Re: [PATCH v2 -next] spi: fix spi-sprd-adi build errors when SPI_SPRD_ADI=y and HWSPINLOCK=m

[Randy Dunlap]

On 09/25/17 14:11, Mark Brown wrote:
> On Mon, Sep 25, 2017 at 12:53:20PM -0700, Randy Dunlap wrote:
>> On 09/25/17 12:44, Mark Brown wrote:
> 
>>> That would then get in the way of build test coverage.  
> 
>> I don't agree, but whatever. I give up.
> 
> OK.  Just as a general thing it's really important that dependency
> changes be thought through and express what the constraints are in ways
> that reflect what's actually going on and what's useful for users, it
> seems to be a frequent problem and it does get in the way.

Yes, it's a frequent problem that drivers do not build due to some kconfig
dependency issue(s).

I would prefer to see the driver maintainer(s) fix build errors themselves,
but they are not always aware of the idiosyncrasies of kconfig.

I think that people often submit drivers (or driver changes) after building
it in one (working) configuration only, without regard for the other possible
combinations.


-- 
~Randy