Hi Ciara,

> -----Original Message-----
> From: Power, Ciara <ciara.po...@intel.com>
> Sent: Wednesday, September 21, 2022 1:51 PM
> To: Akhil Goyal <gak...@marvell.com>; Zhang, Roy Fan
> <roy.fan.zh...@intel.com>
> Cc: dev@dpdk.org; Ji, Kai <kai...@intel.com>; De Lara Guarch, Pablo
> <pablo.de.lara.gua...@intel.com>; Power, Ciara <ciara.po...@intel.com>
> Subject: [PATCH v3 1/5] test/crypto: fix wireless auth digest segment
> 
> The segment size for some tests was too small to hold the auth digest.
> This caused issues when using op->sym->auth.digest.data for comparisons
> in AESNI_MB PMD after a subsequent patch enables SGL.
> 
> For example, if segment size is 2, and digest size is 4, then 4 bytes
> are read from op->sym->auth.digest.data, which overflows into the memory
> after the segment, rather than using the second segment that contains
> the remaining half of the digest.
> 
> Fixes: 11c5485bb276 ("test/crypto: add scatter-gather tests for IP and OOP")
> 
> Signed-off-by: Ciara Power <ciara.po...@intel.com>
> ---
>  app/test/test_cryptodev.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
> 
> diff --git a/app/test/test_cryptodev.c b/app/test/test_cryptodev.c
> index 6ee4480399..5533c135b0 100644
> --- a/app/test/test_cryptodev.c
> +++ b/app/test/test_cryptodev.c
> @@ -3040,6 +3040,14 @@ create_wireless_algo_auth_cipher_operation(
>                       remaining_off -= rte_pktmbuf_data_len(sgl_buf);
>                       sgl_buf = sgl_buf->next;
>               }
> +
> +             /* The last segment should be large enough to hold full digest
> */
> +             if (sgl_buf->data_len < auth_tag_len) {
> +                     rte_pktmbuf_free(sgl_buf->next);
> +                     sgl_buf->next = NULL;
> +                     rte_pktmbuf_append(sgl_buf, auth_tag_len -
> sgl_buf->data_len);

The append shall work once the mbufs are correctly set and mempool is
allocated well. But we should not assume that - it is better to add a simple
check here to make sure the mbuf are appended.

Other than that,
Acked-by: Fan Zhang <roy.fan.zh...@intel.com>

Reply via email to