DPDK patches and discussions
 help / color / mirror / Atom feed
* [PATCH] event/dlb2: fix invalid shift value
@ 2022-03-02 15:12 Timothy McDaniel
  2022-03-03  9:25 ` Jerin Jacob
  0 siblings, 1 reply; 2+ messages in thread
From: Timothy McDaniel @ 2022-03-02 15:12 UTC (permalink / raw)
  To: jerinj; +Cc: john.mcnamara, dev

Add a check to ensure that all shift counts are valid.
Shifting by more than 63 bits may result in undefined behavior, as
noted during coverity scan.

Fixes: e697f35dbdd1 ("event/dlb2: update rolling mask used for dequeue")
Coverity issue: 376527

Signed-off-by: Timothy McDaniel <timothy.mcdaniel@intel.com>
---
 drivers/event/dlb2/dlb2.c | 18 ++++++++++--------
 1 file changed, 10 insertions(+), 8 deletions(-)

diff --git a/drivers/event/dlb2/dlb2.c b/drivers/event/dlb2/dlb2.c
index 09abdd1660..7789dd74e0 100644
--- a/drivers/event/dlb2/dlb2.c
+++ b/drivers/event/dlb2/dlb2.c
@@ -3916,15 +3916,17 @@ dlb2_hw_dequeue_sparse(struct dlb2_eventdev *dlb2,
 								&events[num],
 								&qes[0],
 								num_avail);
-			num += n_iter;
-			/* update rolling_mask for vector code support */
-			m_rshift = qm_port->cq_rolling_mask >> n_iter;
-			m_lshift = qm_port->cq_rolling_mask << (64 - n_iter);
-			m2_rshift = qm_port->cq_rolling_mask_2 >> n_iter;
-			m2_lshift = qm_port->cq_rolling_mask_2 <<
+			if (n_iter != 0) {
+				num += n_iter;
+				/* update rolling_mask for vector code support */
+				m_rshift = qm_port->cq_rolling_mask >> n_iter;
+				m_lshift = qm_port->cq_rolling_mask << (64 - n_iter);
+				m2_rshift = qm_port->cq_rolling_mask_2 >> n_iter;
+				m2_lshift = qm_port->cq_rolling_mask_2 <<
 					(64 - n_iter);
-			qm_port->cq_rolling_mask = (m_rshift | m2_lshift);
-			qm_port->cq_rolling_mask_2 = (m2_rshift | m_lshift);
+				qm_port->cq_rolling_mask = (m_rshift | m2_lshift);
+				qm_port->cq_rolling_mask_2 = (m2_rshift | m_lshift);
+			}
 		} else { /* !use_scalar */
 			num_avail = dlb2_recv_qe_sparse_vec(qm_port,
 							    &events[num],
-- 
2.25.1


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [PATCH] event/dlb2: fix invalid shift value
  2022-03-02 15:12 [PATCH] event/dlb2: fix invalid shift value Timothy McDaniel
@ 2022-03-03  9:25 ` Jerin Jacob
  0 siblings, 0 replies; 2+ messages in thread
From: Jerin Jacob @ 2022-03-03  9:25 UTC (permalink / raw)
  To: Timothy McDaniel; +Cc: Jerin Jacob, John McNamara, dpdk-dev

On Wed, Mar 2, 2022 at 8:42 PM Timothy McDaniel
<timothy.mcdaniel@intel.com> wrote:
>
> Add a check to ensure that all shift counts are valid.
> Shifting by more than 63 bits may result in undefined behavior, as
> noted during coverity scan.
>
> Fixes: e697f35dbdd1 ("event/dlb2: update rolling mask used for dequeue")
> Coverity issue: 376527


Applied to dpdk-next-net-eventdev/for-main. Thanks


> Signed-off-by: Timothy McDaniel <timothy.mcdaniel@intel.com>
> ---
>  drivers/event/dlb2/dlb2.c | 18 ++++++++++--------
>  1 file changed, 10 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/event/dlb2/dlb2.c b/drivers/event/dlb2/dlb2.c
> index 09abdd1660..7789dd74e0 100644
> --- a/drivers/event/dlb2/dlb2.c
> +++ b/drivers/event/dlb2/dlb2.c
> @@ -3916,15 +3916,17 @@ dlb2_hw_dequeue_sparse(struct dlb2_eventdev *dlb2,
>                                                                 &events[num],
>                                                                 &qes[0],
>                                                                 num_avail);
> -                       num += n_iter;
> -                       /* update rolling_mask for vector code support */
> -                       m_rshift = qm_port->cq_rolling_mask >> n_iter;
> -                       m_lshift = qm_port->cq_rolling_mask << (64 - n_iter);
> -                       m2_rshift = qm_port->cq_rolling_mask_2 >> n_iter;
> -                       m2_lshift = qm_port->cq_rolling_mask_2 <<
> +                       if (n_iter != 0) {
> +                               num += n_iter;
> +                               /* update rolling_mask for vector code support */
> +                               m_rshift = qm_port->cq_rolling_mask >> n_iter;
> +                               m_lshift = qm_port->cq_rolling_mask << (64 - n_iter);
> +                               m2_rshift = qm_port->cq_rolling_mask_2 >> n_iter;
> +                               m2_lshift = qm_port->cq_rolling_mask_2 <<
>                                         (64 - n_iter);
> -                       qm_port->cq_rolling_mask = (m_rshift | m2_lshift);
> -                       qm_port->cq_rolling_mask_2 = (m2_rshift | m_lshift);
> +                               qm_port->cq_rolling_mask = (m_rshift | m2_lshift);
> +                               qm_port->cq_rolling_mask_2 = (m2_rshift | m_lshift);
> +                       }
>                 } else { /* !use_scalar */
>                         num_avail = dlb2_recv_qe_sparse_vec(qm_port,
>                                                             &events[num],
> --
> 2.25.1
>

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2022-03-03  9:25 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-03-02 15:12 [PATCH] event/dlb2: fix invalid shift value Timothy McDaniel
2022-03-03  9:25 ` Jerin Jacob

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).