OMAP4: PM: Declare idle modules as functional too
Subsystems (3)
PATCH 1/1 OMAP4: PM: Declare idle modules as functional too
The omap4_cm_wait_module_ready function would only check for the modules to be completely functional before declaring them ready to be accessed.
There might also be instances where in the module is actually in idle (under h/w control) but should still be declared accessible, as the h/w control would make it functional when needed.
Hence make omap4_cm_wait_module_ready return true in case the module is fully functional *or* in idle state. Fail only if the module is fully disabled or stuck intransition.
The explaination from the TRM for the idlest bits on OMAP4 is as below for quick reference
Module idle state:0x0 func: Module is fully functional, including OCP0x1 trans: Module is performing transition: wakeup, or sleep, or sleep
abortion0x2 idle: Module is in Idle mode (only OCP part). It is functional if
using separate functional clock0x3 disabled: Module is disabled and cannot be accessed
--- arch/arm/mach-omap2/cm4xxx.c | 9 +++++---- 1 files changed, 5 insertions(+), 4 deletions(-)
arch/arm/mach-omap2/cm4xxx.c +5 -4
@@ -43,7 +43,6 @@* using separate functional clock * 0x3 disabled: Module is disabled and cannot be accessed * -* TODO: Need to handle module accessible in idle state */ int omap4_cm_wait_module_ready (void __iomem *clkctrl_reg) {
@@ -52,9 +51,11 @@ int omap4_cm_wait_module_ready(void __iomem *clkctrl_reg)if (!clkctrl_reg)return 0 ; -omap_test_timeout (((__raw_readl (clkctrl_reg) & -OMAP4430_IDLEST_MASK ) ==0 ), -MAX_MODULE_READY_TIME , i); +omap_test_timeout (( + ((__raw_readl (clkctrl_reg) &OMAP4430_IDLEST_MASK ) ==0 ) || + (((__raw_readl (clkctrl_reg) &OMAP4430_IDLEST_MASK ) >> +OMAP4430_IDLEST_SHIFT ) ==0x2 )), +MAX_MODULE_READY_TIME , i);return (i <MAX_MODULE_READY_TIME ) ?0 : -EBUSY ; }
Signature · 1 line
1.7.0.4
Hi Rajendra, Partha, Benoît,
Quoted 15 lines · Rajendra Nayak
The
omap4_cm_wait_module_readyfunction would only check for the modules to be completely functional before declaring them ready to be accessed.
There might also be instances where in the module is actually in idle (under h/w control) but should still be declared accessible, as the h/w control would make it functional when needed.Hence make
omap4_cm_wait_module_readyreturn true in case the module is fully functional *or* in idle state. Fail only if the module is fully disabled or stuck intransition.The explaination from the TRM for the idlest bits on OMAP4 is as below for quick reference
Quoted 6 lines
Module idle state:
0x0func: Module is fully functional, including OCP0x1trans: Module is performing transition: wakeup, or sleep, or sleep
abortion0x2idle: Module is in Idle mode (only OCP part). It is functional if
using separate functional clock
Thanks for the detailed changelog. Do we need to check any other registers to ensure that the OCP port is in autoidle, or to ensure that the separate functional clock is active?
- Paul
Quoted 26 lines · Paul Walmsley
Hi Rajendra, Partha, Benoît,
Quoted 15 lines · Rajendra Nayak
The
omap4_cm_wait_module_readyfunction would only check for the modules to be completely functional before declaring them ready to be accessed.
There might also be instances where in the module is actually in idle (under h/w control) but should still be declared accessible, as the h/w control would make it functional when needed.Hence make
omap4_cm_wait_module_readyreturn true in case the module is fully functional *or* in idle state. Fail only if the module is fully disabled or stuck intransition.The explaination from the TRM for the idlest bits on OMAP4 is as below for quick reference
Quoted 6 lines
Module idle state:
0x0func: Module is fully functional, including OCP0x1trans: Module is performing transition: wakeup, or sleep, or sleep
abortion0x2idle: Module is in Idle mode (only OCP part). It is functional if
using separate functional clock
Quoted 3 lines
Thanks for the detailed changelog. Do we need to check any other registers to ensure that the OCP port is in autoidle, or to ensure that the separate functional clock is active?
Unfortunately, we do not have any other status than that in OMAP4 for the modules clocks activity :-(
Benoit
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Quoted 27 lines · Benoit
Quoted 26 lines · Paul Walmsley
Hi Rajendra, Partha, Benoît,
Quoted 21 lines · Rajendra Nayak
The
omap4_cm_wait_module_readyfunction would only check for the modules to be completely functional before declaring them ready to be accessed.
There might also be instances where in the module is actually in idle (under h/w control) but should still be declared accessible, as the h/w control would make it functional when needed.Hence make
omap4_cm_wait_module_readyreturn true in case the module is fully functional *or* in idle state. Fail only if the module is fully disabled or stuck intransition.The explaination from the TRM for the idlest bits on OMAP4 is as below for quick reference
Module idle state:
0x0func: Module is fully functional, including OCP0x1trans: Module is performing transition: wakeup, or sleep, or sleep
abortion0x2idle: Module is in Idle mode (only OCP part). It is functional if
using separate functional clock
Quoted 6 lines
Quoted 3 lines
Thanks for the detailed changelog. Do we need to check any other registers to ensure that the OCP port is in autoidle, or to ensure that the separate functional clock is active?
Unfortunately, we do not have any other status than that in OMAP4 for the modules clocks activity :-(
OK, thanks Rajendra and Benoît. Queued for 2.6.37.
- Paul