When the gadget driver hasn't been (yet) configured, and the cable is
connected to a HOST, the SFTDISCON gets cleared unconditionally, so the
HOST tries to enumerate it.
At the host side, this can result in a stuck USB port or worse. When
getting lucky, some dmesg can be observed at the host side:
new high-speed USB device number ...
device descriptor read/64, error -110
Fix it in drd, by checking the enabled flag before calling
dwc2_hsotg_core_connect(). It will be called later, once configured,
by the normal flow:
- udc_bind_to_driver
- usb_gadget_connect
- dwc2_hsotg_pullup
- dwc2_hsotg_core_connect
Fixes: 17f934024e84 ("usb: dwc2: override PHY input signals with usb role switch support")
Signed-off-by: Fabrice Gasnier <[email protected]>
---
Changes in v2:
- Fix build error: 'struct dwc2_hsotg' has no member named 'enabled';
as reported by the kernel test robot.
https://lore.kernel.org/all/[email protected]/
Add dwc2_is_device_enabled() macro to handle this.
---
drivers/usb/dwc2/core.h | 2 ++
drivers/usb/dwc2/drd.c | 6 ++++--
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/drivers/usb/dwc2/core.h b/drivers/usb/dwc2/core.h
index 8a63da3..8a7751b 100644
--- a/drivers/usb/dwc2/core.h
+++ b/drivers/usb/dwc2/core.h
@@ -1418,6 +1418,7 @@ void dwc2_hsotg_core_connect(struct dwc2_hsotg *hsotg);
void dwc2_hsotg_disconnect(struct dwc2_hsotg *dwc2);
int dwc2_hsotg_set_test_mode(struct dwc2_hsotg *hsotg, int testmode);
#define dwc2_is_device_connected(hsotg) (hsotg->connected)
+#define dwc2_is_device_enabled(hsotg) ((hsotg)->enabled)
int dwc2_backup_device_registers(struct dwc2_hsotg *hsotg);
int dwc2_restore_device_registers(struct dwc2_hsotg *hsotg, int remote_wakeup);
int dwc2_gadget_enter_hibernation(struct dwc2_hsotg *hsotg);
@@ -1454,6 +1455,7 @@ static inline int dwc2_hsotg_set_test_mode(struct dwc2_hsotg *hsotg,
int testmode)
{ return 0; }
#define dwc2_is_device_connected(hsotg) (0)
+#define dwc2_is_device_enabled(hsotg) (0)
static inline int dwc2_backup_device_registers(struct dwc2_hsotg *hsotg)
{ return 0; }
static inline int dwc2_restore_device_registers(struct dwc2_hsotg *hsotg,
diff --git a/drivers/usb/dwc2/drd.c b/drivers/usb/dwc2/drd.c
index 1b39c47..d8d6493 100644
--- a/drivers/usb/dwc2/drd.c
+++ b/drivers/usb/dwc2/drd.c
@@ -130,8 +130,10 @@ static int dwc2_drd_role_sw_set(struct usb_role_switch *sw, enum usb_role role)
already = dwc2_ovr_avalid(hsotg, true);
} else if (role == USB_ROLE_DEVICE) {
already = dwc2_ovr_bvalid(hsotg, true);
- /* This clear DCTL.SFTDISCON bit */
- dwc2_hsotg_core_connect(hsotg);
+ if (dwc2_is_device_enabled(hsotg)) {
+ /* This clear DCTL.SFTDISCON bit */
+ dwc2_hsotg_core_connect(hsotg);
+ }
} else {
if (dwc2_is_device_mode(hsotg)) {
if (!dwc2_ovr_bvalid(hsotg, false))
--
2.7.4
On Tue, Feb 15, 2022 at 12:04:19PM +0100, Fabrice Gasnier wrote:
> When the gadget driver hasn't been (yet) configured, and the cable is
> connected to a HOST, the SFTDISCON gets cleared unconditionally, so the
> HOST tries to enumerate it.
> At the host side, this can result in a stuck USB port or worse. When
> getting lucky, some dmesg can be observed at the host side:
> new high-speed USB device number ...
> device descriptor read/64, error -110
>
> Fix it in drd, by checking the enabled flag before calling
> dwc2_hsotg_core_connect(). It will be called later, once configured,
> by the normal flow:
> - udc_bind_to_driver
> - usb_gadget_connect
> - dwc2_hsotg_pullup
> - dwc2_hsotg_core_connect
>
> Fixes: 17f934024e84 ("usb: dwc2: override PHY input signals with usb role switch support")
> Signed-off-by: Fabrice Gasnier <[email protected]>
> ---
> Changes in v2:
> - Fix build error: 'struct dwc2_hsotg' has no member named 'enabled';
> as reported by the kernel test robot.
> https://lore.kernel.org/all/[email protected]/
> Add dwc2_is_device_enabled() macro to handle this.
> ---
> drivers/usb/dwc2/core.h | 2 ++
> drivers/usb/dwc2/drd.c | 6 ++++--
> 2 files changed, 6 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/usb/dwc2/core.h b/drivers/usb/dwc2/core.h
> index 8a63da3..8a7751b 100644
> --- a/drivers/usb/dwc2/core.h
> +++ b/drivers/usb/dwc2/core.h
> @@ -1418,6 +1418,7 @@ void dwc2_hsotg_core_connect(struct dwc2_hsotg *hsotg);
> void dwc2_hsotg_disconnect(struct dwc2_hsotg *dwc2);
> int dwc2_hsotg_set_test_mode(struct dwc2_hsotg *hsotg, int testmode);
> #define dwc2_is_device_connected(hsotg) (hsotg->connected)
> +#define dwc2_is_device_enabled(hsotg) ((hsotg)->enabled)
Why the extra ()? dwc2_is_device_connected does not have it, so this
one probably should not either, right?
thanks,
greg k-h
On Tue, Feb 15, 2022 at 04:42:46PM +0100, Fabrice Gasnier wrote:
> On 2/15/22 3:04 PM, Greg KH wrote:
> > On Tue, Feb 15, 2022 at 12:04:19PM +0100, Fabrice Gasnier wrote:
> >> When the gadget driver hasn't been (yet) configured, and the cable is
> >> connected to a HOST, the SFTDISCON gets cleared unconditionally, so the
> >> HOST tries to enumerate it.
> >> At the host side, this can result in a stuck USB port or worse. When
> >> getting lucky, some dmesg can be observed at the host side:
> >> new high-speed USB device number ...
> >> device descriptor read/64, error -110
> >>
> >> Fix it in drd, by checking the enabled flag before calling
> >> dwc2_hsotg_core_connect(). It will be called later, once configured,
> >> by the normal flow:
> >> - udc_bind_to_driver
> >> - usb_gadget_connect
> >> - dwc2_hsotg_pullup
> >> - dwc2_hsotg_core_connect
> >>
> >> Fixes: 17f934024e84 ("usb: dwc2: override PHY input signals with usb role switch support")
> >> Signed-off-by: Fabrice Gasnier <[email protected]>
> >> ---
> >> Changes in v2:
> >> - Fix build error: 'struct dwc2_hsotg' has no member named 'enabled';
> >> as reported by the kernel test robot.
> >> https://lore.kernel.org/all/[email protected]/
> >> Add dwc2_is_device_enabled() macro to handle this.
> >> ---
> >> drivers/usb/dwc2/core.h | 2 ++
> >> drivers/usb/dwc2/drd.c | 6 ++++--
> >> 2 files changed, 6 insertions(+), 2 deletions(-)
> >>
> >> diff --git a/drivers/usb/dwc2/core.h b/drivers/usb/dwc2/core.h
> >> index 8a63da3..8a7751b 100644
> >> --- a/drivers/usb/dwc2/core.h
> >> +++ b/drivers/usb/dwc2/core.h
> >> @@ -1418,6 +1418,7 @@ void dwc2_hsotg_core_connect(struct dwc2_hsotg *hsotg);
> >> void dwc2_hsotg_disconnect(struct dwc2_hsotg *dwc2);
> >> int dwc2_hsotg_set_test_mode(struct dwc2_hsotg *hsotg, int testmode);
> >> #define dwc2_is_device_connected(hsotg) (hsotg->connected)
> >> +#define dwc2_is_device_enabled(hsotg) ((hsotg)->enabled)
> >
> > Why the extra ()? dwc2_is_device_connected does not have it, so this
> > one probably should not either, right?
>
> Hi Greg,
>
> I was wondering the same, checkpatch complains without it:
>
> CHECK: Macro argument 'hsotg' may be better as '(hsotg)' to avoid
> precedence issues
checkpatch is wrong here, this is a structure pointer, not anything you
could ever use that could be evaluated any other way.
> I can remove the extra () in a v3 if you wish ?
Please do.
thanks,
greg k-h
On 2/15/22 3:04 PM, Greg KH wrote:
> On Tue, Feb 15, 2022 at 12:04:19PM +0100, Fabrice Gasnier wrote:
>> When the gadget driver hasn't been (yet) configured, and the cable is
>> connected to a HOST, the SFTDISCON gets cleared unconditionally, so the
>> HOST tries to enumerate it.
>> At the host side, this can result in a stuck USB port or worse. When
>> getting lucky, some dmesg can be observed at the host side:
>> new high-speed USB device number ...
>> device descriptor read/64, error -110
>>
>> Fix it in drd, by checking the enabled flag before calling
>> dwc2_hsotg_core_connect(). It will be called later, once configured,
>> by the normal flow:
>> - udc_bind_to_driver
>> - usb_gadget_connect
>> - dwc2_hsotg_pullup
>> - dwc2_hsotg_core_connect
>>
>> Fixes: 17f934024e84 ("usb: dwc2: override PHY input signals with usb role switch support")
>> Signed-off-by: Fabrice Gasnier <[email protected]>
>> ---
>> Changes in v2:
>> - Fix build error: 'struct dwc2_hsotg' has no member named 'enabled';
>> as reported by the kernel test robot.
>> https://lore.kernel.org/all/[email protected]/
>> Add dwc2_is_device_enabled() macro to handle this.
>> ---
>> drivers/usb/dwc2/core.h | 2 ++
>> drivers/usb/dwc2/drd.c | 6 ++++--
>> 2 files changed, 6 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/usb/dwc2/core.h b/drivers/usb/dwc2/core.h
>> index 8a63da3..8a7751b 100644
>> --- a/drivers/usb/dwc2/core.h
>> +++ b/drivers/usb/dwc2/core.h
>> @@ -1418,6 +1418,7 @@ void dwc2_hsotg_core_connect(struct dwc2_hsotg *hsotg);
>> void dwc2_hsotg_disconnect(struct dwc2_hsotg *dwc2);
>> int dwc2_hsotg_set_test_mode(struct dwc2_hsotg *hsotg, int testmode);
>> #define dwc2_is_device_connected(hsotg) (hsotg->connected)
>> +#define dwc2_is_device_enabled(hsotg) ((hsotg)->enabled)
>
> Why the extra ()? dwc2_is_device_connected does not have it, so this
> one probably should not either, right?
Hi Greg,
I was wondering the same, checkpatch complains without it:
CHECK: Macro argument 'hsotg' may be better as '(hsotg)' to avoid
precedence issues
I can remove the extra () in a v3 if you wish ?
Thanks for reviewing,
Best Regards,
Fabrice
>
> thanks,
>
> greg k-h
>