Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit bc12ff7f authored by Sean Tranchetti's avatar Sean Tranchetti Committed by Gerrit - the friendly Code Review server
Browse files

net: qualcomm: rmnet: flush uplink aggregation on priority tx 



If the RmNet driver needs to transmit a packet marked as needing
prioritization, it should skip uplink aggregation of the packet to avoid
adding additional delay. Additionally, any aggregated packets should be
flushed prior to sending the prioritized packet to avoid any ordering
issues.

Change-Id: Ia3063ec3f0553693a0f66c8eb5bbb88d250f7523
Signed-off-by: default avatarSean Tranchetti <stranche@codeaurora.org>
parent 6a26cfeb

drivers/net/ethernet/qualcomm/rmnet/rmnet_map_data.c

+32 −11
Original line number Diff line number Diff line
@@ -1401,11 +1401,30 @@ static struct sk_buff *rmnet_map_build_skb(struct rmnet_port *port) 
	return skb;

}



static void rmnet_map_send_agg_skb(struct rmnet_port *port, unsigned long flags)

{

	struct sk_buff *agg_skb;



	if (!port->agg_skb) {

		spin_unlock_irqrestore(&port->agg_lock, flags);

		return;

	}



	agg_skb = port->agg_skb;

	/* Reset the aggregation state */

	port->agg_skb = NULL;

	port->agg_count = 0;

	memset(&port->agg_time, 0, sizeof(struct timespec));

	port->agg_state = 0;

	spin_unlock_irqrestore(&port->agg_lock, flags);

	hrtimer_cancel(&port->hrtimer);

	dev_queue_xmit(agg_skb);

}



void rmnet_map_tx_aggregate(struct sk_buff *skb, struct rmnet_port *port)

{

	struct timespec diff, last;

	int size, agg_count = 0;

	struct sk_buff *agg_skb;

	int size;

	unsigned long flags;



new_packet:
@@ -1413,6 +1432,16 @@ void rmnet_map_tx_aggregate(struct sk_buff *skb, struct rmnet_port *port) 
	memcpy(&last, &port->agg_last, sizeof(struct timespec));

	getnstimeofday(&port->agg_last);



	if ((port->data_format & RMNET_EGRESS_FORMAT_PRIORITY) &&

	    skb->priority) {

		/* Send out any aggregated SKBs we have */

		rmnet_map_send_agg_skb(port, flags);

		/* Send out the priority SKB. Not holding agg_lock anymore */

		skb->protocol = htons(ETH_P_MAP);

		dev_queue_xmit(skb);

		return;

	}



	if (!port->agg_skb) {

		/* Check to see if we should agg first. If the traffic is very

		 * sparse, don't aggregate. We will need to tune this later
@@ -1452,15 +1481,7 @@ void rmnet_map_tx_aggregate(struct sk_buff *skb, struct rmnet_port *port) 
	if (skb->len > size ||

	    port->agg_count >= port->egress_agg_params.agg_count ||

	    diff.tv_sec > 0 || diff.tv_nsec > rmnet_agg_time_limit) {

		agg_skb = port->agg_skb;

		agg_count = port->agg_count;

		port->agg_skb = 0;

		port->agg_count = 0;

		memset(&port->agg_time, 0, sizeof(struct timespec));

		port->agg_state = 0;

		spin_unlock_irqrestore(&port->agg_lock, flags);

		hrtimer_cancel(&port->hrtimer);

		dev_queue_xmit(agg_skb);

		rmnet_map_send_agg_skb(port, flags);

		goto new_packet;

	}