Re: [RFC PATCH] mm : using bdi->ra_pages instead of ra->ra_pages within readahead

[Matthew Wilcox]

On Thu, Aug 13, 2020 at 04:11:22PM +0800, Zhaoyang Huang wrote:
> file->f_ra->ra_pages will remain the initialized value since it opend, which may
> be NOT equal to bdi->ra_pages as the latter one is updated somehow(etc,
> echo xxx > /sys/block/dm/queue/read_ahead_kb).So having readahead use
> bdi->ra_pages.

But now it ignores the work done by shrink_readahead_size_eio()
and fadvise(POSIX_FADV_SEQUENTIAL).  And you missed the use of ra_size
on the previous line, miscalculating ra->start.

> Signed-off-by: Zhaoyang Huang <zhaoyang.huang@xxxxxxxxxx>
> ---
>  mm/filemap.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/mm/filemap.c b/mm/filemap.c
> index d78f577..259dcfd 100644
> --- a/mm/filemap.c
> +++ b/mm/filemap.c
> @@ -2499,8 +2499,8 @@ static struct file *do_sync_mmap_readahead(struct vm_fault *vmf)
>  	 */
>  	fpin = maybe_unlock_mmap_for_io(vmf, fpin);
>  	ra->start = max_t(long, 0, offset - ra->ra_pages / 2);
> -	ra->size = ra->ra_pages;
> -	ra->async_size = ra->ra_pages / 4;
> +	ra->size = inode_to_bdi(mapping->host)->ra_pages;
> +	ra->async_size = ra->size / 4;
>  	ra_submit(ra, mapping, file);
>  	return fpin;
>  }
> -- 
> 1.9.1
> 
>