Message-ID: <b25f37e4e11d9da5b6d61cbfaa0cafe9889c3926.camel@redhat.com>
Subject: Re: [net-next PATCH V3 4/4] octeontx2-pf: Add support for HTB
 offload
From:   Paolo Abeni <pabeni@redhat.com>
To:     Hariprasad Kelam <hkelam@marvell.com>, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
Cc:     kuba@kernel.org, davem@davemloft.net, edumazet@google.com,
        sgoutham@marvell.com, lcherian@marvell.com, gakula@marvell.com,
        jerinj@marvell.com, sbhatta@marvell.com, jhs@mojatatu.com,
        xiyou.wangcong@gmail.com, jiri@resnulli.us, saeedm@nvidia.com,
        richardcochran@gmail.com, tariqt@nvidia.com,
        linux-rdma@vger.kernel.org, maxtram95@gmail.com,
        naveenm@marvell.com, hariprasad.netdev@gmail.com
Date:   Tue, 07 Feb 2023 13:48:11 +0100
In-Reply-To: <20230206054640.5854-5-hkelam@marvell.com>
References: <20230206054640.5854-1-hkelam@marvell.com>
         <20230206054640.5854-5-hkelam@marvell.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.46.3 (3.46.3-1.fc37) 
MIME-Version: 1.0
Precedence: bulk

On Mon, 2023-02-06 at 11:16 +0530, Hariprasad Kelam wrote:
> +static int otx2_qos_txschq_alloc(struct otx2_nic *pfvf,
> +				 struct otx2_qos_cfg *cfg)
> +{
> +	struct nix_txsch_alloc_req *req;
> +	struct nix_txsch_alloc_rsp *rsp;
> +	struct mbox *mbox =3D &pfvf->mbox;
> +	int lvl, rc, schq;
> +
> +	mutex_lock(&mbox->lock);
> +	req =3D otx2_mbox_alloc_msg_nix_txsch_alloc(&pfvf->mbox);
> +	if (!req)
> +		return -ENOMEM;

This does not releases the mbox->lock mutex on error (another
occurrence below).

[...]


> +static int otx2_qos_txschq_update_config(struct otx2_nic *pfvf,
> +					 struct otx2_qos_node *node,
> +					 struct otx2_qos_cfg *cfg)
> +{
> +	int ret =3D 0;
> +
> +	otx2_qos_txschq_fill_cfg(pfvf, node, cfg);
> +	ret =3D otx2_qos_txschq_push_cfg(pfvf, node, cfg);
> +
> +	return ret;

I personally find the plain:

	return <function>

more easy to read - more instances below.

[...]

> +static void otx2_reset_qdisc(struct net_device *dev, u16 qid)
> +{
> +	struct netdev_queue *dev_queue =3D netdev_get_tx_queue(dev, qid);
> +	struct Qdisc *qdisc =3D dev_queue->qdisc_sleeping;
> +
> +	if (!qdisc)
> +		return;
> +
> +	spin_lock_bh(qdisc_lock(qdisc));
> +	qdisc_reset(qdisc);
> +	spin_unlock_bh(qdisc_lock(qdisc));
> +}

The above looks like a possible shared helper, as mlx code implements a
quite identical function.

Cheers,

Paolo