[dpdk-dev] net/mlx5: fix MTU update Changing the MTU is not related to changing the number of segments, activating or not the multi-segment support should be handled by the application.
Checks
Commit Message
This is a backport based on commit
a0edafe4099b1ef139242abb7baa2c2a48b83fd2
Signed-off-by: Erez Ferber <erezf@mellanox.com>
---
drivers/net/mlx5/mlx5_ethdev.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
Comments
Hi Erez,
On Sun, Jan 28, 2018 at 03:39:47PM +0200, Erez Ferber wrote:
> This is a backport based on commit
> a0edafe4099b1ef139242abb7baa2c2a48b83fd2
There is some issues in this commit, first the title line exceed the
75 characters.
Second it is submitted on the dev mailing list whereas it seems to be
targeting a stable branch, in should have been posted on
stable@dpdk.org and according to the sha-1 I would say the target is
the 17.08.
Can you make things clearer?
> Signed-off-by: Erez Ferber <erezf@mellanox.com>
> ---
> drivers/net/mlx5/mlx5_ethdev.c | 4 +---
> 1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/drivers/net/mlx5/mlx5_ethdev.c b/drivers/net/mlx5/mlx5_ethdev.c
> index c0f73e9..f73d26e 100644
> --- a/drivers/net/mlx5/mlx5_ethdev.c
> +++ b/drivers/net/mlx5/mlx5_ethdev.c
> @@ -930,9 +930,7 @@ recover:
> /* Provide new values to rxq_setup(). */
> dev->data->dev_conf.rxmode.jumbo_frame = sp;
> dev->data->dev_conf.rxmode.max_rx_pkt_len = max_frame_len;
> - if (rehash)
> - ret = rxq_rehash(dev, rxq_ctrl);
> - else
> + if (!rehash)
> ret = rxq_ctrl_setup(dev, rxq_ctrl, 1 << rxq->elts_n,
> rxq_ctrl->socket, NULL, rxq->mp);
> if (!ret)
> --
> 1.8.2.3
@@ -930,9 +930,7 @@ recover:
/* Provide new values to rxq_setup(). */
dev->data->dev_conf.rxmode.jumbo_frame = sp;
dev->data->dev_conf.rxmode.max_rx_pkt_len = max_frame_len;
- if (rehash)
- ret = rxq_rehash(dev, rxq_ctrl);
- else
+ if (!rehash)
ret = rxq_ctrl_setup(dev, rxq_ctrl, 1 << rxq->elts_n,
rxq_ctrl->socket, NULL, rxq->mp);
if (!ret)