Merge tag 'io_uring-6.3-2023-03-03' of git://git.kernel.dk/linux

Pull more io_uring updates from Jens Axboe:
 "Here's a set of fixes/changes that didn't make the first cut, either
  because they got queued before I sent the early merge request, or
  fixes that came in afterwards. In detail:

   - Don't set MSG_NOSIGNAL on recv/recvmsg opcodes, as AF_PACKET will
     error out (David)

   - Fix for spurious poll wakeups (me)

   - Fix for a file leak for buffered reads in certain conditions
     (Joseph)

   - Don't allow registered buffers of mixed types (Pavel)

   - Improve handling of huge pages for registered buffers (Pavel)

   - Provided buffer ring size calculation fix (Wojciech)

   - Minor cleanups (me)"

* tag 'io_uring-6.3-2023-03-03' of git://git.kernel.dk/linux:
  io_uring/poll: don't pass in wake func to io_init_poll_iocb()
  io_uring: fix fget leak when fs don't support nowait buffered read
  io_uring/poll: allow some retries for poll triggering spuriously
  io_uring: remove MSG_NOSIGNAL from recvmsg
  io_uring/rsrc: always initialize 'folio' to NULL
  io_uring/rsrc: optimise registered huge pages
  io_uring/rsrc: optimise single entry advance
  io_uring/rsrc: disallow multi-source reg buffers
  io_uring: remove unused wq_list_merge
  io_uring: fix size calculation when registering buf ring
  io_uring/rsrc: fix a comment in io_import_fixed()
  io_uring: rename 'in_idle' to 'in_cancel'
  io_uring: consolidate the put_ref-and-return section of adding work

8 files changed, 84 insertions, 61 deletions

diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index 1df68da89f99..fd1cc35a1c00 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -719,7 +719,7 @@ static void io_put_task_remote(struct task_struct *task, int nr)
 	struct io_uring_task *tctx = task->io_uring;
 
 	percpu_counter_sub(&tctx->inflight, nr);
-	if (unlikely(atomic_read(&tctx->in_idle)))
+	if (unlikely(atomic_read(&tctx->in_cancel)))
 		wake_up(&tctx->wait);
 	put_task_struct_many(task, nr);
 }
@@ -1258,8 +1258,8 @@ void tctx_task_work(struct callback_head *cb)
 
 	ctx_flush_and_put(ctx, &uring_locked);
 
-	/* relaxed read is enough as only the task itself sets ->in_idle */
-	if (unlikely(atomic_read(&tctx->in_idle)))
+	/* relaxed read is enough as only the task itself sets ->in_cancel */
+	if (unlikely(atomic_read(&tctx->in_cancel)))
 		io_uring_drop_tctx_refs(current);
 
 	trace_io_uring_task_work_run(tctx, count, loops);
@@ -1285,17 +1285,15 @@ static void io_req_local_work_add(struct io_kiocb *req)
 
 	percpu_ref_get(&ctx->refs);
 
-	if (!llist_add(&req->io_task_work.node, &ctx->work_llist)) {
-		percpu_ref_put(&ctx->refs);
-		return;
-	}
+	if (!llist_add(&req->io_task_work.node, &ctx->work_llist))
+		goto put_ref;
+
 	/* needed for the following wake up */
 	smp_mb__after_atomic();
 
-	if (unlikely(atomic_read(&req->task->io_uring->in_idle))) {
+	if (unlikely(atomic_read(&req->task->io_uring->in_cancel))) {
 		io_move_task_work_from_local(ctx);
-		percpu_ref_put(&ctx->refs);
-		return;
+		goto put_ref;
 	}
 
 	if (ctx->flags & IORING_SETUP_TASKRUN_FLAG)
@@ -1305,6 +1303,8 @@ static void io_req_local_work_add(struct io_kiocb *req)
 
 	if (READ_ONCE(ctx->cq_waiting))
 		wake_up_state(ctx->submitter_task, TASK_INTERRUPTIBLE);
+
+put_ref:
 	percpu_ref_put(&ctx->refs);
 }
 
@@ -1777,7 +1777,7 @@ int io_req_prep_async(struct io_kiocb *req)
 	const struct io_issue_def *def = &io_issue_defs[req->opcode];
 
 	/* assign early for deferred execution for non-fixed file */
-	if (def->needs_file && !(req->flags & REQ_F_FIXED_FILE))
+	if (def->needs_file && !(req->flags & REQ_F_FIXED_FILE) && !req->file)
 		req->file = io_file_get_normal(req, req->cqe.fd);
 	if (!cdef->prep_async)
 		return 0;
@@ -2937,12 +2937,12 @@ static __cold void io_tctx_exit_cb(struct callback_head *cb)
 
 	work = container_of(cb, struct io_tctx_exit, task_work);
 	/*
-	 * When @in_idle, we're in cancellation and it's racy to remove the
+	 * When @in_cancel, we're in cancellation and it's racy to remove the
 	 * node. It'll be removed by the end of cancellation, just ignore it.
 	 * tctx can be NULL if the queueing of this task_work raced with
 	 * work cancelation off the exec path.
 	 */
-	if (tctx && !atomic_read(&tctx->in_idle))
+	if (tctx && !atomic_read(&tctx->in_cancel))
 		io_uring_del_tctx_node((unsigned long)work->ctx);
 	complete(&work->completion);
 }
@@ -3210,7 +3210,7 @@ __cold void io_uring_cancel_generic(bool cancel_all, struct io_sq_data *sqd)
 	if (tctx->io_wq)
 		io_wq_exit_start(tctx->io_wq);
 
-	atomic_inc(&tctx->in_idle);
+	atomic_inc(&tctx->in_cancel);
 	do {
 		bool loop = false;
 
@@ -3261,9 +3261,9 @@ __cold void io_uring_cancel_generic(bool cancel_all, struct io_sq_data *sqd)
 	if (cancel_all) {
 		/*
 		 * We shouldn't run task_works after cancel, so just leave
-		 * ->in_idle set for normal exit.
+		 * ->in_cancel set for normal exit.
 		 */
-		atomic_dec(&tctx->in_idle);
+		atomic_dec(&tctx->in_cancel);
 		/* for exec all current's requests should be gone, kill tctx */
 		__io_uring_free(current);
 	}
diff --git a/io_uring/kbuf.c b/io_uring/kbuf.c
index 4a6401080c1f..3002dc827195 100644
--- a/io_uring/kbuf.c
+++ b/io_uring/kbuf.c
@@ -505,7 +505,7 @@ int io_register_pbuf_ring(struct io_ring_ctx *ctx, void __user *arg)
 	}
 
 	pages = io_pin_pages(reg.ring_addr,
-			     struct_size(br, bufs, reg.ring_entries),
+			     flex_array_size(br, bufs, reg.ring_entries),
 			     &nr_pages);
 	if (IS_ERR(pages)) {
 		kfree(free_bl);
diff --git a/io_uring/net.c b/io_uring/net.c
index cbd4b725f58c..b7f190ca528e 100644
--- a/io_uring/net.c
+++ b/io_uring/net.c
@@ -567,7 +567,7 @@ int io_recvmsg_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 	sr->flags = READ_ONCE(sqe->ioprio);
 	if (sr->flags & ~(RECVMSG_FLAGS))
 		return -EINVAL;
-	sr->msg_flags = READ_ONCE(sqe->msg_flags) | MSG_NOSIGNAL;
+	sr->msg_flags = READ_ONCE(sqe->msg_flags);
 	if (sr->msg_flags & MSG_DONTWAIT)
 		req->flags |= REQ_F_NOWAIT;
 	if (sr->msg_flags & MSG_ERRQUEUE)
diff --git a/io_uring/poll.c b/io_uring/poll.c
index 8339a92b4510..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)
@@ -508,7 +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);
@@ -569,7 +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;
 
@@ -650,6 +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)
 {
@@ -665,14 +675,18 @@ 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;
 }
 
@@ -694,8 +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;
 
diff --git a/io_uring/poll.h b/io_uring/poll.h
index 5f3bae50fc81..b2393b403a2c 100644
--- a/io_uring/poll.h
+++ b/io_uring/poll.h
@@ -12,6 +12,7 @@ struct io_poll {
 	struct file			*file;
 	struct wait_queue_head		*head;
 	__poll_t			events;
+	int				retries;
 	struct wait_queue_entry		wait;
 };
 
diff --git a/io_uring/rsrc.c b/io_uring/rsrc.c
index a59fc02de598..056f40946ff6 100644
--- a/io_uring/rsrc.c
+++ b/io_uring/rsrc.c
@@ -1162,14 +1162,17 @@ struct page **io_pin_pages(unsigned long ubuf, unsigned long len, int *npages)
 	pret = pin_user_pages(ubuf, nr_pages, FOLL_WRITE | FOLL_LONGTERM,
 			      pages, vmas);
 	if (pret == nr_pages) {
+		struct file *file = vmas[0]->vm_file;
+
 		/* don't support file backed memory */
 		for (i = 0; i < nr_pages; i++) {
-			struct vm_area_struct *vma = vmas[i];
-
-			if (vma_is_shmem(vma))
+			if (vmas[i]->vm_file != file) {
+				ret = -EINVAL;
+				break;
+			}
+			if (!file)
 				continue;
-			if (vma->vm_file &&
-			    !is_file_hugepages(vma->vm_file)) {
+			if (!vma_is_shmem(vmas[i]) && !is_file_hugepages(file)) {
 				ret = -EOPNOTSUPP;
 				break;
 			}
@@ -1207,6 +1210,7 @@ static int io_sqe_buffer_register(struct io_ring_ctx *ctx, struct iovec *iov,
 	unsigned long off;
 	size_t size;
 	int ret, nr_pages, i;
+	struct folio *folio = NULL;
 
 	*pimu = ctx->dummy_ubuf;
 	if (!iov->iov_base)
@@ -1221,6 +1225,21 @@ static int io_sqe_buffer_register(struct io_ring_ctx *ctx, struct iovec *iov,
 		goto done;
 	}
 
+	/* If it's a huge page, try to coalesce them into a single bvec entry */
+	if (nr_pages > 1) {
+		folio = page_folio(pages[0]);
+		for (i = 1; i < nr_pages; i++) {
+			if (page_folio(pages[i]) != folio) {
+				folio = NULL;
+				break;
+			}
+		}
+		if (folio) {
+			folio_put_refs(folio, nr_pages - 1);
+			nr_pages = 1;
+		}
+	}
+
 	imu = kvmalloc(struct_size(imu, bvec, nr_pages), GFP_KERNEL);
 	if (!imu)
 		goto done;
@@ -1233,6 +1252,17 @@ static int io_sqe_buffer_register(struct io_ring_ctx *ctx, struct iovec *iov,
 
 	off = (unsigned long) iov->iov_base & ~PAGE_MASK;
 	size = iov->iov_len;
+	/* store original address for later verification */
+	imu->ubuf = (unsigned long) iov->iov_base;
+	imu->ubuf_end = imu->ubuf + iov->iov_len;
+	imu->nr_bvecs = nr_pages;
+	*pimu = imu;
+	ret = 0;
+
+	if (folio) {
+		bvec_set_page(&imu->bvec[0], pages[0], size, off);
+		goto done;
+	}
 	for (i = 0; i < nr_pages; i++) {
 		size_t vec_len;
 
@@ -1241,12 +1271,6 @@ static int io_sqe_buffer_register(struct io_ring_ctx *ctx, struct iovec *iov,
 		off = 0;
 		size -= vec_len;
 	}
-	/* store original address for later verification */
-	imu->ubuf = (unsigned long) iov->iov_base;
-	imu->ubuf_end = imu->ubuf + iov->iov_len;
-	imu->nr_bvecs = nr_pages;
-	*pimu = imu;
-	ret = 0;
 done:
 	if (ret)
 		kvfree(imu);
@@ -1335,7 +1359,7 @@ int io_import_fixed(int ddir, struct iov_iter *iter,
 		return -EFAULT;
 
 	/*
-	 * May not be a start of buffer, set size appropriately
+	 * Might not be a start of buffer, set size appropriately
 	 * and advance us to the beginning.
 	 */
 	offset = buf_addr - imu->ubuf;
@@ -1361,7 +1385,15 @@ int io_import_fixed(int ddir, struct iov_iter *iter,
 		const struct bio_vec *bvec = imu->bvec;
 
 		if (offset <= bvec->bv_len) {
-			iov_iter_advance(iter, offset);
+			/*
+			 * Note, huge pages buffers consists of one large
+			 * bvec entry and should always go this way. The other
+			 * branch doesn't expect non PAGE_SIZE'd chunks.
+			 */
+			iter->bvec = bvec;
+			iter->nr_segs = bvec->bv_len;
+			iter->count -= offset;
+			iter->iov_offset = offset;
 		} else {
 			unsigned long seg_skip;
 
diff --git a/io_uring/slist.h b/io_uring/slist.h
index f27601fa4660..7c198a40d5f1 100644
--- a/io_uring/slist.h
+++ b/io_uring/slist.h
@@ -27,28 +27,6 @@ static inline void wq_list_add_after(struct io_wq_work_node *node,
 		list->last = node;
 }
 
-/**
- * wq_list_merge - merge the second list to the first one.
- * @list0: the first list
- * @list1: the second list
- * Return the first node after mergence.
- */
-static inline struct io_wq_work_node *wq_list_merge(struct io_wq_work_list *list0,
-						    struct io_wq_work_list *list1)
-{
-	struct io_wq_work_node *ret;
-
-	if (!list0->first) {
-		ret = list1->first;
-	} else {
-		ret = list0->first;
-		list0->last->next = list1->first;
-	}
-	INIT_WQ_LIST(list0);
-	INIT_WQ_LIST(list1);
-	return ret;
-}
-
 static inline void wq_list_add_tail(struct io_wq_work_node *node,
 				    struct io_wq_work_list *list)
 {
diff --git a/io_uring/tctx.c b/io_uring/tctx.c
index 4324b1cf1f6a..3a8d1dd97e1b 100644
--- a/io_uring/tctx.c
+++ b/io_uring/tctx.c
@@ -83,7 +83,7 @@ __cold int io_uring_alloc_task_context(struct task_struct *task,
 
 	xa_init(&tctx->xa);
 	init_waitqueue_head(&tctx->wait);
-	atomic_set(&tctx->in_idle, 0);
+	atomic_set(&tctx->in_cancel, 0);
 	atomic_set(&tctx->inflight_tracked, 0);
 	task->io_uring = tctx;
 	init_llist_head(&tctx->task_list);