linux-kernel - Re: [PATCH net-next v1 3/3] net: stmmac: Optimize cache prefetch in RX path

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]

Message-ID: <f20c339f-5286-477c-9255-e2e1fbeba57c@intel.com>
Date: Mon, 13 Jan 2025 13:10:46 +0100
From: Alexander Lobakin <aleksander.lobakin@...el.com>
To: Furong Xu <0x1207@...il.com>
CC: <netdev@...r.kernel.org>, <linux-stm32@...md-mailman.stormreply.com>,
	<linux-arm-kernel@...ts.infradead.org>, <linux-kernel@...r.kernel.org>,
	Andrew Lunn <andrew+netdev@...n.ch>, "David S. Miller" <davem@...emloft.net>,
	Eric Dumazet <edumazet@...gle.com>, Jakub Kicinski <kuba@...nel.org>, "Paolo
 Abeni" <pabeni@...hat.com>, Maxime Coquelin <mcoquelin.stm32@...il.com>,
	<xfr@...look.com>
Subject: Re: [PATCH net-next v1 3/3] net: stmmac: Optimize cache prefetch in
 RX path

From: Furong Xu <0x1207@...il.com>
Date: Fri, 10 Jan 2025 17:53:59 +0800

> Current code prefetches cache lines for the received frame first, and
> then dma_sync_single_for_cpu() against this frame, this is wrong.
> Cache prefetch should be triggered after dma_sync_single_for_cpu().
> 
> This patch brings ~2.8% driver performance improvement in a TCP RX
> throughput test with iPerf tool on a single isolated Cortex-A65 CPU
> core, 2.84 Gbits/sec increased to 2.92 Gbits/sec.
> 
> Signed-off-by: Furong Xu <0x1207@...il.com>
> ---
>  drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 7 +++----
>  1 file changed, 3 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> index c1aeaec53b4c..1b4e8b035b1a 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> @@ -5497,10 +5497,6 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit, u32 queue)
>  
>  		/* Buffer is good. Go on. */
>  
> -		prefetch(page_address(buf->page) + buf->page_offset);
> -		if (buf->sec_page)
> -			prefetch(page_address(buf->sec_page));
> -
>  		buf1_len = stmmac_rx_buf1_len(priv, p, status, len);
>  		len += buf1_len;
>  		buf2_len = stmmac_rx_buf2_len(priv, p, status, len);
> @@ -5522,6 +5518,7 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit, u32 queue)
>  
>  			dma_sync_single_for_cpu(priv->device, buf->addr,
>  						buf1_len, dma_dir);
> +			prefetch(page_address(buf->page) + buf->page_offset);
>  
>  			xdp_init_buff(&ctx.xdp, buf_sz, &rx_q->xdp_rxq);
>  			xdp_prepare_buff(&ctx.xdp, page_address(buf->page),
> @@ -5596,6 +5593,7 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit, u32 queue)
>  		} else if (buf1_len) {
>  			dma_sync_single_for_cpu(priv->device, buf->addr,
>  						buf1_len, dma_dir);
> +			prefetch(page_address(buf->page) + buf->page_offset);
>  			skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags,
>  					buf->page, buf->page_offset, buf1_len,
>  					priv->dma_conf.dma_buf_sz);

Are you sure you need to prefetch frags as well? I'd say this is a waste
of cycles, as the kernel core stack barely looks at payload...
Probably prefetching only header buffers would be enough.

> @@ -5608,6 +5606,7 @@ static int stmmac_rx(struct stmmac_priv *priv, int limit, u32 queue)
>  		if (buf2_len) {
>  			dma_sync_single_for_cpu(priv->device, buf->sec_addr,
>  						buf2_len, dma_dir);
> +			prefetch(page_address(buf->sec_page));
>  			skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags,
>  					buf->sec_page, 0, buf2_len,
>  					priv->dma_conf.dma_buf_sz);

Thanks,
Olek