[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20200615145115.GA23038@Red>
Date: Mon, 15 Jun 2020 16:51:15 +0200
From: LABBE Corentin <clabbe@...libre.com>
To: Charles Keepax <ckeepax@...nsource.cirrus.com>
Cc: davem@...emloft.net, nicolas.ferre@...rochip.com, kuba@...nel.org,
netdev@...r.kernel.org
Subject: Re: [PATCH] net: macb: Only disable NAPI on the actual error path
On Mon, Jun 15, 2020 at 02:18:54PM +0100, Charles Keepax wrote:
> A recent change added a disable to NAPI into macb_open, this was
> intended to only happen on the error path but accidentally applies
> to all paths. This causes NAPI to be disabled on the success path, which
> leads to the network to no longer functioning.
>
> Fixes: 014406babc1f ("net: cadence: macb: disable NAPI on error")
> Signed-off-by: Charles Keepax <ckeepax@...nsource.cirrus.com>
> ---
> drivers/net/ethernet/cadence/macb_main.c | 9 ++++-----
> 1 file changed, 4 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c
> index 5b9d7c60eebc0..67933079aeea5 100644
> --- a/drivers/net/ethernet/cadence/macb_main.c
> +++ b/drivers/net/ethernet/cadence/macb_main.c
> @@ -2565,15 +2565,14 @@ static int macb_open(struct net_device *dev)
> if (bp->ptp_info)
> bp->ptp_info->ptp_init(dev);
>
> + return 0;
> +
> napi_exit:
> for (q = 0, queue = bp->queues; q < bp->num_queues; ++q, ++queue)
> napi_disable(&queue->napi);
> pm_exit:
> - if (err) {
> - pm_runtime_put_sync(&bp->pdev->dev);
> - return err;
> - }
> - return 0;
> + pm_runtime_put_sync(&bp->pdev->dev);
> + return err;
> }
>
> static int macb_close(struct net_device *dev)
> --
> 2.11.0
>
Shame on me to not have see this.
With the fix it works.
Tested-by: Corentin Labbe <clabbe@...libre.com>
Tested-on: qemu-xilinx-zynq-a9
Powered by blists - more mailing lists