Re: [PATCH 5/9] blk-crypto: optimize bio splitting in blk_crypto_fallback_encrypt_bio

[Eric Biggers <ebiggers@kernel.org>]

On Wed, Dec 17, 2025 at 07:06:48AM +0100, Christoph Hellwig wrote:
> +		if (++enc_idx == enc_bio->bi_max_vecs) {
> +			/*
> +			 * For each additional encrypted bio submitted,
> +			 * increment the source bio's remaining count.  Each
> +			 * encrypted bio's completion handler calls bio_endio on
> +			 * the source bio, so this keeps the source bio from
> +			 * completing until the last encrypted bio does.
> +			 */
> +			bio_inc_remaining(src_bio);
> +			submit_bio(enc_bio);
> +			goto new_bio;
> +		}

Actually I think using bi_max_vecs is broken.

This code assumes that bi_max_vecs matches the nr_segs that was passed
to bio_alloc_bioset().

That assumption is incorrect, though.  If nr_segs > 0 && nr_segs <
BIO_INLINE_VECS, bio_alloc_bioset() sets bi_max_vecs to BIO_INLINE_VECS.
BIO_INLINE_VECS is 4.

I think blk_crypto_alloc_enc_bio() will need to return a nr_enc_pages
value.  That value will need to be used above as well as at
out_free_enc_bio, instead of bi_max_vecs.

- Eric