Re: [PATCH] spi: Remove CONFIG_ prefix from Kconfig select

[Joe Perches <joe@xxxxxxxxxxx>]

On Thu, 2020-03-05 at 08:21 -0800, Randy Dunlap wrote:
> On 3/5/20 8:20 AM, John Garry wrote:
> > On 05/03/2020 16:01, Joe Perches wrote:
> > > On Thu, 2020-03-05 at 15:53 +0000, John Garry wrote:
> > > > On 05/03/2020 15:15, Joe Perches wrote:
> > > > > commit a2ca53b52e00 ("spi: Add HiSilicon v3xx SPI NOR flash
> > > > > controller driver") likely inadvertently used a select statement
> > > > > with a CONFIG_ prefix, remove the prefix.
> > > > > 
> > > > > Signed-off-by: Joe Perches <joe@xxxxxxxxxxx>
> > > > 
> > > > It's a pity checkpatch can't pick this stuff up...
> > > 
> > > You are free to write a rule and submit a patch.
> > > 
> > 
> > I'm not opposed to the idea...

It'd be better I think to update the grammar
verification in the tool that parses these files.

see: scripts/kconfig/...

> > > The grammar is pretty hard to verify from a patch
> > > fragment though.
> > > 
> > > There are 2 existing nominal false positives.
> > > 
> > > $ git grep 'select\b.*\bCONFIG_' -- '*/Kconfig*'
> > > drivers/edac/Kconfig:     select CONFIG_ACPI_NFIT.
> > > drivers/edac/Kconfig:     select CONFIG_ACPI_NFIT.
> > 
> > Ah, that's in the help text. I guess that those can be ignored simply based on the indentation.

The grammar is per line whitespace independent, so not necessarily.

> > drivers/platform/chrome/Kconfig:        select CONFIG_MFD_CROS_EC_DEV
> > 
> > The 2nd looks like it's incorrect.
> 
> Yes, we just got a patch for that one today.

That's the reason I submitted this one.