ up_read(&mm->mmap_sem);
+ return ret;
}
/*
rcu_read_lock();
p = find_task_by_pid(pid);
- if (!p)
- goto out_unlock;
- if ((current->euid != p->euid) && (current->euid != p->uid)) {
- p = NULL;
- goto out_unlock;
- }
- if (p->exit_state != 0) {
- p = NULL;
- goto out_unlock;
- }
- get_task_struct(p);
-out_unlock:
+
+ if (!p || ((current->euid != p->euid) && (current->euid != p->uid)))
+ p = ERR_PTR(-ESRCH);
+ else
+ get_task_struct(p);
+
rcu_read_unlock();
return p;
struct futex_q *this, *next;
struct plist_head *head;
struct task_struct *p;
- pid_t pid;
+ pid_t pid = uval & FUTEX_TID_MASK;
head = &hb->chain;
return -EINVAL;
WARN_ON(!atomic_read(&pi_state->refcount));
+ WARN_ON(pid && pi_state->owner &&
+ pi_state->owner->pid != pid);
atomic_inc(&pi_state->refcount);
*ps = pi_state;
/*
* We are the first waiter - try to look up the real owner and attach
- * the new pi_state to it, but bail out when the owner died bit is set
- * and TID = 0:
+ * the new pi_state to it, but bail out when TID = 0
*/
- pid = uval & FUTEX_TID_MASK;
- if (!pid && (uval & FUTEX_OWNER_DIED))
+ if (!pid)
return -ESRCH;
p = futex_find_get_task(pid);
- if (!p)
- return -ESRCH;
+ if (IS_ERR(p))
+ return PTR_ERR(p);
+
+ /*
+ * We need to look at the task state flags to figure out,
+ * whether the task is exiting. To protect against the do_exit
+ * change of the task flags, we do this protected by
+ * p->pi_lock:
+ */
+ spin_lock_irq(&p->pi_lock);
+ if (unlikely(p->flags & PF_EXITING)) {
+ /*
+ * The task is on the way out. When PF_EXITPIDONE is
+ * set, we know that the task has finished the
+ * cleanup:
+ */
+ int ret = (p->flags & PF_EXITPIDONE) ? -ESRCH : -EAGAIN;
+
+ spin_unlock_irq(&p->pi_lock);
+ put_task_struct(p);
+ return ret;
+ }
pi_state = alloc_pi_state();
/* Store the key for possible exit cleanups: */
pi_state->key = *key;
- spin_lock_irq(&p->pi_lock);
WARN_ON(!list_empty(&pi_state->list));
list_add(&pi_state->list, &p->pi_state_list);
pi_state->owner = p;
* preserve the owner died bit.)
*/
if (!(uval & FUTEX_OWNER_DIED)) {
+ int ret = 0;
+
newval = FUTEX_WAITERS | new_owner->pid;
- /* Keep the FUTEX_WAITER_REQUEUED flag if it was set */
- newval |= (uval & FUTEX_WAITER_REQUEUED);
- pagefault_disable();
- curval = futex_atomic_cmpxchg_inatomic(uaddr, uval, newval);
- pagefault_enable();
+ curval = cmpxchg_futex_value_locked(uaddr, uval, newval);
+
if (curval == -EFAULT)
- return -EFAULT;
+ ret = -EFAULT;
if (curval != uval)
- return -EINVAL;
+ ret = -EINVAL;
+ if (ret) {
+ spin_unlock(&pi_state->pi_mutex.wait_lock);
+ return ret;
+ }
}
spin_lock_irq(&pi_state->owner->pi_lock);
* There is no waiter, so we unlock the futex. The owner died
* bit has not to be preserved here. We are the owner:
*/
- pagefault_disable();
- oldval = futex_atomic_cmpxchg_inatomic(uaddr, uval, 0);
- pagefault_enable();
+ oldval = cmpxchg_futex_value_locked(uaddr, uval, 0);
if (oldval == -EFAULT)
return oldval;
* Wake up all waiters hashed on the physical page that is mapped
* to this virtual address:
*/
-static int futex_wake(u32 __user *uaddr, int nr_wake)
+static int futex_wake(u32 __user *uaddr, struct rw_semaphore *fshared,
+ int nr_wake)
{
struct futex_hash_bucket *hb;
struct futex_q *this, *next;
union futex_key key;
int ret;
- down_read(¤t->mm->mmap_sem);
+ futex_lock_mm(fshared);
- ret = get_futex_key(uaddr, &key);
+ ret = get_futex_key(uaddr, fshared, &key);
if (unlikely(ret != 0))
goto out;
spin_unlock(&hb->lock);
out:
- up_read(¤t->mm->mmap_sem);
- return ret;
-}
-
-/*
- * Called from futex_requeue_pi.
- * Set FUTEX_WAITERS and FUTEX_WAITER_REQUEUED flags on the
- * PI-futex value; search its associated pi_state if an owner exist
- * or create a new one without owner.
- */
-static inline int
-lookup_pi_state_for_requeue(u32 __user *uaddr, struct futex_hash_bucket *hb,
- union futex_key *key,
- struct futex_pi_state **pi_state)
-{
- u32 curval, uval, newval;
-
-retry:
- /*
- * We can't handle a fault cleanly because we can't
- * release the locks here. Simply return the fault.
- */
- if (get_futex_value_locked(&curval, uaddr))
- return -EFAULT;
-
- /* set the flags FUTEX_WAITERS and FUTEX_WAITER_REQUEUED */
- if ((curval & (FUTEX_WAITERS | FUTEX_WAITER_REQUEUED))
- != (FUTEX_WAITERS | FUTEX_WAITER_REQUEUED)) {
- /*
- * No waiters yet, we prepare the futex to have some waiters.
- */
-
- uval = curval;
- newval = uval | FUTEX_WAITERS | FUTEX_WAITER_REQUEUED;
-
- pagefault_disable();
- curval = futex_atomic_cmpxchg_inatomic(uaddr, uval, newval);
- pagefault_enable();
-
- if (unlikely(curval == -EFAULT))
- return -EFAULT;
- if (unlikely(curval != uval))
- goto retry;
- }
-
- if (!(curval & FUTEX_TID_MASK)
- || lookup_pi_state(curval, hb, key, pi_state)) {
- /* the futex has no owner (yet) or the lookup failed:
- allocate one pi_state without owner */
-
- *pi_state = alloc_pi_state();
-
- /* Already stores the key: */
- (*pi_state)->key = *key;
-
- /* init the mutex without owner */
- __rt_mutex_init(&(*pi_state)->pi_mutex, NULL);
- }
-
- return 0;
-}
-
-/*
- * Keep the first nr_wake waiter from futex1, wake up one,
- * and requeue the next nr_requeue waiters following hashed on
- * one physical page to another physical page (PI-futex uaddr2)
- */
-static int futex_requeue_pi(u32 __user *uaddr1, u32 __user *uaddr2,
- int nr_wake, int nr_requeue, u32 *cmpval)
-{
- union futex_key key1, key2;
- struct futex_hash_bucket *hb1, *hb2;
- struct plist_head *head1;
- struct futex_q *this, *next;
- struct futex_pi_state *pi_state2 = NULL;
- struct rt_mutex_waiter *waiter, *top_waiter = NULL;
- struct rt_mutex *lock2 = NULL;
- int ret, drop_count = 0;
-
- if (refill_pi_state_cache())
- return -ENOMEM;
-
-retry:
- /*
- * First take all the futex related locks:
- */
- down_read(¤t->mm->mmap_sem);
-
- ret = get_futex_key(uaddr1, &key1);
- if (unlikely(ret != 0))
- goto out;
- ret = get_futex_key(uaddr2, &key2);
- if (unlikely(ret != 0))
- goto out;
-
- hb1 = hash_futex(&key1);
- hb2 = hash_futex(&key2);
-
- double_lock_hb(hb1, hb2);
-
- if (likely(cmpval != NULL)) {
- u32 curval;
-
- ret = get_futex_value_locked(&curval, uaddr1);
-
- if (unlikely(ret)) {
- spin_unlock(&hb1->lock);
- if (hb1 != hb2)
- spin_unlock(&hb2->lock);
-
- /*
- * If we would have faulted, release mmap_sem, fault
- * it in and start all over again.
- */
- up_read(¤t->mm->mmap_sem);
-
- ret = get_user(curval, uaddr1);
-
- if (!ret)
- goto retry;
-
- return ret;
- }
- if (curval != *cmpval) {
- ret = -EAGAIN;
- goto out_unlock;
- }
- }
-
- head1 = &hb1->chain;
- plist_for_each_entry_safe(this, next, head1, list) {
- if (!match_futex (&this->key, &key1))
- continue;
- if (++ret <= nr_wake) {
- wake_futex(this);
- } else {
- /*
- * FIRST: get and set the pi_state
- */
- if (!pi_state2) {
- int s;
- /* do this only the first time we requeue someone */
- s = lookup_pi_state_for_requeue(uaddr2, hb2,
- &key2, &pi_state2);
- if (s) {
- ret = s;
- goto out_unlock;
- }
-
- lock2 = &pi_state2->pi_mutex;
- spin_lock(&lock2->wait_lock);
-
- /* Save the top waiter of the wait_list */
- if (rt_mutex_has_waiters(lock2))
- top_waiter = rt_mutex_top_waiter(lock2);
- } else
- atomic_inc(&pi_state2->refcount);
-
-
- this->pi_state = pi_state2;
-
- /*
- * SECOND: requeue futex_q to the correct hashbucket
- */
-
- /*
- * If key1 and key2 hash to the same bucket, no need to
- * requeue.
- */
- if (likely(head1 != &hb2->chain)) {
- plist_del(&this->list, &hb1->chain);
- plist_add(&this->list, &hb2->chain);
- this->lock_ptr = &hb2->lock;
-#ifdef CONFIG_DEBUG_PI_LIST
- this->list.plist.lock = &hb2->lock;
-#endif
- }
- this->key = key2;
- get_futex_key_refs(&key2);
- drop_count++;
-
-
- /*
- * THIRD: queue it to lock2
- */
- spin_lock_irq(&this->task->pi_lock);
- waiter = &this->waiter;
- waiter->task = this->task;
- waiter->lock = lock2;
- plist_node_init(&waiter->list_entry, this->task->prio);
- plist_node_init(&waiter->pi_list_entry, this->task->prio);
- plist_add(&waiter->list_entry, &lock2->wait_list);
- this->task->pi_blocked_on = waiter;
- spin_unlock_irq(&this->task->pi_lock);
-
- if (ret - nr_wake >= nr_requeue)
- break;
- }
- }
-
- /* If we've requeued some tasks and the top_waiter of the rt_mutex
- has changed, we must adjust the priority of the owner, if any */
- if (drop_count) {
- struct task_struct *owner = rt_mutex_owner(lock2);
- if (owner &&
- (top_waiter != (waiter = rt_mutex_top_waiter(lock2)))) {
- int chain_walk = 0;
-
- spin_lock_irq(&owner->pi_lock);
- if (top_waiter)
- plist_del(&top_waiter->pi_list_entry, &owner->pi_waiters);
- else
- /*
- * There was no waiters before the requeue,
- * the flag must be updated
- */
- mark_rt_mutex_waiters(lock2);
-
- plist_add(&waiter->pi_list_entry, &owner->pi_waiters);
- __rt_mutex_adjust_prio(owner);
- if (owner->pi_blocked_on) {
- chain_walk = 1;
- get_task_struct(owner);
- }
-
- spin_unlock_irq(&owner->pi_lock);
- spin_unlock(&lock2->wait_lock);
-
- if (chain_walk)
- rt_mutex_adjust_prio_chain(owner, 0, lock2, NULL,
- current);
- } else {
- /* No owner or the top_waiter does not change */
- mark_rt_mutex_waiters(lock2);
- spin_unlock(&lock2->wait_lock);
- }
- }
-
-out_unlock:
- spin_unlock(&hb1->lock);
- if (hb1 != hb2)
- spin_unlock(&hb2->lock);
-
- /* drop_futex_key_refs() must be called outside the spinlocks. */
- while (--drop_count >= 0)
- drop_futex_key_refs(&key1);
-
-out:
- up_read(¤t->mm->mmap_sem);
+ futex_unlock_mm(fshared);
return ret;
}
* to this virtual address:
*/
static int
-futex_wake_op(u32 __user *uaddr1, u32 __user *uaddr2,
+futex_wake_op(u32 __user *uaddr1, struct rw_semaphore *fshared,
+ u32 __user *uaddr2,
int nr_wake, int nr_wake2, int op)
{
union futex_key key1, key2;
int ret, op_ret, attempt = 0;
retryfull:
- down_read(¤t->mm->mmap_sem);
+ futex_lock_mm(fshared);
- ret = get_futex_key(uaddr1, &key1);
+ ret = get_futex_key(uaddr1, fshared, &key1);
if (unlikely(ret != 0))
goto out;
- ret = get_futex_key(uaddr2, &key2);
+ ret = get_futex_key(uaddr2, fshared, &key2);
if (unlikely(ret != 0))
goto out;
* still holding the mmap_sem.
*/
if (attempt++) {
- if (futex_handle_fault((unsigned long)uaddr2,
- attempt)) {
- ret = -EFAULT;
+ ret = futex_handle_fault((unsigned long)uaddr2,
+ fshared, attempt);
+ if (ret)
goto out;
- }
goto retry;
}
* If we would have faulted, release mmap_sem,
* fault it in and start all over again.
*/
- up_read(¤t->mm->mmap_sem);
+ futex_unlock_mm(fshared);
ret = get_user(dummy, uaddr2);
if (ret)
if (hb1 != hb2)
spin_unlock(&hb2->lock);
out:
- up_read(¤t->mm->mmap_sem);
+ futex_unlock_mm(fshared);
+
return ret;
}
* Requeue all waiters hashed on one physical page to another
* physical page.
*/
-static int futex_requeue(u32 __user *uaddr1, u32 __user *uaddr2,
+static int futex_requeue(u32 __user *uaddr1, struct rw_semaphore *fshared,
+ u32 __user *uaddr2,
int nr_wake, int nr_requeue, u32 *cmpval)
{
union futex_key key1, key2;
int ret, drop_count = 0;
retry:
- down_read(¤t->mm->mmap_sem);
+ futex_lock_mm(fshared);
- ret = get_futex_key(uaddr1, &key1);
+ ret = get_futex_key(uaddr1, fshared, &key1);
if (unlikely(ret != 0))
goto out;
- ret = get_futex_key(uaddr2, &key2);
+ ret = get_futex_key(uaddr2, fshared, &key2);
if (unlikely(ret != 0))
goto out;
* If we would have faulted, release mmap_sem, fault
* it in and start all over again.
*/
- up_read(¤t->mm->mmap_sem);
+ futex_unlock_mm(fshared);
ret = get_user(curval, uaddr1);
#ifdef CONFIG_DEBUG_PI_LIST
this->list.plist.lock = &hb2->lock;
#endif
- }
+ }
this->key = key2;
get_futex_key_refs(&key2);
drop_count++;
drop_futex_key_refs(&key1);
out:
- up_read(¤t->mm->mmap_sem);
+ futex_unlock_mm(fshared);
return ret;
}
/*
* Fixup the pi_state owner with current.
*
- * The cur->mm semaphore must be held, it is released at return of this
- * function.
+ * Must be called with hash bucket lock held and mm->sem held for non
+ * private futexes.
*/
static int fixup_pi_state_owner(u32 __user *uaddr, struct futex_q *q,
- struct futex_hash_bucket *hb,
struct task_struct *curr)
{
u32 newtid = curr->pid | FUTEX_WAITERS;
list_add(&pi_state->list, &curr->pi_state_list);
spin_unlock_irq(&curr->pi_lock);
- /* Unqueue and drop the lock */
- unqueue_me_pi(q);
- up_read(&curr->mm->mmap_sem);
/*
* We own it, so we have to replace the pending owner
* TID. This must be atomic as we have preserve the
* owner died bit here.
*/
- ret = get_user(uval, uaddr);
+ ret = get_futex_value_locked(&uval, uaddr);
+
while (!ret) {
newval = (uval & FUTEX_OWNER_DIED) | newtid;
- newval |= (uval & FUTEX_WAITER_REQUEUED);
- curval = futex_atomic_cmpxchg_inatomic(uaddr,
- uval, newval);
+
+ curval = cmpxchg_futex_value_locked(uaddr, uval, newval);
+
if (curval == -EFAULT)
- ret = -EFAULT;
+ ret = -EFAULT;
if (curval == uval)
break;
uval = curval;
return ret;
}
+/*
+ * In case we must use restart_block to restart a futex_wait,
+ * we encode in the 'arg3' shared capability
+ */
+#define ARG3_SHARED 1
+
static long futex_wait_restart(struct restart_block *restart);
-static int futex_wait(u32 __user *uaddr, u32 val, ktime_t *abs_time)
+
+static int futex_wait(u32 __user *uaddr, struct rw_semaphore *fshared,
+ u32 val, ktime_t *abs_time)
{
struct task_struct *curr = current;
DECLARE_WAITQUEUE(wait, curr);
struct futex_q q;
u32 uval;
int ret;
- struct hrtimer_sleeper t, *to = NULL;
+ struct hrtimer_sleeper t;
int rem = 0;
q.pi_state = NULL;
retry:
- down_read(&curr->mm->mmap_sem);
+ futex_lock_mm(fshared);
- ret = get_futex_key(uaddr, &q.key);
+ ret = get_futex_key(uaddr, fshared, &q.key);
if (unlikely(ret != 0))
goto out_release_sem;
* a wakeup when *uaddr != val on entry to the syscall. This is
* rare, but normal.
*
- * We hold the mmap semaphore, so the mapping cannot have changed
- * since we looked it up in get_futex_key.
+ * for shared futexes, we hold the mmap semaphore, so the mapping
+ * cannot have changed since we looked it up in get_futex_key.
*/
ret = get_futex_value_locked(&uval, uaddr);
* If we would have faulted, release mmap_sem, fault it in and
* start all over again.
*/
- up_read(&curr->mm->mmap_sem);
+ futex_unlock_mm(fshared);
ret = get_user(uval, uaddr);
if (uval != val)
goto out_unlock_release_sem;
- /*
- * This rt_mutex_waiter structure is prepared here and will
- * be used only if this task is requeued from a normal futex to
- * a PI-futex with futex_requeue_pi.
- */
- debug_rt_mutex_init_waiter(&q.waiter);
- q.waiter.task = NULL;
-
/* Only actually queue if *uaddr contained val. */
__queue_me(&q, hb);
* Now the futex is queued and we have checked the data, we
* don't want to hold mmap_sem while we sleep.
*/
- up_read(&curr->mm->mmap_sem);
+ futex_unlock_mm(fshared);
/*
* There might have been scheduling since the queue_me(), as we
if (!abs_time)
schedule();
else {
- to = &t;
hrtimer_init(&t.timer, CLOCK_MONOTONIC, HRTIMER_MODE_ABS);
hrtimer_init_sleeper(&t, current);
t.timer.expires = *abs_time;
* we are the only user of it.
*/
- if (q.pi_state) {
- /*
- * We were woken but have been requeued on a PI-futex.
- * We have to complete the lock acquisition by taking
- * the rtmutex.
- */
-
- struct rt_mutex *lock = &q.pi_state->pi_mutex;
-
- spin_lock(&lock->wait_lock);
- if (unlikely(q.waiter.task)) {
- remove_waiter(lock, &q.waiter);
- }
- spin_unlock(&lock->wait_lock);
-
- if (rem)
- ret = -ETIMEDOUT;
- else
- ret = rt_mutex_timed_lock(lock, to, 1);
-
- down_read(&curr->mm->mmap_sem);
- spin_lock(q.lock_ptr);
-
- /*
- * Got the lock. We might not be the anticipated owner if we
- * did a lock-steal - fix up the PI-state in that case.
- */
- if (!ret && q.pi_state->owner != curr) {
- /*
- * We MUST play with the futex we were requeued on,
- * NOT the current futex.
- * We can retrieve it from the key of the pi_state
- */
- uaddr = q.pi_state->key.uaddr;
-
- /* mmap_sem and hash_bucket lock are unlocked at
- return of this function */
- ret = fixup_pi_state_owner(uaddr, &q, hb, curr);
- } else {
- /*
- * Catch the rare case, where the lock was released
- * when we were on the way back before we locked
- * the hash bucket.
- */
- if (ret && q.pi_state->owner == curr) {
- if (rt_mutex_trylock(&q.pi_state->pi_mutex))
- ret = 0;
- }
- /* Unqueue and drop the lock */
- unqueue_me_pi(&q);
- up_read(&curr->mm->mmap_sem);
- }
-
- debug_rt_mutex_free_waiter(&q.waiter);
-
- return ret;
- }
-
- debug_rt_mutex_free_waiter(&q.waiter);
-
/* If we were woken (and unqueued), we succeeded, whatever. */
if (!unqueue_me(&q))
return 0;
restart->arg0 = (unsigned long)uaddr;
restart->arg1 = (unsigned long)val;
restart->arg2 = (unsigned long)abs_time;
+ restart->arg3 = 0;
+ if (fshared)
+ restart->arg3 |= ARG3_SHARED;
return -ERESTART_RESTARTBLOCK;
}
queue_unlock(&q, hb);
out_release_sem:
- up_read(&curr->mm->mmap_sem);
+ futex_unlock_mm(fshared);
return ret;
}
u32 __user *uaddr = (u32 __user *)restart->arg0;
u32 val = (u32)restart->arg1;
ktime_t *abs_time = (ktime_t *)restart->arg2;
+ struct rw_semaphore *fshared = NULL;
restart->fn = do_no_restart_syscall;
- return (long)futex_wait(uaddr, val, abs_time);
+ if (restart->arg3 & ARG3_SHARED)
+ fshared = ¤t->mm->mmap_sem;
+ return (long)futex_wait(uaddr, fshared, val, abs_time);
}
-static void set_pi_futex_owner(struct futex_hash_bucket *hb,
- union futex_key *key, struct task_struct *p)
-{
- struct plist_head *head;
- struct futex_q *this, *next;
- struct futex_pi_state *pi_state = NULL;
- struct rt_mutex *lock;
-
- /* Search a waiter that should already exists */
-
- head = &hb->chain;
-
- plist_for_each_entry_safe(this, next, head, list) {
- if (match_futex (&this->key, key)) {
- pi_state = this->pi_state;
- break;
- }
- }
-
- BUG_ON(!pi_state);
-
- /* set p as pi_state's owner */
- lock = &pi_state->pi_mutex;
-
- spin_lock(&lock->wait_lock);
- spin_lock_irq(&p->pi_lock);
-
- list_add(&pi_state->list, &p->pi_state_list);
- pi_state->owner = p;
-
-
- /* set p as pi_mutex's owner */
- debug_rt_mutex_proxy_lock(lock, p);
- WARN_ON(rt_mutex_owner(lock));
- rt_mutex_set_owner(lock, p, 0);
- rt_mutex_deadlock_account_lock(lock, p);
-
- plist_add(&rt_mutex_top_waiter(lock)->pi_list_entry,
- &p->pi_waiters);
- __rt_mutex_adjust_prio(p);
-
- spin_unlock_irq(&p->pi_lock);
- spin_unlock(&lock->wait_lock);
-}
-
/*
* Userspace tried a 0 -> TID atomic transition of the futex value
* and failed. The kernel side here does the whole locking operation:
* if there are waiters then it will block, it does PI, etc. (Due to
* races the kernel might see a 0 value of the futex too.)
*/
-static int futex_lock_pi(u32 __user *uaddr, int detect, ktime_t *time,
- int trylock)
+static int futex_lock_pi(u32 __user *uaddr, struct rw_semaphore *fshared,
+ int detect, ktime_t *time, int trylock)
{
struct hrtimer_sleeper timeout, *to = NULL;
struct task_struct *curr = current;
struct futex_hash_bucket *hb;
u32 uval, newval, curval;
struct futex_q q;
- int ret, lock_held, attempt = 0;
+ int ret, lock_taken, ownerdied = 0, attempt = 0;
if (refill_pi_state_cache())
return -ENOMEM;
q.pi_state = NULL;
retry:
- down_read(&curr->mm->mmap_sem);
+ futex_lock_mm(fshared);
- ret = get_futex_key(uaddr, &q.key);
+ ret = get_futex_key(uaddr, fshared, &q.key);
if (unlikely(ret != 0))
goto out_release_sem;
+ retry_unlocked:
hb = queue_lock(&q, -1, NULL);
retry_locked:
- lock_held = 0;
+ ret = lock_taken = 0;
/*
* To avoid races, we attempt to take the lock here again
*/
newval = current->pid;
- pagefault_disable();
- curval = futex_atomic_cmpxchg_inatomic(uaddr, 0, newval);
- pagefault_enable();
+ curval = cmpxchg_futex_value_locked(uaddr, 0, newval);
if (unlikely(curval == -EFAULT))
goto uaddr_faulted;
- /* We own the lock already */
+ /*
+ * Detect deadlocks. In case of REQUEUE_PI this is a valid
+ * situation and we return success to user space.
+ */
if (unlikely((curval & FUTEX_TID_MASK) == current->pid)) {
- if (!detect && 0)
- force_sig(SIGKILL, current);
- /*
- * Normally, this check is done in user space.
- * In case of requeue, the owner may attempt to lock this futex,
- * even if the ownership has already been given by the previous
- * waker.
- * In the usual case, this is a case of deadlock, but not in case
- * of REQUEUE_PI.
- */
- if (!(curval & FUTEX_WAITER_REQUEUED))
- ret = -EDEADLK;
+ ret = -EDEADLK;
goto out_unlock_release_sem;
}
/*
- * Surprise - we got the lock. Just return
- * to userspace:
+ * Surprise - we got the lock. Just return to userspace:
*/
if (unlikely(!curval))
goto out_unlock_release_sem;
uval = curval;
+
/*
- * In case of a requeue, check if there already is an owner
- * If not, just take the futex.
+ * Set the WAITERS flag, so the owner will know it has someone
+ * to wake at next unlock
*/
- if ((curval & FUTEX_WAITER_REQUEUED) && !(curval & FUTEX_TID_MASK)) {
- /* set current as futex owner */
- newval = curval | current->pid;
- lock_held = 1;
- } else
- /* Set the WAITERS flag, so the owner will know it has someone
- to wake at next unlock */
- newval = curval | FUTEX_WAITERS;
+ newval = curval | FUTEX_WAITERS;
- pagefault_disable();
- curval = futex_atomic_cmpxchg_inatomic(uaddr, uval, newval);
- pagefault_enable();
+ /*
+ * There are two cases, where a futex might have no owner (the
+ * owner TID is 0): OWNER_DIED. We take over the futex in this
+ * case. We also do an unconditional take over, when the owner
+ * of the futex died.
+ *
+ * This is safe as we are protected by the hash bucket lock !
+ */
+ if (unlikely(ownerdied || !(curval & FUTEX_TID_MASK))) {
+ /* Keep the OWNER_DIED bit */
+ newval = (curval & ~FUTEX_TID_MASK) | current->pid;
+ ownerdied = 0;
+ lock_taken = 1;
+ }
+
+ curval = cmpxchg_futex_value_locked(uaddr, uval, newval);
if (unlikely(curval == -EFAULT))
goto uaddr_faulted;
if (unlikely(curval != uval))
goto retry_locked;
- if (lock_held) {
- set_pi_futex_owner(hb, &q.key, curr);
+ /*
+ * We took the lock due to owner died take over.
+ */
+ if (unlikely(lock_taken))
goto out_unlock_release_sem;
- }
/*
* We dont have the lock. Look up the PI state (or create it if
ret = lookup_pi_state(uval, hb, &q.key, &q.pi_state);
if (unlikely(ret)) {
- /*
- * There were no waiters and the owner task lookup
- * failed. When the OWNER_DIED bit is set, then we
- * know that this is a robust futex and we actually
- * take the lock. This is safe as we are protected by
- * the hash bucket lock. We also set the waiters bit
- * unconditionally here, to simplify glibc handling of
- * multiple tasks racing to acquire the lock and
- * cleanup the problems which were left by the dead
- * owner.
- */
- if (curval & FUTEX_OWNER_DIED) {
- uval = newval;
- newval = current->pid |
- FUTEX_OWNER_DIED | FUTEX_WAITERS;
+ switch (ret) {
- pagefault_disable();
- curval = futex_atomic_cmpxchg_inatomic(uaddr,
- uval, newval);
- pagefault_enable();
+ case -EAGAIN:
+ /*
+ * Task is exiting and we just wait for the
+ * exit to complete.
+ */
+ queue_unlock(&q, hb);
+ futex_unlock_mm(fshared);
+ cond_resched();
+ goto retry;
- if (unlikely(curval == -EFAULT))
+ case -ESRCH:
+ /*
+ * No owner found for this futex. Check if the
+ * OWNER_DIED bit is set to figure out whether
+ * this is a robust futex or not.
+ */
+ if (get_futex_value_locked(&curval, uaddr))
goto uaddr_faulted;
- if (unlikely(curval != uval))
+
+ /*
+ * We simply start over in case of a robust
+ * futex. The code above will take the futex
+ * and return happy.
+ */
+ if (curval & FUTEX_OWNER_DIED) {
+ ownerdied = 1;
goto retry_locked;
- ret = 0;
+ }
+ default:
+ goto out_unlock_release_sem;
}
- goto out_unlock_release_sem;
}
/*
* Now the futex is queued and we have checked the data, we
* don't want to hold mmap_sem while we sleep.
*/
- up_read(&curr->mm->mmap_sem);
+ futex_unlock_mm(fshared);
WARN_ON(!q.pi_state);
/*
ret = ret ? 0 : -EWOULDBLOCK;
}
- down_read(&curr->mm->mmap_sem);
+ futex_lock_mm(fshared);
spin_lock(q.lock_ptr);
- /*
- * Got the lock. We might not be the anticipated owner if we
- * did a lock-steal - fix up the PI-state in that case.
- */
- if (!ret && q.pi_state->owner != curr)
- /* mmap_sem is unlocked at return of this function */
- ret = fixup_pi_state_owner(uaddr, &q, hb, curr);
- else {
+ if (!ret) {
+ /*
+ * Got the lock. We might not be the anticipated owner
+ * if we did a lock-steal - fix up the PI-state in
+ * that case:
+ */
+ if (q.pi_state->owner != curr)
+ ret = fixup_pi_state_owner(uaddr, &q, curr);
+ } else {
/*
* Catch the rare case, where the lock was released
- * when we were on the way back before we locked
- * the hash bucket.
+ * when we were on the way back before we locked the
+ * hash bucket.
*/
- if (ret && q.pi_state->owner == curr) {
- if (rt_mutex_trylock(&q.pi_state->pi_mutex))
- ret = 0;
+ if (q.pi_state->owner == curr &&
+ rt_mutex_trylock(&q.pi_state->pi_mutex)) {
+ ret = 0;
+ } else {
+ /*
+ * Paranoia check. If we did not take the lock
+ * in the trylock above, then we should not be
+ * the owner of the rtmutex, neither the real
+ * nor the pending one:
+ */
+ if (rt_mutex_owner(&q.pi_state->pi_mutex) == curr)
+ printk(KERN_ERR "futex_lock_pi: ret = %d "
+ "pi-mutex: %p pi-state %p\n", ret,
+ q.pi_state->pi_mutex.owner,
+ q.pi_state->owner);
}
- /* Unqueue and drop the lock */
- unqueue_me_pi(&q);
- up_read(&curr->mm->mmap_sem);
}
- if (!detect && ret == -EDEADLK && 0)
- force_sig(SIGKILL, current);
+ /* Unqueue and drop the lock */
+ unqueue_me_pi(&q);
+ futex_unlock_mm(fshared);
return ret != -EINTR ? ret : -ERESTARTNOINTR;
queue_unlock(&q, hb);
out_release_sem:
- up_read(&curr->mm->mmap_sem);
+ futex_unlock_mm(fshared);
return ret;
uaddr_faulted:
* non-atomically. Therefore, if get_user below is not
* enough, we need to handle the fault ourselves, while
* still holding the mmap_sem.
+ *
+ * ... and hb->lock. :-) --ANK
*/
+ queue_unlock(&q, hb);
+
if (attempt++) {
- if (futex_handle_fault((unsigned long)uaddr, attempt)) {
- ret = -EFAULT;
- goto out_unlock_release_sem;
- }
- goto retry_locked;
+ ret = futex_handle_fault((unsigned long)uaddr, fshared,
+ attempt);
+ if (ret)
+ goto out_release_sem;
+ goto retry_unlocked;
}
- queue_unlock(&q, hb);
- up_read(&curr->mm->mmap_sem);
+ futex_unlock_mm(fshared);
ret = get_user(uval, uaddr);
if (!ret && (uval != -EFAULT))
* This is the in-kernel slowpath: we look up the PI state (if any),
* and do the rt-mutex unlock.
*/
-static int futex_unlock_pi(u32 __user *uaddr)
+static int futex_unlock_pi(u32 __user *uaddr, struct rw_semaphore *fshared)
{
struct futex_hash_bucket *hb;
struct futex_q *this, *next;
/*
* First take all the futex related locks:
*/
- down_read(¤t->mm->mmap_sem);
+ futex_lock_mm(fshared);
- ret = get_futex_key(uaddr, &key);
+ ret = get_futex_key(uaddr, fshared, &key);
if (unlikely(ret != 0))
goto out;
hb = hash_futex(&key);
+retry_unlocked:
spin_lock(&hb->lock);
-retry_locked:
/*
* To avoid races, try to do the TID -> 0 atomic transition
* again. If it succeeds then we can return without waking
* anyone else up:
*/
- if (!(uval & FUTEX_OWNER_DIED)) {
- pagefault_disable();
- uval = futex_atomic_cmpxchg_inatomic(uaddr, current->pid, 0);
- pagefault_enable();
- }
+ if (!(uval & FUTEX_OWNER_DIED))
+ uval = cmpxchg_futex_value_locked(uaddr, current->pid, 0);
+
if (unlikely(uval == -EFAULT))
goto pi_faulted;
out_unlock:
spin_unlock(&hb->lock);
out:
- up_read(¤t->mm->mmap_sem);
+ futex_unlock_mm(fshared);
return ret;
* non-atomically. Therefore, if get_user below is not
* enough, we need to handle the fault ourselves, while
* still holding the mmap_sem.
+ *
+ * ... and hb->lock. --ANK
*/
+ spin_unlock(&hb->lock);
+
if (attempt++) {
- if (futex_handle_fault((unsigned long)uaddr, attempt)) {
- ret = -EFAULT;
- goto out_unlock;
- }
- goto retry_locked;
+ ret = futex_handle_fault((unsigned long)uaddr, fshared,
+ attempt);
+ if (ret)
+ goto out;
+ goto retry_unlocked;
}
- spin_unlock(&hb->lock);
- up_read(¤t->mm->mmap_sem);
+ futex_unlock_mm(fshared);
ret = get_user(uval, uaddr);
if (!ret && (uval != -EFAULT))
struct futex_q *q;
struct file *filp;
int ret, err;
+ struct rw_semaphore *fshared;
static unsigned long printk_interval;
if (printk_timed_ratelimit(&printk_interval, 60 * 60 * 1000)) {
printk(KERN_WARNING "Process `%s' used FUTEX_FD, which "
- "will be removed from the kernel in June 2007\n",
- current->comm);
+ "will be removed from the kernel in June 2007\n",
+ current->comm);
}
ret = -EINVAL;
}
q->pi_state = NULL;
- down_read(¤t->mm->mmap_sem);
- err = get_futex_key(uaddr, &q->key);
+ fshared = ¤t->mm->mmap_sem;
+ down_read(fshared);
+ err = get_futex_key(uaddr, fshared, &q->key);
if (unlikely(err != 0)) {
- up_read(¤t->mm->mmap_sem);
+ up_read(fshared);
kfree(q);
goto error;
}
filp->private_data = q;
queue_me(q, ret, filp);
- up_read(¤t->mm->mmap_sem);
+ up_read(fshared);
/* Now we map fd to filp, so userspace can access it */
fd_install(ret, filp);
* userspace.
*/
mval = (uval & FUTEX_WAITERS) | FUTEX_OWNER_DIED;
- /* Also keep the FUTEX_WAITER_REQUEUED flag if set */
- mval |= (uval & FUTEX_WAITER_REQUEUED);
nval = futex_atomic_cmpxchg_inatomic(uaddr, uval, mval);
if (nval == -EFAULT)
* Wake robust non-PI futexes here. The wakeup of
* PI futexes happens in exit_pi_state():
*/
- if (!pi) {
- if (uval & FUTEX_WAITERS)
- futex_wake(uaddr, 1);
- }
+ if (!pi && (uval & FUTEX_WAITERS))
+ futex_wake(uaddr, &curr->mm->mmap_sem, 1);
}
return 0;
}
return;
if (pending)
- handle_futex_death((void __user *)pending + futex_offset, curr, pip);
+ handle_futex_death((void __user *)pending + futex_offset,
+ curr, pip);
while (entry != &head->list) {
/*
u32 __user *uaddr2, u32 val2, u32 val3)
{
int ret;
+ int cmd = op & FUTEX_CMD_MASK;
+ struct rw_semaphore *fshared = NULL;
- switch (op) {
+ if (!(op & FUTEX_PRIVATE_FLAG))
+ fshared = ¤t->mm->mmap_sem;
+
+ switch (cmd) {
case FUTEX_WAIT:
- ret = futex_wait(uaddr, val, timeout);
+ ret = futex_wait(uaddr, fshared, val, timeout);
break;
case FUTEX_WAKE:
- ret = futex_wake(uaddr, val);
+ ret = futex_wake(uaddr, fshared, val);
break;
case FUTEX_FD:
/* non-zero val means F_SETOWN(getpid()) & F_SETSIG(val) */
ret = futex_fd(uaddr, val);
break;
case FUTEX_REQUEUE:
- ret = futex_requeue(uaddr, uaddr2, val, val2, NULL);
+ ret = futex_requeue(uaddr, fshared, uaddr2, val, val2, NULL);
break;
case FUTEX_CMP_REQUEUE:
- ret = futex_requeue(uaddr, uaddr2, val, val2, &val3);
+ ret = futex_requeue(uaddr, fshared, uaddr2, val, val2, &val3);
break;
case FUTEX_WAKE_OP:
- ret = futex_wake_op(uaddr, uaddr2, val, val2, val3);
+ ret = futex_wake_op(uaddr, fshared, uaddr2, val, val2, val3);
break;
case FUTEX_LOCK_PI:
- ret = futex_lock_pi(uaddr, val, timeout, 0);
+ ret = futex_lock_pi(uaddr, fshared, val, timeout, 0);
break;
case FUTEX_UNLOCK_PI:
- ret = futex_unlock_pi(uaddr);
+ ret = futex_unlock_pi(uaddr, fshared);
break;
case FUTEX_TRYLOCK_PI:
- ret = futex_lock_pi(uaddr, 0, timeout, 1);
- break;
- case FUTEX_CMP_REQUEUE_PI:
- ret = futex_requeue_pi(uaddr, uaddr2, val, val2, &val3);
+ ret = futex_lock_pi(uaddr, fshared, 0, timeout, 1);
break;
default:
ret = -ENOSYS;
struct timespec ts;
ktime_t t, *tp = NULL;
u32 val2 = 0;
+ int cmd = op & FUTEX_CMD_MASK;
- if (utime && (op == FUTEX_WAIT || op == FUTEX_LOCK_PI)) {
+ if (utime && (cmd == FUTEX_WAIT || cmd == FUTEX_LOCK_PI)) {
if (copy_from_user(&ts, utime, sizeof(ts)) != 0)
return -EFAULT;
if (!timespec_valid(&ts))
return -EINVAL;
t = timespec_to_ktime(ts);
- if (op == FUTEX_WAIT)
+ if (cmd == FUTEX_WAIT)
t = ktime_add(ktime_get(), t);
tp = &t;
}
/*
- * requeue parameter in 'utime' if op == FUTEX_REQUEUE.
+ * requeue parameter in 'utime' if cmd == FUTEX_REQUEUE.
*/
- if (op == FUTEX_REQUEUE || op == FUTEX_CMP_REQUEUE
- || op == FUTEX_CMP_REQUEUE_PI)
+ if (cmd == FUTEX_REQUEUE || cmd == FUTEX_CMP_REQUEUE)
val2 = (u32) (unsigned long) utime;
return do_futex(uaddr, op, val, tp, uaddr2, val2, val3);
git.openpandora.org / pandora-kernel.git / blobdiff