#endif
/*------ sysctl variables----*/
-atomic_t aio_nr = ATOMIC_INIT(0); /* current system wide number of aio requests */
-unsigned aio_max_nr = 0x10000; /* system wide maximum number of aio requests */
+static DEFINE_SPINLOCK(aio_nr_lock);
+unsigned long aio_nr; /* current system wide number of aio requests */
+unsigned long aio_max_nr = 0x10000; /* system wide maximum number of aio requests */
/*----end sysctl variables---*/
static kmem_cache_t *kiocb_cachep;
return ERR_PTR(-EINVAL);
}
- if (nr_events > aio_max_nr)
+ if ((unsigned long)nr_events > aio_max_nr)
return ERR_PTR(-EAGAIN);
ctx = kmem_cache_alloc(kioctx_cachep, GFP_KERNEL);
goto out_freectx;
/* limit the number of system wide aios */
- atomic_add(ctx->max_reqs, &aio_nr); /* undone by __put_ioctx */
- if (unlikely(atomic_read(&aio_nr) > aio_max_nr))
+ spin_lock(&aio_nr_lock);
+ if (aio_nr + ctx->max_reqs > aio_max_nr ||
+ aio_nr + ctx->max_reqs < aio_nr)
+ ctx->max_reqs = 0;
+ else
+ aio_nr += ctx->max_reqs;
+ spin_unlock(&aio_nr_lock);
+ if (ctx->max_reqs == 0)
goto out_cleanup;
/* now link into global list. kludge. FIXME */
return ctx;
out_cleanup:
- atomic_sub(ctx->max_reqs, &aio_nr);
- ctx->max_reqs = 0; /* prevent __put_ioctx from sub'ing aio_nr */
__put_ioctx(ctx);
return ERR_PTR(-EAGAIN);
pr_debug("__put_ioctx: freeing %p\n", ctx);
kmem_cache_free(kioctx_cachep, ctx);
- atomic_sub(nr_events, &aio_nr);
+ if (nr_events) {
+ spin_lock(&aio_nr_lock);
+ BUG_ON(aio_nr - nr_events > aio_nr);
+ aio_nr -= nr_events;
+ spin_unlock(&aio_nr_lock);
+ }
}
/* aio_get_req
if (unlikely(!req))
return NULL;
- req->ki_flags = 1 << KIF_LOCKED;
+ req->ki_flags = 0;
req->ki_users = 2;
req->ki_key = 0;
req->ki_ctx = ctx;
static inline void really_put_req(struct kioctx *ctx, struct kiocb *req)
{
+ assert_spin_locked(&ctx->ctx_lock);
+
if (req->ki_dtor)
req->ki_dtor(req);
kmem_cache_free(kiocb_cachep, req);
dprintk(KERN_DEBUG "aio_put(%p): f_count=%d\n",
req, atomic_read(&req->ki_filp->f_count));
+ assert_spin_locked(&ctx->ctx_lock);
+
req->ki_users --;
if (unlikely(req->ki_users < 0))
BUG();
return ioctx;
}
-static int lock_kiocb_action(void *param)
-{
- schedule();
- return 0;
-}
-
-static inline void lock_kiocb(struct kiocb *iocb)
-{
- wait_on_bit_lock(&iocb->ki_flags, KIF_LOCKED, lock_kiocb_action,
- TASK_UNINTERRUPTIBLE);
-}
-
-static inline void unlock_kiocb(struct kiocb *iocb)
-{
- kiocbClearLocked(iocb);
- smp_mb__after_clear_bit();
- wake_up_bit(&iocb->ki_flags, KIF_LOCKED);
-}
-
/*
* use_mm
* Makes the calling kernel thread take on the specified
* the kiocb (to tell the caller to activate the work
* queue to process it), or 0, if it found that it was
* already queued.
- *
- * Should be called with the spin lock iocb->ki_ctx->ctx_lock
- * held
*/
static inline int __queue_kicked_iocb(struct kiocb *iocb)
{
struct kioctx *ctx = iocb->ki_ctx;
+ assert_spin_locked(&ctx->ctx_lock);
+
if (list_empty(&iocb->ki_run_list)) {
list_add_tail(&iocb->ki_run_list,
&ctx->run_list);
* Process all pending retries queued on the ioctx
* run list.
* Assumes it is operating within the aio issuer's mm
- * context. Expects to be called with ctx->ctx_lock held
+ * context.
*/
static int __aio_run_iocbs(struct kioctx *ctx)
{
struct kiocb *iocb;
LIST_HEAD(run_list);
+ assert_spin_locked(&ctx->ctx_lock);
+
list_splice_init(&ctx->run_list, &run_list);
while (!list_empty(&run_list)) {
iocb = list_entry(run_list.next, struct kiocb,
* Hold an extra reference while retrying i/o.
*/
iocb->ki_users++; /* grab extra reference */
- lock_kiocb(iocb);
aio_run_iocb(iocb);
- unlock_kiocb(iocb);
if (__aio_put_req(ctx, iocb)) /* drop extra ref */
put_ioctx(ctx);
}
unsigned long tail;
int ret;
- /* Special case handling for sync iocbs: events go directly
- * into the iocb for fast handling. Note that this will not
- * work if we allow sync kiocbs to be cancelled. in which
- * case the usage count checks will have to move under ctx_lock
- * for all cases.
+ /*
+ * Special case handling for sync iocbs:
+ * - events go directly into the iocb for fast handling
+ * - the sync task with the iocb in its stack holds the single iocb
+ * ref, no other paths have a way to get another ref
+ * - the sync task helpfully left a reference to itself in the iocb
*/
if (is_sync_kiocb(iocb)) {
- int ret;
-
+ BUG_ON(iocb->ki_users != 1);
iocb->ki_user_data = res;
- if (iocb->ki_users == 1) {
- iocb->ki_users = 0;
- ret = 1;
- } else {
- spin_lock_irq(&ctx->ctx_lock);
- iocb->ki_users--;
- ret = (0 == iocb->ki_users);
- spin_unlock_irq(&ctx->ctx_lock);
- }
- /* sync iocbs put the task here for us */
+ iocb->ki_users = 0;
wake_up_process(iocb->ki_obj.tsk);
- return ret;
+ return 1;
}
info = &ctx->ring_info;
goto out;
ret = -EINVAL;
- if (unlikely(ctx || (int)nr_events <= 0)) {
- pr_debug("EINVAL: io_setup: ctx or nr_events > max\n");
+ if (unlikely(ctx || nr_events == 0)) {
+ pr_debug("EINVAL: io_setup: ctx %lu nr_events %u\n",
+ ctx, nr_events);
goto out;
}
if (unlikely(!access_ok(VERIFY_WRITE, kiocb->ki_buf,
kiocb->ki_left)))
break;
+ ret = security_file_permission(file, MAY_READ);
+ if (unlikely(ret))
+ break;
ret = -EINVAL;
if (file->f_op->aio_read)
kiocb->ki_retry = aio_pread;
if (unlikely(!access_ok(VERIFY_READ, kiocb->ki_buf,
kiocb->ki_left)))
break;
+ ret = security_file_permission(file, MAY_WRITE);
+ if (unlikely(ret))
+ break;
ret = -EINVAL;
if (file->f_op->aio_write)
kiocb->ki_retry = aio_pwrite;
spin_lock_irq(&ctx->ctx_lock);
aio_run_iocb(req);
- unlock_kiocb(req);
if (!list_empty(&ctx->run_list)) {
/* drain the run list */
while (__aio_run_iocbs(ctx))
/* lookup_kiocb
* Finds a given iocb for cancellation.
- * MUST be called with ctx->ctx_lock held.
*/
static struct kiocb *lookup_kiocb(struct kioctx *ctx, struct iocb __user *iocb,
u32 key)
{
struct list_head *pos;
+
+ assert_spin_locked(&ctx->ctx_lock);
+
/* TODO: use a hash or array, this sucks. */
list_for_each(pos, &ctx->active_reqs) {
struct kiocb *kiocb = list_kiocb(pos);
if (NULL != cancel) {
struct io_event tmp;
pr_debug("calling cancel\n");
- lock_kiocb(kiocb);
memset(&tmp, 0, sizeof(tmp));
tmp.obj = (u64)(unsigned long)kiocb->ki_obj.user;
tmp.data = kiocb->ki_user_data;
if (copy_to_user(result, &tmp, sizeof(tmp)))
ret = -EFAULT;
}
- unlock_kiocb(kiocb);
} else
ret = -EINVAL;