Message ID | 20160918165018.24547-1-hdegoede@redhat.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Hi, On Sunday 18 September 2016 10:20 PM, Hans de Goede wrote: > The sunxi musb has a bug where sometimes it will generate a babble > error on device disconnect instead of a disconnect irq. When this > happens the musb-controller switches from host mode to device mode > (it clears MUSB_DEVCTL_SESSION and sets MUSB_DEVCTL_BDEVICE) and > gets stuck in this state. > > Clearing this requires reporting Vbus low for 200 or more ms, but > on some devices Vbus is simply always high (host-only mode, no Vbus > control). The phy-sun4i-usb code already has code to force a session > end for devices without Vbus control. > > This commit adds a sun4i_usb_phy_force_session_end() function exporting > this functionality to the sunxi-musb glue, so that it can force a session > end to fixup the stuck state after a babble error. > > Signed-off-by: Hans de Goede <hdegoede@redhat.com> > --- > drivers/phy/phy-sun4i-usb.c | 11 +++++++++++ > include/linux/phy/phy-sun4i-usb.h | 7 +++++++ > 2 files changed, 18 insertions(+) > > diff --git a/drivers/phy/phy-sun4i-usb.c b/drivers/phy/phy-sun4i-usb.c > index 43c0d98..06f4e11a 100644 > --- a/drivers/phy/phy-sun4i-usb.c > +++ b/drivers/phy/phy-sun4i-usb.c > @@ -470,6 +470,17 @@ void sun4i_usb_phy_set_squelch_detect(struct phy *_phy, bool enabled) > } > EXPORT_SYMBOL_GPL(sun4i_usb_phy_set_squelch_detect); > > +void sun4i_usb_phy_force_session_end(struct phy *_phy) > +{ > + struct sun4i_usb_phy *phy = phy_get_drvdata(_phy); > + struct sun4i_usb_phy_data *data = to_sun4i_usb_phy_data(phy); > + > + data->id_det = -1; > + data->force_session_end = true; > + queue_delayed_work(system_wq, &data->detect, 0); > +} > +EXPORT_SYMBOL_GPL(sun4i_usb_phy_force_session_end); Er.. one more export symbol :-( > + > static const struct phy_ops sun4i_usb_phy_ops = { > .init = sun4i_usb_phy_init, > .exit = sun4i_usb_phy_exit, > diff --git a/include/linux/phy/phy-sun4i-usb.h b/include/linux/phy/phy-sun4i-usb.h > index 50aed92..3bb773f 100644 > --- a/include/linux/phy/phy-sun4i-usb.h > +++ b/include/linux/phy/phy-sun4i-usb.h > @@ -23,4 +23,11 @@ > */ > void sun4i_usb_phy_set_squelch_detect(struct phy *phy, bool enabled); > > +/** > + * sun4i_usb_force_session_end() - Force the current session to end > + * by reporting VBus low for 200+ ms > + * @phy: reference to a sun4i usb phy > + */ > +void sun4i_usb_phy_force_session_end(struct phy *phy); Should we include a static inline function if sun4i phy is not defined? Thanks Kishon
Hi, On 09/20/2016 07:45 AM, Kishon Vijay Abraham I wrote: > Hi, > > On Sunday 18 September 2016 10:20 PM, Hans de Goede wrote: >> The sunxi musb has a bug where sometimes it will generate a babble >> error on device disconnect instead of a disconnect irq. When this >> happens the musb-controller switches from host mode to device mode >> (it clears MUSB_DEVCTL_SESSION and sets MUSB_DEVCTL_BDEVICE) and >> gets stuck in this state. >> >> Clearing this requires reporting Vbus low for 200 or more ms, but >> on some devices Vbus is simply always high (host-only mode, no Vbus >> control). The phy-sun4i-usb code already has code to force a session >> end for devices without Vbus control. >> >> This commit adds a sun4i_usb_phy_force_session_end() function exporting >> this functionality to the sunxi-musb glue, so that it can force a session >> end to fixup the stuck state after a babble error. >> >> Signed-off-by: Hans de Goede <hdegoede@redhat.com> >> --- >> drivers/phy/phy-sun4i-usb.c | 11 +++++++++++ >> include/linux/phy/phy-sun4i-usb.h | 7 +++++++ >> 2 files changed, 18 insertions(+) >> >> diff --git a/drivers/phy/phy-sun4i-usb.c b/drivers/phy/phy-sun4i-usb.c >> index 43c0d98..06f4e11a 100644 >> --- a/drivers/phy/phy-sun4i-usb.c >> +++ b/drivers/phy/phy-sun4i-usb.c >> @@ -470,6 +470,17 @@ void sun4i_usb_phy_set_squelch_detect(struct phy *_phy, bool enabled) >> } >> EXPORT_SYMBOL_GPL(sun4i_usb_phy_set_squelch_detect); >> >> +void sun4i_usb_phy_force_session_end(struct phy *_phy) >> +{ >> + struct sun4i_usb_phy *phy = phy_get_drvdata(_phy); >> + struct sun4i_usb_phy_data *data = to_sun4i_usb_phy_data(phy); >> + >> + data->id_det = -1; >> + data->force_session_end = true; >> + queue_delayed_work(system_wq, &data->detect, 0); >> +} >> +EXPORT_SYMBOL_GPL(sun4i_usb_phy_force_session_end); > > Er.. one more export symbol :-( Yes unfortunately we need one more to work around sunxi musb / phy bugs. >> + >> static const struct phy_ops sun4i_usb_phy_ops = { >> .init = sun4i_usb_phy_init, >> .exit = sun4i_usb_phy_exit, >> diff --git a/include/linux/phy/phy-sun4i-usb.h b/include/linux/phy/phy-sun4i-usb.h >> index 50aed92..3bb773f 100644 >> --- a/include/linux/phy/phy-sun4i-usb.h >> +++ b/include/linux/phy/phy-sun4i-usb.h >> @@ -23,4 +23,11 @@ >> */ >> void sun4i_usb_phy_set_squelch_detect(struct phy *phy, bool enabled); >> >> +/** >> + * sun4i_usb_force_session_end() - Force the current session to end >> + * by reporting VBus low for 200+ ms >> + * @phy: reference to a sun4i usb phy >> + */ >> +void sun4i_usb_phy_force_session_end(struct phy *phy); > > Should we include a static inline function if sun4i phy is not defined? No, we're also not doing that for the already exported sun4i_usb_phy_set_squelch_detect() And it is not necessary since the only caller is drivers/usb/musb/sunxi.c, and drivers/usb/musb/Kconfig has: config USB_MUSB_SUNXI tristate "Allwinner (sunxi)" depends on PHY_SUN4I_USB Regards, Hans
Hi, On Wed, Sep 21, 2016 at 11:05:33AM +0300, Hans de Goede wrote: > Hi, > > On 09/20/2016 07:45 AM, Kishon Vijay Abraham I wrote: > >Hi, > > > >On Sunday 18 September 2016 10:20 PM, Hans de Goede wrote: > >>The sunxi musb has a bug where sometimes it will generate a babble > >>error on device disconnect instead of a disconnect irq. When this > >>happens the musb-controller switches from host mode to device mode > >>(it clears MUSB_DEVCTL_SESSION and sets MUSB_DEVCTL_BDEVICE) and > >>gets stuck in this state. > >> > >>Clearing this requires reporting Vbus low for 200 or more ms, but > >>on some devices Vbus is simply always high (host-only mode, no Vbus > >>control). The phy-sun4i-usb code already has code to force a session > >>end for devices without Vbus control. > >> > >>This commit adds a sun4i_usb_phy_force_session_end() function exporting > >>this functionality to the sunxi-musb glue, so that it can force a session > >>end to fixup the stuck state after a babble error. > >> > >>Signed-off-by: Hans de Goede <hdegoede@redhat.com> > >>--- > >> drivers/phy/phy-sun4i-usb.c | 11 +++++++++++ > >> include/linux/phy/phy-sun4i-usb.h | 7 +++++++ > >> 2 files changed, 18 insertions(+) > >> > >>diff --git a/drivers/phy/phy-sun4i-usb.c b/drivers/phy/phy-sun4i-usb.c > >>index 43c0d98..06f4e11a 100644 > >>--- a/drivers/phy/phy-sun4i-usb.c > >>+++ b/drivers/phy/phy-sun4i-usb.c > >>@@ -470,6 +470,17 @@ void sun4i_usb_phy_set_squelch_detect(struct phy *_phy, bool enabled) > >> } > >> EXPORT_SYMBOL_GPL(sun4i_usb_phy_set_squelch_detect); > >> > >>+void sun4i_usb_phy_force_session_end(struct phy *_phy) > >>+{ > >>+ struct sun4i_usb_phy *phy = phy_get_drvdata(_phy); > >>+ struct sun4i_usb_phy_data *data = to_sun4i_usb_phy_data(phy); > >>+ > >>+ data->id_det = -1; > >>+ data->force_session_end = true; > >>+ queue_delayed_work(system_wq, &data->detect, 0); > >>+} > >>+EXPORT_SYMBOL_GPL(sun4i_usb_phy_force_session_end); > > > >Er.. one more export symbol :-( > > Yes unfortunately we need one more to work around sunxi musb / phy bugs. Instead, can you somehow reuse sun4i_usb_phy_set_mode()? Regards, -Bin. > > >>+ > >> static const struct phy_ops sun4i_usb_phy_ops = { > >> .init = sun4i_usb_phy_init, > >> .exit = sun4i_usb_phy_exit, > >>diff --git a/include/linux/phy/phy-sun4i-usb.h b/include/linux/phy/phy-sun4i-usb.h > >>index 50aed92..3bb773f 100644 > >>--- a/include/linux/phy/phy-sun4i-usb.h > >>+++ b/include/linux/phy/phy-sun4i-usb.h > >>@@ -23,4 +23,11 @@ > >> */ > >> void sun4i_usb_phy_set_squelch_detect(struct phy *phy, bool enabled); > >> > >>+/** > >>+ * sun4i_usb_force_session_end() - Force the current session to end > >>+ * by reporting VBus low for 200+ ms > >>+ * @phy: reference to a sun4i usb phy > >>+ */ > >>+void sun4i_usb_phy_force_session_end(struct phy *phy); > > > >Should we include a static inline function if sun4i phy is not defined? > > No, we're also not doing that for the already exported > sun4i_usb_phy_set_squelch_detect() > > And it is not necessary since the only caller is drivers/usb/musb/sunxi.c, > and drivers/usb/musb/Kconfig has: > > config USB_MUSB_SUNXI > tristate "Allwinner (sunxi)" > depends on PHY_SUN4I_USB > > Regards, > > Hans
Hi, On 09/22/2016 05:28 AM, Bin Liu wrote: > Hi, > > On Wed, Sep 21, 2016 at 11:05:33AM +0300, Hans de Goede wrote: >> Hi, >> >> On 09/20/2016 07:45 AM, Kishon Vijay Abraham I wrote: >>> Hi, >>> >>> On Sunday 18 September 2016 10:20 PM, Hans de Goede wrote: >>>> The sunxi musb has a bug where sometimes it will generate a babble >>>> error on device disconnect instead of a disconnect irq. When this >>>> happens the musb-controller switches from host mode to device mode >>>> (it clears MUSB_DEVCTL_SESSION and sets MUSB_DEVCTL_BDEVICE) and >>>> gets stuck in this state. >>>> >>>> Clearing this requires reporting Vbus low for 200 or more ms, but >>>> on some devices Vbus is simply always high (host-only mode, no Vbus >>>> control). The phy-sun4i-usb code already has code to force a session >>>> end for devices without Vbus control. >>>> >>>> This commit adds a sun4i_usb_phy_force_session_end() function exporting >>>> this functionality to the sunxi-musb glue, so that it can force a session >>>> end to fixup the stuck state after a babble error. >>>> >>>> Signed-off-by: Hans de Goede <hdegoede@redhat.com> >>>> --- >>>> drivers/phy/phy-sun4i-usb.c | 11 +++++++++++ >>>> include/linux/phy/phy-sun4i-usb.h | 7 +++++++ >>>> 2 files changed, 18 insertions(+) >>>> >>>> diff --git a/drivers/phy/phy-sun4i-usb.c b/drivers/phy/phy-sun4i-usb.c >>>> index 43c0d98..06f4e11a 100644 >>>> --- a/drivers/phy/phy-sun4i-usb.c >>>> +++ b/drivers/phy/phy-sun4i-usb.c >>>> @@ -470,6 +470,17 @@ void sun4i_usb_phy_set_squelch_detect(struct phy *_phy, bool enabled) >>>> } >>>> EXPORT_SYMBOL_GPL(sun4i_usb_phy_set_squelch_detect); >>>> >>>> +void sun4i_usb_phy_force_session_end(struct phy *_phy) >>>> +{ >>>> + struct sun4i_usb_phy *phy = phy_get_drvdata(_phy); >>>> + struct sun4i_usb_phy_data *data = to_sun4i_usb_phy_data(phy); >>>> + >>>> + data->id_det = -1; >>>> + data->force_session_end = true; >>>> + queue_delayed_work(system_wq, &data->detect, 0); >>>> +} >>>> +EXPORT_SYMBOL_GPL(sun4i_usb_phy_force_session_end); >>> >>> Er.. one more export symbol :-( >> >> Yes unfortunately we need one more to work around sunxi musb / phy bugs. > > Instead, can you somehow reuse sun4i_usb_phy_set_mode()? That was my first thought too, and I looked into this, but it did not work out. Regards, Hans > > Regards, > -Bin. > >> >>>> + >>>> static const struct phy_ops sun4i_usb_phy_ops = { >>>> .init = sun4i_usb_phy_init, >>>> .exit = sun4i_usb_phy_exit, >>>> diff --git a/include/linux/phy/phy-sun4i-usb.h b/include/linux/phy/phy-sun4i-usb.h >>>> index 50aed92..3bb773f 100644 >>>> --- a/include/linux/phy/phy-sun4i-usb.h >>>> +++ b/include/linux/phy/phy-sun4i-usb.h >>>> @@ -23,4 +23,11 @@ >>>> */ >>>> void sun4i_usb_phy_set_squelch_detect(struct phy *phy, bool enabled); >>>> >>>> +/** >>>> + * sun4i_usb_force_session_end() - Force the current session to end >>>> + * by reporting VBus low for 200+ ms >>>> + * @phy: reference to a sun4i usb phy >>>> + */ >>>> +void sun4i_usb_phy_force_session_end(struct phy *phy); >>> >>> Should we include a static inline function if sun4i phy is not defined? >> >> No, we're also not doing that for the already exported >> sun4i_usb_phy_set_squelch_detect() >> >> And it is not necessary since the only caller is drivers/usb/musb/sunxi.c, >> and drivers/usb/musb/Kconfig has: >> >> config USB_MUSB_SUNXI >> tristate "Allwinner (sunxi)" >> depends on PHY_SUN4I_USB >> >> Regards, >> >> Hans
diff --git a/drivers/phy/phy-sun4i-usb.c b/drivers/phy/phy-sun4i-usb.c index 43c0d98..06f4e11a 100644 --- a/drivers/phy/phy-sun4i-usb.c +++ b/drivers/phy/phy-sun4i-usb.c @@ -470,6 +470,17 @@ void sun4i_usb_phy_set_squelch_detect(struct phy *_phy, bool enabled) } EXPORT_SYMBOL_GPL(sun4i_usb_phy_set_squelch_detect); +void sun4i_usb_phy_force_session_end(struct phy *_phy) +{ + struct sun4i_usb_phy *phy = phy_get_drvdata(_phy); + struct sun4i_usb_phy_data *data = to_sun4i_usb_phy_data(phy); + + data->id_det = -1; + data->force_session_end = true; + queue_delayed_work(system_wq, &data->detect, 0); +} +EXPORT_SYMBOL_GPL(sun4i_usb_phy_force_session_end); + static const struct phy_ops sun4i_usb_phy_ops = { .init = sun4i_usb_phy_init, .exit = sun4i_usb_phy_exit, diff --git a/include/linux/phy/phy-sun4i-usb.h b/include/linux/phy/phy-sun4i-usb.h index 50aed92..3bb773f 100644 --- a/include/linux/phy/phy-sun4i-usb.h +++ b/include/linux/phy/phy-sun4i-usb.h @@ -23,4 +23,11 @@ */ void sun4i_usb_phy_set_squelch_detect(struct phy *phy, bool enabled); +/** + * sun4i_usb_force_session_end() - Force the current session to end + * by reporting VBus low for 200+ ms + * @phy: reference to a sun4i usb phy + */ +void sun4i_usb_phy_force_session_end(struct phy *phy); + #endif
The sunxi musb has a bug where sometimes it will generate a babble error on device disconnect instead of a disconnect irq. When this happens the musb-controller switches from host mode to device mode (it clears MUSB_DEVCTL_SESSION and sets MUSB_DEVCTL_BDEVICE) and gets stuck in this state. Clearing this requires reporting Vbus low for 200 or more ms, but on some devices Vbus is simply always high (host-only mode, no Vbus control). The phy-sun4i-usb code already has code to force a session end for devices without Vbus control. This commit adds a sun4i_usb_phy_force_session_end() function exporting this functionality to the sunxi-musb glue, so that it can force a session end to fixup the stuck state after a babble error. Signed-off-by: Hans de Goede <hdegoede@redhat.com> --- drivers/phy/phy-sun4i-usb.c | 11 +++++++++++ include/linux/phy/phy-sun4i-usb.h | 7 +++++++ 2 files changed, 18 insertions(+)