#include <linux/bio.h>
#include <linux/blkdev.h>
#include <linux/bootmem.h> /* for max_pfn/max_low_pfn */
+#include <linux/slab.h>
#include "blk.h"
*/
static struct kmem_cache *iocontext_cachep;
-static void cfq_dtor(struct io_context *ioc)
+/**
+ * get_io_context - increment reference count to io_context
+ * @ioc: io_context to get
+ *
+ * Increment reference count to @ioc.
+ */
+void get_io_context(struct io_context *ioc)
{
- if (!hlist_empty(&ioc->cic_list)) {
- struct cfq_io_context *cic;
+ BUG_ON(atomic_long_read(&ioc->refcount) <= 0);
+ atomic_long_inc(&ioc->refcount);
+}
+EXPORT_SYMBOL(get_io_context);
- cic = list_entry(ioc->cic_list.first, struct cfq_io_context,
- cic_list);
- cic->dtor(ioc);
- }
+static void icq_free_icq_rcu(struct rcu_head *head)
+{
+ struct io_cq *icq = container_of(head, struct io_cq, __rcu_head);
+
+ kmem_cache_free(icq->__rcu_icq_cache, icq);
+}
+
+/* Exit an icq. Called with both ioc and q locked. */
+static void ioc_exit_icq(struct io_cq *icq)
+{
+ struct elevator_type *et = icq->q->elevator->type;
+
+ if (icq->flags & ICQ_EXITED)
+ return;
+
+ if (et->ops.elevator_exit_icq_fn)
+ et->ops.elevator_exit_icq_fn(icq);
+
+ icq->flags |= ICQ_EXITED;
+}
+
+/* Release an icq. Called with both ioc and q locked. */
+static void ioc_destroy_icq(struct io_cq *icq)
+{
+ struct io_context *ioc = icq->ioc;
+ struct request_queue *q = icq->q;
+ struct elevator_type *et = q->elevator->type;
+
+ lockdep_assert_held(&ioc->lock);
+ lockdep_assert_held(q->queue_lock);
+
+ radix_tree_delete(&ioc->icq_tree, icq->q->id);
+ hlist_del_init(&icq->ioc_node);
+ list_del_init(&icq->q_node);
+
+ /*
+ * Both setting lookup hint to and clearing it from @icq are done
+ * under queue_lock. If it's not pointing to @icq now, it never
+ * will. Hint assignment itself can race safely.
+ */
+ if (rcu_dereference_raw(ioc->icq_hint) == icq)
+ rcu_assign_pointer(ioc->icq_hint, NULL);
+
+ ioc_exit_icq(icq);
+
+ /*
+ * @icq->q might have gone away by the time RCU callback runs
+ * making it impossible to determine icq_cache. Record it in @icq.
+ */
+ icq->__rcu_icq_cache = et->icq_cache;
+ call_rcu(&icq->__rcu_head, icq_free_icq_rcu);
}
/*
- * IO Context helper functions. put_io_context() returns 1 if there are no
- * more users of this io context, 0 otherwise.
+ * Slow path for ioc release in put_io_context(). Performs double-lock
+ * dancing to unlink all icq's and then frees ioc.
*/
-int put_io_context(struct io_context *ioc)
+static void ioc_release_fn(struct work_struct *work)
{
- if (ioc == NULL)
- return 1;
+ struct io_context *ioc = container_of(work, struct io_context,
+ release_work);
+ unsigned long flags;
- BUG_ON(atomic_long_read(&ioc->refcount) == 0);
+ /*
+ * Exiting icq may call into put_io_context() through elevator
+ * which will trigger lockdep warning. The ioc's are guaranteed to
+ * be different, use a different locking subclass here. Use
+ * irqsave variant as there's no spin_lock_irq_nested().
+ */
+ spin_lock_irqsave_nested(&ioc->lock, flags, 1);
- if (atomic_long_dec_and_test(&ioc->refcount)) {
- rcu_read_lock();
- cfq_dtor(ioc);
- rcu_read_unlock();
+ while (!hlist_empty(&ioc->icq_list)) {
+ struct io_cq *icq = hlist_entry(ioc->icq_list.first,
+ struct io_cq, ioc_node);
+ struct request_queue *q = icq->q;
- kmem_cache_free(iocontext_cachep, ioc);
- return 1;
+ if (spin_trylock(q->queue_lock)) {
+ ioc_destroy_icq(icq);
+ spin_unlock(q->queue_lock);
+ } else {
+ spin_unlock_irqrestore(&ioc->lock, flags);
+ cpu_relax();
+ spin_lock_irqsave_nested(&ioc->lock, flags, 1);
+ }
}
- return 0;
+
+ spin_unlock_irqrestore(&ioc->lock, flags);
+
+ kmem_cache_free(iocontext_cachep, ioc);
}
-EXPORT_SYMBOL(put_io_context);
-static void cfq_exit(struct io_context *ioc)
+/**
+ * put_io_context - put a reference of io_context
+ * @ioc: io_context to put
+ *
+ * Decrement reference count of @ioc and release it if the count reaches
+ * zero.
+ */
+void put_io_context(struct io_context *ioc)
{
- rcu_read_lock();
+ unsigned long flags;
+ bool free_ioc = false;
+
+ if (ioc == NULL)
+ return;
- if (!hlist_empty(&ioc->cic_list)) {
- struct cfq_io_context *cic;
+ BUG_ON(atomic_long_read(&ioc->refcount) <= 0);
- cic = list_entry(ioc->cic_list.first, struct cfq_io_context,
- cic_list);
- cic->exit(ioc);
+ /*
+ * Releasing ioc requires reverse order double locking and we may
+ * already be holding a queue_lock. Do it asynchronously from wq.
+ */
+ if (atomic_long_dec_and_test(&ioc->refcount)) {
+ spin_lock_irqsave(&ioc->lock, flags);
+ if (!hlist_empty(&ioc->icq_list))
+ schedule_work(&ioc->release_work);
+ else
+ free_ioc = true;
+ spin_unlock_irqrestore(&ioc->lock, flags);
}
- rcu_read_unlock();
+
+ if (free_ioc)
+ kmem_cache_free(iocontext_cachep, ioc);
}
+EXPORT_SYMBOL(put_io_context);
-/* Called by the exitting task */
+/* Called by the exiting task */
void exit_io_context(struct task_struct *task)
{
struct io_context *ioc;
+ struct io_cq *icq;
+ struct hlist_node *n;
+ unsigned long flags;
task_lock(task);
ioc = task->io_context;
task->io_context = NULL;
task_unlock(task);
- if (atomic_dec_and_test(&ioc->nr_tasks)) {
- cfq_exit(ioc);
+ if (!atomic_dec_and_test(&ioc->nr_tasks)) {
+ put_io_context(ioc);
+ return;
+ }
+ /*
+ * Need ioc lock to walk icq_list and q lock to exit icq. Perform
+ * reverse double locking. Read comment in ioc_release_fn() for
+ * explanation on the nested locking annotation.
+ */
+retry:
+ spin_lock_irqsave_nested(&ioc->lock, flags, 1);
+ hlist_for_each_entry(icq, n, &ioc->icq_list, ioc_node) {
+ if (icq->flags & ICQ_EXITED)
+ continue;
+ if (spin_trylock(icq->q->queue_lock)) {
+ ioc_exit_icq(icq);
+ spin_unlock(icq->q->queue_lock);
+ } else {
+ spin_unlock_irqrestore(&ioc->lock, flags);
+ cpu_relax();
+ goto retry;
+ }
}
+ spin_unlock_irqrestore(&ioc->lock, flags);
+
put_io_context(ioc);
}
-struct io_context *alloc_io_context(gfp_t gfp_flags, int node)
+/**
+ * ioc_clear_queue - break any ioc association with the specified queue
+ * @q: request_queue being cleared
+ *
+ * Walk @q->icq_list and exit all io_cq's. Must be called with @q locked.
+ */
+void ioc_clear_queue(struct request_queue *q)
{
- struct io_context *ret;
+ lockdep_assert_held(q->queue_lock);
- ret = kmem_cache_alloc_node(iocontext_cachep, gfp_flags, node);
- if (ret) {
- atomic_long_set(&ret->refcount, 1);
- atomic_set(&ret->nr_tasks, 1);
- spin_lock_init(&ret->lock);
- ret->ioprio_changed = 0;
- ret->ioprio = 0;
- ret->last_waited = 0; /* doesn't matter... */
- ret->nr_batch_requests = 0; /* because this is 0 */
- INIT_RADIX_TREE(&ret->radix_root, GFP_ATOMIC | __GFP_HIGH);
- INIT_HLIST_HEAD(&ret->cic_list);
- ret->ioc_data = NULL;
+ while (!list_empty(&q->icq_list)) {
+ struct io_cq *icq = list_entry(q->icq_list.next,
+ struct io_cq, q_node);
+ struct io_context *ioc = icq->ioc;
+
+ spin_lock(&ioc->lock);
+ ioc_destroy_icq(icq);
+ spin_unlock(&ioc->lock);
}
+}
- return ret;
+void create_io_context_slowpath(struct task_struct *task, gfp_t gfp_flags,
+ int node)
+{
+ struct io_context *ioc;
+
+ ioc = kmem_cache_alloc_node(iocontext_cachep, gfp_flags | __GFP_ZERO,
+ node);
+ if (unlikely(!ioc))
+ return;
+
+ /* initialize */
+ atomic_long_set(&ioc->refcount, 1);
+ atomic_set(&ioc->nr_tasks, 1);
+ spin_lock_init(&ioc->lock);
+ INIT_RADIX_TREE(&ioc->icq_tree, GFP_ATOMIC | __GFP_HIGH);
+ INIT_HLIST_HEAD(&ioc->icq_list);
+ INIT_WORK(&ioc->release_work, ioc_release_fn);
+
+ /*
+ * Try to install. ioc shouldn't be installed if someone else
+ * already did or @task, which isn't %current, is exiting. Note
+ * that we need to allow ioc creation on exiting %current as exit
+ * path may issue IOs from e.g. exit_files(). The exit path is
+ * responsible for not issuing IO after exit_io_context().
+ */
+ task_lock(task);
+ if (!task->io_context &&
+ (task == current || !(task->flags & PF_EXITING)))
+ task->io_context = ioc;
+ else
+ kmem_cache_free(iocontext_cachep, ioc);
+ task_unlock(task);
}
-/*
- * If the current task has no IO context then create one and initialise it.
- * Otherwise, return its existing IO context.
+/**
+ * get_task_io_context - get io_context of a task
+ * @task: task of interest
+ * @gfp_flags: allocation flags, used if allocation is necessary
+ * @node: allocation node, used if allocation is necessary
+ *
+ * Return io_context of @task. If it doesn't exist, it is created with
+ * @gfp_flags and @node. The returned io_context has its reference count
+ * incremented.
*
- * This returned IO context doesn't have a specifically elevated refcount,
- * but since the current task itself holds a reference, the context can be
- * used in general code, so long as it stays within `current` context.
+ * This function always goes through task_lock() and it's better to use
+ * %current->io_context + get_io_context() for %current.
*/
-struct io_context *current_io_context(gfp_t gfp_flags, int node)
+struct io_context *get_task_io_context(struct task_struct *task,
+ gfp_t gfp_flags, int node)
{
- struct task_struct *tsk = current;
- struct io_context *ret;
+ struct io_context *ioc;
- ret = tsk->io_context;
- if (likely(ret))
- return ret;
+ might_sleep_if(gfp_flags & __GFP_WAIT);
- ret = alloc_io_context(gfp_flags, node);
- if (ret) {
- /* make sure set_task_ioprio() sees the settings above */
- smp_wmb();
- tsk->io_context = ret;
- }
+ do {
+ task_lock(task);
+ ioc = task->io_context;
+ if (likely(ioc)) {
+ get_io_context(ioc);
+ task_unlock(task);
+ return ioc;
+ }
+ task_unlock(task);
+ } while (create_io_context(task, gfp_flags, node));
- return ret;
+ return NULL;
}
+EXPORT_SYMBOL(get_task_io_context);
-/*
- * If the current task has no IO context then create one and initialise it.
- * If it does have a context, take a ref on it.
+/**
+ * ioc_lookup_icq - lookup io_cq from ioc
+ * @ioc: the associated io_context
+ * @q: the associated request_queue
*
- * This is always called in the context of the task which submitted the I/O.
+ * Look up io_cq associated with @ioc - @q pair from @ioc. Must be called
+ * with @q->queue_lock held.
*/
-struct io_context *get_io_context(gfp_t gfp_flags, int node)
+struct io_cq *ioc_lookup_icq(struct io_context *ioc, struct request_queue *q)
{
- struct io_context *ret = NULL;
+ struct io_cq *icq;
+
+ lockdep_assert_held(q->queue_lock);
/*
- * Check for unlikely race with exiting task. ioc ref count is
- * zero when ioc is being detached.
+ * icq's are indexed from @ioc using radix tree and hint pointer,
+ * both of which are protected with RCU. All removals are done
+ * holding both q and ioc locks, and we're holding q lock - if we
+ * find a icq which points to us, it's guaranteed to be valid.
*/
- do {
- ret = current_io_context(gfp_flags, node);
- if (unlikely(!ret))
- break;
- } while (!atomic_long_inc_not_zero(&ret->refcount));
+ rcu_read_lock();
+ icq = rcu_dereference(ioc->icq_hint);
+ if (icq && icq->q == q)
+ goto out;
- return ret;
+ icq = radix_tree_lookup(&ioc->icq_tree, q->id);
+ if (icq && icq->q == q)
+ rcu_assign_pointer(ioc->icq_hint, icq); /* allowed to race */
+ else
+ icq = NULL;
+out:
+ rcu_read_unlock();
+ return icq;
}
-EXPORT_SYMBOL(get_io_context);
+EXPORT_SYMBOL(ioc_lookup_icq);
+
+/**
+ * ioc_create_icq - create and link io_cq
+ * @q: request_queue of interest
+ * @gfp_mask: allocation mask
+ *
+ * Make sure io_cq linking %current->io_context and @q exists. If either
+ * io_context and/or icq don't exist, they will be created using @gfp_mask.
+ *
+ * The caller is responsible for ensuring @ioc won't go away and @q is
+ * alive and will stay alive until this function returns.
+ */
+struct io_cq *ioc_create_icq(struct request_queue *q, gfp_t gfp_mask)
+{
+ struct elevator_type *et = q->elevator->type;
+ struct io_context *ioc;
+ struct io_cq *icq;
+
+ /* allocate stuff */
+ ioc = create_io_context(current, gfp_mask, q->node);
+ if (!ioc)
+ return NULL;
-void copy_io_context(struct io_context **pdst, struct io_context **psrc)
+ icq = kmem_cache_alloc_node(et->icq_cache, gfp_mask | __GFP_ZERO,
+ q->node);
+ if (!icq)
+ return NULL;
+
+ if (radix_tree_preload(gfp_mask) < 0) {
+ kmem_cache_free(et->icq_cache, icq);
+ return NULL;
+ }
+
+ icq->ioc = ioc;
+ icq->q = q;
+ INIT_LIST_HEAD(&icq->q_node);
+ INIT_HLIST_NODE(&icq->ioc_node);
+
+ /* lock both q and ioc and try to link @icq */
+ spin_lock_irq(q->queue_lock);
+ spin_lock(&ioc->lock);
+
+ if (likely(!radix_tree_insert(&ioc->icq_tree, q->id, icq))) {
+ hlist_add_head(&icq->ioc_node, &ioc->icq_list);
+ list_add(&icq->q_node, &q->icq_list);
+ if (et->ops.elevator_init_icq_fn)
+ et->ops.elevator_init_icq_fn(icq);
+ } else {
+ kmem_cache_free(et->icq_cache, icq);
+ icq = ioc_lookup_icq(ioc, q);
+ if (!icq)
+ printk(KERN_ERR "cfq: icq link failed!\n");
+ }
+
+ spin_unlock(&ioc->lock);
+ spin_unlock_irq(q->queue_lock);
+ radix_tree_preload_end();
+ return icq;
+}
+
+void ioc_set_icq_flags(struct io_context *ioc, unsigned int flags)
+{
+ struct io_cq *icq;
+ struct hlist_node *n;
+
+ hlist_for_each_entry(icq, n, &ioc->icq_list, ioc_node)
+ icq->flags |= flags;
+}
+
+/**
+ * ioc_ioprio_changed - notify ioprio change
+ * @ioc: io_context of interest
+ * @ioprio: new ioprio
+ *
+ * @ioc's ioprio has changed to @ioprio. Set %ICQ_IOPRIO_CHANGED for all
+ * icq's. iosched is responsible for checking the bit and applying it on
+ * request issue path.
+ */
+void ioc_ioprio_changed(struct io_context *ioc, int ioprio)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&ioc->lock, flags);
+ ioc->ioprio = ioprio;
+ ioc_set_icq_flags(ioc, ICQ_IOPRIO_CHANGED);
+ spin_unlock_irqrestore(&ioc->lock, flags);
+}
+
+/**
+ * ioc_cgroup_changed - notify cgroup change
+ * @ioc: io_context of interest
+ *
+ * @ioc's cgroup has changed. Set %ICQ_CGROUP_CHANGED for all icq's.
+ * iosched is responsible for checking the bit and applying it on request
+ * issue path.
+ */
+void ioc_cgroup_changed(struct io_context *ioc)
+{
+ unsigned long flags;
+
+ spin_lock_irqsave(&ioc->lock, flags);
+ ioc_set_icq_flags(ioc, ICQ_CGROUP_CHANGED);
+ spin_unlock_irqrestore(&ioc->lock, flags);
+}
+EXPORT_SYMBOL(ioc_cgroup_changed);
+
+/**
+ * icq_get_changed - fetch and clear icq changed mask
+ * @icq: icq of interest
+ *
+ * Fetch and clear ICQ_*_CHANGED bits from @icq. Grabs and releases
+ * @icq->ioc->lock.
+ */
+unsigned icq_get_changed(struct io_cq *icq)
{
- struct io_context *src = *psrc;
- struct io_context *dst = *pdst;
+ unsigned int changed = 0;
+ unsigned long flags;
- if (src) {
- BUG_ON(atomic_long_read(&src->refcount) == 0);
- atomic_long_inc(&src->refcount);
- put_io_context(dst);
- *pdst = src;
+ if (unlikely(icq->flags & ICQ_CHANGED_MASK)) {
+ spin_lock_irqsave(&icq->ioc->lock, flags);
+ changed = icq->flags & ICQ_CHANGED_MASK;
+ icq->flags &= ~ICQ_CHANGED_MASK;
+ spin_unlock_irqrestore(&icq->ioc->lock, flags);
}
+ return changed;
}
-EXPORT_SYMBOL(copy_io_context);
+EXPORT_SYMBOL(icq_get_changed);
static int __init blk_ioc_init(void)
{
projects / linux-flexiantxendom0-3.2.10.git / blobdiff