/*
* OMAP2/3-common clockdomains
+ *
+ * Even though the 2420 has a single PRCM module from the
+ * interconnect's perspective, internally it does appear to have
+ * separate PRM and CM clockdomains. The usual test case is
+ * sys_clkout/sys_clkout2.
*/
/* This is an implicit clockdomain - it is never defined as such in TRM */
.omap_chip = OMAP_CHIP_INIT(CHIP_IS_OMAP24XX | CHIP_IS_OMAP3430),
};
+static struct clockdomain prm_clkdm = {
+ .name = "prm_clkdm",
+ .pwrdm = { .name = "wkup_pwrdm" },
+ .omap_chip = OMAP_CHIP_INIT(CHIP_IS_OMAP24XX | CHIP_IS_OMAP3430),
+};
+
+static struct clockdomain cm_clkdm = {
+ .name = "cm_clkdm",
+ .pwrdm = { .name = "core_pwrdm" },
+ .omap_chip = OMAP_CHIP_INIT(CHIP_IS_OMAP24XX | CHIP_IS_OMAP3430),
+};
+
/*
* 2420-only clockdomains
*/
static struct clockdomain *clockdomains_omap[] = {
&wkup_clkdm,
+ &cm_clkdm,
+ &prm_clkdm,
#ifdef CONFIG_ARCH_OMAP2420
&mpu_2420_clkdm,
/*
* Maximum number of clockdomains that can be associated with a powerdomain.
- * CORE powerdomain is probably the worst case.
+ * CORE powerdomain on OMAP3 is the worst case
*/
-#define PWRDM_MAX_CLKDMS 3
+#define PWRDM_MAX_CLKDMS 4
/* XXX A completely arbitrary number. What is reasonable here? */
#define PWRDM_TRANSITION_BAILOUT 100000