io_uring: open/close and SQPOLL

io_uring: open/close and SQPOLL

[Pavel Begunkov]

Do I miss something, or open/close/etc need an explicit check
for IORING_SETUP_SQPOLL? E.g., now an open() request somehow
returns -EFAULT, and that doesn't seem good.


diff --git a/fs/io_uring.c b/fs/io_uring.c
index 732ec73ec3c0..e841fd03d976 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -3023,6 +3023,8 @@ static int io_openat2_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 	size_t len;
 	int ret;
 
+	if (unlikely(req->ctx->flags & (IORING_SETUP_IOPOLL|IORING_SETUP_SQPOLL)))
+		return -EINVAL;
 	if (sqe->ioprio || sqe->buf_index)
 		return -EINVAL;
 	if (req->flags & REQ_F_FIXED_FILE)
@@ -3417,6 +3419,8 @@ static int io_close_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
 	 */
 	req->work.flags |= IO_WQ_WORK_NO_CANCEL;
 
+	if (unlikely(req->ctx->flags & (IORING_SETUP_IOPOLL|IORING_SETUP_SQPOLL)))
+		return -EINVAL;
 	if (sqe->ioprio || sqe->off || sqe->addr || sqe->len ||
 	    sqe->rw_flags || sqe->buf_index)
 		return -EINVAL;

-- 
Pavel Begunkov

Re: io_uring: open/close and SQPOLL

[Jens Axboe]

On 5/30/20 1:45 PM, Pavel Begunkov wrote:
> Do I miss something, or open/close/etc need an explicit check
> for IORING_SETUP_SQPOLL? E.g., now an open() request somehow
> returns -EFAULT, and that doesn't seem good.

Probably best just to disable them - can you send a real patch?

-- 
Jens Axboe