> -----Original Message-----
> From: David Marchand <david.march...@redhat.com>
> Sent: Friday, January 6, 2023 6:27 PM
> To: Kai Ji <kai...@intel.com>; Ruifeng Wang <ruifeng.w...@arm.com>
> Cc: Fan Zhang <fanzhang....@gmail.com>; Akhil Goyal <gak...@marvell.com>;
> dev@dpdk.org; nd
> <n...@arm.com>; sta...@dpdk.org; Feifei Wang <feifei.wa...@arm.com>
> Subject: Re: [PATCH] crypto/openssl: fix warning on copy length
>
> On Fri, Jan 6, 2023 at 11:16 AM Ruifeng Wang <ruifeng.w...@arm.com> wrote:
> >
> > When building with gcc 11.2.0, the compiler warns as follows:
> > In function 'memcpy',
> > inlined from 'openssl_set_session_auth_parameters'
> at ../drivers/crypto/openssl/rte_openssl_pmd.c:699:3,
> > inlined from 'openssl_set_session_parameters'
> at ../drivers/crypto/openssl/rte_openssl_pmd.c:826:9:
> > /usr/include/aarch64-linux-gnu/bits/string_fortified.h:29:10: warning:
> '__builtin_memcpy' forming offset [4, 8] is out of the bounds [0, 4]
> [-Warray-bounds]
> >
> > Fixed the warning by copying up to string size.
> >
> > Fixes: 75adf1eae44f ("crypto/openssl: update HMAC routine with 3.0 EVP API")
> > Cc: sta...@dpdk.org
> > Cc: kai...@intel.com
> >
> > Signed-off-by: Ruifeng Wang <ruifeng.w...@arm.com>
> > Reviewed-by: Feifei Wang <feifei.wa...@arm.com>
> > ---
> > drivers/crypto/openssl/rte_openssl_pmd.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/crypto/openssl/rte_openssl_pmd.c
> b/drivers/crypto/openssl/rte_openssl_pmd.c
> > index 05449b6e98..8458ad487a 100644
> > --- a/drivers/crypto/openssl/rte_openssl_pmd.c
> > +++ b/drivers/crypto/openssl/rte_openssl_pmd.c
> > @@ -696,7 +696,7 @@ openssl_set_session_auth_parameters(struct
> > openssl_session *sess,
> > algo = digest_name_get(xform->auth.algo);
> > if (!algo)
> > return -EINVAL;
> > - rte_memcpy(algo_name, algo, (sizeof(algo)+1));
> > + rte_memcpy(algo_name, algo, strlen(algo) + 1);
>
> Why is there a need for copying such a string to a local storage?
From OpenSSL document, I can see OSSL_PARAM_construct_utf8_string() takes a
buffer
as input. But I'm not sure if a const is acceptable.
I would keep the fix simple for backport. Removing copy is an improvement that
can be done by someone who has more knowledge of OpenSSL library.
>
> If it is really needed, we are dealing with strings, so I suggest:
> strlcpy(algo_name, algo, sizeof(algo_name));
Thanks for the suggestion. Will update in v2.
Regards,
Ruifeng
>
>
> --
> David Marchand
