Re: [PATCH] gpio: syscon: do not use raw "set" callback in syscon_gpio_dir_out

[Vladimir Murzin <vladimir.murzin@xxxxxxx>]

On 22/09/17 14:47, Linus Walleij wrote:
> On Fri, Sep 22, 2017 at 12:23 PM, Vladimir Murzin
> <vladimir.murzin@xxxxxxx> wrote:
> 
>> I tried to use gpio-syscon to fit into my case which is very
>> similar to those pseudo-GPIOs in drivers/mfd/vexpress-sysreg.c
> 
> I do not like what that driver is doing and it should not be
> taken as inspiration.
> 
> I have several times slammed down on people trying to shoehorn
> things that are not GPIO into the GPIO subsystem just out of
> convenience.
> 

It is why I submitted only this patch, if you think there is no
issue I'm fine :)

> The question to ask is always: is this bit/line/pin really
> "general purpose input/output"?

In my case definitely it is not. These are some bits which 
control simple CLCD panel [1]. The only reason I've looked
into gpio-syscon is that I saw you forced keystone bits in
there with that "it is not general purpose" reason, but
from your response it seems it is not the right place for
such stuff.

> 
> Not "how can I quickly code up something that makes this
> thing work?"
> 
> See for example drivers/leds/leds-syscon.c
> That was my response to someone trying to first use
> syscon-gpio on a register and then gpio-leds on top of
> that, hilarious layers of indirection!
> 
> If the use case is MMC card detect, we need card detection
> using syscon directly in the MMC subsystem, not hacks
> like what the Vexpress MFD driver is doing.

No, it is not MMC card detect.

> 
> I have the Versatile Express in my office and one day I will
> fix up this thing.

Understood.

[1] http://infocenter.arm.com/help/topic/com.arm.doc.dai0399c/index.html#arm_toc19 (FPGAIO->MISC)

Cheers
Vladimir

> 
> Yours,
> Linus Walleij
> 

--
To unsubscribe from this list: send the line "unsubscribe linux-gpio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html