netdev - Re: [PATCH net-next 1/1] e1000e: Enable Link Partner Advertised Support

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: <Y7U/4Q0QKtkuexLu@gmx.fr>
Date:   Wed, 4 Jan 2023 04:59:13 -0400
From:   Jamie Gloudon <jamie.gloudon@....fr>
To:     Andrew Lunn <andrew@...n.ch>
Cc:     Tony Nguyen <anthony.l.nguyen@...el.com>, davem@...emloft.net,
        kuba@...nel.org, pabeni@...hat.com, edumazet@...gle.com,
        netdev@...r.kernel.org, sasha.neftin@...el.com,
        Naama Meir <naamax.meir@...ux.intel.com>
Subject: Re: [PATCH net-next 1/1] e1000e: Enable Link Partner Advertised
 Support

On Wed, Jan 04, 2023 at 01:47:01AM +0100, Andrew Lunn wrote:
> > --- a/drivers/net/ethernet/intel/e1000e/phy.c
> > +++ b/drivers/net/ethernet/intel/e1000e/phy.c
> > @@ -2,6 +2,7 @@
> >  /* Copyright(c) 1999 - 2018 Intel Corporation. */
> >  
> >  #include "e1000.h"
> > +#include <linux/ethtool.h>
> >  
> >  static s32 e1000_wait_autoneg(struct e1000_hw *hw);
> >  static s32 e1000_access_phy_wakeup_reg_bm(struct e1000_hw *hw, u32 offset,
> > @@ -1011,6 +1012,8 @@ static s32 e1000_phy_setup_autoneg(struct e1000_hw *hw)
> >  		 */
> >  		mii_autoneg_adv_reg &=
> >  		    ~(ADVERTISE_PAUSE_ASYM | ADVERTISE_PAUSE_CAP);
> > +		phy->autoneg_advertised &=
> > +		    ~(ADVERTISED_Pause | ADVERTISED_Asym_Pause);
> >  		break;
> >  	case e1000_fc_rx_pause:
> >  		/* Rx Flow control is enabled, and Tx Flow control is
> > @@ -1024,6 +1027,8 @@ static s32 e1000_phy_setup_autoneg(struct e1000_hw *hw)
> >  		 */
> >  		mii_autoneg_adv_reg |=
> >  		    (ADVERTISE_PAUSE_ASYM | ADVERTISE_PAUSE_CAP);
> > +		phy->autoneg_advertised |=
> > +		    (ADVERTISED_Pause | ADVERTISED_Asym_Pause);
> >  		break;
> >  	case e1000_fc_tx_pause:
> >  		/* Tx Flow control is enabled, and Rx Flow control is
> > @@ -1031,6 +1036,8 @@ static s32 e1000_phy_setup_autoneg(struct e1000_hw *hw)
> >  		 */
> >  		mii_autoneg_adv_reg |= ADVERTISE_PAUSE_ASYM;
> >  		mii_autoneg_adv_reg &= ~ADVERTISE_PAUSE_CAP;
> > +		phy->autoneg_advertised |= ADVERTISED_Asym_Pause;
> > +		phy->autoneg_advertised &= ~ADVERTISED_Pause;
> >  		break;
> >  	case e1000_fc_full:
> >  		/* Flow control (both Rx and Tx) is enabled by a software
> > @@ -1038,6 +1045,8 @@ static s32 e1000_phy_setup_autoneg(struct e1000_hw *hw)
> >  		 */
> >  		mii_autoneg_adv_reg |=
> >  		    (ADVERTISE_PAUSE_ASYM | ADVERTISE_PAUSE_CAP);
> > +		phy->autoneg_advertised |=
> > +		    (ADVERTISED_Pause | ADVERTISED_Asym_Pause);
> >  		break;
> >  	default:
> >  		e_dbg("Flow control param set incorrectly\n");
> > -- 
> 
> I don't know this driver at all. What i don't see anywhere here is
> using the results of the pause auto-neg. Is there some code somewhere
> that looks at the local and link peer advertising values and runs a
> resolve algorithm to determine what pause should be used, and program
> it into the MAC?
> 
>     Andrew
This is a old patch i had laying around, If i remember correctly, phy->autoneg_advertised plugs in "Link partner
advertised pause frame use link" line in ethtool everytime the nic renegotiate.