linux-kernel - Re: [PATCH net 09/13] net: stmmac: Remove default maxmtu DT-platform setting

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 for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date:   Tue, 14 Mar 2023 15:03:47 +0300
From:   Serge Semin <fancer.lancer@...il.com>
To:     Piotr Raczynski <piotr.raczynski@...el.com>
Cc:     Serge Semin <Sergey.Semin@...kalelectronics.ru>,
        Giuseppe Cavallaro <peppe.cavallaro@...com>,
        Alexandre Torgue <alexandre.torgue@...s.st.com>,
        Jose Abreu <joabreu@...opsys.com>,
        "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>,
        Russell King <linux@...linux.org.uk>,
        Andrew Lunn <andrew@...n.ch>,
        Heiner Kallweit <hkallweit1@...il.com>,
        Alexey Malahov <Alexey.Malahov@...kalelectronics.ru>,
        Pavel Parkhomenko <Pavel.Parkhomenko@...kalelectronics.ru>,
        Christian Marangi <ansuelsmth@...il.com>,
        Biao Huang <biao.huang@...iatek.com>,
        Yang Yingliang <yangyingliang@...wei.com>,
        netdev@...r.kernel.org, linux-stm32@...md-mailman.stormreply.com,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
        Jose Abreu <Jose.Abreu@...opsys.com>
Subject: Re: [PATCH net 09/13] net: stmmac: Remove default maxmtu DT-platform
 setting

On Tue, Mar 14, 2023 at 12:20:18PM +0100, Piotr Raczynski wrote:
> On Tue, Mar 14, 2023 at 01:42:33AM +0300, Serge Semin wrote:
> > Initializing maxmtu platform parameter in the stmmac_probe_config_dt()
> > method by default makes being pointless the DW MAC-specific maximum MTU
> > selection algorithm implemented in the stmmac_dvr_probe() method. At least
> > for xGMAC we'll always have a frame MTU limited with 9000 while it
> > supports units up to 16KB. Let's remove the default initialization of
> > the maxmtu platform setting then. We don't replace it with setting the
> > maxmtu with some greater value because a default maximum MTU is
> > calculated later in the stmmac_dvr_probe() anyway. That would have been a
> > pointless limitation too. Instead from now the main STMMAC driver code
> > will consider the out of bounds maxmtu value as invalid and will silently
> > replace it with a maximum MTU value specific to the corresponding DW MAC.
> > 
> > Note this alteration will only affect the xGMAC IP-cores due to the way
> > the MTU autodetecion algorithm is implemented. So from now the driver will
> > permit DW xGMACs to handle frames up to 16KB length (XGMAC_JUMBO_LEN). As
> > before DW GMAC IP-cores of v4.0 and higher and IP-cores with enhanced
> > descriptor support will be able to work with frames up to 8KB (JUMBO_LEN).
> > The rest of the NICs will support frames of SKB_MAX_HEAD(NET_SKB_PAD +
> > NET_IP_ALIGN) size.
> > 
> > Fixes: 7d9e6c5afab6 ("net: stmmac: Integrate XGMAC into main driver flow")
> > Signed-off-by: Serge Semin <Sergey.Semin@...kalelectronics.ru>
> > ---
> >  drivers/net/ethernet/stmicro/stmmac/stmmac_main.c     | 4 ----
> >  drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c | 5 -----
> >  2 files changed, 9 deletions(-)
> > 
> > diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> > index 32aa7953d296..e5cb4edc4e23 100644
> > --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> > +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> > @@ -7252,10 +7252,6 @@ int stmmac_dvr_probe(struct device *device,
> >  	if ((priv->plat->maxmtu < ndev->max_mtu) &&
> >  	    (priv->plat->maxmtu >= ndev->min_mtu))
> >  		ndev->max_mtu = priv->plat->maxmtu;

> > -	else if (priv->plat->maxmtu < ndev->min_mtu)
> > -		dev_warn(priv->device,
> > -			 "%s: warning: maxmtu having invalid value (%d)\n",
> > -			 __func__, priv->plat->maxmtu);
> 
> Looks fine but by removing plat->maxmtu = JUMBO_LEN; you eliminate the
> case of dev_warn here or you remove dev_warn since the driver will be
> able to fix the mtu value?

That warning is kind of odd if not to say contradicting. Why having
max_mtu being less than the HW-specific min-value is more dangerous
than having it being greater than the max-value for which there is no
warning? The driver gets to the HW-specific min and max MTU values in
anyway. Why to warn in one case and leaving unnoticed another?..

Anyway getting back to this patch change I remove the warning since
plat_stmmaceth_data.maxmtu is now will be initialized with zero most
of the time (if glue-drivers leave it uninitialized or there is no
"max-frame-size" DT-property specified) which will falsely trigger
that warning.

-Serge(y)

> >  
> >  	if (flow_ctrl)
> >  		priv->flow_ctrl = FLOW_AUTO;	/* RX/TX pause on */
> > diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c
> > index 067a40fe0a23..857411105a0a 100644
> > --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c
> > +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c
> > @@ -468,11 +468,6 @@ stmmac_probe_config_dt(struct platform_device *pdev, u8 *mac)
> >  	plat->en_tx_lpi_clockgating =
> >  		of_property_read_bool(np, "snps,en-tx-lpi-clockgating");
> >  
> > -	/* Set the maxmtu to a default of JUMBO_LEN in case the
> > -	 * parameter is not present in the device tree.
> > -	 */
> > -	plat->maxmtu = JUMBO_LEN;
> > -
> >  	/* Set default value for multicast hash bins */
> >  	plat->multicast_filter_bins = HASH_TABLE_SIZE;
> >  
> > -- 
> > 2.39.2
> > 
> >