netdev - Re: [PATCH v3 net-next 5/5] net: ethernet: mtk_eth_soc: add support for page_pool_get

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: <d432c897a8eef451bdd65cfdf5b1da0d866a9a5b.camel@redhat.com>
Date:   Tue, 19 Jul 2022 15:58:11 +0200
From:   Paolo Abeni <pabeni@...hat.com>
To:     Lorenzo Bianconi <lorenzo.bianconi@...hat.com>
Cc:     Lorenzo Bianconi <lorenzo@...nel.org>, netdev@...r.kernel.org,
        nbd@....name, john@...ozen.org, sean.wang@...iatek.com,
        Mark-MC.Lee@...iatek.com, davem@...emloft.net, edumazet@...gle.com,
        kuba@...nel.org, matthias.bgg@...il.com,
        linux-mediatek@...ts.infradead.org, ilias.apalodimas@...aro.org,
        jbrouer@...hat.com
Subject: Re: [PATCH v3 net-next 5/5] net: ethernet: mtk_eth_soc: add support
 for page_pool_get_stats

On Tue, 2022-07-19 at 12:18 +0200, Lorenzo Bianconi wrote:
> > On Sat, 2022-07-16 at 09:34 +0200, Lorenzo Bianconi wrote:
> > > Introduce support for the page_pool stats API into mtk_eth_soc
> > > driver.
> > > Report page_pool stats through ethtool.
> > > 
> > > Signed-off-by: Lorenzo Bianconi <lorenzo@...nel.org>
> > > ---
> > >  drivers/net/ethernet/mediatek/Kconfig       |  1 +
> > >  drivers/net/ethernet/mediatek/mtk_eth_soc.c | 40
> > > +++++++++++++++++++--
> > >  2 files changed, 38 insertions(+), 3 deletions(-)
> > > 
> > > diff --git a/drivers/net/ethernet/mediatek/Kconfig
> > > b/drivers/net/ethernet/mediatek/Kconfig
> > > index d2422c7b31b0..97374fb3ee79 100644
> > > --- a/drivers/net/ethernet/mediatek/Kconfig
> > > +++ b/drivers/net/ethernet/mediatek/Kconfig
> > > @@ -18,6 +18,7 @@ config NET_MEDIATEK_SOC
> > >  	select PHYLINK
> > >  	select DIMLIB
> > >  	select PAGE_POOL
> > > +	select PAGE_POOL_STATS
> > >  	help
> > >  	  This driver supports the gigabit ethernet MACs in the
> > >  	  MediaTek SoC family.
> > > diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> > > b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> > > index abb8bc281015..eba95a86086d 100644
> > > --- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> > > +++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
> > > @@ -3517,11 +3517,19 @@ static void mtk_get_strings(struct
> > > net_device *dev, u32 stringset, u8 *data)
> > >  	int i;
> > >  
> > >  	switch (stringset) {
> > > -	case ETH_SS_STATS:
> > > +	case ETH_SS_STATS: {
> > > +		struct mtk_mac *mac = netdev_priv(dev);
> > > +		struct mtk_eth *eth = mac->hw;
> > > +
> > >  		for (i = 0; i < ARRAY_SIZE(mtk_ethtool_stats);
> > > i++) {
> > >  			memcpy(data, mtk_ethtool_stats[i].str,
> > > ETH_GSTRING_LEN);
> > >  			data += ETH_GSTRING_LEN;
> > >  		}
> > > +		if (!eth->hwlro)
> > 
> > I see the page_pool is enabled if and only if !hwlro, but I think
> > it
> > would be more clear if you explicitly check for page_pool here (and
> > in
> > a few other places below), so that if the condition to enable
> > page_pool
> > someday will change, this code will still be fine.
> 
> Hi Paolo,
> 
> page_pool pointer is defined in mtk_rx_ring structure, so
> theoretically we can have a
> page_pool defined for queue 0 but not for queues {1, 2, 3}.

I see. I missed hwlro is a per device setting.

> "!eth->hwlro" means
> there is at least one page_pool allocated. Do you prefer to do
> something like:
> 
> bool mtk_is_pp_enabled(struct mtk_eth *eth)
> {

> 	for (i = 0; i < ARRAY_SIZE(eth->rx_ring); i++) {
> 		struct mtk_rx_ring *ring = &eth->rx_ring[i];
> 
> 		if (ring->page_pool)
> 			return true;
> 	}
> 	return false;
> }

Even:

bool mtk_is_pp_enabled(struct mtk_eth *eth)
{
	return !eth->hwlro;
}

will suffice to encaspulate the logic behind page pool enabling in a
single place.

/P