Re: [PATCH v2 3/3] usb: dwc3: gadget: return error if command sent to DEPCMD register fails

[Felipe Balbi]

Hi,

On Mon, Jul 13, 2015 at 05:50:49PM +0000, John Youn wrote:
> On 7/11/2015 12:29 PM, Felipe Balbi wrote:
> > Hi,
> > 
> > On Sat, Jul 11, 2015 at 05:17:32PM +0000, Subbaraya Sundeep Bhatta wrote:
> >>>>>> Hi Felipe,
> >>>>>>
> >>>>>> Just an update on this.
> >>>>>>
> >>>>>> I'm trying to get this working with our latest IP with dwc3 from
> >>>>>> your testing/next branch. It fails the usbtest with a problem
> >>>>>> unrelated to this patch.
> >>>>>> .
> >>>>>> It passes on 4.1.1.
> >>>>>>
> >>>>>> I'll have to look into the failure but I won't get to it until next
> >>>>>> week as I'm off the rest of this week.
> >>>>>
> >>>>> interesting... If you could post failure signature, I can help
> >>>>> looking at it, but I guess it's too late to ask :-)
> >>>>>
> >>>>> thanks for helping though
> >>>>>
> >>>>
> >>>>
> >>>> Hi Felipe,
> >>>>
> >>>> Nevermind about my issue, it ended up being a setup-related problem.
> >>>>
> >>>> I actually do see the same error as you due to this series of patches.
> >>>> Except I see it happening before even the first iteration. I get a
> >>>> completion status of 1 for the Set Endpoint Transfer Resources
> >>>> command. I'm not sure why this is.
> >>>>
> >>>> I don't see any conflict with any previous Transfer Complete.
> >>
> >> Same behavior at my end too. Fails before first iteration and I get
> >> completion status of 1 for Set Endpoint Resource command. Attached the
> >> logs of testing done with this patch and without this patch.
> >> Without this patch I often see completion status of 1 for Set Endpoint
> >> Transfer Resources command for Bulk and Isoc endpoints but test
> >> proceeds because driver just logs command completion status and moves
> >> on. We can revert this patch for time being. IP version is 2.90a.
> > 
> > yeah, that's what I mean, it really seems like it's the IP misbehaving.
> > 
> > John, let's try to figure out what's the root cause of this, we really
> > want to use command completion status at some point, but for now we need
> > to revert the patch :-(
> > 
> > Let me know if you want me to log STARS ticket on your solvnet system.
> > 
> > cheers
> > 
> 
> Hi Felipe,
> 
> We found the issue last week. 
> 
> The start config command isn't getting called during SET_INTERFACE.
> Thus the transfer resource index isn't getting reset, and with
> multiple SET_INTERFACE commands it will eventually exhaust the
> resources.
> 
> I tried out a fix and it works for me. I'll send it out separately for
> review.

Thanks a lot John. Not sure how come we missed that for such a long time
:-) Let's Cc stable and get it plugged ASAP :-)

> Also, I noticed that the trace message that shows control transfers
> doesn't show the SET_INTERFACE properly. Any idea why this is?
> 
> For example, here is the line in the trace that corresponds to the
> SET_INTERFACE:
> irq/33-dwc3-10808 [003] d...  2443.494368: dwc3_ctrl_req: bRequestType 01 bRequest 0b wValue 0001 wIndex 0000 wLength 0

I'll have a look at this when I'm back in the office (Jul 18th).

-- 
balbi

Attachment: signature.asc
Description: Digital signature