netdev - Re: [PATCH net] net: txgbe: fix the issue of TX failture

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] [day] [month] [year] [list]

Message-ID: <20250625165920.GH1562@horms.kernel.org>
Date: Wed, 25 Jun 2025 17:59:20 +0100
From: Simon Horman <horms@...nel.org>
To: Jiawen Wu <jiawenwu@...stnetic.com>
Cc: netdev@...r.kernel.org, andrew+netdev@...n.ch, davem@...emloft.net,
	edumazet@...gle.com, kuba@...nel.org, pabeni@...hat.com,
	mengyuanlou@...-swift.com
Subject: Re: [PATCH net] net: txgbe: fix the issue of TX failture

On Wed, Jun 25, 2025 at 05:51:19PM +0800, Jiawen Wu wrote:
> On Wed, Jun 25, 2025 5:36 PM, Simon Horman wrote:
> > On Tue, Jun 24, 2025 at 05:30:21PM +0800, Jiawen Wu wrote:
> > > There is a occasional problem that ping is failed between AML devices.
> > > That is because the manual enablement of the security Tx path on the
> > > hardware is missing, no matter what its previous state was.
> > >
> > > Fixes: 6f8b4c01a8cd ("net: txgbe: Implement PHYLINK for AML 25G/10G devices")
> > > Signed-off-by: Jiawen Wu <jiawenwu@...stnetic.com>
> > 
> > nit: failure is misspelt in the subject
> 
> Thanks. :)
> 
> > 
> > > ---
> > >  drivers/net/ethernet/wangxun/txgbe/txgbe_aml.c | 1 +
> > >  1 file changed, 1 insertion(+)
> > >
> > > diff --git a/drivers/net/ethernet/wangxun/txgbe/txgbe_aml.c b/drivers/net/ethernet/wangxun/txgbe/txgbe_aml.c
> > > index 7dbcf41750c1..dc87ccad9652 100644
> > > --- a/drivers/net/ethernet/wangxun/txgbe/txgbe_aml.c
> > > +++ b/drivers/net/ethernet/wangxun/txgbe/txgbe_aml.c
> > > @@ -294,6 +294,7 @@ static void txgbe_mac_link_up_aml(struct phylink_config *config,
> > >  	wx_fc_enable(wx, tx_pause, rx_pause);
> > >
> > >  	txgbe_reconfig_mac(wx);
> > > +	txgbe_enable_sec_tx_path(wx);
> > >
> > >  	txcfg = rd32(wx, TXGBE_AML_MAC_TX_CFG);
> > >  	txcfg &= ~TXGBE_AML_MAC_TX_CFG_SPEED_MASK;
> > 
> > Hi Jiawen,
> > 
> > I am unsure if it is important, but I notice that:
> > 
> > * txgbe_mac_link_up_aml is the mac_link_up callback for txgbe_aml
> > 
> > * Whereas for txgbe_phy txgbe_enable_sec_tx_path() is called in
> >   txgbe_mac_finish(), which is the mac_finish callback
> > 
> > Could you comment on this asymmetry?
> 
> For txgbe_sp, the configuration of PCS is completed in the driver.
> Disable sec_tx_path -> configure PCS -> enable sec_tx_path, it is
> the necessary sequence. So these MAC operations were added in
> txgbe_mac_prepare() and txgbe_mac_finish().
> 
> For txgbe_aml, the configuration of PCS is completed in the firmware.
> So I didn't implement .mac_prepare and .mac_finish.

Thanks, that seems reasonable to me.

Reviewed-by: Simon Horman <horms@...nel.org>