linux-kernel - Re: [PATCH RFC net-next] net: pcs: pcs-mtk-lynxi fix mtk_pcs_lynxi_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]

Date: Wed, 3 Jan 2024 00:10:54 +0100
From: Daniel Golle <daniel@...rotopia.org>
To: Eric Woudstra <ericwouds@...il.com>
Cc: "Russell King (Oracle)" <linux@...linux.org.uk>,
	Alexander Couzens <lynxis@...0.eu>, Andrew Lunn <andrew@...n.ch>,
	Heiner Kallweit <hkallweit1@...il.com>,
	"David S. Miller" <davem@...emloft.net>,
	Eric Dumazet <edumazet@...gle.com>,
	Jakub Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>,
	Matthias Brugger <matthias.bgg@...il.com>,
	AngeloGioacchino Del Regno <angelogioacchino.delregno@...labora.com>,
	Frank Wunderlich <frank-w@...lic-files.de>, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
	linux-mediatek@...ts.infradead.org
Subject: Re: [PATCH RFC net-next] net: pcs: pcs-mtk-lynxi fix
 mtk_pcs_lynxi_get_state() for 2500base-x

On Tue, Jan 02, 2024 at 11:13:58PM +0100, Eric Woudstra wrote:
> I believe the general idea is that phylink should be aware wether to use inband or outband negotiation in order to setup the hardware correctly. Speaking of a situation where there is a PHY attached.

Well, SGMII speed/duplex AN is not defined for 2500Base-X by any
standard and not supported by the hardware (unlike e.g. RealTek
which came up with their own proprietary extension called HiSGMII).

So we should simply never set the SGMII_SPEED_DUPLEX_AN bit if using
2500Base-X on MediaTek LynxI PCS -- in-band link status will still work,
and as 2500Base-X anyway only supports 2500M speed at full duplex it is
not a problem that speed and duplex are hard-coded in the driver in this
case imho. And that works fine for SFPs with and without
present/discoverable/accessible PHY.

Surely, having phylink take care whether SGMII_SPEED_DUPLEX_AN should be
set would be even nicer.

I believe that source of confusion here is simply that

in-band-status != SGMII_SPEED_DUPLEX_AN

We *do* have in-band-status even without having SGMII_SPEED_DUPLEX_AN set
with 2500Base-X link mode (as in: link being up or down and link, duplex
and speed is fixed anyway for 2500Base-X).

> 
> On January 2, 2024 9:01:23 PM GMT+01:00, Daniel Golle <daniel@...rotopia.org> wrote:
> >On Tue, Jan 02, 2024 at 08:33:32PM +0100, Eric Woudstra wrote:
> >> [...]
> >> 
> >> So if phylink_mii_c22_pcs_decode_state() should not set the speed, then it is not correctly set somewhere else.
> >
> >Yes, but the fix should go to pcs-mtk-lynxi.c and you don't need to
> >change phylink for it to work.
> >This should be enough:
> >https://patchwork.kernel.org/project/netdevbpf/patch/091e466912f1333bb76d23e95dc6019c9b71645f.1699565880.git.daniel@makrotopia.org/
> >