netdev - Re: [PATCH V2] net: phy: tja11xx: Add TJA11xx PHY driver

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: <08d66ea6-481c-94d0-373b-a3f987232951@gmail.com>
Date:   Sat, 22 Dec 2018 18:24:35 +0100
From:   Heiner Kallweit <hkallweit1@...il.com>
To:     Marek Vasut <marex@...x.de>, Andrew Lunn <andrew@...n.ch>
Cc:     netdev@...r.kernel.org, f.fainelli@...il.com
Subject: Re: [PATCH V2] net: phy: tja11xx: Add TJA11xx PHY driver

On 22.12.2018 00:22, Marek Vasut wrote:
> On 12/15/2018 06:38 PM, Heiner Kallweit wrote:
>> On 15.12.2018 18:01, Andrew Lunn wrote:
>>>> +static struct tja11xx_phy_stats tja11xx_hw_stats[] = {
>>>> +	{ "phy_symbol_error_count", 20, 0, 0xffff },
>>>> +	{ "phy_overtemp_error", 21, 1, BIT(1) },
>>>> +	{ "phy_undervolt_error", 21, 3, BIT(3) },
>>>> +	{ "phy_polarity_detect", 25, 6, BIT(6) },
>>>> +	{ "phy_open_detect", 25, 7, BIT(7) },
>>>> +	{ "phy_short_detect", 25, 8, BIT(8) },
>>>
>>> Hi Marek
>>>
>>> You have a number of one bit counters here, which is pretty unusual.
>>> The names also don't really suggest they are counters.
>>>
>> Apart from few counters the values seem to be flags. I just think
>> it could be done in a little bit more readable form, e.g. instead of
>> { "phy_short_detect", 25, 8, BIT(8) } use
>> { "phy_short_detect", 25, BIT(8) } and in tja11xx_get_stats() then
>> use FIELD_GET (see linux/bitfield.h).
> 
> This doesn't work with the counters, it only works with flags. The array
> contains both.
> 
Maybe we have a misunderstanding here, but FIELD_GET would work also
with the counters. FIELD_GET calculates the field offset, so you don't
need to specify it. Let's say you would have an entry in your array like
this: { "phy_xxx_error_count", 20, 4, 0xfff0 }

Then you would do:
#define XXX_ERROR_CNT_MASK GENMASK(15, 4)
{ "phy_xxx_error_count", 20, XXX_ERROR_CNT_MASK }
and access the value by
val = FIELD_GET(XXX_ERROR_CNT_MASK, reg)


>> The idea of HWMON attributes sounds good to me because it allows
>> to use the flags to trigger actions in a structured way. And I
>> assume in case of e.g. "PHY undervolt" some monitoring system
>> would like to be informed (especially because we talk about
>> automotive here).
> 
> I added HWMON_T_CRIT_ALARM and HWMON_I_LCRIT_ALARM for
> phy_overtemp_error and phy_undervolt_error respectively.
> Are there any other hwmon attributes I can use to replace the flags in
> the tja11xx_hw_stats array ?
> 
I think that's it.