netdev - Re: [PATCH v2 net-next] net: socionext: get rid of huge dma sync in netsec_alloc_rx

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] [thread-next>] [day] [month] [year] [list]
Date:   Fri, 10 Jan 2020 16:56:31 +0200
From:   Ilias Apalodimas <ilias.apalodimas@...aro.org>
To:     Lorenzo Bianconi <lorenzo@...nel.org>
Cc:     netdev@...r.kernel.org, brouer@...hat.com, davem@...emloft.net,
        lorenzo.bianconi@...hat.com
Subject: Re: [PATCH v2 net-next] net: socionext: get rid of huge dma sync in
 netsec_alloc_rx_data

On Fri, Jan 10, 2020 at 02:57:44PM +0100, Lorenzo Bianconi wrote:
> Socionext driver can run on dma coherent and non-coherent devices.
> Get rid of huge dma_sync_single_for_device in netsec_alloc_rx_data since
> now the driver can let page_pool API to managed needed DMA sync
> 
> Signed-off-by: Lorenzo Bianconi <lorenzo@...nel.org>
> ---
> Changes since v1:
> - rely on original frame size for dma sync
> ---
>  drivers/net/ethernet/socionext/netsec.c | 43 +++++++++++++++----------
>  1 file changed, 26 insertions(+), 17 deletions(-)
> 
> diff --git a/drivers/net/ethernet/socionext/netsec.c b/drivers/net/ethernet/socionext/netsec.c
> index b5a9e947a4a8..45c76b437457 100644
> --- a/drivers/net/ethernet/socionext/netsec.c
> +++ b/drivers/net/ethernet/socionext/netsec.c
> @@ -243,6 +243,7 @@
>  			       NET_IP_ALIGN)
>  #define NETSEC_RX_BUF_NON_DATA (NETSEC_RXBUF_HEADROOM + \
>  				SKB_DATA_ALIGN(sizeof(struct skb_shared_info)))
> +#define NETSEC_RX_BUF_SIZE	(PAGE_SIZE - NETSEC_RX_BUF_NON_DATA)
>  
>  #define DESC_SZ	sizeof(struct netsec_de)
>  
> @@ -719,7 +720,6 @@ static void *netsec_alloc_rx_data(struct netsec_priv *priv,
>  {
>  
>  	struct netsec_desc_ring *dring = &priv->desc_ring[NETSEC_RING_RX];
> -	enum dma_data_direction dma_dir;
>  	struct page *page;
>  
>  	page = page_pool_dev_alloc_pages(dring->page_pool);
> @@ -734,9 +734,7 @@ static void *netsec_alloc_rx_data(struct netsec_priv *priv,
>  	/* Make sure the incoming payload fits in the page for XDP and non-XDP
>  	 * cases and reserve enough space for headroom + skb_shared_info
>  	 */
> -	*desc_len = PAGE_SIZE - NETSEC_RX_BUF_NON_DATA;
> -	dma_dir = page_pool_get_dma_dir(dring->page_pool);
> -	dma_sync_single_for_device(priv->dev, *dma_handle, *desc_len, dma_dir);
> +	*desc_len = NETSEC_RX_BUF_SIZE;
>  
>  	return page_address(page);
>  }
> @@ -883,6 +881,8 @@ static u32 netsec_xdp_xmit_back(struct netsec_priv *priv, struct xdp_buff *xdp)
>  static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog,
>  			  struct xdp_buff *xdp)
>  {
> +	struct netsec_desc_ring *dring = &priv->desc_ring[NETSEC_RING_RX];
> +	unsigned int len = xdp->data_end - xdp->data;

We need to account for XDP expanding the headers as well here. 
So something like max(xdp->data_end(before bpf), xdp->data_end(after bpf)) -
xdp->data (original)

>  	u32 ret = NETSEC_XDP_PASS;
>  	int err;
>  	u32 act;
> @@ -896,7 +896,9 @@ static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog,
>  	case XDP_TX:
>  		ret = netsec_xdp_xmit_back(priv, xdp);
>  		if (ret != NETSEC_XDP_TX)
> -			xdp_return_buff(xdp);
> +			__page_pool_put_page(dring->page_pool,
> +				     virt_to_head_page(xdp->data),
> +				     len, true);
>  		break;
>  	case XDP_REDIRECT:
>  		err = xdp_do_redirect(priv->ndev, xdp, prog);
> @@ -904,7 +906,9 @@ static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog,
>  			ret = NETSEC_XDP_REDIR;
>  		} else {
>  			ret = NETSEC_XDP_CONSUMED;
> -			xdp_return_buff(xdp);
> +			__page_pool_put_page(dring->page_pool,
> +				     virt_to_head_page(xdp->data),
> +				     len, true);
>  		}
>  		break;
>  	default:
> @@ -915,7 +919,9 @@ static u32 netsec_run_xdp(struct netsec_priv *priv, struct bpf_prog *prog,
>  		/* fall through -- handle aborts by dropping packet */
>  	case XDP_DROP:
>  		ret = NETSEC_XDP_CONSUMED;
> -		xdp_return_buff(xdp);
> +		__page_pool_put_page(dring->page_pool,
> +				     virt_to_head_page(xdp->data),
> +				     len, true);
>  		break;
>  	}
>  
> @@ -1014,7 +1020,8 @@ static int netsec_process_rx(struct netsec_priv *priv, int budget)
>  			 * cache state. Since we paid the allocation cost if
>  			 * building an skb fails try to put the page into cache
>  			 */
> -			page_pool_recycle_direct(dring->page_pool, page);
> +			__page_pool_put_page(dring->page_pool, page,
> +					     pkt_len, true);

Same here, a bpf prog with XDP_PASS verdict might change lenghts

>  			netif_err(priv, drv, priv->ndev,
>  				  "rx failed to build skb\n");
>  			break;
> @@ -1272,17 +1279,19 @@ static int netsec_setup_rx_dring(struct netsec_priv *priv)
>  {
>  	struct netsec_desc_ring *dring = &priv->desc_ring[NETSEC_RING_RX];
>  	struct bpf_prog *xdp_prog = READ_ONCE(priv->xdp_prog);
> -	struct page_pool_params pp_params = { 0 };
> +	struct page_pool_params pp_params = {
> +		.order = 0,
> +		/* internal DMA mapping in page_pool */
> +		.flags = PP_FLAG_DMA_MAP | PP_FLAG_DMA_SYNC_DEV,
> +		.pool_size = DESC_NUM,
> +		.nid = NUMA_NO_NODE,
> +		.dev = priv->dev,
> +		.dma_dir = xdp_prog ? DMA_BIDIRECTIONAL : DMA_FROM_DEVICE,
> +		.offset = NETSEC_RXBUF_HEADROOM,
> +		.max_len = NETSEC_RX_BUF_SIZE,
> +	};
>  	int i, err;
>  
> -	pp_params.order = 0;
> -	/* internal DMA mapping in page_pool */
> -	pp_params.flags = PP_FLAG_DMA_MAP;
> -	pp_params.pool_size = DESC_NUM;
> -	pp_params.nid = NUMA_NO_NODE;
> -	pp_params.dev = priv->dev;
> -	pp_params.dma_dir = xdp_prog ? DMA_BIDIRECTIONAL : DMA_FROM_DEVICE;
> -
>  	dring->page_pool = page_pool_create(&pp_params);
>  	if (IS_ERR(dring->page_pool)) {
>  		err = PTR_ERR(dring->page_pool);
> -- 
> 2.21.1
> 

Thanks
/Ilias