netdev - Re: [PATCH v2 net-next 14/19] ionic: Add Tx and Rx handling

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:   Mon, 1 Jul 2019 11:17:08 -0700
From:   Shannon Nelson <snelson@...sando.io>
To:     Jakub Kicinski <jakub.kicinski@...ronome.com>
Cc:     netdev@...r.kernel.org
Subject: Re: [PATCH v2 net-next 14/19] ionic: Add Tx and Rx handling

On 6/29/19 11:57 AM, Jakub Kicinski wrote:
> On Fri, 28 Jun 2019 14:39:29 -0700, Shannon Nelson wrote:
>> +static int ionic_tx(struct queue *q, struct sk_buff *skb)
>> +{
>> +	struct tx_stats *stats = q_to_tx_stats(q);
>> +	int err;
>> +
>> +	if (skb->ip_summed == CHECKSUM_PARTIAL)
>> +		err = ionic_tx_calc_csum(q, skb);
>> +	else
>> +		err = ionic_tx_calc_no_csum(q, skb);
>> +	if (err)
>> +		return err;
>> +
>> +	err = ionic_tx_skb_frags(q, skb);
>> +	if (err)
>> +		return err;
>> +
>> +	skb_tx_timestamp(skb);
>> +	stats->pkts++;
>> +	stats->bytes += skb->len;
> Presumably this is 64bit so you should use
> u64_stats_update_begin()
> u64_stats_update_end()
> around it (and all other stats).

Since the device won't work in a 32-bit arch and I have the the Kconfig 
set to depend on 64BIT, I wasn't sure I needed to bother with the extra 
syntactic sugar.

>> +
>> +	ionic_txq_post(q, !netdev_xmit_more(), ionic_tx_clean, skb);
>> +
>> +	return 0;
>> +}
>> +
>> +static int ionic_tx_descs_needed(struct queue *q, struct sk_buff *skb)
>> +{
>> +	struct tx_stats *stats = q_to_tx_stats(q);
>> +	int err;
>> +
>> +	/* If TSO, need roundup(skb->len/mss) descs */
>> +	if (skb_is_gso(skb))
>> +		return (skb->len / skb_shinfo(skb)->gso_size) + 1;
>> +
>> +	/* If non-TSO, just need 1 desc and nr_frags sg elems */
>> +	if (skb_shinfo(skb)->nr_frags <= IONIC_TX_MAX_SG_ELEMS)
>> +		return 1;
>> +
>> +	/* Too many frags, so linearize */
>> +	err = skb_linearize(skb);
>> +	if (err)
>> +		return err;
>> +
>> +	stats->linearize++;
>> +
>> +	/* Need 1 desc and zero sg elems */
>> +	return 1;
>> +}
>> +
>> +netdev_tx_t ionic_start_xmit(struct sk_buff *skb, struct net_device *netdev)
>> +{
>> +	u16 queue_index = skb_get_queue_mapping(skb);
>> +	struct lif *lif = netdev_priv(netdev);
>> +	struct queue *q;
>> +	int ndescs;
>> +	int err;
>> +
>> +	if (unlikely(!test_bit(LIF_UP, lif->state))) {
>> +		dev_kfree_skb(skb);
>> +		return NETDEV_TX_OK;
>> +	}
>> +
>> +	if (likely(lif_to_txqcq(lif, queue_index)))
>> +		q = lif_to_txq(lif, queue_index);
>> +	else
>> +		q = lif_to_txq(lif, 0);
>> +
>> +	ndescs = ionic_tx_descs_needed(q, skb);
>> +	if (ndescs < 0)
>> +		goto err_out_drop;
>> +
>> +	if (!ionic_q_has_space(q, ndescs)) {
>> +		netif_stop_subqueue(netdev, queue_index);
>> +		q->stop++;
>> +
>> +		/* Might race with ionic_tx_clean, check again */
>> +		smp_rmb();
>> +		if (ionic_q_has_space(q, ndescs)) {
>> +			netif_wake_subqueue(netdev, queue_index);
>> +			q->wake++;
>> +		} else {
>> +			return NETDEV_TX_BUSY;
> This should never really happen..

Couldn't we have an Rx interrupt and a call to ionic_tx_clean() in the 
middle of this?

>
>> +		}
>> +	}
>> +
>> +	if (skb_is_gso(skb))
>> +		err = ionic_tx_tso(q, skb);
>> +	else
>> +		err = ionic_tx(q, skb);
>> +
>> +	if (err)
>> +		goto err_out_drop;
> .. at this point if you can't guarantee fitting biggest possible frame
> in, you have to stop the ring.

Yep, that would work.

Thanks,
sln