Re: [PATCH] io_uring: Optimizing return value

[Jens Axboe <axboe@kernel.dk>]

On 8/11/22 8:00 PM, Zhang chunchao wrote:
> Delete return value ret Initialize assignment, change return value ret
> to EOPNOTSUPP when IO_IS_URING_FOPS failed.
> 
> Signed-off-by: Zhang chunchao <chunchao@nfschina.com>
> ---
>  io_uring/io_uring.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
> index b54218da075c..1b56f3d1a47b 100644
> --- a/io_uring/io_uring.c
> +++ b/io_uring/io_uring.c
> @@ -3859,16 +3859,17 @@ SYSCALL_DEFINE4(io_uring_register, unsigned int, fd, unsigned int, opcode,
>  		void __user *, arg, unsigned int, nr_args)
>  {
>  	struct io_ring_ctx *ctx;
> -	long ret = -EBADF;
> +	long ret;
>  	struct fd f;
>  
>  	f = fdget(fd);
>  	if (!f.file)
>  		return -EBADF;
>  
> -	ret = -EOPNOTSUPP;
> -	if (!io_is_uring_fops(f.file))
> +	if (!io_is_uring_fops(f.file)) {
> +		ret = -EOPNOTSUPP;
>  		goto out_fput;
> +	}
>  
>  	ctx = f.file->private_data;

As mentioned in the other reply, just do the first part. We don't need
to move the other one, I think you'll find the generated code is the
same.

I'd title the commit as "io_uring: remove useless setting of 'ret'".

-- 
Jens Axboe