summaryrefslogtreecommitdiff
path: root/io_uring/poll.c
diff options
context:
space:
mode:
Diffstat (limited to 'io_uring/poll.c')
-rw-r--r--io_uring/poll.c80
1 files changed, 56 insertions, 24 deletions
diff --git a/io_uring/poll.c b/io_uring/poll.c
index ee7da6150ec4..795facbd0e9f 100644
--- a/io_uring/poll.c
+++ b/io_uring/poll.c
@@ -51,6 +51,9 @@ struct io_poll_table {
#define IO_WQE_F_DOUBLE 1
+static int io_poll_wake(struct wait_queue_entry *wait, unsigned mode, int sync,
+ void *key);
+
static inline struct io_kiocb *wqe_to_req(struct wait_queue_entry *wqe)
{
unsigned long priv = (unsigned long)wqe->private;
@@ -164,15 +167,14 @@ static void io_poll_tw_hash_eject(struct io_kiocb *req, bool *locked)
}
}
-static void io_init_poll_iocb(struct io_poll *poll, __poll_t events,
- wait_queue_func_t wake_func)
+static void io_init_poll_iocb(struct io_poll *poll, __poll_t events)
{
poll->head = NULL;
#define IO_POLL_UNMASK (EPOLLERR|EPOLLHUP|EPOLLNVAL|EPOLLRDHUP)
/* mask in events that we always want/need */
poll->events = events | IO_POLL_UNMASK;
INIT_LIST_HEAD(&poll->wait.entry);
- init_waitqueue_func_entry(&poll->wait, wake_func);
+ init_waitqueue_func_entry(&poll->wait, io_poll_wake);
}
static inline void io_poll_remove_entry(struct io_poll *poll)
@@ -223,21 +225,22 @@ enum {
IOU_POLL_DONE = 0,
IOU_POLL_NO_ACTION = 1,
IOU_POLL_REMOVE_POLL_USE_RES = 2,
+ IOU_POLL_REISSUE = 3,
};
/*
* All poll tw should go through this. Checks for poll events, manages
* references, does rewait, etc.
*
- * Returns a negative error on failure. IOU_POLL_NO_ACTION when no action require,
- * which is either spurious wakeup or multishot CQE is served.
- * IOU_POLL_DONE when it's done with the request, then the mask is stored in req->cqe.res.
- * IOU_POLL_REMOVE_POLL_USE_RES indicates to remove multishot poll and that the result
- * is stored in req->cqe.
+ * Returns a negative error on failure. IOU_POLL_NO_ACTION when no action
+ * require, which is either spurious wakeup or multishot CQE is served.
+ * IOU_POLL_DONE when it's done with the request, then the mask is stored in
+ * req->cqe.res. IOU_POLL_REMOVE_POLL_USE_RES indicates to remove multishot
+ * poll and that the result is stored in req->cqe.
*/
static int io_poll_check_events(struct io_kiocb *req, bool *locked)
{
- int v, ret;
+ int v;
/* req->task == current here, checking PF_EXITING is safe */
if (unlikely(req->task->flags & PF_EXITING))
@@ -276,10 +279,19 @@ static int io_poll_check_events(struct io_kiocb *req, bool *locked)
if (!req->cqe.res) {
struct poll_table_struct pt = { ._key = req->apoll_events };
req->cqe.res = vfs_poll(req->file, &pt) & req->apoll_events;
+ /*
+ * We got woken with a mask, but someone else got to
+ * it first. The above vfs_poll() doesn't add us back
+ * to the waitqueue, so if we get nothing back, we
+ * should be safe and attempt a reissue.
+ */
+ if (unlikely(!req->cqe.res)) {
+ /* Multishot armed need not reissue */
+ if (!(req->apoll_events & EPOLLONESHOT))
+ continue;
+ return IOU_POLL_REISSUE;
+ }
}
-
- if ((unlikely(!req->cqe.res)))
- continue;
if (req->apoll_events & EPOLLONESHOT)
return IOU_POLL_DONE;
@@ -294,7 +306,7 @@ static int io_poll_check_events(struct io_kiocb *req, bool *locked)
return IOU_POLL_REMOVE_POLL_USE_RES;
}
} else {
- ret = io_poll_issue(req, locked);
+ int ret = io_poll_issue(req, locked);
if (ret == IOU_STOP_MULTISHOT)
return IOU_POLL_REMOVE_POLL_USE_RES;
if (ret < 0)
@@ -330,6 +342,9 @@ static void io_poll_task_func(struct io_kiocb *req, bool *locked)
poll = io_kiocb_to_cmd(req, struct io_poll);
req->cqe.res = mangle_poll(req->cqe.res & poll->events);
+ } else if (ret == IOU_POLL_REISSUE) {
+ io_req_task_submit(req, locked);
+ return;
} else if (ret != IOU_POLL_REMOVE_POLL_USE_RES) {
req->cqe.res = ret;
req_set_fail(req);
@@ -342,7 +357,7 @@ static void io_poll_task_func(struct io_kiocb *req, bool *locked)
if (ret == IOU_POLL_REMOVE_POLL_USE_RES)
io_req_task_complete(req, locked);
- else if (ret == IOU_POLL_DONE)
+ else if (ret == IOU_POLL_DONE || ret == IOU_POLL_REISSUE)
io_req_task_submit(req, locked);
else
io_req_defer_failed(req, ret);
@@ -495,7 +510,7 @@ static void __io_queue_proc(struct io_poll *poll, struct io_poll_table *pt,
/* mark as double wq entry */
wqe_private |= IO_WQE_F_DOUBLE;
- io_init_poll_iocb(poll, first->events, first->wait.func);
+ io_init_poll_iocb(poll, first->events);
if (!io_poll_double_prepare(req)) {
/* the request is completing, just back off */
kfree(poll);
@@ -533,6 +548,14 @@ static bool io_poll_can_finish_inline(struct io_kiocb *req,
return pt->owning || io_poll_get_ownership(req);
}
+static void io_poll_add_hash(struct io_kiocb *req)
+{
+ if (req->flags & REQ_F_HASH_LOCKED)
+ io_poll_req_insert_locked(req);
+ else
+ io_poll_req_insert(req);
+}
+
/*
* Returns 0 when it's handed over for polling. The caller owns the requests if
* it returns non-zero, but otherwise should not touch it. Negative values
@@ -548,7 +571,7 @@ static int __io_arm_poll_handler(struct io_kiocb *req,
INIT_HLIST_NODE(&req->hash_node);
req->work.cancel_seq = atomic_read(&ctx->cancel_seq);
- io_init_poll_iocb(poll, mask, io_poll_wake);
+ io_init_poll_iocb(poll, mask);
poll->file = req->file;
req->apoll_events = poll->events;
@@ -591,18 +614,17 @@ static int __io_arm_poll_handler(struct io_kiocb *req,
if (mask &&
((poll->events & (EPOLLET|EPOLLONESHOT)) == (EPOLLET|EPOLLONESHOT))) {
- if (!io_poll_can_finish_inline(req, ipt))
+ if (!io_poll_can_finish_inline(req, ipt)) {
+ io_poll_add_hash(req);
return 0;
+ }
io_poll_remove_entries(req);
ipt->result_mask = mask;
/* no one else has access to the req, forget about the ref */
return 1;
}
- if (req->flags & REQ_F_HASH_LOCKED)
- io_poll_req_insert_locked(req);
- else
- io_poll_req_insert(req);
+ io_poll_add_hash(req);
if (mask && (poll->events & EPOLLET) &&
io_poll_can_finish_inline(req, ipt)) {
@@ -630,6 +652,14 @@ static void io_async_queue_proc(struct file *file, struct wait_queue_head *head,
__io_queue_proc(&apoll->poll, pt, head, &apoll->double_poll);
}
+/*
+ * We can't reliably detect loops in repeated poll triggers and issue
+ * subsequently failing. But rather than fail these immediately, allow a
+ * certain amount of retries before we give up. Given that this condition
+ * should _rarely_ trigger even once, we should be fine with a larger value.
+ */
+#define APOLL_MAX_RETRY 128
+
static struct async_poll *io_req_alloc_apoll(struct io_kiocb *req,
unsigned issue_flags)
{
@@ -645,20 +675,24 @@ static struct async_poll *io_req_alloc_apoll(struct io_kiocb *req,
if (entry == NULL)
goto alloc_apoll;
apoll = container_of(entry, struct async_poll, cache);
+ apoll->poll.retries = APOLL_MAX_RETRY;
} else {
alloc_apoll:
apoll = kmalloc(sizeof(*apoll), GFP_ATOMIC);
if (unlikely(!apoll))
return NULL;
+ apoll->poll.retries = APOLL_MAX_RETRY;
}
apoll->double_poll = NULL;
req->apoll = apoll;
+ if (unlikely(!--apoll->poll.retries))
+ return NULL;
return apoll;
}
int io_arm_poll_handler(struct io_kiocb *req, unsigned issue_flags)
{
- const struct io_op_def *def = &io_op_defs[req->opcode];
+ const struct io_issue_def *def = &io_issue_defs[req->opcode];
struct async_poll *apoll;
struct io_poll_table ipt;
__poll_t mask = POLLPRI | POLLERR | EPOLLET;
@@ -674,8 +708,6 @@ int io_arm_poll_handler(struct io_kiocb *req, unsigned issue_flags)
return IO_APOLL_ABORTED;
if (!file_can_poll(req->file))
return IO_APOLL_ABORTED;
- if ((req->flags & (REQ_F_POLLED|REQ_F_PARTIAL_IO)) == REQ_F_POLLED)
- return IO_APOLL_ABORTED;
if (!(req->flags & REQ_F_APOLL_MULTISHOT))
mask |= EPOLLONESHOT;
generated by cgit v1.2.3 (git 2.46.0) at 2025-07-09 06:56:40 +0000