Re: [PATCH 1/3] fs,block: Introduce IOCB_ZONE_APPEND and direct-io handling

[Pavel Begunkov <asml.silence@gmail.com>]

On 17/06/2020 20:23, Kanchan Joshi wrote:
> From: Selvakumar S <selvakuma.s1@samsung.com>
> 
> Introduce IOCB_ZONE_APPEND flag, which is set in kiocb->ki_flags for
> zone-append. Direct I/O submission path uses this flag to send bio with
> append op. And completion path uses the same to return zone-relative
> offset to upper layer.
> 
> Signed-off-by: SelvaKumar S <selvakuma.s1@samsung.com>
> Signed-off-by: Kanchan Joshi <joshi.k@samsung.com>
> Signed-off-by: Nitesh Shetty <nj.shetty@samsung.com>
> Signed-off-by: Javier Gonzalez <javier.gonz@samsung.com>
> ---
>  fs/block_dev.c     | 19 ++++++++++++++++++-
>  include/linux/fs.h |  1 +
>  2 files changed, 19 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/block_dev.c b/fs/block_dev.c
> index 47860e5..4c84b4d0 100644
> --- a/fs/block_dev.c
> +++ b/fs/block_dev.c
> @@ -185,6 +185,10 @@ static unsigned int dio_bio_write_op(struct kiocb *iocb)
>  	/* avoid the need for a I/O completion work item */
>  	if (iocb->ki_flags & IOCB_DSYNC)
>  		op |= REQ_FUA;
> +#ifdef CONFIG_BLK_DEV_ZONED
> +	if (iocb->ki_flags & IOCB_ZONE_APPEND)
> +		op |= REQ_OP_ZONE_APPEND | REQ_NOMERGE;
> +#endif
>  	return op;
>  }
>  
> @@ -295,6 +299,14 @@ static int blkdev_iopoll(struct kiocb *kiocb, bool wait)
>  	return blk_poll(q, READ_ONCE(kiocb->ki_cookie), wait);
>  }
>  
> +#ifdef CONFIG_BLK_DEV_ZONED
> +static inline long blkdev_bio_end_io_append(struct bio *bio)
> +{
> +	return (bio->bi_iter.bi_sector %
> +		blk_queue_zone_sectors(bio->bi_disk->queue)) << SECTOR_SHIFT;

IIRC, zone_size is pow2 and bit operations should suffice.
Anyway, this can use a temporary variable.

> +}
> +#endif
> +
>  static void blkdev_bio_end_io(struct bio *bio)
>  {
>  	struct blkdev_dio *dio = bio->bi_private;
> @@ -307,15 +319,20 @@ static void blkdev_bio_end_io(struct bio *bio)
>  		if (!dio->is_sync) {
>  			struct kiocb *iocb = dio->iocb;
>  			ssize_t ret;
> +			long res = 0;
>  
>  			if (likely(!dio->bio.bi_status)) {
>  				ret = dio->size;
>  				iocb->ki_pos += ret;
> +#ifdef CONFIG_BLK_DEV_ZONED
> +				if (iocb->ki_flags & IOCB_ZONE_APPEND)
> +					res = blkdev_bio_end_io_append(bio);
> +#endif
>  			} else {
>  				ret = blk_status_to_errno(dio->bio.bi_status);
>  			}
>  
> -			dio->iocb->ki_complete(iocb, ret, 0);
> +			dio->iocb->ki_complete(iocb, ret, res);
>  			if (dio->multi_bio)
>  				bio_put(&dio->bio);
>  		} else {
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index 6c4ab4d..dc547b9 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -315,6 +315,7 @@ enum rw_hint {
>  #define IOCB_SYNC		(1 << 5)
>  #define IOCB_WRITE		(1 << 6)
>  #define IOCB_NOWAIT		(1 << 7)
> +#define IOCB_ZONE_APPEND	(1 << 8)
>  
>  struct kiocb {
>  	struct file		*ki_filp;
> 

-- 
Pavel Begunkov