> -----Original Message----- > From: "Högander" Jouni [mailto:[EMAIL PROTECTED] > Sent: Tuesday, September 30, 2008 3:19 PM > To: ext Kevin Hilman > Cc: Rajendra Nayak; [email protected] > Subject: Re: [PATCH 15/16] OMAP3: Dynamic enable/disable of > OFF support > > "ext Kevin Hilman" <[EMAIL PROTECTED]> writes: > > > "Rajendra Nayak" <[EMAIL PROTECTED]> writes: > > > >> I just managed to see that this patch seems to break suspend > >> functionality. If after bootup I enable OFF mode the subsequent > >> suspend tries to put a few power domains to OFF which are currently > >> in RET, and since there is no code in place today to handle RET to > >> OFF transitions, they don't transition to OFF and remain in RET. > > > > See commit a974addcfa23181667fe32e5032820917bf4a2b2 from > Tero Kristo. > > This patch was meant to address these kinds of transitions, but it > > seems it's not working exactly right. I'm debugging now. > > > >> Supporting run time enable/disable of OFF functionality is kind of > >> becoming complicated. Do we really see a need to have a run-time > >> option to enable/disable OFF mode or can we have a compile time > >> option for this? > > > > Run-time option is required. > > Whole cpuidle + srf concept is about changing between OFF/RET/ON > dynamically. So if implementing it through one sysfs entry is > discarded because it's complicated, how do you think you could > implement it in cpuidle and srf?
What I meant was that hacking around the *existing* code (which neither has cpuidle nor SRF) just to support runtime enable/disable of OFF mode is getting complicated. > > > > > Kevin > > > >>> -----Original Message----- > >>> From: [EMAIL PROTECTED] > >>> [mailto:[EMAIL PROTECTED] On Behalf Of > Rajendra Nayak > >>> Sent: Friday, September 26, 2008 5:50 PM > >>> To: [email protected] > >>> Cc: 'Kevin Hilman' > >>> Subject: [PATCH 15/16] OMAP3: Dynamic enable/disable of > OFF support > >>> > >>> This patch adds a runtime sysfs knob (/sys/power/enable_off_mode) > >>> to enable/disbale CORE OFF support for OMAP3. > >>> > >>> Signed-off-by: Rajendra Nayak <[EMAIL PROTECTED]> > >>> --- > >>> arch/arm/mach-omap2/pm.c | 21 +++++++++++++++++---- > >>> arch/arm/mach-omap2/pm.h | 2 ++ > >>> arch/arm/mach-omap2/pm34xx.c | 26 ++++++++++++++++++++++++++ > >>> 3 files changed, 45 insertions(+), 4 deletions(-) > >>> > >>> Index: linux-omap-2.6/arch/arm/mach-omap2/pm34xx.c > >>> > =================================================================== > >>> --- linux-omap-2.6.orig/arch/arm/mach-omap2/pm34xx.c > >>> 2008-09-26 16:39:30.000000000 +0530 > >>> +++ linux-omap-2.6/arch/arm/mach-omap2/pm34xx.c > >>> 2008-09-26 16:39:34.000000000 +0530 > >>> @@ -395,6 +395,32 @@ static int omap3_can_sleep(void) > >>> return 1; > >>> } > >>> > >>> +int set_next_pwrst(struct powerdomain *pwrdm) > >>> +{ > >>> + struct power_state *pwrst; > >>> + int ret = 0; > >>> + u32 state; > >>> + > >>> + if (!pwrdm->pwrsts) > >>> + return 0; > >>> + > >>> + if (enable_off_mode) > >>> + state = PWRDM_POWER_OFF; > >>> + else > >>> + state = PWRDM_POWER_RET; > >>> + > >>> + ret = pwrdm_set_next_pwrst(pwrdm, state); > >>> + if (ret) { > >>> + printk(KERN_ERR "Unable to set state of > >>> powerdomain: %s\n", > >>> + pwrdm->name); > >>> + goto err; > >>> + } > >>> + list_for_each_entry(pwrst, &pwrst_list, node) > >>> + pwrst->next_state = state; > >>> +err: > >>> + return ret; > >>> +} > >>> + > >>> /* This sets pwrdm state (other than mpu & core. > Currently only ON & > >>> * RET are supported. Function is assuming that clkdm > doesn't have > >>> * hw_sup mode enabled. */ > >>> Index: linux-omap-2.6/arch/arm/mach-omap2/pm.c > >>> > =================================================================== > >>> --- linux-omap-2.6.orig/arch/arm/mach-omap2/pm.c > >>> 2008-09-26 16:39:30.000000000 +0530 > >>> +++ linux-omap-2.6/arch/arm/mach-omap2/pm.c 2008-09-26 > >>> 16:39:34.000000000 +0530 > >>> @@ -35,6 +35,7 @@ > >>> > >>> unsigned short enable_dyn_sleep; > >>> unsigned short clocks_off_while_idle; > >>> +unsigned short enable_off_mode; > >>> atomic_t sleep_block = ATOMIC_INIT(0); > >>> > >>> static ssize_t idle_show(struct kobject *, struct > >>> kobj_attribute *, char *); > >>> @@ -47,6 +48,9 @@ static struct kobj_attribute sleep_while > >>> static struct kobj_attribute clocks_off_while_idle_attr = > >>> __ATTR(clocks_off_while_idle, 0644, idle_show, idle_store); > >>> > >>> +static struct kobj_attribute enable_off_mode_attr = > >>> + __ATTR(enable_off_mode, 0644, idle_show, idle_store); > >>> + > >>> static ssize_t idle_show(struct kobject *kobj, struct > >>> kobj_attribute *attr, > >>> char *buf) > >>> { > >>> @@ -54,6 +58,8 @@ static ssize_t idle_show(struct kobject > >>> return sprintf(buf, "%hu\n", enable_dyn_sleep); > >>> else if (attr == &clocks_off_while_idle_attr) > >>> return sprintf(buf, "%hu\n", clocks_off_while_idle); > >>> + else if (attr == &enable_off_mode_attr) > >>> + return sprintf(buf, "%hu\n", enable_off_mode); > >>> else > >>> return -EINVAL; > >>> } > >>> @@ -69,13 +75,16 @@ static ssize_t idle_store(struct kobject > >>> return -EINVAL; > >>> } > >>> > >>> - if (attr == &sleep_while_idle_attr) > >>> + if (attr == &sleep_while_idle_attr) { > >>> enable_dyn_sleep = value; > >>> - else if (attr == &clocks_off_while_idle_attr) > >>> + } else if (attr == &clocks_off_while_idle_attr) { > >>> clocks_off_while_idle = value; > >>> - else > >>> + } else if (attr == &enable_off_mode_attr) { > >>> + enable_off_mode = value; > >>> + pwrdm_for_each(set_next_pwrst); > >>> + } else { > >>> return -EINVAL; > >>> - > >>> + } > >>> return n; > >>> } > >>> > >>> @@ -114,6 +123,10 @@ static int __init omap_pm_init(void) > >>> &clocks_off_while_idle_attr.attr); > >>> if (error) > >>> printk(KERN_ERR "sysfs_create_file failed: > >>> %d\n", error); > >>> + error = sysfs_create_file(power_kobj, > >>> + &enable_off_mode_attr.attr); > >>> + if (error) > >>> + printk(KERN_ERR "sysfs_create_file failed: > >>> %d\n", error); > >>> > >>> return error; > >>> } > >>> Index: linux-omap-2.6/arch/arm/mach-omap2/pm.h > >>> > =================================================================== > >>> --- linux-omap-2.6.orig/arch/arm/mach-omap2/pm.h > >>> 2008-09-26 16:39:30.000000000 +0530 > >>> +++ linux-omap-2.6/arch/arm/mach-omap2/pm.h 2008-09-26 > >>> 16:39:34.000000000 +0530 > >>> @@ -18,10 +18,12 @@ extern int omap3_pm_init(void); > >>> > >>> extern unsigned short enable_dyn_sleep; > >>> extern unsigned short clocks_off_while_idle; > >>> +extern unsigned short enable_off_mode; > >>> extern atomic_t sleep_block; > >>> > >>> extern void omap2_block_sleep(void); > >>> extern void omap2_allow_sleep(void); > >>> +extern int set_next_pwrst(struct powerdomain *pwrdm); > >>> > >>> #define OMAP343X_TABLE_ADDRESS_OFFSET 0x31 > >>> #define OMAP343X_TABLE_VALUE_OFFSET 0x30 > >>> > >>> -- > >>> 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 > >>> > >>> > > -- > > 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 > > > > > > -- > Jouni Högander > > > -- 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
