clk: tango4: undefined CONFIG_ARCH_TANGOX

clk: tango4: undefined CONFIG_ARCH_TANGOX

[Valentin Rothberg]

Hi Marc,

your commit ed12dfc92f01 ("clk: tango4: clkgen driver for Tango4
platforms") has shown up in today's linux-next tree (i.e.,
next-20151120) adding the following build condition to the tango4 clk
driver:

drivers/clk/Makefile:45:obj-$(CONFIG_ARCH_TANGOX) += clk-tango4.o

However, ARCH_TANGOX is nowhere defined in Kconfig so that the driver
cannot be compiled at the current state.  I checked the LKML, and found
a bunch of patches referencing ARCH_TANGOX as well, but I could not find
any patch adding this option.

Is there a patch queued somewhere that adds ARCH_TANGOX?

I detected the issue with scripts/checkkconfigsymbols.py by diffing
yesterday's and today's linux-next.

Kind regards,
 Valentin

Re: clk: tango4: undefined CONFIG_ARCH_TANGOX

[Marc Gonzalez]

On 20/11/2015 09:50, Valentin Rothberg wrote:

> your commit ed12dfc92f01 ("clk: tango4: clkgen driver for Tango4
> platforms") has shown up in today's linux-next tree (i.e.,
> next-20151120) adding the following build condition to the tango4 clk
> driver:
> 
> drivers/clk/Makefile:45:obj-$(CONFIG_ARCH_TANGOX) += clk-tango4.o
> 
> However, ARCH_TANGOX is nowhere defined in Kconfig so that the driver
> cannot be compiled at the current state.  I checked the LKML, and found
> a bunch of patches referencing ARCH_TANGOX as well, but I could not find
> any patch adding this option.
> 
> Is there a patch queued somewhere that adds ARCH_TANGOX?

Hello Valentin,

Platform support has not been accepted yet.

http://thread.gmane.org/gmane.linux.ports.arm.kernel/456280

In fact, Kevin Hilman has pointed out that the arch should not
be called TANGOX, because X is a wildcard.

(However, several unrelated drivers have been submitted with
TANGOX in the name. Is that a problem?)

tango3 was a MIPS-based design
tango4 is an ARM-based design (with one MIPS-based outlier).
tango5 is an ARM-based design

Although Mans is against the idea, I believe there should be one
different clk driver for each arch.

I'm not sure how to handle situations where there's
A) a separate driver for 3,4,5
B) a   single driver for 3,4,5
C) one driver for 2 arches, another driver for the other arch

Regards.

Re: clk: tango4: undefined CONFIG_ARCH_TANGOX

[Måns Rullgård]

Marc Gonzalez <marc_gonzalez@sigmadesigns.com> writes:

> On 20/11/2015 09:50, Valentin Rothberg wrote:
>
>> your commit ed12dfc92f01 ("clk: tango4: clkgen driver for Tango4
>> platforms") has shown up in today's linux-next tree (i.e.,
>> next-20151120) adding the following build condition to the tango4 clk
>> driver:
>> 
>> drivers/clk/Makefile:45:obj-$(CONFIG_ARCH_TANGOX) += clk-tango4.o
>> 
>> However, ARCH_TANGOX is nowhere defined in Kconfig so that the driver
>> cannot be compiled at the current state.  I checked the LKML, and found
>> a bunch of patches referencing ARCH_TANGOX as well, but I could not find
>> any patch adding this option.
>> 
>> Is there a patch queued somewhere that adds ARCH_TANGOX?
>
> Hello Valentin,
>
> Platform support has not been accepted yet.
>
> http://thread.gmane.org/gmane.linux.ports.arm.kernel/456280
>
> In fact, Kevin Hilman has pointed out that the arch should not
> be called TANGOX, because X is a wildcard.
>
> (However, several unrelated drivers have been submitted with
> TANGOX in the name. Is that a problem?)
>
> tango3 was a MIPS-based design
> tango4 is an ARM-based design (with one MIPS-based outlier).
> tango5 is an ARM-based design
>
> Although Mans is against the idea, I believe there should be one
> different clk driver for each arch.

It's essentially the same clock generator.  It should be a single
driver.

-- 
Måns Rullgård
mans@mansr.com