On Mon, Feb 1, 2021 at 2:54 AM Geert Uytterhoeven [off-list ref] wrote:

Hi Saravana,

On Thu, Jan 28, 2021 at 6:08 PM Saravana Kannan [off-list ref] wrote:

quoted

I already accounted for early clocks like this when I designed
fw_devlink. Each driver shouldn't need to set OF_POPULATED.
drivers/clk/clk.c already does this for you.

I think the problem is that your driver is using
CLK_OF_DECLARE_DRIVER() instead of CLK_OF_DECLARE(). The comments for
CLK_OF_DECLARE_DRIVER() says:
/*
 * Use this macro when you have a driver that requires two initialization
 * routines, one at of_clk_init(), and one at platform device probe
 */

In your case, you are explicitly NOT having a driver bind to this
clock later. So you shouldn't be using CLK_OF_DECLARE() instead.

Typo. I meant to say this driver SHOULD be using CLK_OF_DECLARE()
instead. I wonder if this is what caused you to send the email --
because we are saying the same thing.

Unless I'm missing something, name##_of_clk_init_driver() clearing
OF_POPULATED again causes consumer driver probing to be postponed by
fw_devlink until the second initialization phase of the provider has been
completed?

Right, if they use CLK_OF_DECLARE_DRIVER() what you said above will
happen and that's the issue they are trying to fix.

This is wrong if the consumer only needs a clock instantiated during the
first phase, and may cause issues if the consumer is a critical device.
E.g. a timer, on ARM SoCs lacking an architecture timer (pre-Cortex
A7/A15) or global timer (pre-Cortex A9, or single-core Cortex A9).
Probably there are more examples.

So, needing devices like IRQ, timer and clocks early is fine.
fw_devlink can handle that correctly if the proper macros are used
(since most frameworks set the OF_POPULATED flag for these devices).

-Saravana

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel