netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Florian Fainelli <f.fainelli@gmail.com>
To: Petri Gynther <pgynther@google.com>, netdev@vger.kernel.org
Cc: davem@davemloft.net, jaedon.shin@gmail.com
Subject: Re: [PATCH net-next v2] net: bcmgenet: simplify __bcmgenet_tx_reclaim()
Date: Wed, 04 Mar 2015 15:18:24 -0800	[thread overview]
Message-ID: <54F792C0.9060204@gmail.com> (raw)
In-Reply-To: <20150304223002.0233E22021A@puck.mtv.corp.google.com>

On 04/03/15 14:30, Petri Gynther wrote:
> 1. Use c_index and ring->c_index to determine how many TxCBs/TxBDs are
>    ready for cleanup
>    - c_index = the current value of TDMA_CONS_INDEX
>    - TDMA_CONS_INDEX is HW-incremented and auto-wraparound (0x0-0xFFFF)
>    - ring->c_index = __bcmgenet_tx_reclaim() cleaned up to this point on
>      the previous invocation
> 
> 2. Add bcmgenet_tx_ring->clean_ptr
>    - index of the next TxCB to be cleaned
>    - incremented as TxCBs/TxBDs are processed
>    - value always in range [ring->cb_ptr, ring->end_ptr]
> 
> 3. Fix incrementing of dev->stats.tx_packets
>    - should be incremented only when tx_cb_ptr->skb != NULL
> 
> These changes simplify __bcmgenet_tx_reclaim(). Furthermore, Tx ring size
> can now be any value.
> 
> With the old code, Tx ring size had to be a power-of-2:
>    num_tx_bds = ring->size;
>    c_index &= (num_tx_bds - 1);
>    last_c_index &= (num_tx_bds - 1);
> 
> Signed-off-by: Petri Gynther <pgynther@google.com>

Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>

Thanks!

> ---
>  drivers/net/ethernet/broadcom/genet/bcmgenet.c | 45 ++++++++++++--------------
>  drivers/net/ethernet/broadcom/genet/bcmgenet.h |  1 +
>  2 files changed, 22 insertions(+), 24 deletions(-)
> 
> diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.c b/drivers/net/ethernet/broadcom/genet/bcmgenet.c
> index 84feb24..9137187 100644
> --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.c
> +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.c
> @@ -978,39 +978,32 @@ static unsigned int __bcmgenet_tx_reclaim(struct net_device *dev,
>  					  struct bcmgenet_tx_ring *ring)
>  {
>  	struct bcmgenet_priv *priv = netdev_priv(dev);
> -	int last_tx_cn, last_c_index, num_tx_bds;
>  	struct enet_cb *tx_cb_ptr;
>  	struct netdev_queue *txq;
>  	unsigned int pkts_compl = 0;
> -	unsigned int bds_compl;
>  	unsigned int c_index;
> +	unsigned int txbds_ready;
> +	unsigned int txbds_processed = 0;
>  
>  	/* Compute how many buffers are transmitted since last xmit call */
>  	c_index = bcmgenet_tdma_ring_readl(priv, ring->index, TDMA_CONS_INDEX);
> -	txq = netdev_get_tx_queue(dev, ring->queue);
> -
> -	last_c_index = ring->c_index;
> -	num_tx_bds = ring->size;
> -
> -	c_index &= (num_tx_bds - 1);
> +	c_index &= DMA_C_INDEX_MASK;
>  
> -	if (c_index >= last_c_index)
> -		last_tx_cn = c_index - last_c_index;
> +	if (likely(c_index >= ring->c_index))
> +		txbds_ready = c_index - ring->c_index;
>  	else
> -		last_tx_cn = num_tx_bds - last_c_index + c_index;
> +		txbds_ready = (DMA_C_INDEX_MASK + 1) - ring->c_index + c_index;
>  
>  	netif_dbg(priv, tx_done, dev,
> -		  "%s ring=%d index=%d last_tx_cn=%d last_index=%d\n",
> -		  __func__, ring->index,
> -		  c_index, last_tx_cn, last_c_index);
> +		  "%s ring=%d old_c_index=%u c_index=%u txbds_ready=%u\n",
> +		  __func__, ring->index, ring->c_index, c_index, txbds_ready);
>  
>  	/* Reclaim transmitted buffers */
> -	while (last_tx_cn-- > 0) {
> -		tx_cb_ptr = ring->cbs + last_c_index;
> -		bds_compl = 0;
> +	while (txbds_processed < txbds_ready) {
> +		tx_cb_ptr = &priv->tx_cbs[ring->clean_ptr];
>  		if (tx_cb_ptr->skb) {
>  			pkts_compl++;
> -			bds_compl = skb_shinfo(tx_cb_ptr->skb)->nr_frags + 1;
> +			dev->stats.tx_packets++;
>  			dev->stats.tx_bytes += tx_cb_ptr->skb->len;
>  			dma_unmap_single(&dev->dev,
>  					 dma_unmap_addr(tx_cb_ptr, dma_addr),
> @@ -1026,20 +1019,23 @@ static unsigned int __bcmgenet_tx_reclaim(struct net_device *dev,
>  				       DMA_TO_DEVICE);
>  			dma_unmap_addr_set(tx_cb_ptr, dma_addr, 0);
>  		}
> -		dev->stats.tx_packets++;
> -		ring->free_bds += bds_compl;
>  
> -		last_c_index++;
> -		last_c_index &= (num_tx_bds - 1);
> +		txbds_processed++;
> +		if (likely(ring->clean_ptr < ring->end_ptr))
> +			ring->clean_ptr++;
> +		else
> +			ring->clean_ptr = ring->cb_ptr;
>  	}
>  
> +	ring->free_bds += txbds_processed;
> +	ring->c_index = (ring->c_index + txbds_processed) & DMA_C_INDEX_MASK;
> +
>  	if (ring->free_bds > (MAX_SKB_FRAGS + 1)) {
> +		txq = netdev_get_tx_queue(dev, ring->queue);
>  		if (netif_tx_queue_stopped(txq))
>  			netif_tx_wake_queue(txq);
>  	}
>  
> -	ring->c_index = c_index;
> -
>  	return pkts_compl;
>  }
>  
> @@ -1734,6 +1730,7 @@ static void bcmgenet_init_tx_ring(struct bcmgenet_priv *priv,
>  	}
>  	ring->cbs = priv->tx_cbs + start_ptr;
>  	ring->size = size;
> +	ring->clean_ptr = start_ptr;
>  	ring->c_index = 0;
>  	ring->free_bds = size;
>  	ring->write_ptr = start_ptr;
> diff --git a/drivers/net/ethernet/broadcom/genet/bcmgenet.h b/drivers/net/ethernet/broadcom/genet/bcmgenet.h
> index 016bd12..548b7e9 100644
> --- a/drivers/net/ethernet/broadcom/genet/bcmgenet.h
> +++ b/drivers/net/ethernet/broadcom/genet/bcmgenet.h
> @@ -525,6 +525,7 @@ struct bcmgenet_tx_ring {
>  	unsigned int	queue;		/* queue index */
>  	struct enet_cb	*cbs;		/* tx ring buffer control block*/
>  	unsigned int	size;		/* size of each tx ring */
> +	unsigned int    clean_ptr;      /* Tx ring clean pointer */
>  	unsigned int	c_index;	/* last consumer index of each ring*/
>  	unsigned int	free_bds;	/* # of free bds for each ring */
>  	unsigned int	write_ptr;	/* Tx ring write pointer SW copy */
> 


-- 
Florian

  reply	other threads:[~2015-03-04 23:18 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-03-04 22:30 [PATCH net-next v2] net: bcmgenet: simplify __bcmgenet_tx_reclaim() Petri Gynther
2015-03-04 23:18 ` Florian Fainelli [this message]
2015-03-05  4:55 ` David Miller

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=54F792C0.9060204@gmail.com \
    --to=f.fainelli@gmail.com \
    --cc=davem@davemloft.net \
    --cc=jaedon.shin@gmail.com \
    --cc=netdev@vger.kernel.org \
    --cc=pgynther@google.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).