Add power management support for chipidea
Subsystems (4)
cover Add power management support for chipidea
Hi Greg,
This serial adds power management (system & runtime) for chipidea core. With this, the chipidea controller can be at low power mode when it is not in use, and the chipidea controller can be the system wakeup source. It needs to depend on my patch[1], since [1] adds some common PHY APIs, and this serial uses it.
It has been verified at Freescale i.mx6Q SabreSD platform, I will verify it at other FSL platforms during the patch review.
Hi Alan,
Due to chipidea core and imx concontroller needs some special operations during the standard ehci routine, I override .hub_control, .bus_suspend, and .bus_resume.
There is one special thing is I use flag ehci->bus_suspended to know it was a global suspend before due to I need to notify PHY when the suspend has finished (portsc.suspendM is set) and the resume signal has finished (portsc.fpr is cleared) for high speed device, but there are two places will send suspend/resume, and I don't want to patch ehci-hub.c (if you think patch ehci-hub.c is a good way, I can do it).
The related host patches:
usb: chipidea: host: add quirk for ehci operation
usb: chipidea: host: add ehci quirk for imx controller
Hi Shawn,
The last two are devicetree related to enable runtime pm.
Peter Chen (11):
usb: chipidea: Add power management support
usb: chipidea: imx: add power management support
usb: chipidea: usbmisc_imx: remove the controller's clock information
usb: chipidea: add wakeup interrupt handler
usb: chipidea: usbmisc_imx: add set_wakup API
usb: chipidea: imx: call set_wakeup when necessary
usb: chipidea: host: add quirk for ehci operation
usb: chipidea: host: add ehci quirk for imx controller
usb: chipidea: imx: Enable CI_HDRC_IMX_EHCI_QUIRK if the phy has
notify APIs
usb: chipidea: imx: add binding for supporting runtime pm
ARM: dts: imx6qdl-sabresd: Enable runtime pm for usbotg and usb host
1
.../devicetree/bindings/usb/ci13xxx-imx.txt | 2 + arch/arm/boot/dts/imx6qdl-sabresd.dtsi | 2 + drivers/usb/chipidea/ci.h | 3 + drivers/usb/chipidea/ci_hdrc_imx.c | 140 +++++++++++++++- drivers/usb/chipidea/ci_hdrc_imx.h | 1 + drivers/usb/chipidea/core.c | 139 +++++++++++++++ drivers/usb/chipidea/host.c | 180 ++++++++++++++++++++ drivers/usb/chipidea/otg.c | 5 + drivers/usb/chipidea/usbmisc_imx.c | 59 +++++-- include/linux/usb/chipidea.h | 2 + 10 files changed, 512 insertions(+), 21 deletions(-)
[1] Add power management support for MXS PHY:
http://marc.info/?l=linux-usb&m=138156975205686&w=2
PATCH 1/11 usb: chipidea: Add power management support
This commit adds runtime and system power management support for chipidea core. The runtime pm support is controlled by glue layer, it can be enabled by flag CI_HDRC_SUPPORTS_RUNTIME_PM.
--- drivers/usb/chipidea/ci.h | 2 + drivers/usb/chipidea/core.c | 119 ++++++++++++++++++++++++++++++++++++++++++ include/linux/usb/chipidea.h | 1 + 3 files changed, 122 insertions(+), 0 deletions(-)
drivers/usb/chipidea/ci.h +2 -0
@@ -173,6 +173,8 @@ struct ci_hdrc {struct dentry *debugfs;bool id_event;bool b_sess_valid_event; +bool supports_runtime_pm; +atomic_t in_lpm; };static inline struct ci_role_driver *ci_role (struct ci_hdrc *ci)
drivers/usb/chipidea/core.c +119 -0
148 lines not on this page · raw message
include/linux/usb/chipidea.h +1 -0
@@ -18,6 +18,7 @@ struct ci_hdrc_platform_data {unsigned long flags;#define CI_HDRC_REGS_SHARED BIT(0)#define CI_HDRC_REQUIRE_TRANSCEIVER BIT(1) +#define CI_HDRC_SUPPORTS_RUNTIME_PM BIT(2)#define CI_HDRC_DISABLE_STREAMING BIT(3)/* * Only set it when DCCPARAMS.DC==1 and DCCPARAMS.HC==1,
Signature · 1 line
1.7.1
PATCH 1/11 usb: chipidea: Add power management support
Hi,
Quoted 4 lines · Peter Chen
This commit adds runtime and system power management support for chipidea core. The runtime pm support is controlled by glue layer, it can be enabled by flag
CI_HDRC_SUPPORTS_RUNTIME_PM.
[...]
drivers/usb/chipidea/core.c static int ci_hdrc_remove(struct platform_device *pdev) · line 706
Quoted 10 lines
drivers/usb/chipidea/core.c +10
-0+#ifdef CONFIG_PM +static int ci_controller_suspend (struct device *dev) +{ +struct ci_hdrc *ci =dev_get_drvdata (dev); + +dev_dbg (dev,"at %s\n" , __func__); + +if (atomic_read (&ci->in_lpm)) +return 0 ; +
What does this 'atomic_read()' buy you over just testing/assinging a simple integer. Note that just because the function has 'atomic' in its name the sequence:
atomic_read(); ... atomic_set();
does not magically become an atomic operation.
Lothar Wa?mann
Signature · 9 lines
___________________________________________________________ Ka-Ro electronics GmbH | Pascalstra?e 22 | D - 52076 Aachen Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10 Gesch?ftsf?hrer: Matthias Kaussen Handelsregistereintrag: Amtsgericht Aachen, HRB 4996 www.karo-electronics.de | info at karo-electronics.de ___________________________________________________________
PATCH 1/11 usb: chipidea: Add power management support
Quoted 21 lines · Lothar Wa?mann
Hi,
Quoted 4 lines · Peter Chen
This commit adds runtime and system power management support for chipidea core. The runtime pm support is controlled by glue layer, it can be enabled by flag
CI_HDRC_SUPPORTS_RUNTIME_PM.[...]
drivers/usb/chipidea/core.c static int ci_hdrc_remove(struct platform_device *pdev) · line 706
Quoted 10 lines
drivers/usb/chipidea/core.c +10
-0+#ifdef CONFIG_PM +static int ci_controller_suspend (struct device *dev) +{ +struct ci_hdrc *ci =dev_get_drvdata (dev); + +dev_dbg (dev,"at %s\n" , __func__); + +if (atomic_read (&ci->in_lpm)) +return 0 ; +What does this '
atomic_read()' buy you over just testing/assinging a simple integer. Note that just because the function has 'atomic' in its name the sequence:
Quoted 4 lines
atomic_read(); ... atomic_set();does not magically become an atomic operation.
I just want the read and set are atomic, not the operations between atomic_read and atomic_set.
Signature · 2 lines
Best Regards, Peter Chen
Sascha Hauer
maintainerPATCH 1/11 usb: chipidea: Add power management support
Quoted 27 lines · Peter Chen
Quoted 21 lines · Lothar Wa?mann
Hi,
Quoted 4 lines · Peter Chen
This commit adds runtime and system power management support for chipidea core. The runtime pm support is controlled by glue layer, it can be enabled by flag
CI_HDRC_SUPPORTS_RUNTIME_PM.[...]
drivers/usb/chipidea/core.c static int ci_hdrc_remove(struct platform_device *pdev) · line 706
Quoted 10 lines
drivers/usb/chipidea/core.c +10
-0+#ifdef CONFIG_PM +static int ci_controller_suspend (struct device *dev) +{ +struct ci_hdrc *ci =dev_get_drvdata (dev); + +dev_dbg (dev,"at %s\n" , __func__); + +if (atomic_read (&ci->in_lpm)) +return 0 ; +What does this '
atomic_read()' buy you over just testing/assinging a simple integer. Note that just because the function has 'atomic' in its name the sequence:Quoted 4 lines
atomic_read(); ... atomic_set();does not magically become an atomic operation.
Quoted 2 lines
I just want the read and set are atomic, not the operations between
atomic_readandatomic_set.
It makes no sense to use atomic operations here. atomic types are to atomically read/modify/write variables. If all you ever do is atomic_read and atomic_set then you can use a simple bool variable.
Sascha
Signature · 4 lines
Pengutronix e.K. | | Industrial Linux Solutions | http://www.pengutronix.de/ | Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
PATCH 1/11 usb: chipidea: Add power management support
Quoted 28 lines · Sascha Hauer
Quoted 27 lines · Peter Chen
Quoted 25 lines · Lothar Wa?mann
Hi,
Quoted 4 lines · Peter Chen
This commit adds runtime and system power management support for chipidea core. The runtime pm support is controlled by glue layer, it can be enabled by flag
CI_HDRC_SUPPORTS_RUNTIME_PM.[...]
drivers/usb/chipidea/core.c static int ci_hdrc_remove(struct platform_device *pdev) · line 706
Quoted 10 lines
drivers/usb/chipidea/core.c +10
-0+#ifdef CONFIG_PM +static int ci_controller_suspend (struct device *dev) +{ +struct ci_hdrc *ci =dev_get_drvdata (dev); + +dev_dbg (dev,"at %s\n" , __func__); + +if (atomic_read (&ci->in_lpm)) +return 0 ; +What does this '
atomic_read()' buy you over just testing/assinging a simple integer. Note that just because the function has 'atomic' in its name the sequence:atomic_read(); ... atomic_set();does not magically become an atomic operation.
Quoted 7 lines
Quoted 2 lines
I just want the read and set are atomic, not the operations between
atomic_readandatomic_set.It makes no sense to use atomic operations here. atomic types are to atomically read/modify/write variables. If all you ever do is
atomic_readandatomic_setthen you can use a simple bool variable.
It is for ARM, but for other platforms, it may not.
Signature · 2 lines
Best Regards, Peter Chen
Sascha Hauer
maintainerPATCH 1/11 usb: chipidea: Add power management support
Quoted 32 lines · Peter Chen
Quoted 31 lines · Sascha Hauer
Quoted 29 lines · Peter Chen
Quoted 21 lines · Lothar Wa?mann
Hi,
Quoted 4 lines · Peter Chen
This commit adds runtime and system power management support for chipidea core. The runtime pm support is controlled by glue layer, it can be enabled by flag
CI_HDRC_SUPPORTS_RUNTIME_PM.[...]
drivers/usb/chipidea/core.c static int ci_hdrc_remove(struct platform_device *pdev) · line 706
Quoted 10 lines
drivers/usb/chipidea/core.c +10
-0+#ifdef CONFIG_PM +static int ci_controller_suspend (struct device *dev) +{ +struct ci_hdrc *ci =dev_get_drvdata (dev); + +dev_dbg (dev,"at %s\n" , __func__); + +if (atomic_read (&ci->in_lpm)) +return 0 ; +What does this '
atomic_read()' buy you over just testing/assinging a simple integer. Note that just because the function has 'atomic' in its name the sequence:Quoted 4 lines
atomic_read(); ... atomic_set();does not magically become an atomic operation.
I just want the read and set are atomic, not the operations between
atomic_readandatomic_set.
Quoted 6 lines
Quoted 4 lines
It makes no sense to use atomic operations here. atomic types are to atomically read/modify/write variables. If all you ever do is
atomic_readandatomic_setthen you can use a simple bool variable.It is for ARM, but for other platforms, it may not.
It is for other platforms aswell. Otherwise the kernel would break into pieces.
You only need atomic ops when you want to atomically read/modify/write a variable.
Sascha
Signature · 4 lines
Pengutronix e.K. | | Industrial Linux Solutions | http://www.pengutronix.de/ | Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
PATCH 1/11 usb: chipidea: Add power management support
In reply to Peter ChenQuoted 1 line · Peter Chen
It is for ARM, but for other platforms, it may not.
Wrong. atomic_read() and atomic_set() are defined the same way and have the same lack of atomicity across all architectures. There is nothing special about these over a standard load/store. Your code using this is quite simply broken and founded on false assumptions about these macros.
PATCH 1/11 usb: chipidea: Add power management support
In reply to Peter ChenQuoted 27 lines · Peter Chen
Quoted 21 lines · Lothar Wa?mann
Hi,
Quoted 4 lines · Peter Chen
This commit adds runtime and system power management support for chipidea core. The runtime pm support is controlled by glue layer, it can be enabled by flag
CI_HDRC_SUPPORTS_RUNTIME_PM.[...]
drivers/usb/chipidea/core.c static int ci_hdrc_remove(struct platform_device *pdev) · line 706
Quoted 10 lines
drivers/usb/chipidea/core.c +10
-0+#ifdef CONFIG_PM +static int ci_controller_suspend (struct device *dev) +{ +struct ci_hdrc *ci =dev_get_drvdata (dev); + +dev_dbg (dev,"at %s\n" , __func__); + +if (atomic_read (&ci->in_lpm)) +return 0 ; +What does this '
atomic_read()' buy you over just testing/assinging a simple integer. Note that just because the function has 'atomic' in its name the sequence:Quoted 4 lines
atomic_read(); ... atomic_set();does not magically become an atomic operation.
Quoted 2 lines
I just want the read and set are atomic, not the operations between
atomic_readandatomic_set.
There is nothing magical about atomic_read() or atomic_set():
#define atomic_read(v) (*(volatile int *)&(v)->counter) #define atomic_set(v,i) (((v)->counter) = (i))
You might as well just read and write the variable directly if you're going to do this. The atomicity of atomic_t comes from the *other* operations which can be done on an atomic_t, not from some magical macro which starts with the name "atomic_".
Lesson 1: whenever you see atomic_read() and atomic_set() in a patch be very very very suspicious; it's 99% of times plainly wrong and a sign of something being totally broken.
PATCH 1/11 usb: chipidea: Add power management support
In reply to Peter ChenQuoted 3 lines · Peter Chen
This commit adds runtime and system power management support for chipidea core. The runtime pm support is controlled by glue layer, it can be enabled by flag
CI_HDRC_SUPPORTS_RUNTIME_PM.
Let's look at the locking.
1. Runtime PM. These callbacks are locked with a spinlock, which holds dev->power.lock. This lock is taken either with or without disabling IRQs depending on whether runtime PM is IRQ safe or not.
2. Normal PM. These callbacks are locked by holding dev->mutex.
Now, there's a little bit of protection between these two operations - when normal PM places a device into a low power state, it 'gets' a reference on the runtime PM to ensure no runtime PM transitions occur while normal PM is active. (See pm_runtime_get_noresume() in device_prepare().) This is only dropped when the normal PM resumes the device.
Moreover, all runtime PM events are flushed before the suspend callback occurs (see the pm_runtime_barrier() in __device_suspend()).
What that means is that you can't receive any runtime PM events while you are in your suspend/resume callbacks. So, each call is mutually exclusive.
So, runtime PM callbacks vs normal PM callbacks for any single device are all called with mutual exclusion - you won't have two running at any time.
Hence, for the reasons stated previously about the non-atomic nature of atomic_read()/atomic_set(), there's even more reasons that their use here is just mere obfuscation: the accesses to this state tracking variable is already guaranteed to be single-threaded by core code, so the use of atomic_read()/atomic_set() just adds additional needless confusion to this code.
PATCH 1/11 usb: chipidea: Add power management support
Quoted 31 lines · Russell King - ARM Linux
Quoted 3 lines · Peter Chen
This commit adds runtime and system power management support for chipidea core. The runtime pm support is controlled by glue layer, it can be enabled by flag
CI_HDRC_SUPPORTS_RUNTIME_PM.Let's look at the locking.
1. Runtime PM. These callbacks are locked with a spinlock, which holds
dev->power.lock. This lock is taken either with or without disabling IRQs depending on whether runtime PM is IRQ safe or not.2. Normal PM. These callbacks are locked by holding
dev->mutex.Now, there's a little bit of protection between these two operations - when normal PM places a device into a low power state, it 'gets' a reference on the runtime PM to ensure no runtime PM transitions occur while normal PM is active. (See
pm_runtime_get_noresume()indevice_prepare().) This is only dropped when the normal PM resumes the device.Moreover, all runtime PM events are flushed before the suspend callback occurs (see the
pm_runtime_barrier()in__device_suspend()).What that means is that you can't receive any runtime PM events while you are in your suspend/resume callbacks. So, each call is mutually exclusive.
So, runtime PM callbacks vs normal PM callbacks for any single device are all called with mutual exclusion - you won't have two running at any time.
Quoted 7 lines
Hence, for the reasons stated previously about the non-atomic nature of atomic_read()/atomic_set(), there's even more reasons that their use here is just mere obfuscation: the accesses to this state tracking variable is already guaranteed to be single-threaded by core code, so the use of atomic_read()/atomic_set() just adds additional needless confusion to this code.
Many Thanks, Russell.
So, the lessons for this topic are:
- If one atomic variable's operation only includes one instruction like atomic_read and atomic_set, it is not meaningful for using atomic operation, we can just use bool instead of it.
- The runtime pm itself, normal pm itself, runtime pm and normal pm are all already exclusion, we don't need protection for variables who are only used at pm situation.
Signature · 2 lines
Best Regards, Peter Chen
PATCH 1/11 usb: chipidea: Add power management support
Quoted 5 lines · Peter Chen
So, the lessons for this topic are:
- If one atomic variable's operation only includes one instruction like
atomic_readandatomic_set, it is not meaningful for using atomic operation, we can just use bool instead of it.
The lesson here is that these are 100% equivalent as far as safety from races is concerned:
a = atomic_read(&v); a = v->counter; atomic_set(&v, b); v->counter = b;
and in general, whenever atomic_read() gets used it's almost certainly a sign of a bug.
Consider this (similar has been submitted):
a = atomic_read(&v); if (a != 0) a += 1; atomic_set(&v, a);
and people have thought that somehow this is magically safe from races because they're using atomic_t, and somehow that saves the universe. The above is in fact no safer than:
a = *v; if (a != 0) a += 1; *v = a;
The only thing that using atomic_* does is add a false sense of security and a level of obfuscation to catch the unwary reviewer.
The reason is quite simple: a single access read in itself is atomic. Either it has read the value, or it hasn't. A single access store is itself atomic. Either the data has been written, or it hasn't. The issue is _always_ what you do around it.
PATCH 2/11 usb: chipidea: imx: add power management support
In reply to Peter ChenAdd system and runtime power management support for imx gluy layer.
--- drivers/usb/chipidea/ci_hdrc_imx.c | 99 ++++++++++++++++++++++++++++++++++- 1 files changed, 96 insertions(+), 3 deletions(-)
drivers/usb/chipidea/ci_hdrc_imx.c +96 -3
@@ -28,6 +28,8 @@ struct ci_hdrc_imx_data {struct platform_device *ci_pdev;struct clk *clk;struct imx_usbmisc_data *usbmisc_data; +atomic_t in_lpm; +bool supports_runtime_pm; };/* Common functions shared by usbmisc drivers */
@@ -82,6 +84,7 @@ static int ci_hdrc_imx_probe(struct platform_device *pdev) CI_HDRC_DISABLE_STREAMING, };int ret; +struct device_node *np = pdev->dev.of_node; data =devm_kzalloc (&pdev->dev,sizeof (*data),GFP_KERNEL );if (!data) {
@@ -115,6 +118,11 @@ static int ci_hdrc_imx_probe(struct platform_device *pdev) pdata.phy = data->phy; +if (of_find_property (np,"supports_runtime_pm" ,NULL )) { + pdata.flags |=CI_HDRC_SUPPORTS_RUNTIME_PM ; + data->supports_runtime_pm = true; + } +if (!pdev->dev.dma_mask) pdev->dev.dma_mask = &pdev->dev.coherent_dma_mask;if (!pdev->dev.coherent_dma_mask)
@@ -151,8 +159,12 @@ static int ci_hdrc_imx_probe(struct platform_device *pdev)platform_set_drvdata (pdev, data); -pm_runtime_no_callbacks (&pdev->dev); -pm_runtime_enable (&pdev->dev); +device_set_wakeup_capable (&pdev->dev, true); + +if (data->supports_runtime_pm) { +pm_runtime_set_active (&pdev->dev); +pm_runtime_enable (&pdev->dev); + }return 0 ;
@@ -167,13 +179,93 @@ static int ci_hdrc_imx_remove(struct platform_device *pdev) {struct ci_hdrc_imx_data *data =platform_get_drvdata (pdev); -pm_runtime_disable (&pdev->dev);ci_hdrc_remove_device (data->ci_pdev); +if (data->supports_runtime_pm) { +pm_runtime_get_sync (&pdev->dev); +pm_runtime_disable (&pdev->dev); +pm_runtime_put_noidle (&pdev->dev); + } +clk_disable_unprepare (data->clk); + +return 0 ; +} + +#ifdef CONFIG_PM +static int imx_controller_suspend (struct device *dev) +{ +struct ci_hdrc_imx_data *data =dev_get_drvdata (dev); + +dev_dbg (dev,"at %s\n" , __func__); + +if (atomic_read (&data->in_lpm)) +return 0 ; +clk_disable_unprepare (data->clk); +atomic_set (&data->in_lpm,1 ); +return 0 ; } +static int imx_controller_resume (struct device *dev) +{ +struct ci_hdrc_imx_data *data =dev_get_drvdata (dev); +int ret =0 ; + +dev_dbg (dev,"at %s\n" , __func__); + +if (!atomic_read (&data->in_lpm)) +return 0 ; + + ret =clk_prepare_enable (data->clk); +if (ret) +return ret; + +atomic_set (&data->in_lpm,0 ); + +return ret; +} + +#ifdef CONFIG_PM_SLEEP +static int ci_hdrc_imx_suspend (struct device *dev) +{ +return imx_controller_suspend (dev); +} + +static int ci_hdrc_imx_resume (struct device *dev) +{ +struct ci_hdrc_imx_data *data =dev_get_drvdata (dev); +int ret; + + ret =imx_controller_resume (dev); +if (!ret && data->supports_runtime_pm) { +pm_runtime_disable (dev); +pm_runtime_set_active (dev); +pm_runtime_enable (dev); + } + +return ret; +} +#endif /* CONFIG_PM_SLEEP */ + +#ifdef CONFIG_PM_RUNTIME +static int ci_hdrc_imx_runtime_suspend (struct device *dev) +{ +return imx_controller_suspend (dev); +} + +static int ci_hdrc_imx_runtime_resume (struct device *dev) +{ +return imx_controller_resume (dev); +} +#endif /* CONFIG_PM_RUNTIME */ + +#endif /* CONFIG_PM */ +static const struct dev_pm_ops ci_hdrc_imx_pm_ops = { + SET_SYSTEM_SLEEP_PM_OPS(ci_hdrc_imx_suspend, ci_hdrc_imx_resume) +SET_RUNTIME_PM_OPS (ci_hdrc_imx_runtime_suspend, + ci_hdrc_imx_runtime_resume,NULL ) +};static const struct of_device_id ci_hdrc_imx_dt_ids[] = { { .compatible ="fsl,imx27-usb" , }, {/* sentinel */ }
@@ -187,6 +279,7 @@ static struct platform_driver ci_hdrc_imx_driver = { .name = "imx_usb", .owner = THIS_MODULE, .of_match_table = ci_hdrc_imx_dt_ids, + .pm = &ci_hdrc_imx_pm_ops, }, };
Signature · 1 line
1.7.1
PATCH 3/11 usb: chipidea: usbmisc_imx: remove the controller's clock information
In reply to Peter ChenCurrently, the usbmisc is just an API supplier for controller driver, the controller calls related APIs to handle different things among the SoCs, before calling it, the clock must be on. So the clock operation is useless for usbmisc, it also increases the difficulties to manage the clock, especially at runtime power management situation.
--- drivers/usb/chipidea/usbmisc_imx.c | 18 ------------------ 1 files changed, 0 insertions(+), 18 deletions(-)
drivers/usb/chipidea/usbmisc_imx.c +0 -18
@@ -11,7 +11,6 @@#include <linux/module.h> #include <linux/of_platform.h> -#include <linux/clk.h> #include <linux/err.h> #include <linux/io.h> #include <linux/delay.h>
@@ -40,7 +39,6 @@ struct usbmisc_ops {struct imx_usbmisc {void __iomem *base;spinlock_t lock; -struct clk *clk;const struct usbmisc_ops *ops; };
@@ -177,7 +175,6 @@ static int usbmisc_imx_probe(struct platform_device *pdev) {struct resource *res;struct imx_usbmisc *data; -int ret;struct of_device_id *tmp_dev;if (usbmisc)
@@ -194,20 +191,6 @@ static int usbmisc_imx_probe(struct platform_device *pdev)if (IS_ERR (data->base))return PTR_ERR (data->base); - data->clk =devm_clk_get (&pdev->dev,NULL ); -if (IS_ERR (data->clk)) { -dev_err (&pdev->dev, -"failed to get clock, err=%ld\n" ,PTR_ERR (data->clk)); -return PTR_ERR (data->clk); - } - - ret =clk_prepare_enable (data->clk); -if (ret) { -dev_err (&pdev->dev, -"clk_prepare_enable failed, err=%d\n" , ret); -return ret; - } - tmp_dev = (struct of_device_id *)of_match_device (usbmisc_imx_dt_ids, &pdev->dev); data->ops = (const struct usbmisc_ops *)tmp_dev->data;
@@ -218,7 +201,6 @@ static int usbmisc_imx_probe(struct platform_device *pdev)static int usbmisc_imx_remove (struct platform_device *pdev) { -clk_disable_unprepare (usbmisc->clk); usbmisc =NULL ;return 0 ; }
Signature · 1 line
1.7.1
PATCH 4/11 usb: chipidea: add wakeup interrupt handler
In reply to Peter ChenWhen the controller is at suspend mode, it can be waken up by external events (like vbus, dp/dm or id change). Once we receive the wakeup interrupt, we need to resume the controller first, eg open clocks, disable some wakeup settings, etc. After that, the controller can receive the normal USB interrupts.
--- drivers/usb/chipidea/ci.h | 1 + drivers/usb/chipidea/core.c | 22 +++++++++++++++++++++- drivers/usb/chipidea/otg.c | 5 +++++ 3 files changed, 27 insertions(+), 1 deletions(-)
drivers/usb/chipidea/ci.h +1 -0
@@ -175,6 +175,7 @@ struct ci_hdrc {bool b_sess_valid_event;bool supports_runtime_pm;atomic_t in_lpm; +bool wakeup_int; };static inline struct ci_role_driver *ci_role (struct ci_hdrc *ci)
drivers/usb/chipidea/core.c +21 -1
@@ -190,6 +190,13 @@ static void ci_hdrc_enter_lpm(struct ci_hdrc *ci, bool enable)* than 1ms) to leave low power mode. */ usleep_range (1500 ,2000 ); + }else if (!enable) { +/* +* At wakeup interrupt, the phcd will be cleared by hardware +* automatically, but the controller needs at least 1ms +* to reflect PHY's status. +*/ +usleep_range (1200 ,1800 ); } }
@@ -355,6 +362,13 @@ static irqreturn_t ci_irq(int irq, void *data)irqreturn_t ret =IRQ_NONE ;u32 otgsc =0 ; +if (atomic_read (&ci->in_lpm)) { +disable_irq_nosync (irq); + ci->wakeup_int = true; +pm_runtime_get (ci->dev); +return IRQ_HANDLED ; + } +if (ci->is_otg) otgsc =hw_read (ci,OP_OTGSC , ~0 );
@@ -737,7 +751,13 @@ static int ci_controller_resume(struct device *dev)usb_phy_set_wakeup (ci->transceiver, false); } -atomic_set (&ci->in_lpm,0 ); +if (ci->wakeup_int) { + ci->wakeup_int = false; +atomic_set (&ci->in_lpm,0 ); +enable_irq (ci->irq); +pm_runtime_put (ci->dev); + }else +atomic_set (&ci->in_lpm,0 );return 0 ; }
drivers/usb/chipidea/otg.c +5 -0
@@ -78,10 +78,15 @@ static void ci_otg_work(struct work_struct *work)if (ci->id_event) { ci->id_event = false; +/* Keep controller active during id switch */ +pm_runtime_get_sync (ci->dev);ci_handle_id_switch (ci); +pm_runtime_put_sync (ci->dev); }else if (ci->b_sess_valid_event) { ci->b_sess_valid_event = false; +pm_runtime_get_sync (ci->dev);ci_handle_vbus_change (ci); +pm_runtime_put_sync (ci->dev); }else dev_err (ci->dev,"unexpected event occurs at %s\n" , __func__);
Signature · 1 line
1.7.1
PATCH 5/11 usb: chipidea: usbmisc_imx: add set_wakup API
In reply to Peter ChenIt is used to enable USB wakeup
--- drivers/usb/chipidea/ci_hdrc_imx.h | 1 + drivers/usb/chipidea/usbmisc_imx.c | 41 ++++++++++++++++++++++++++++++++++++ 2 files changed, 42 insertions(+), 0 deletions(-)
drivers/usb/chipidea/ci_hdrc_imx.h +1 -0
@@ -18,3 +18,4 @@ struct imx_usbmisc_data {int imx_usbmisc_init (struct imx_usbmisc_data *);int imx_usbmisc_init_post (struct imx_usbmisc_data *); +int imx_usbmisc_set_wakeup (struct imx_usbmisc_data *,bool );
drivers/usb/chipidea/usbmisc_imx.c +41 -0
71 lines not on this page · raw message
Signature · 1 line
1.7.1
PATCH 6/11 usb: chipidea: imx: call set_wakeup when necessary
In reply to Peter Chen- Disable wakeup after probe
- Enable wakeup during the suspend
- Disable wakeup after controller is active
--- drivers/usb/chipidea/ci_hdrc_imx.c | 36 ++++++++++++++++++++++++++++++++++++ 1 files changed, 36 insertions(+), 0 deletions(-)
drivers/usb/chipidea/ci_hdrc_imx.c +36 -0
@@ -157,6 +157,15 @@ static int ci_hdrc_imx_probe(struct platform_device *pdev) } } +if (data->usbmisc_data) { + ret =imx_usbmisc_set_wakeup (data->usbmisc_data, false); +if (ret) { +dev_err (&pdev->dev,"usbmisc set_wakeup failed, ret=%d\n" , + ret); + goto disable_device; + } + } +platform_set_drvdata (pdev, data);device_set_wakeup_capable (&pdev->dev, true);
@@ -194,12 +203,23 @@ static int ci_hdrc_imx_remove(struct platform_device *pdev)static int imx_controller_suspend (struct device *dev) {struct ci_hdrc_imx_data *data =dev_get_drvdata (dev); +int ret;dev_dbg (dev,"at %s\n" , __func__);if (atomic_read (&data->in_lpm))return 0 ; +if (data->usbmisc_data) { + ret =imx_usbmisc_set_wakeup (data->usbmisc_data, true); +if (ret) { +dev_err (dev, +"usbmisc set_wakeup failed, ret=%d\n" , + ret); +return ret; + } + } +clk_disable_unprepare (data->clk);atomic_set (&data->in_lpm,1 );
@@ -221,8 +241,24 @@ static int imx_controller_resume(struct device *dev)if (ret)return ret; +if (data->usbmisc_data) { + ret =imx_usbmisc_set_wakeup (data->usbmisc_data, false); +if (ret) { +dev_err (dev, +"usbmisc set_wakeup failed, ret=%d\n" , + ret); + ret = -EINVAL ; + goto clk_disable; + } + } +atomic_set (&data->in_lpm,0 ); +return 0 ; + +clk_disable: +clk_disable_unprepare (data->clk); +return ret; }
Signature · 1 line
1.7.1
PATCH 7/11 usb: chipidea: host: add quirk for ehci operation
In reply to Peter ChenFor chipidea controller, it does not follow ehci spec strictly. Taking resume signal as an example, it will stop resume signal about 20-21ms later automatically, but standard ehci spec says, the resume signal is controlled by software (clear portsc.PORT_RESUME).
This operation causes some remote wakeup problems for high speed devices due to host controller does not send SOF in time since software can't guarantee set run/stop bit in time (run/stop bit was cleared at the ehci suspend routine).
When software sets run/stop bit, it needs 1 SoF time to make it effect. If we close the PHY clock just after setting run/stop bit, it does not be set in practice, so a software delay is needed.
--- drivers/usb/chipidea/host.c | 51 +++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 51 insertions(+), 0 deletions(-)
drivers/usb/chipidea/host.c +51 -0
@@ -33,6 +33,53 @@#include "host.h" static struct hc_driver __read_mostly ci_ehci_hc_driver; +static int (*orig_bus_suspend)(struct usb_hcd *hcd); + +static int ci_ehci_bus_suspend (struct usb_hcd *hcd) +{ +struct ehci_hcd *ehci =hcd_to_ehci (hcd); +int port; +u32 tmp; + +int ret =orig_bus_suspend (hcd); + +if (ret) +return ret; + + port =HCS_N_PORTS (ehci->hcs_params); +while (port--) { +u32 __iomem *reg = &ehci->regs->port_status[port]; +u32 portsc =ehci_readl (ehci, reg); + +if (portsc &PORT_CONNECT ) { +/* +* For chipidea, the resume signal will be ended +* automatically, so for remote wakeup case, the +* usbcmd.rs may not be set before the resume has +* ended if other resume path consumes too much +* time (~23ms-24ms), in that case, the SOF will not +* send out within 3ms after resume ends, then the +* device will enter suspend again. +*/ +if (hcd->self.root_hub->do_remote_wakeup) { +ehci_dbg (ehci, +"Remote wakeup is enabled, " +"and device is on the port\n" ); + + tmp =ehci_readl (ehci, &ehci->regs->command); + tmp |=CMD_RUN ; +ehci_writel (ehci, tmp, &ehci->regs->command); +/* +* It needs a short delay between set RUNSTOP +* and set PHCD. +*/ +udelay (125 ); + } + } + } + +return 0 ; +}static irqreturn_t host_irq (struct ci_hdrc *ci) {
@@ -132,5 +179,9 @@ int ci_hdrc_host_init(struct ci_hdrc *ci)ehci_init_driver (&ci_ehci_hc_driver,NULL ); + orig_bus_suspend = ci_ehci_hc_driver.bus_suspend; + + ci_ehci_hc_driver.bus_suspend = ci_ehci_bus_suspend; +return 0 ; }
Signature · 1 line
1.7.1
PATCH 8/11 usb: chipidea: host: add ehci quirk for imx controller
In reply to Peter ChenWhen the port goes to suspend or finishes resme, it needs to notify PHY, it is not a standard EHCI operation, so we add a quirk for it.
--- drivers/usb/chipidea/host.c | 129 ++++++++++++++++++++++++++++++++++++++++++ include/linux/usb/chipidea.h | 1 + 2 files changed, 130 insertions(+), 0 deletions(-)
drivers/usb/chipidea/host.c +129 -0
@@ -34,6 +34,10 @@static struct hc_driver __read_mostly ci_ehci_hc_driver;static int (*orig_bus_suspend)(struct usb_hcd *hcd); +static int (*orig_bus_resume)(struct usb_hcd *hcd); +static int (*orig_hub_control)(struct usb_hcd *hcd, +u16 typeReq,u16 wValue,u16 wIndex, +char *buf,u16 wLength);static int ci_ehci_bus_suspend (struct usb_hcd *hcd) {
@@ -75,12 +79,131 @@ static int ci_ehci_bus_suspend(struct usb_hcd *hcd)*/ udelay (125 ); } +if (hcd->phy &&test_bit (port, &ehci->bus_suspended) + && (ehci_port_speed (ehci, portsc) == +USB_PORT_STAT_HIGH_SPEED )) +/* +* notify the USB PHY, it is for global +* suspend case. +*/ +usb_phy_notify_suspend (hcd->phy, +USB_SPEED_HIGH ); } }return 0 ; } +static int ci_imx_ehci_bus_resume (struct usb_hcd *hcd) +{ +struct ehci_hcd *ehci =hcd_to_ehci (hcd); +int port; + +int ret =orig_bus_resume (hcd); + +if (ret) +return ret; + + port =HCS_N_PORTS (ehci->hcs_params); +while (port--) { +u32 __iomem *reg = &ehci->regs->port_status[port]; +u32 portsc =ehci_readl (ehci, reg); +/* +* Notify PHY after resume signal has finished, it is +* for global suspend case. +*/ +if (hcd->phy + &&test_bit (port, &ehci->bus_suspended) + && (portsc &PORT_CONNECT ) + && (ehci_port_speed (ehci, portsc) == +USB_PORT_STAT_HIGH_SPEED )) +/* notify the USB PHY */ +usb_phy_notify_resume (hcd->phy,USB_SPEED_HIGH ); + } + +return 0 ; +} + +/* The below code is based on tegra ehci driver */ +static int ci_imx_ehci_hub_control ( +struct usb_hcd *hcd, +u16 typeReq, +u16 wValue, +u16 wIndex, +char *buf, +u16 wLength +) +{ +struct ehci_hcd *ehci =hcd_to_ehci (hcd); +u32 __iomem *status_reg; +u32 temp; +unsigned long flags; +int retval =0 ; + + status_reg = &ehci->regs->port_status[(wIndex &0xff ) -1 ]; + +spin_lock_irqsave (&ehci->lock, flags); + +if (typeReq == SetPortFeature && wValue ==USB_PORT_FEAT_SUSPEND ) { + temp =ehci_readl (ehci, status_reg); +if ((temp &PORT_PE ) ==0 || (temp &PORT_RESET ) !=0 ) { + retval = -EPIPE ; + goto done; + } + + temp &= ~(PORT_RWC_BITS |PORT_WKCONN_E ); + temp |=PORT_WKDISC_E |PORT_WKOC_E ; +ehci_writel (ehci, temp |PORT_SUSPEND , status_reg); + +/* +* If a transaction is in progress, there may be a delay in +* suspending the port. Poll until the port is suspended. +*/ +if (ehci_handshake (ehci, status_reg,PORT_SUSPEND , +PORT_SUSPEND ,5000 )) +ehci_err (ehci,"timeout waiting for SUSPEND\n" ); + +spin_unlock_irqrestore (&ehci->lock, flags); +if (ehci_port_speed (ehci, temp) == +USB_PORT_STAT_HIGH_SPEED && hcd->phy) { +/* notify the USB PHY */ +usb_phy_notify_suspend (hcd->phy,USB_SPEED_HIGH ); + } +spin_lock_irqsave (&ehci->lock, flags); + +set_bit ((wIndex &0xff ) -1 , &ehci->suspended_ports); + goto done; + } + +/* +* After resume has finished, it needs do some post resume +* operation for some SoCs. +*/ +else if (typeReq == ClearPortFeature && + wValue ==USB_PORT_FEAT_C_SUSPEND ) { + +/* Make sure the resume has finished, it should be finished */ +if (ehci_handshake (ehci, status_reg,PORT_RESUME ,0 ,25000 )) +ehci_err (ehci,"timeout waiting for resume\n" ); + + temp =ehci_readl (ehci, status_reg); + +if (ehci_port_speed (ehci, temp) == +USB_PORT_STAT_HIGH_SPEED && hcd->phy) { +/* notify the USB PHY */ +usb_phy_notify_resume (hcd->phy,USB_SPEED_HIGH ); + } + } + +spin_unlock_irqrestore (&ehci->lock, flags); + +/* Handle the hub control events here */ +return orig_hub_control (hcd, typeReq, wValue, wIndex, buf, wLength); +done: +spin_unlock_irqrestore (&ehci->lock, flags); +return retval; +} +static irqreturn_t host_irq (struct ci_hdrc *ci) {return usb_hcd_irq (ci->irq, ci->hcd);
@@ -180,8 +303,14 @@ int ci_hdrc_host_init(struct ci_hdrc *ci)ehci_init_driver (&ci_ehci_hc_driver,NULL ); orig_bus_suspend = ci_ehci_hc_driver.bus_suspend; + orig_bus_resume = ci_ehci_hc_driver.bus_resume; + orig_hub_control = ci_ehci_hc_driver.hub_control; ci_ehci_hc_driver.bus_suspend = ci_ehci_bus_suspend; +if (ci->platdata->flags &CI_HDRC_IMX_EHCI_QUIRK ) { + ci_ehci_hc_driver.bus_resume = ci_imx_ehci_bus_resume; + ci_ehci_hc_driver.hub_control = ci_imx_ehci_hub_control; + }return 0 ; }
include/linux/usb/chipidea.h +1 -0
@@ -25,6 +25,7 @@ struct ci_hdrc_platform_data {* but otg is not supported (no register otgsc). */ #define CI_HDRC_DUAL_ROLE_NOT_OTG BIT(4) +#define CI_HDRC_IMX_EHCI_QUIRK BIT(5)enum usb_dr_mode dr_mode;#define CI_HDRC_CONTROLLER_RESET_EVENT 0#define CI_HDRC_CONTROLLER_STOPPED_EVENT 1
Signature · 1 line
1.7.1
PATCH 9/11 usb: chipidea: imx: Enable CI_HDRC_IMX_EHCI_QUIRK if the phy has notify APIs
In reply to Peter ChenIf the PHY has .notify_suspend and .notify_resume, it means this imx usb controller needs to add quirks for standard ehci routine.
--- drivers/usb/chipidea/ci_hdrc_imx.c | 5 +++++ 1 files changed, 5 insertions(+), 0 deletions(-)
drivers/usb/chipidea/ci_hdrc_imx.c +5 -0
@@ -118,6 +118,11 @@ static int ci_hdrc_imx_probe(struct platform_device *pdev) pdata.phy = data->phy; +/* notify APIs are used for adding quirks at standard EHCI routine */ +if (pdata.phy && pdata.phy->notify_suspend + && pdata.phy->notify_resume) + pdata.flags |=CI_HDRC_IMX_EHCI_QUIRK ; +if (of_find_property (np,"supports_runtime_pm" ,NULL )) { pdata.flags |=CI_HDRC_SUPPORTS_RUNTIME_PM ; data->supports_runtime_pm = true;
Signature · 1 line
1.7.1
PATCH 10/11 usb: chipidea: imx: add binding for supporting runtime pm
In reply to Peter ChenAdd property for supporting runtime power management
--- .../devicetree/bindings/usb/ci13xxx-imx.txt | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-)
Documentation/devicetree/bindings/usb/ci13xxx-imx.txt +2 -0
@@ -18,6 +18,7 @@ Optional properties: - vbus-supply: regulator for vbus - disable-over-current: disable over current detect - external-vbus-divider: enables off-chip resistor divider for Vbus +- supports_runtime_pm: enable runtime pm support Examples: usb at 02184000 { /* USB OTG */
@@ -28,4 +29,5 @@ usb at 02184000 { /* USB OTG */ fsl,usbmisc = <&usbmisc 0>; disable-over-current; external-vbus-divider; + supports_runtime_pm; };
Signature · 1 line
1.7.1
PATCH 10/11 usb: chipidea: imx: add binding for supporting runtime pm
Quoted 20 lines · Peter Chen
Add property for supporting runtime power management
Signed-off-by Peter Chen <[email protected]>--- .../devicetree/bindings/usb/ci13xxx-imx.txt | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-)Documentation/devicetree/bindings/usb/ci13xxx-imx.txt +1
-0@@ -18,6 +18,7 @@ Optional properties: - vbus-supply: regulator for vbus - disable-over-current: disable over current detect - external-vbus-divider: enables off-chip resistor divider for Vbus +- supports_runtime_pm: enable runtime pm support Examples: usb at 02184000 { /* USB OTG */
Documentation/devicetree/bindings/usb/ci13xxx-imx.txt usb at 02184000 { /* USB OTG */ · line 33
Quoted 5 lines
Documentation/devicetree/bindings/usb/ci13xxx-imx.txt +1
-0@@ -28,4 +29,5 @@ usb at 02184000 { /* USB OTG */ fsl,usbmisc = <&usbmisc 0>; disable-over-current; external-vbus-divider; + supports_runtime_pm; };
This does not sound like a property of the hardware. What's the _hardware_ difference between parts that support runtime PM and parts that don't?
Alan Stern
PATCH 10/11 usb: chipidea: imx: add binding for supporting runtime pm
Quoted 19 lines · Alan Stern
Quoted 17 lines · Peter Chen
Add property for supporting runtime power management
Signed-off-by Peter Chen <[email protected]>--- .../devicetree/bindings/usb/ci13xxx-imx.txt | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-)Documentation/devicetree/bindings/usb/ci13xxx-imx.txt +1
-0@@ -18,6 +18,7 @@ Optional properties: - vbus-supply: regulator for vbus - disable-over-current: disable over current detect - external-vbus-divider: enables off-chip resistor divider for Vbus +- supports_runtime_pm: enable runtime pm support
Quoted 12 lines
Documentation/devicetree/bindings/usb/ci13xxx-imx.txt usb at 02184000 { /* USB OTG */ · line 33
Quoted 8 lines
Documentation/devicetree/bindings/usb/ci13xxx-imx.txt +1
-0Examples: usb at 02184000 { /* USB OTG */@@ -28,4 +29,5 @@ usb at 02184000 { /* USB OTG */ fsl,usbmisc = <&usbmisc 0>; disable-over-current; external-vbus-divider; + supports_runtime_pm; };This does not sound like a property of the hardware. What's the _hardware_ difference between parts that support runtime PM and parts that don't?
Thanks.
Quoted 1 line
From my point, all hardware using chipidea core should support runtime pm.
But some of platforms need special glue layer operations to support it, it will break other platforms if enable chipidea core runtime pm. Since device tree describes hardware property, maybe I should move it to glue layer, or do you have any suggestions?
Signature · 2 lines
Best Regards, Peter Chen
PATCH 10/11 usb: chipidea: imx: add binding for supporting runtime pm
Dear Peter Chen,
Quoted 40 lines · Alan Stern
Quoted 31 lines · Peter Chen
Quoted 30 lines
Add property for supporting runtime power management
Signed-off-by Peter Chen <[email protected]>---
.../devicetree/bindings/usb/ci13xxx-imx.txt | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-)(quoted diff) +0
-0
b/Documentation/devicetree/bindings/usb/ci13xxx-imx.txtindex b4b5b79..f666598 100644Documentation/devicetree/bindings/usb/ci13xxx-imx.txt +1
-0@@ -18,6 +18,7 @@ Optional properties: - vbus-supply: regulator for vbus - disable-over-current: disable over current detect - external-vbus-divider: enables off-chip resistor divider for Vbus +- supports_runtime_pm: enable runtime pm support Examples:usb at 02184000 { /* USB OTG */
Documentation/devicetree/bindings/usb/ci13xxx-imx.txt +0
-0@@ -28,4 +29,5 @@ usb at 02184000 { /* USB OTG */ fsl,usbmisc = <&usbmisc 0>; disable-over-current; external-vbus-divider;Quoted 7 lines
Quoted 3 lines
Documentation/devicetree/bindings/usb/ci13xxx-imx.txt +1
-0+ supports_runtime_pm;};
This does not sound like a property of the hardware. What's the _hardware_ difference between parts that support runtime PM and parts that don't?
Quoted 7 lines
Thanks.
From my point, all hardware using chipidea core should support runtime pm. But some of platforms need special glue layer operations to support it, it will break other platforms if enable chipidea core runtime pm. Since device tree describes hardware property, maybe I should move it to glue layer, or do you have any suggestions?
You should certainly move it out of DT. This is linux-specific property, it has nothing to do with HW. The best course of action would be to fix those platforms that are broken by runtime PM.
Best regards,
Marek Vasut
PATCH 10/11 usb: chipidea: imx: add binding for supporting runtime pm
Quoted 50 lines · Marek Vasut
Dear Peter Chen,
Quoted 40 lines · Alan Stern
Quoted 31 lines · Peter Chen
Quoted 30 lines
Add property for supporting runtime power management
Signed-off-by Peter Chen <[email protected]>---
.../devicetree/bindings/usb/ci13xxx-imx.txt | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-)(quoted diff) +0
-0
b/Documentation/devicetree/bindings/usb/ci13xxx-imx.txtindex b4b5b79..f666598 100644Documentation/devicetree/bindings/usb/ci13xxx-imx.txt +1
-0@@ -18,6 +18,7 @@ Optional properties: - vbus-supply: regulator for vbus - disable-over-current: disable over current detect - external-vbus-divider: enables off-chip resistor divider for Vbus +- supports_runtime_pm: enable runtime pm support Examples:usb at 02184000 { /* USB OTG */
Documentation/devicetree/bindings/usb/ci13xxx-imx.txt +0
-0@@ -28,4 +29,5 @@ usb at 02184000 { /* USB OTG */ fsl,usbmisc = <&usbmisc 0>; disable-over-current; external-vbus-divider;Quoted 7 lines
Quoted 3 lines
Documentation/devicetree/bindings/usb/ci13xxx-imx.txt +1
-0+ supports_runtime_pm;};
This does not sound like a property of the hardware. What's the _hardware_ difference between parts that support runtime PM and parts that don't?
Quoted 7 lines
Thanks.
From my point, all hardware using chipidea core should support runtime pm. But some of platforms need special glue layer operations to support it, it will break other platforms if enable chipidea core runtime pm. Since device tree describes hardware property, maybe I should move it to glue layer, or do you have any suggestions?
Quoted 4 lines
You should certainly move it out of DT. This is linux-specific property, it has nothing to do with HW. The best course of action would be to fix those platforms that are broken by runtime PM.
Thanks, I move it out of DT. But I can only verify it at i.mx platform, I will add a flag at glue layer.
Signature · 2 lines
Best Regards, Peter Chen
PATCH 11/11 ARM: dts: imx6qdl-sabresd: Enable runtime pm for usbotg and usb host 1
In reply to Peter ChenEnable runtime power management support for usbotg and usb host 1.
--- arch/arm/boot/dts/imx6qdl-sabresd.dtsi | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-)
arch/arm/boot/dts/imx6qdl-sabresd.dtsi +2 -0
@@ -215,6 +215,7 @@ &usbh1 { vbus-supply = <®_usb_h1_vbus>; + supports_runtime_pm; status = "okay"; };
@@ -223,6 +224,7 @@ pinctrl-names = "default"; pinctrl-0 = <&pinctrl_usbotg_2>; disable-over-current; + supports_runtime_pm; status = "okay"; };
Signature · 1 line
1.7.1