On Wed, 26 Oct 2016, Thomas Gleixner wrote:

On Wed, 26 Oct 2016, Lee Jones wrote:

quoted

On Fri, 14 Oct 2016, Guenter Roeck wrote:

quoted

The call to irq_set_parent() causes the following build error if tps65217
is built as module.

ERROR: ".irq_set_parent" [drivers/mfd/tps65217.ko] undefined!

The problem was introduced with commit 6556bdacf646f ("mfd: tps65217: Add
support for IRQs").

The author states: "I have added irq_set_parent() similarly as in
drivers/base/regmap/regmap-irq.c. But to be honest I am not sure what it
really does in case of tps65217."

So let's drop it.

Fixes: 6556bdacf646f ("mfd: tps65217: Add support for IRQs")
Cc: Marcin Niestroj <redacted>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Thomas Gleixner <redacted>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
---
 drivers/mfd/tps65217.c | 1 -
 1 file changed, 1 deletion(-)

This has been fixed now.

It was not fixed. The export was a work around as everyone was bitching
about the build robots failing forever.
 
So if the irq_set_parent() call is not required for functionality of the
driver then it should not be there in the first place.

Ah, I thought this was just another one of the many hacks I received
in response to the auto-builder's complains.  I've just been NACKing
them out of habit.

However, if is it true that we can simply just remove the call to
irq_set_parent(), then I'm happy to apply the patch.

Even happier with an Ack from you!

-- 
Lee Jones
Linaro STMicroelectronics Landing Team Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog