netdev - Re: [PATCH net-next v1 4/9] net: ethtool: Add attributes for cable test reports

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: <9e8fa21bd1b51104038dbdc54fb61674@walle.cc>
Date:   Wed, 29 Apr 2020 21:32:51 +0200
From:   Michael Walle <michael@...le.cc>
To:     Florian Fainelli <f.fainelli@...il.com>
Cc:     Andrew Lunn <andrew@...n.ch>, cphealy@...il.com,
        davem@...emloft.net, hkallweit1@...il.com, mkubecek@...e.cz,
        netdev@...r.kernel.org
Subject: Re: [PATCH net-next v1 4/9] net: ethtool: Add attributes for cable
 test reports

Am 2020-04-29 20:58, schrieb Florian Fainelli:
> On 4/29/20 11:57 AM, Andrew Lunn wrote:
>> On Wed, Apr 29, 2020 at 06:16:05PM +0200, Michael Walle wrote:
>>> Hi,
>>> 
>>>>>>> +enum {
>>>>>>> +	ETHTOOL_A_CABLE_PAIR_0,
>>>>>>> +	ETHTOOL_A_CABLE_PAIR_1,
>>>>>>> +	ETHTOOL_A_CABLE_PAIR_2,
>>>>>>> +	ETHTOOL_A_CABLE_PAIR_3,
>>>>>>> +};
>>>>>> 
>>>>>> Do we really need this enum, couldn't we simply use a number 
>>>>>> (possibly
>>>>>> with a sanity check of maximum value)?
>>>>> 
>>>>> They are not strictly required. But it helps with consistence. Are 
>>>>> the
>>>>> pairs numbered 0, 1, 2, 3, or 1, 2, 3, 4?
>>>> 
>>>> OK, I'm not strictly opposed to it, it just felt a bit weird.
>>> 
>>> Speaking of the pairs. What is PAIR_0 and what is PAIR_3? Maybe
>>> it is specified somewhere in a standard, but IMHO an example for
>>> a normal TP cable would help to prevent wild growth amongst the
>>> PHY drivers and would help to provide consistent reporting towards
>>> the user space.
>> 
>> Hi Michael
>> 
>> Good question
>> 
>> Section 25.4.3 gives the pin out for 100BaseT. There is no pair
>> numbering, just transmit+, transmit- and receive+, receive- signals.
>> 
>> 1000BaseT calls the signals BI_DA+, BI_DA-, BI_DB+, BI_DB-, BI_DC+,
>> BI_DC-, BI_DDA+, BI_DD-. Comparing the pinout 100BaseT would use
>> BI_DA+, BI_DA-, BI_DB+, BI_DB. But 1000BaseT does not really have
>> transmit and receive pairs due to Auto MDI-X.
>> 
>> BroadReach calls the one pair it has BI_DA+/BI_DA-.
>> 
>> Maybe it would be better to have:
>> 
>> enum {
>> 	ETHTOOL_A_CABLE_PAIR_A,
>> 	ETHTOOL_A_CABLE_PAIR_B,
>> 	ETHTOOL_A_CABLE_PAIR_C,
>> 	ETHTOOL_A_CABLE_PAIR_D,
>> };
> 
> Yes, that would be clearer IMHO. Broadcom PHYs tend to refer to pairs 
> A,
> B, C and D in their datasheets.

Qualcomm Atheros calls them MDI[0], MDI[1], etc.. maybe mention
the corresponding pin on an RJ45 connector for reference?

-michael