[PATCH] usb: dwc3: gadget: Avoid duplicate requests to enable Run/Stop

[PATCH] usb: dwc3: gadget: Avoid duplicate requests to enable Run/Stop

[Wesley Cheng]

Relocate the pullups_connected check until after it is ensured that there
are no runtime PM transitions.  If another context triggered the DWC3
core's runtime resume, it may have already enabled the Run/Stop.  Do not
re-run the entire pullup sequence again, as it may issue a core soft
reset while Run/Stop is already set.

Signed-off-by: Wesley Cheng <quic_wcheng@quicinc.com>
---
 drivers/usb/dwc3/gadget.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c
index aeeec751c53c..67b1039c664c 100644
--- a/drivers/usb/dwc3/gadget.c
+++ b/drivers/usb/dwc3/gadget.c
@@ -2539,9 +2539,6 @@ static int dwc3_gadget_pullup(struct usb_gadget *g, int is_on)
 
 	is_on = !!is_on;
 
-	if (dwc->pullups_connected == is_on)
-		return 0;
-
 	dwc->softconnect = is_on;
 
 	/*
@@ -2566,6 +2563,9 @@ static int dwc3_gadget_pullup(struct usb_gadget *g, int is_on)
 		return 0;
 	}
 
+	if (dwc->pullups_connected == is_on)
+		return 0;
+
 	if (!is_on) {
 		ret = dwc3_gadget_soft_disconnect(dwc);
 	} else {

Re: [PATCH] usb: dwc3: gadget: Avoid duplicate requests to enable Run/Stop

[Greg KH]

On Mon, Jul 25, 2022 at 01:11:36PM -0700, Wesley Cheng wrote:
> Relocate the pullups_connected check until after it is ensured that there
> are no runtime PM transitions.  If another context triggered the DWC3
> core's runtime resume, it may have already enabled the Run/Stop.  Do not
> re-run the entire pullup sequence again, as it may issue a core soft
> reset while Run/Stop is already set.
> 
> Signed-off-by: Wesley Cheng <quic_wcheng@quicinc.com>

What commit does this fix?

How far back should it go to stable kernels?

thanks,

greg k-h

Re: [PATCH] usb: dwc3: gadget: Avoid duplicate requests to enable Run/Stop

[Wesley Cheng]

Hi Greg,

On 7/26/2022 8:43 AM, Greg KH wrote:
> On Mon, Jul 25, 2022 at 01:11:36PM -0700, Wesley Cheng wrote:
>> Relocate the pullups_connected check until after it is ensured that there
>> are no runtime PM transitions.  If another context triggered the DWC3
>> core's runtime resume, it may have already enabled the Run/Stop.  Do not
>> re-run the entire pullup sequence again, as it may issue a core soft
>> reset while Run/Stop is already set.
>>
>> Signed-off-by: Wesley Cheng <quic_wcheng@quicinc.com>
> 
> What commit does this fix?
> 
> How far back should it go to stable kernels?
> 
Let met add a fixes tag and resubmit.  Thanks!

Thanks
Wesley Cheng