Re: [PATCH v1 6/7] nfs: Optimize direct I/O to use folios for requests

[Anna Schumaker]

Hi Pranjal,

On Wed, Jun 3, 2026, at 1:30 AM, Pranjal Shrivastava wrote:
> Optimize nfs_direct_extract_pages() to group contiguous pages from the
> same folio into single nfs_page structures. This effectively migrates
> NFS Direct I/O from being page-based to being folio-based.
>
> Reduce the number of nfs_page allocations and subsequent iterations
> by utilizing nfs_page_create_from_folio() to create aggregated
> requests.

I am seeing a LOT of failing xfstests after applying this patch (testing
against various NFS versions over TCP with AUTH_SYS):

+-------------+-----------+-------------+-------------+-------------+
|    testcase | tcp-sys-3 | tcp-sys-4.0 | tcp-sys-4.1 | tcp-sys-4.2 |
+-------------+-----------+-------------+-------------+-------------+
| generic/091 | failure   | failure     | failure     | failure     |
| generic/130 | failure   | failure     | failure     | failure     |
| generic/139 | skipped   | skipped     | skipped     | failure     |
| generic/143 | skipped   | skipped     | skipped     | failure     |
| generic/154 | skipped   | skipped     | skipped     | failure     |
| generic/155 | skipped   | skipped     | skipped     | failure     |
| generic/183 | skipped   | skipped     | skipped     | failure     |
| generic/188 | skipped   | skipped     | skipped     | failure     |
| generic/190 | skipped   | skipped     | skipped     | failure     |
| generic/196 | skipped   | skipped     | skipped     | failure     |
| generic/198 | failure   | failure     | failure     | failure     |
| generic/203 | skipped   | skipped     | skipped     | failure     |
| generic/214 | skipped   | skipped     | skipped     | failure     |
| generic/240 | failure   | failure     | failure     | failure     |
| generic/263 | failure   | failure     | failure     | failure     |
| generic/287 | skipped   | skipped     | skipped     | failure     |
| generic/290 | skipped   | skipped     | skipped     | failure     |
| generic/292 | skipped   | skipped     | skipped     | failure     |
| generic/330 | skipped   | skipped     | skipped     | failure     |
| generic/444 | failure   | skipped     | skipped     | skipped     |
| generic/450 | failure   | failure     | failure     | failure     |
| generic/451 | failure   | failure     | failure     | failure     |
| generic/586 | skipped   | skipped     | skipped     | failure     |
| generic/647 | failure   | failure     | failure     | failure     |
| generic/708 | failure   | failure     | failure     | failure     |
| generic/729 | failure   | failure     | failure     | failure     |
| generic/760 | failure   | failure     | failure     | failure     |
+-------------+-----------+-------------+-------------+-------------+

I'm curious if you've run xfstests against your changes, and if you
see the same failures?

Thanks,
Anna

>
> Signed-off-by: Pranjal Shrivastava <praan@xxxxxxxxxx>
> ---
>  fs/nfs/direct.c | 36 +++++++++++++++++++++++++++++-------
>  1 file changed, 29 insertions(+), 7 deletions(-)
>
> diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c
> index 59002c150f23..93e1af9ec36b 100644
> --- a/fs/nfs/direct.c
> +++ b/fs/nfs/direct.c
> @@ -194,23 +194,45 @@ static ssize_t nfs_direct_extract_pages(struct 
> nfs_direct_req *dreq,
>  		return result;
> 
>  	npages = (result + pgbase + PAGE_SIZE - 1) >> PAGE_SHIFT;
> -	for (i = 0; i < npages; i++) {
> +	for (i = 0; i < npages; ) {
> +		unsigned int chunk_len, folio_offset;
> +		unsigned int nr_to_add = 1;
>  		struct nfs_page *req;
> -		unsigned int req_len = min_t(size_t, result - bytes, PAGE_SIZE - pgbase);
> +		struct folio *folio;
> 
> -		req = nfs_page_create_from_page(dreq->ctx, pagevec[i],
> -						pinned, pgbase, *pos,
> -						req_len);
> +		folio = page_folio(pagevec[i]);
> +		folio_offset = (folio_page_idx(folio, pagevec[i]) << PAGE_SHIFT) + pgbase;
> +		chunk_len = min_t(size_t, result - bytes, PAGE_SIZE - pgbase);
> +
> +		while (i + nr_to_add < npages) {
> +			struct page *next_page = pagevec[i + nr_to_add];
> +			struct page *prev_page = pagevec[i + nr_to_add - 1];
> +
> +			if (page_folio(next_page) != folio ||
> +			    next_page != prev_page + 1)
> +				break;
> +
> +			chunk_len += min_t(size_t, result - bytes - chunk_len, PAGE_SIZE);
> +			nr_to_add++;
> +		}
> +
> +		req = nfs_page_create_from_folio(dreq->ctx, folio,
> +						  pinned, folio_offset,
> +						  chunk_len);
>  		if (IS_ERR(req)) {
>  			if (!bytes)
>  				bytes = PTR_ERR(req);
>  			break;
>  		}
> 
> +		req->wb_index = *pos >> PAGE_SHIFT;
> +		req->wb_offset = *pos;
> +
>  		list_add_tail(&req->wb_list, list);
>  		pgbase = 0;
> -		bytes += req_len;
> -		*pos += req_len;
> +		bytes += chunk_len;
> +		*pos += chunk_len;
> +		i += nr_to_add;
>  	}
> 
>  	if (i < npages) {
> -- 
> 2.54.0.1013.g208068f2d8-goog