Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 25 Jul 2022 20:06:01 -0700
From:   Eric Biggers <ebiggers@kernel.org>
To:     Mike Snitzer <snitzer@redhat.com>
Cc:     Nathan Huckleberry <nhuck@google.com>,
        linux-kernel@vger.kernel.org, dm-devel@redhat.com,
        Alasdair Kergon <agk@redhat.com>,
        Mike Snitzer <snitzer@kernel.org>,
        Sami Tolvanen <samitolvanen@google.com>
Subject: Re: [PATCH 3/3] dm-verity: Add try_verify_in_tasklet
Message-ID: <Yt9aGRXNNPGZFvld@sol.localdomain>
References: <20220722093823.4158756-1-nhuck@google.com>
 <20220722093823.4158756-4-nhuck@google.com>
 <Yt9KTzXUeA8xAiGv@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <Yt9KTzXUeA8xAiGv@redhat.com>
Precedence: bulk

On Mon, Jul 25, 2022 at 09:58:39PM -0400, Mike Snitzer wrote:
> 
> > @@ -1156,7 +1217,7 @@ static int verity_ctr(struct dm_target *ti, unsigned argc, char **argv)
> >  		goto bad;
> >  	}
> >  
> > -	v->tfm = crypto_alloc_ahash(v->alg_name, 0, 0);
> > +	v->tfm = crypto_alloc_ahash(v->alg_name, 0, CRYPTO_ALG_ASYNC);
> >  	if (IS_ERR(v->tfm)) {
> >  		ti->error = "Cannot initialize hash function";
> >  		r = PTR_ERR(v->tfm);
> 
> This hunk that adds the CRYPTO_ALG_ASYNC flag _seems_ unrelated.

I believe it's needed to ensure that only a synchronous algorithm is allocated,
so that verity_hash_update() doesn't have to sleep during the tasklet.  It
should be conditional on v->use_tasklet, though.

> @@ -321,14 +320,12 @@ static int verity_verify_level(struct dm_verity *v, struct dm_verity_io *io,
>  		if (likely(memcmp(verity_io_real_digest(v, io), want_digest,
>  				  v->digest_size) == 0))
>  			aux->hash_verified = 1;
> -		else if (io->in_tasklet) {
> +		else if (io->in_tasklet)
>  			/*
>  			 * FEC code cannot be run in a tasklet since it may
> -			 * sleep.  We need to resume execution in a work-queue
> -			 * to handle FEC.
> +			 * sleep, so fallback to using a work-queue.
>  			 */
>  			return -EAGAIN;
> -		}


Doesn't this need to be:

			r = -EAGAIN;
			goto release_ret_r;

- Eric