Re: [PATCH 1/3] mtd: nand: pxa3xx_nand: add register access debug

From: Ezequiel Garcia
Date: Mon Aug 24 2015 - 09:46:29 EST


On 23 August 2015 at 16:09, Robert Jarzmik <robert.jarzmik@xxxxxxx> wrote:
> Robert Jarzmik <robert.jarzmik@xxxxxxx> writes:
>
>> Ezequiel Garcia <ezequiel@xxxxxxxxxxxxxxxxxxxx> writes:
>>
>>> On 11 Aug 09:57 PM, Robert Jarzmik wrote:
>>>> Add verbose debug for register accesses. This enables easier debugging
>>>> by following where and how hardware is stimulated, and how it answers.
>>>>
>>>
>>> I really don't see why we want this patch. It's probably an useful hack to
>>> use in some cases, but can't see why we would want it in mainline.
>>>
>>> Feel free to prove me wrong.
>> Why not.
>>
>> Imagine that there is a bug in the driver, such as the one with the status bits
>> clearing. You can't reproduce it, and the person who has the hardware to
>> reproduce it is not skilled enough to do any debug.
>>

So let's assume this person is not skilled enough to debug it, but it is skilled
enough to rebuild a kernel (to activate the verbose debug).

Then this person can also apply a simple patch (this patch), which seems
a pretty trivial step to do.

Since you might also need some other debug traces, I'm not convinced about
pushing this particular group to mainline.

>> What are the tools and what will you ask this tester to do in order to debug and
>> solve his problem, as a good driver maintainer ?
>>
>> Amongst the drivers I wrote in the past, this approach was the most reliable to
>> capture traces to debug remotely.
>
> Ezequiel, do you agree with my point ?
>

I agree that the hack is useful for debugging purposes, it's just that I don't
see why we want it in mainline - where we usually avoid clutter.

Also, your argument applies to all the other drivers, but that doesn't mean
we will patch them all.

Anyway, this is just my opinion. If Brian thinks differently, I have no problem
with it.

Thanks!
--
Ezequiel GarcÃa, VanguardiaSur
www.vanguardiasur.com.ar
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/