Message-ID: <cb2915c5-831a-f2cf-55be-633ea21fa048@redhat.com>
Date:   Thu, 23 Dec 2021 18:21:11 +0100
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.3.0
Subject: Re: [PATCH v5 2/4] leds: simatic-ipc-leds: add new driver for Siemens
 Industial PCs
Content-Language: en-US
To:     Henning Schild <henning.schild@siemens.com>,
        Pavel Machek <pavel@ucw.cz>
Cc:     linux-kernel@vger.kernel.org, linux-leds@vger.kernel.org,
        platform-driver-x86@vger.kernel.org,
        linux-watchdog@vger.kernel.org,
        Srikanth Krishnakar <skrishnakar@gmail.com>,
        Jan Kiszka <jan.kiszka@siemens.com>,
        Gerd Haeussler <gerd.haeussler.ext@siemens.com>,
        Guenter Roeck <linux@roeck-us.net>,
        Wim Van Sebroeck <wim@linux-watchdog.org>,
        Mark Gross <mgross@linux.intel.com>,
        Andy Shevchenko <andy.shevchenko@gmail.com>,
        Enrico Weigelt <lkml@metux.net>
References: <20211213120502.20661-1-henning.schild@siemens.com>
 <20211213120502.20661-3-henning.schild@siemens.com>
 <20211215201800.GA28336@duo.ucw.cz>
 <61983e62-bc3d-a711-c197-7436a2e33531@redhat.com>
 <20211219164903.GB28481@duo.ucw.cz>
 <20211220085220.23753298@md1za8fc.ad001.siemens.net>
 <20211220091434.2f696d76@md1za8fc.ad001.siemens.net>
From:   Hans de Goede <hdegoede@redhat.com>
In-Reply-To: <20211220091434.2f696d76@md1za8fc.ad001.siemens.net>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

Hi,

On 12/20/21 09:14, Henning Schild wrote:
> Am Mon, 20 Dec 2021 08:53:55 +0100
> schrieb Henning Schild <henning.schild@siemens.com>:
> 
>> Am Sun, 19 Dec 2021 17:49:03 +0100
>> schrieb Pavel Machek <pavel@ucw.cz>:
>>
>>> On Wed 2021-12-15 21:53:56, Hans de Goede wrote:  
>>>> Hi,
>>>>
>>>> On 12/15/21 21:18, Pavel Machek wrote:    
>>>>> On Mon 2021-12-13 13:05:00, Henning Schild wrote:    
>>>>>> This driver adds initial support for several devices from
>>>>>> Siemens. It is based on a platform driver introduced in an
>>>>>> earlier commit.
>>>>>>
>>>>>> One of the supported machines has GPIO connected LEDs, here we
>>>>>> poke GPIO memory directly because pinctrl does not come up.
>>>>>>
>>>>>> Signed-off-by: Henning Schild <henning.schild@siemens.com>    
>>>>>
>>>>> Acked-by: Pavel Machek <pavel@ucw.cz>    
>>>>
>>>> I see that this patch #includes
>>>> linux/platform_data/x86/simatic-ipc-base.h which gets added by
>>>> patch 1/4.
>>>>
>>>> Pavel, can I take this patch upstream through the pdx86 tree (with
>>>> you Ack added)? Or shall I prepare an immutable branch with patch
>>>> 1 for you to merge ?    
>>>
>>> Yes, you can.
>>>
>>>   
>>>>>> +
>>>>>> +static struct simatic_ipc_led simatic_ipc_leds_io[] = {
>>>>>> +	{1 << 15, "green:" LED_FUNCTION_STATUS "-1" },
>>>>>> +	{1 << 7,  "yellow:" LED_FUNCTION_STATUS "-1" },
>>>>>> +	{1 << 14, "red:" LED_FUNCTION_STATUS "-2" },
>>>>>> +	{1 << 6,  "yellow:" LED_FUNCTION_STATUS "-2" },
>>>>>> +	{1 << 13, "red:" LED_FUNCTION_STATUS "-3" },
>>>>>> +	{1 << 5,  "yellow:" LED_FUNCTION_STATUS "-3" },
>>>>>> +	{ }
>>>>>> +};    
>>>
>>> But I'd still like better naming than red:status-2.  
>>
>> We had the name discussion already several times, and i have to admit
>> i am not too happy either.
>>
>> But my impression was that this is an acceptable compromise. I am not
>> happy because the names lack scope, which i had in the first round
>> with "simatic-ipc:red:...".
>>
>> Function is also a bit unclear, but with the numbers and the user
>> manual, or looking at the chassis it kind of adds up and should be
>> clear to users which is which.
>>
>> But i agree with Hans that we should sort this out before merge. So
>> please say what makes you unhappy, maybe that can be fixed ... might
>> even make me happier about the names i feel i had to choose.
>>
>> The LEDs are per definition of the manuals meant for
>> users/applications to signal whatever the use-case might want to
>> signal. There are 3 of them numbered 1-3 on the chassis, and next to
>> the number can often (not always) be found a string like "error",
>> "maint", "run-stop" So a function suggestion i would say.
>>
>> I could envision to use "fault" or "alarm" instead of "status" for the
>> one labeled "error". And maybe "standby" for the one called "maint"
>> but i would really like to keep the numbers.
>>
>> Which would look like
>>
>> status-1
>> alarm-2
>> standby-3
>>
>> But still i have to clue what those names stand for and choosing
>> and of those "undefined" names could just suggest things and break
>> expectations. Calling them all "status" is neutral ... 
>>
>> Or can you explain the difference between "fault", "panic" and
>> "alarm". Ask 5 people and get at least 3 different expectations ... i
>> guess.
> 
> Long story short, i am also not happy but the current suggestion is the
> most generic and least "expectation-creating" i could come up with.
> While keeping a mapping between the name and the chassis/manual.
> 
> So i will stick with it, unless i get concrete suggestions on how to
> improve.

Ok, given the above I've gone ahead and merged this series.

I too believe the current status names are fine, but if someone disagrees,
they still have the entire 5.17 cycle to come up with something better.

Regards,

Hans