I am working with a board that deviates from what seems to be common OMAP-as-clock-slave-from-the-TWL4030-PMIC-clock-master practice. In order to get this board to work correctly when used with SmartReflex-enabled, I had to add the following to the board initialization function:
/* In this architecture, we use the OMAP3 as a 19.2 MHz clock
* master and the TWL4030 as a clock slave, fed via the OMAP3's
* sys_clkout1 pin.
*
* We need to request the 'sys_clkout1' clock source to at least
* give it a reference count of one (1); otherwise, if the SmartReflex
* power management code has been configured, it'll get unceremoniously
* turned off with a message:
*
* Disabling unused clock "sys_clkout1"
*
* and general system functionality fails thereafter.
*/
sys_clkout1 = clk_get(NULL, "sys_clkout1");
BUG_ON(IS_ERR(sys_clkout1));
clk_enable(sys_clkout1);
However, now working with suspend/resume functionality, I see that this clock is preventing the system from fully going down to sleep. This leads me to believe that perhaps board-specific initialization was/is not the right place for this as there is no board-specific suspend/resume handling.
Any recommendations on a more architecturally-correct way to address OMAP-as-clock-master-to-the-TWL4030-PMIC-clock-slave in a suspend/resume-friendly way?
Best,
Grant--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
