{
struct blk_queue_tag *bqt = q->queue_tags;
- if (unlikely(bqt == NULL || tag >= bqt->max_depth))
+ if (unlikely(bqt == NULL || tag >= bqt->real_max_depth))
return NULL;
return bqt->tag_index[tag];
memset(tag_index, 0, depth * sizeof(struct request *));
memset(tag_map, 0, nr_ulongs * sizeof(unsigned long));
+ tags->real_max_depth = depth;
tags->max_depth = depth;
tags->tag_index = tag_index;
tags->tag_map = tag_map;
if (!bqt)
return -ENXIO;
+ /*
+ * if we already have large enough real_max_depth. just
+ * adjust max_depth. *NOTE* as requests with tag value
+ * between new_depth and real_max_depth can be in-flight, tag
+ * map can not be shrunk blindly here.
+ */
+ if (new_depth <= bqt->real_max_depth) {
+ bqt->max_depth = new_depth;
+ return 0;
+ }
+
/*
* save the old state info, so we can copy it back
*/
tag_index = bqt->tag_index;
tag_map = bqt->tag_map;
- max_depth = bqt->max_depth;
+ max_depth = bqt->real_max_depth;
if (init_tag_map(q, bqt, new_depth))
return -ENOMEM;
BUG_ON(tag == -1);
- if (unlikely(tag >= bqt->max_depth))
+ if (unlikely(tag >= bqt->real_max_depth))
/*
* This can happen after tag depth has been reduced.
* FIXME: how about a warning or info message here?
#define blkdev_free_rq(list) list_entry((list)->next, struct request, queuelist)
/*
- * Get a free request, queue_lock must not be held
+ * Get a free request, queue_lock must be held.
+ * Returns NULL on failure, with queue_lock held.
+ * Returns !NULL on success, with queue_lock *not held*.
*/
static struct request *get_request(request_queue_t *q, int rw, struct bio *bio,
int gfp_mask)
{
struct request *rq = NULL;
struct request_list *rl = &q->rq;
- struct io_context *ioc = get_io_context(gfp_mask);
+ struct io_context *ioc = current_io_context(GFP_ATOMIC);
if (unlikely(test_bit(QUEUE_FLAG_DRAIN, &q->queue_flags)))
goto out;
- spin_lock_irq(q->queue_lock);
if (rl->count[rw]+1 >= q->nr_requests) {
/*
* The queue will fill after this allocation, so set it as
* The queue is full and the allocating process is not a
* "batcher", and not exempted by the IO scheduler
*/
- spin_unlock_irq(q->queue_lock);
goto out;
}
get_rq:
+ /*
+ * Only allow batching queuers to allocate up to 50% over the defined
+ * limit of requests, otherwise we could have thousands of requests
+ * allocated with any setting of ->nr_requests
+ */
+ if (rl->count[rw] >= (3 * q->nr_requests / 2))
+ goto out;
+
rl->count[rw]++;
rl->starved[rw] = 0;
if (rl->count[rw] >= queue_congestion_on_threshold(q))
if (unlikely(rl->count[rw] == 0))
rl->starved[rw] = 1;
- spin_unlock_irq(q->queue_lock);
goto out;
}
rq_init(q, rq);
rq->rl = rl;
out:
- put_io_context(ioc);
return rq;
}
/*
* No available requests for this queue, unplug the device and wait for some
* requests to become available.
+ *
+ * Called with q->queue_lock held, and returns with it unlocked.
*/
static struct request *get_request_wait(request_queue_t *q, int rw,
struct bio *bio)
{
- DEFINE_WAIT(wait);
struct request *rq;
- do {
+ rq = get_request(q, rw, bio, GFP_NOIO);
+ while (!rq) {
+ DEFINE_WAIT(wait);
struct request_list *rl = &q->rq;
prepare_to_wait_exclusive(&rl->wait[rw], &wait,
if (!rq) {
struct io_context *ioc;
- generic_unplug_device(q);
+ __generic_unplug_device(q);
+ spin_unlock_irq(q->queue_lock);
io_schedule();
/*
* up to a big batch of them for a small period time.
* See ioc_batching, ioc_set_batching
*/
- ioc = get_io_context(GFP_NOIO);
+ ioc = current_io_context(GFP_NOIO);
ioc_set_batching(q, ioc);
- put_io_context(ioc);
+
+ spin_lock_irq(q->queue_lock);
}
finish_wait(&rl->wait[rw], &wait);
- } while (!rq);
+ }
return rq;
}
BUG_ON(rw != READ && rw != WRITE);
- if (gfp_mask & __GFP_WAIT)
+ spin_lock_irq(q->queue_lock);
+ if (gfp_mask & __GFP_WAIT) {
rq = get_request_wait(q, rw, NULL);
- else
+ } else {
rq = get_request(q, rw, NULL, gfp_mask);
+ if (!rq)
+ spin_unlock_irq(q->queue_lock);
+ }
+ /* q->queue_lock is unlocked at this point */
return rq;
}
-
EXPORT_SYMBOL(blk_get_request);
/**
static int __make_request(request_queue_t *q, struct bio *bio)
{
- struct request *req, *freereq = NULL;
+ struct request *req;
int el_ret, rw, nr_sectors, cur_nr_sectors, barrier, err, sync;
unsigned short prio;
sector_t sector;
goto end_io;
}
-again:
spin_lock_irq(q->queue_lock);
- if (elv_queue_empty(q)) {
- blk_plug_device(q);
- goto get_rq;
- }
- if (barrier)
+ if (unlikely(barrier) || elv_queue_empty(q))
goto get_rq;
el_ret = elv_merge(q, &req, bio);
elv_merged_request(q, req);
goto out;
- /*
- * elevator says don't/can't merge. get new request
- */
- case ELEVATOR_NO_MERGE:
- break;
-
+ /* ELV_NO_MERGE: elevator says don't/can't merge. */
default:
- printk("elevator returned crap (%d)\n", el_ret);
- BUG();
+ ;
}
+get_rq:
/*
- * Grab a free request from the freelist - if that is empty, check
- * if we are doing read ahead and abort instead of blocking for
- * a free slot.
+ * Grab a free request. This is might sleep but can not fail.
+ * Returns with the queue unlocked.
+ */
+ req = get_request_wait(q, rw, bio);
+
+ /*
+ * After dropping the lock and possibly sleeping here, our request
+ * may now be mergeable after it had proven unmergeable (above).
+ * We don't worry about that case for efficiency. It won't happen
+ * often, and the elevators are able to handle it.
*/
-get_rq:
- if (freereq) {
- req = freereq;
- freereq = NULL;
- } else {
- spin_unlock_irq(q->queue_lock);
- if ((freereq = get_request(q, rw, bio, GFP_ATOMIC)) == NULL) {
- /*
- * READA bit set
- */
- err = -EWOULDBLOCK;
- if (bio_rw_ahead(bio))
- goto end_io;
-
- freereq = get_request_wait(q, rw, bio);
- }
- goto again;
- }
req->flags |= REQ_CMD;
req->rq_disk = bio->bi_bdev->bd_disk;
req->start_time = jiffies;
+ spin_lock_irq(q->queue_lock);
+ if (elv_queue_empty(q))
+ blk_plug_device(q);
add_request(q, req);
out:
- if (freereq)
- __blk_put_request(q, freereq);
if (sync)
__generic_unplug_device(q);
/*
* 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.
+ * Otherwise, return its existing IO context.
*
- * This is always called in the context of the task which submitted the I/O.
- * But weird things happen, so we disable local interrupts to ensure exclusive
- * access to *current.
+ * 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.
*/
-struct io_context *get_io_context(int gfp_flags)
+struct io_context *current_io_context(int gfp_flags)
{
struct task_struct *tsk = current;
- unsigned long flags;
struct io_context *ret;
- local_irq_save(flags);
ret = tsk->io_context;
- if (ret)
- goto out;
-
- local_irq_restore(flags);
+ if (likely(ret))
+ return ret;
ret = kmem_cache_alloc(iocontext_cachep, gfp_flags);
if (ret) {
ret->nr_batch_requests = 0; /* because this is 0 */
ret->aic = NULL;
ret->cic = NULL;
+ tsk->io_context = ret;
+ }
- local_irq_save(flags);
-
- /*
- * very unlikely, someone raced with us in setting up the task
- * io context. free new context and just grab a reference.
- */
- if (!tsk->io_context)
- tsk->io_context = ret;
- else {
- kmem_cache_free(iocontext_cachep, ret);
- ret = tsk->io_context;
- }
+ return ret;
+}
+EXPORT_SYMBOL(current_io_context);
-out:
+/*
+ * 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.
+ *
+ * This is always called in the context of the task which submitted the I/O.
+ */
+struct io_context *get_io_context(int gfp_flags)
+{
+ struct io_context *ret;
+ ret = current_io_context(gfp_flags);
+ if (likely(ret))
atomic_inc(&ret->refcount);
- local_irq_restore(flags);
- }
-
return ret;
}
EXPORT_SYMBOL(get_io_context);
projects / firefly-linux-kernel-4.4.55.git / blobdiff