Received-SPF: pass (google.com: domain of linux-crypto-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <20210726152724.2281408-1-mudongliangabcd@gmail.com>
 <YQg4lehajLpQjyPd@Red> <CAD-N9QWvR-7caWCnk1CMo8sWPEC=CfKU2_v=YkTVr0o5L7wehA@mail.gmail.com>
In-Reply-To: <CAD-N9QWvR-7caWCnk1CMo8sWPEC=CfKU2_v=YkTVr0o5L7wehA@mail.gmail.com>
From:   Dongliang Mu <mudongliangabcd@gmail.com>
Date:   Tue, 3 Aug 2021 14:06:43 +0800
Message-ID: <CAD-N9QVfreHCV7673DVq6aS3+3RYQ4-x8eZuGF0pvUbRhTL-Cg@mail.gmail.com>
Subject: Re: [PATCH] crypto: sun8i-ce: fix multiple memory leaks in sun8i_ce_hash_run
To:     Corentin Labbe <clabbe.montjoie@gmail.com>
Cc:     Herbert Xu <herbert@gondor.apana.org.au>,
        "David S. Miller" <davem@davemloft.net>,
        Maxime Ripard <mripard@kernel.org>,
        Chen-Yu Tsai <wens@csie.org>,
        Jernej Skrabec <jernej.skrabec@gmail.com>,
        Ard Biesheuvel <ardb@kernel.org>,
        Jonathan Corbet <corbet@lwn.net>,
        Eric Biggers <ebiggers@google.com>,
        Xiang Chen <chenxiang66@hisilicon.com>,
        Mauro Carvalho Chehab <mchehab+huawei@kernel.org>,
        Corentin Labbe <clabbe@baylibre.com>,
        "Jason A. Donenfeld" <Jason@zx2c4.com>,
        linux-crypto@vger.kernel.org, linux-arm-kernel@lists.infradead.org,
        linux-sunxi@lists.linux.dev,
        linux-kernel <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, Aug 3, 2021 at 1:57 PM Dongliang Mu <mudongliangabcd@gmail.com> wro=
te:
>
> On Tue, Aug 3, 2021 at 2:25 AM Corentin Labbe <clabbe.montjoie@gmail.com>=
 wrote:
> >
> > Le Mon, Jul 26, 2021 at 11:27:12PM +0800, Dongliang Mu a =C3=A9crit :
> > > In sun8i_ce_hash_run, all the dma_mmap_sg/single will cause memory le=
ak
> > > due to no corresponding unmap operation if errors happen.
> > >
> > > Fix this by adding error handling part for all the dma_mmap_sg/single=
.
> > >
> >
> > I think it could be better worded, error handling is already there (but=
 bad).
>
> Sure. How about "Fix this by freeing the objects allocated by
> dma_mmap_sg/single when errors occur."?
>
> >
> > > Fixes: 56f6d5aee88d ("crypto: sun8i-ce - support hash algorithms")
> > > Signed-off-by: Dongliang Mu <mudongliangabcd@gmail.com>
> > > ---
> > >  .../crypto/allwinner/sun8i-ce/sun8i-ce-hash.c | 28 +++++++++--------=
--
> > >  1 file changed, 13 insertions(+), 15 deletions(-)
> > >
> > > diff --git a/drivers/crypto/allwinner/sun8i-ce/sun8i-ce-hash.c b/driv=
ers/crypto/allwinner/sun8i-ce/sun8i-ce-hash.c
> > > index 88194718a806..d454ad99deee 100644
> > > --- a/drivers/crypto/allwinner/sun8i-ce/sun8i-ce-hash.c
> > > +++ b/drivers/crypto/allwinner/sun8i-ce/sun8i-ce-hash.c
> > > @@ -286,16 +286,14 @@ int sun8i_ce_hash_run(struct crypto_engine *eng=
ine, void *breq)
> > >
> > >       /* the padding could be up to two block. */
> > >       buf =3D kzalloc(bs * 2, GFP_KERNEL | GFP_DMA);
> > > -     if (!buf) {
> > > -             err =3D -ENOMEM;
> > > -             goto theend;
> >
> > Please keep all goto error for being consistent.
>
> OK, no problem.
>
> BTW, usually for the 1st malloc failure, I prefer returning with errno di=
rectly.

For this case, as crypto_finalize_hash_request must be executed, so
directly return is incorrect. I will add a goto label before
crypto_finalize_hash_request.

>
> >
> > > -     }
> > > +     if (!buf)
> > > +             return -ENOMEM;
> > >       bf =3D (__le32 *)buf;
> > >
> > >       result =3D kzalloc(digestsize, GFP_KERNEL | GFP_DMA);
> > >       if (!result) {
> > > -             err =3D -ENOMEM;
> > > -             goto theend;
> > > +             kfree(buf);
> > > +             return -ENOMEM;
> > >       }
> > >
> > >       flow =3D rctx->flow;
> > > @@ -321,7 +319,7 @@ int sun8i_ce_hash_run(struct crypto_engine *engin=
e, void *breq)
> > >       if (nr_sgs <=3D 0 || nr_sgs > MAX_SG) {
> > >               dev_err(ce->dev, "Invalid sg number %d\n", nr_sgs);
> > >               err =3D -EINVAL;
> > > -             goto theend;
> > > +             goto err_result;
> > >       }
> > >
> > >       len =3D areq->nbytes;
> > > @@ -334,7 +332,7 @@ int sun8i_ce_hash_run(struct crypto_engine *engin=
e, void *breq)
> > >       if (len > 0) {
> > >               dev_err(ce->dev, "remaining len %d\n", len);
> > >               err =3D -EINVAL;
> > > -             goto theend;
> > > +             goto err_unmap_sg;
> > >       }
> > >       addr_res =3D dma_map_single(ce->dev, result, digestsize, DMA_FR=
OM_DEVICE);
> > >       cet->t_dst[0].addr =3D cpu_to_le32(addr_res);
> > > @@ -342,7 +340,7 @@ int sun8i_ce_hash_run(struct crypto_engine *engin=
e, void *breq)
> > >       if (dma_mapping_error(ce->dev, addr_res)) {
> > >               dev_err(ce->dev, "DMA map dest\n");
> > >               err =3D -EINVAL;
> > > -             goto theend;
> > > +             goto err_unmap_sg;
> > >       }
> > >
> > >       byte_count =3D areq->nbytes;
> > > @@ -392,7 +390,7 @@ int sun8i_ce_hash_run(struct crypto_engine *engin=
e, void *breq)
> > >       if (dma_mapping_error(ce->dev, addr_pad)) {
> > >               dev_err(ce->dev, "DMA error on padding SG\n");
> > >               err =3D -EINVAL;
> > > -             goto theend;
> > > +             goto err_addr_res;
> > >       }
> > >
> > >       if (ce->variant->hash_t_dlen_in_bits)
> > > @@ -405,15 +403,15 @@ int sun8i_ce_hash_run(struct crypto_engine *eng=
ine, void *breq)
> > >       err =3D sun8i_ce_run_task(ce, flow, crypto_tfm_alg_name(areq->b=
ase.tfm));
> > >
> > >       dma_unmap_single(ce->dev, addr_pad, j * 4, DMA_TO_DEVICE);
> > > +err_addr_res:
> > > +     dma_unmap_single(ce->dev, addr_res, digestsize, DMA_FROM_DEVICE=
);
> > > +err_unmap_sg:
> > >       dma_unmap_sg(ce->dev, areq->src, sg_nents(areq->src),
> > >                    DMA_TO_DEVICE);
> > > -     dma_unmap_single(ce->dev, addr_res, digestsize, DMA_FROM_DEVICE=
);
> > > -
> > > -
> > >       memcpy(areq->result, result, algt->alg.hash.halg.digestsize);
> >
> > The result should be copied only when everything is ok. Please add a "i=
f (!err)"
>
> Sure.
>
> >
> > Thanks for your work
> > Regards