All of lore.kernel.org
 help / color / mirror / Atom feed
From: Satya Tangirala <satyat@google.com>
To: Eric Biggers <ebiggers@kernel.org>
Cc: linux-block@vger.kernel.org, Jens Axboe <axboe@kernel.dk>,
	dm-devel@redhat.com, Miaohe Lin <linmiaohe@huawei.com>
Subject: Re: [PATCH v2 2/3] block: make blk_crypto_rq_bio_prep() able to fail
Date: Thu, 17 Sep 2020 22:19:29 +0000	[thread overview]
Message-ID: <20200917221929.GB421296@google.com> (raw)
In-Reply-To: <20200916035315.34046-3-ebiggers@kernel.org>

On Tue, Sep 15, 2020 at 08:53:14PM -0700, Eric Biggers wrote:
> From: Eric Biggers <ebiggers@google.com>
> 
> blk_crypto_rq_bio_prep() assumes its gfp_mask argument always includes
> __GFP_DIRECT_RECLAIM, so that the mempool_alloc() will always succeed.
> 
> However, blk_crypto_rq_bio_prep() might be called with GFP_ATOMIC via
> setup_clone() in drivers/md/dm-rq.c.
> 
> This case isn't currently reachable with a bio that actually has an
> encryption context.  However, it's fragile to rely on this.  Just make
> blk_crypto_rq_bio_prep() able to fail.
> 
> Cc: Miaohe Lin <linmiaohe@huawei.com>
> Suggested-by: Satya Tangirala <satyat@google.com>
> Signed-off-by: Eric Biggers <ebiggers@google.com>
> ---
>  block/blk-core.c            |  8 +++++---
>  block/blk-crypto-internal.h | 21 ++++++++++++++++-----
>  block/blk-crypto.c          | 18 +++++++-----------
>  block/blk-mq.c              |  7 ++++++-
>  4 files changed, 34 insertions(+), 20 deletions(-)
> 
> diff --git a/block/blk-core.c b/block/blk-core.c
> index ca3f0f00c9435..fbeaa49f6fe2c 100644
> --- a/block/blk-core.c
> +++ b/block/blk-core.c
> @@ -1620,8 +1620,10 @@ int blk_rq_prep_clone(struct request *rq, struct request *rq_src,
>  		if (rq->bio) {
>  			rq->biotail->bi_next = bio;
>  			rq->biotail = bio;
> -		} else
> +		} else {
>  			rq->bio = rq->biotail = bio;
> +		}
> +		bio = NULL;
>  	}
>  
>  	/* Copy attributes of the original request to the clone request. */
> @@ -1634,8 +1636,8 @@ int blk_rq_prep_clone(struct request *rq, struct request *rq_src,
>  	rq->nr_phys_segments = rq_src->nr_phys_segments;
>  	rq->ioprio = rq_src->ioprio;
>  
> -	if (rq->bio)
> -		blk_crypto_rq_bio_prep(rq, rq->bio, gfp_mask);
> +	if (rq->bio && blk_crypto_rq_bio_prep(rq, rq->bio, gfp_mask) < 0)
> +		goto free_and_out;
>  
>  	return 0;
>  
> diff --git a/block/blk-crypto-internal.h b/block/blk-crypto-internal.h
> index d2b0f565d83cb..0d36aae538d7b 100644
> --- a/block/blk-crypto-internal.h
> +++ b/block/blk-crypto-internal.h
> @@ -142,13 +142,24 @@ static inline void blk_crypto_free_request(struct request *rq)
>  		__blk_crypto_free_request(rq);
>  }
>  
> -void __blk_crypto_rq_bio_prep(struct request *rq, struct bio *bio,
> -			      gfp_t gfp_mask);
> -static inline void blk_crypto_rq_bio_prep(struct request *rq, struct bio *bio,
> -					  gfp_t gfp_mask)
> +int __blk_crypto_rq_bio_prep(struct request *rq, struct bio *bio,
> +			     gfp_t gfp_mask);
> +/**
> + * blk_crypto_rq_bio_prep - Prepare a request's crypt_ctx when its first bio
> + *			    is inserted
> + * @rq: The request to prepare
> + * @bio: The first bio being inserted into the request
> + * @gfp_mask: Memory allocation flags
> + *
> + * Return: 0 on success, -ENOMEM if out of memory.  -ENOMEM is only possible if
> + *	   @gfp_mask doesn't include %__GFP_DIRECT_RECLAIM.
> + */
> +static inline int blk_crypto_rq_bio_prep(struct request *rq, struct bio *bio,
> +					 gfp_t gfp_mask)
>  {
>  	if (bio_has_crypt_ctx(bio))
> -		__blk_crypto_rq_bio_prep(rq, bio, gfp_mask);
> +		return __blk_crypto_rq_bio_prep(rq, bio, gfp_mask);
> +	return 0;
>  }
>  
>  /**
> diff --git a/block/blk-crypto.c b/block/blk-crypto.c
> index a3f27a19067c9..bbe7974fd74f0 100644
> --- a/block/blk-crypto.c
> +++ b/block/blk-crypto.c
> @@ -283,20 +283,16 @@ bool __blk_crypto_bio_prep(struct bio **bio_ptr)
>  	return false;
>  }
>  
> -/**
> - * __blk_crypto_rq_bio_prep - Prepare a request's crypt_ctx when its first bio
> - *			      is inserted
> - *
> - * @rq: The request to prepare
> - * @bio: The first bio being inserted into the request
> - * @gfp_mask: gfp mask
> - */
> -void __blk_crypto_rq_bio_prep(struct request *rq, struct bio *bio,
> -			      gfp_t gfp_mask)
> +int __blk_crypto_rq_bio_prep(struct request *rq, struct bio *bio,
> +			     gfp_t gfp_mask)
>  {
> -	if (!rq->crypt_ctx)
> +	if (!rq->crypt_ctx) {
>  		rq->crypt_ctx = mempool_alloc(bio_crypt_ctx_pool, gfp_mask);
> +		if (!rq->crypt_ctx)
> +			return -ENOMEM;
> +	}
>  	*rq->crypt_ctx = *bio->bi_crypt_context;
> +	return 0;
>  }
>  
>  /**
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index e04b759add758..9ec0e7149ae69 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -1940,13 +1940,18 @@ void blk_mq_flush_plug_list(struct blk_plug *plug, bool from_schedule)
>  static void blk_mq_bio_to_request(struct request *rq, struct bio *bio,
>  		unsigned int nr_segs)
>  {
> +	int err;
> +
>  	if (bio->bi_opf & REQ_RAHEAD)
>  		rq->cmd_flags |= REQ_FAILFAST_MASK;
>  
>  	rq->__sector = bio->bi_iter.bi_sector;
>  	rq->write_hint = bio->bi_write_hint;
>  	blk_rq_bio_prep(rq, bio, nr_segs);
> -	blk_crypto_rq_bio_prep(rq, bio, GFP_NOIO);
> +
> +	/* This can't fail, since GFP_NOIO includes __GFP_DIRECT_RECLAIM. */
> +	err = blk_crypto_rq_bio_prep(rq, bio, GFP_NOIO);
> +	WARN_ON_ONCE(err);
>  
>  	blk_account_io_start(rq);
>  }
Looks good!

Reviewed-by: Satya Tangirala <satyat@google.com>

> -- 
> 2.28.0
> 

  reply	other threads:[~2020-09-17 22:19 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-09-16  3:53 [PATCH v2 0/3] block: fix up bio_crypt_ctx allocation Eric Biggers
2020-09-16  3:53 ` [PATCH v2 1/3] block: make bio_crypt_clone() able to fail Eric Biggers
2020-09-17 22:17   ` Satya Tangirala
2020-09-24  0:56   ` Mike Snitzer
2020-09-16  3:53 ` [PATCH v2 2/3] block: make blk_crypto_rq_bio_prep() " Eric Biggers
2020-09-17 22:19   ` Satya Tangirala [this message]
2020-09-24  0:57   ` Mike Snitzer
2020-09-16  3:53 ` [PATCH v2 3/3] block: warn if !__GFP_DIRECT_RECLAIM in bio_crypt_set_ctx() Eric Biggers
2020-09-17 22:26   ` Satya Tangirala
2020-09-28 20:59 ` [PATCH v2 0/3] block: fix up bio_crypt_ctx allocation Eric Biggers
2020-10-05 16:42   ` Eric Biggers
2020-10-05 16:48 ` Jens Axboe

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20200917221929.GB421296@google.com \
    --to=satyat@google.com \
    --cc=axboe@kernel.dk \
    --cc=dm-devel@redhat.com \
    --cc=ebiggers@kernel.org \
    --cc=linmiaohe@huawei.com \
    --cc=linux-block@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.