Re: [PATCH v5 09/14] ntfs: update runlist handling and cluster allocator

[Christoph Hellwig]

> +	for (index = start_index; index < end_index; index++) {
> +		folio = filemap_lock_folio(vol->lcnbmp_ino->i_mapping, index);
> +		if (IS_ERR(folio)) {
> +			page_cache_sync_readahead(vol->lcnbmp_ino->i_mapping, ra, NULL,

You probably only want to kick off a read for -ENOENT here, and not
any error?

> +					index, end_index - index);
> +			folio = read_mapping_folio(vol->lcnbmp_ino->i_mapping, index, NULL);
> +			if (!IS_ERR(folio))
> +				folio_lock(folio);
> +		}

either way, this seems like a nice primitive for doing reasonably
efficient reads from the page cache, and I could think of a few
other places to it.  Maybe factor it into helper, even if you keep
it in the ntfs code for now?