| Message ID | Pine.LNX.4.44L0.1507081617200.2243-100000@iolanthe.rowland.org (mailing list archive) |
|---|---|
| State | Superseded, archived |
| Headers | show |
On 8 July 2015 at 22:31, Alan Stern <stern@rowland.harvard.edu> wrote: > On Wed, 8 Jul 2015, Rafael J. Wysocki wrote: > >> I seem to have lost the context here, sorry about that. >> >> The idea seems to be to rely on the fact that the RPM status for all devices >> is initially RPM_SUSPENDED and that never changes if runtime PM is never >> enabled for the device, so in that particular case it would be OK to treat >> the "power.direct_complete set + RPM status == RPM_SUSPENDED" combination >> as valid even though runtime PM has never been enabled for the device in >> question (provided that power.direct_complete will never be set for "real" >> devices that don't support runtime PM). Is that correct? > > I would have expressed it slightly differently, but yes, that's correct. > >> That seems to be fragile, but I have no strong opinion. > > In itself it's not all that bad, I think. In the presence of Tomeu's > new direct_complete_default flag, however, it does seem quite fragile. > > We may want to do the direct_complete_default thing in a different way. > For example, the PM core could automatically set the direct_complete > flag if a device has _none_ of the system suspend callbacks (i.e., no > prepare, suspend, suspend_late, suspend_noirq, resume_noirq, > resume_early, resume, or complete). Although it would be a little > awkward to check this, it would be safer than inheriting > direct_complete_default from the parent and it ought to solve Tomeu's > problem just as well. Yeah, I think this is an improvement. Will give it a try. >> Let's do that change if it allows us to make forward progress here. Please >> feel free to submit a documentation patch along the lines you've suggested. > > Here's a proposed patch to illustrate what I have in mind. Since it > removes the only usage of pm_runtime_suspended_if_enabled(), it also > removes the definition of that function. Will this patch be picked up as-is or should I add it to my series with a proper changelog? Thanks, Tomeu > Alan Stern > > > > Index: usb-4.1/drivers/base/power/main.c > =================================================================== > --- usb-4.1.orig/drivers/base/power/main.c > +++ usb-4.1/drivers/base/power/main.c > @@ -1374,7 +1374,7 @@ static int __device_suspend(struct devic > if (dev->power.direct_complete) { > if (pm_runtime_status_suspended(dev)) { > pm_runtime_disable(dev); > - if (pm_runtime_suspended_if_enabled(dev)) > + if (pm_runtime_status_suspended(dev)) > goto Complete; > > pm_runtime_enable(dev); > Index: usb-4.1/Documentation/power/devices.txt > =================================================================== > --- usb-4.1.orig/Documentation/power/devices.txt > +++ usb-4.1/Documentation/power/devices.txt > @@ -341,6 +341,13 @@ the phases are: > and is entirely responsible for bringing the device back to the > functional state as appropriate. > > + Note that this direct-complete procedure applies even if the device is > + disabled for runtime PM; only the runtime-PM status matters. It follows > + that if a device has system-sleep callbacks but does not support runtime > + PM, then its prepare callback must never return a positive value. This > + is because all devices are initially set to runtime-suspended with > + runtime PM disabled. > + > 2. The suspend methods should quiesce the device to stop it from performing > I/O. They also may save the device registers and put it into the > appropriate low-power state, depending on the bus type the device is on, > Index: usb-4.1/Documentation/power/runtime_pm.txt > =================================================================== > --- usb-4.1.orig/Documentation/power/runtime_pm.txt > +++ usb-4.1/Documentation/power/runtime_pm.txt > @@ -445,10 +445,6 @@ drivers/base/power/runtime.c and include > bool pm_runtime_status_suspended(struct device *dev); > - return true if the device's runtime PM status is 'suspended' > > - bool pm_runtime_suspended_if_enabled(struct device *dev); > - - return true if the device's runtime PM status is 'suspended' and its > - 'power.disable_depth' field is equal to 1 > - > void pm_runtime_allow(struct device *dev); > - set the power.runtime_auto flag for the device and decrease its usage > counter (used by the /sys/devices/.../power/control interface to > Index: usb-4.1/include/linux/pm_runtime.h > =================================================================== > --- usb-4.1.orig/include/linux/pm_runtime.h > +++ usb-4.1/include/linux/pm_runtime.h > @@ -98,11 +98,6 @@ static inline bool pm_runtime_status_sus > return dev->power.runtime_status == RPM_SUSPENDED; > } > > -static inline bool pm_runtime_suspended_if_enabled(struct device *dev) > -{ > - return pm_runtime_status_suspended(dev) && dev->power.disable_depth == 1; > -} > - > static inline bool pm_runtime_enabled(struct device *dev) > { > return !dev->power.disable_depth; > @@ -164,7 +159,6 @@ static inline void device_set_run_wake(s > static inline bool pm_runtime_suspended(struct device *dev) { return false; } > static inline bool pm_runtime_active(struct device *dev) { return true; } > static inline bool pm_runtime_status_suspended(struct device *dev) { return false; } > -static inline bool pm_runtime_suspended_if_enabled(struct device *dev) { return false; } > static inline bool pm_runtime_enabled(struct device *dev) { return false; } > > static inline void pm_runtime_no_callbacks(struct device *dev) {} > > -- > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > Please read the FAQ at http://www.tux.org/lkml/ -- To unsubscribe from this list: send the line "unsubscribe linux-pm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, 14 Jul 2015, Tomeu Vizoso wrote: > > We may want to do the direct_complete_default thing in a different way. > > For example, the PM core could automatically set the direct_complete > > flag if a device has _none_ of the system suspend callbacks (i.e., no > > prepare, suspend, suspend_late, suspend_noirq, resume_noirq, > > resume_early, resume, or complete). Although it would be a little > > awkward to check this, it would be safer than inheriting > > direct_complete_default from the parent and it ought to solve Tomeu's > > problem just as well. > > Yeah, I think this is an improvement. Will give it a try. Sounds good. > > Here's a proposed patch to illustrate what I have in mind. Since it > > removes the only usage of pm_runtime_suspended_if_enabled(), it also > > removes the definition of that function. > > Will this patch be picked up as-is or should I add it to my series > with a proper changelog? You can add it to your series with my S-O-B: Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Alan Stern -- To unsubscribe from this list: send the line "unsubscribe linux-pm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Index: usb-4.1/drivers/base/power/main.c =================================================================== --- usb-4.1.orig/drivers/base/power/main.c +++ usb-4.1/drivers/base/power/main.c @@ -1374,7 +1374,7 @@ static int __device_suspend(struct devic if (dev->power.direct_complete) { if (pm_runtime_status_suspended(dev)) { pm_runtime_disable(dev); - if (pm_runtime_suspended_if_enabled(dev)) + if (pm_runtime_status_suspended(dev)) goto Complete; pm_runtime_enable(dev); Index: usb-4.1/Documentation/power/devices.txt =================================================================== --- usb-4.1.orig/Documentation/power/devices.txt +++ usb-4.1/Documentation/power/devices.txt @@ -341,6 +341,13 @@ the phases are: and is entirely responsible for bringing the device back to the functional state as appropriate. + Note that this direct-complete procedure applies even if the device is + disabled for runtime PM; only the runtime-PM status matters. It follows + that if a device has system-sleep callbacks but does not support runtime + PM, then its prepare callback must never return a positive value. This + is because all devices are initially set to runtime-suspended with + runtime PM disabled. + 2. The suspend methods should quiesce the device to stop it from performing I/O. They also may save the device registers and put it into the appropriate low-power state, depending on the bus type the device is on, Index: usb-4.1/Documentation/power/runtime_pm.txt =================================================================== --- usb-4.1.orig/Documentation/power/runtime_pm.txt +++ usb-4.1/Documentation/power/runtime_pm.txt @@ -445,10 +445,6 @@ drivers/base/power/runtime.c and include bool pm_runtime_status_suspended(struct device *dev); - return true if the device's runtime PM status is 'suspended' - bool pm_runtime_suspended_if_enabled(struct device *dev); - - return true if the device's runtime PM status is 'suspended' and its - 'power.disable_depth' field is equal to 1 - void pm_runtime_allow(struct device *dev); - set the power.runtime_auto flag for the device and decrease its usage counter (used by the /sys/devices/.../power/control interface to Index: usb-4.1/include/linux/pm_runtime.h =================================================================== --- usb-4.1.orig/include/linux/pm_runtime.h +++ usb-4.1/include/linux/pm_runtime.h @@ -98,11 +98,6 @@ static inline bool pm_runtime_status_sus return dev->power.runtime_status == RPM_SUSPENDED; } -static inline bool pm_runtime_suspended_if_enabled(struct device *dev) -{ - return pm_runtime_status_suspended(dev) && dev->power.disable_depth == 1; -} - static inline bool pm_runtime_enabled(struct device *dev) { return !dev->power.disable_depth; @@ -164,7 +159,6 @@ static inline void device_set_run_wake(s static inline bool pm_runtime_suspended(struct device *dev) { return false; } static inline bool pm_runtime_active(struct device *dev) { return true; } static inline bool pm_runtime_status_suspended(struct device *dev) { return false; } -static inline bool pm_runtime_suspended_if_enabled(struct device *dev) { return false; } static inline bool pm_runtime_enabled(struct device *dev) { return false; } static inline void pm_runtime_no_callbacks(struct device *dev) {}