Re: [PATCH v2 0/3] uvc gadget performance issues

From: Dan Vacura
Date: Tue Oct 11 2022 - 16:16:31 EST

Hi Michael,

Appreciate the support!

On Tue, Oct 11, 2022 at 09:48:08PM +0200, Michael Grzeschik wrote:
> Hi Dan!
>
> Thanks for the patches.
>
> On Tue, Oct 11, 2022 at 01:34:32PM -0500, Dan Vacura wrote:
> > Hello uvc gadget developers,
> >
> > Please find my V2 series with added patches to disable these performance
> > features at the userspace level for devices that don't work well with
> > the UDC hw, i.e. dwc3 in this case. Also included are updates to
> > comments for the v1 patch.
> >
> > Original note:
> >
> > I'm working on a 5.15.41 based kernel on a qcom chipset with the dwc3
> > controller and I'm encountering two problems related to the recent performance
> > improvement changes:
> >
> > https://patchwork.kernel.org/project/linux-usb/patch/20210628155311.16762-5-m.grzeschik@xxxxxxxxxxxxxx/ and
> > https://patchwork.kernel.org/project/linux-usb/patch/20210628155311.16762-6-m.grzeschik@xxxxxxxxxxxxxx/
> >
> > If I revert these two changes, then I have much improved stability and a
> > transmission problem I'm seeing is gone. Has there been any success from
> > others on 5.15 with this uvc improvement and any recommendations for my
> > current problems? Those being:
> >
> > 1) a smmu panic, snippet here: 
> >
> > <3>[  718.314900][  T803] arm-smmu 15000000.apps-smmu: Unhandled arm-smmu context fault from a600000.dwc3!
> > <3>[  718.314994][  T803] arm-smmu 15000000.apps-smmu: FAR    = 0x00000000efe60800
> > <3>[  718.315023][  T803] arm-smmu 15000000.apps-smmu: PAR    = 0x0000000000000000
> > <3>[  718.315048][  T803] arm-smmu 15000000.apps-smmu: FSR    = 0x40000402 [TF R SS ]
> > <3>[  718.315074][  T803] arm-smmu 15000000.apps-smmu: FSYNR0    = 0x5f0003
> > <3>[  718.315096][  T803] arm-smmu 15000000.apps-smmu: FSYNR1    = 0xaa02
> > <3>[  718.315117][  T803] arm-smmu 15000000.apps-smmu: context bank#    = 0x1b
> > <3>[  718.315141][  T803] arm-smmu 15000000.apps-smmu: TTBR0  = 0x001b0000c2a92000
> > <3>[  718.315165][  T803] arm-smmu 15000000.apps-smmu: TTBR1  = 0x001b000000000000
> > <3>[  718.315192][  T803] arm-smmu 15000000.apps-smmu: SCTLR  = 0x0a5f00e7 ACTLR  = 0x00000003
> > <3>[  718.315245][  T803] arm-smmu 15000000.apps-smmu: CBAR  = 0x0001f300
> > <3>[  718.315274][  T803] arm-smmu 15000000.apps-smmu: MAIR0   = 0xf404ff44 MAIR1   = 0x0000efe4
> > <3>[  718.315297][  T803] arm-smmu 15000000.apps-smmu: SID = 0x40
> > <3>[  718.315318][  T803] arm-smmu 15000000.apps-smmu: Client info: BID=0x5, PID=0xa, MID=0x2
> > <3>[  718.315377][  T803] arm-smmu 15000000.apps-smmu: soft iova-to-phys=0x0000000000000000
> >
> > I can reduce this panic with the proposed patch, but it still happens until I
> > disable the "req->no_interrupt = 1" logic.
>
> This actually smells very much like an race between hardware and
> software, that is probably working on the same memory. I would guess
> that the hardware in the non interrupt case is currently processing
> queued memory, while at the same time the software stack will update
> that same memory with new data.
>
> In my opinion this should be fixed, rather then making the interrupt
> load optional. Also we could discuss if an option to adjust the load
> adds some extra value, but out of this issue scope you describe here.

Yes, I agree this should be fixed and it could just be an issue with the
dwc3 driver. I'd need help on this as an understanding of what the hw is
doing is needed to timely rootcause the issue.

As for giving the option to reduce the load to userspace, I thought it'd
be helpful for others since it seems this is the only gadget driver that
uses this option and the amount that is hardcoded may be desired to be
adjusted, like in my case to always interrupt. Although, if the bug
didn't exist, I'd be fine with the existing hardcode.

As for reproducing, did you try this with the sg functionality disabled?
The crash seems to occur sooner in that setup.

>
> Also, is this issue also being more likely to happen when streaming YUYV?
>
> > 2) The frame is not fully transmitted in dwc3 with sg support enabled.
> >
> > There seems to be a mapping limit I'm seeing where only the roughly first
> > 70% of the total frame is sent. Interestingly, if I allocate a larger
> > size for the buffer upfront, in uvc_queue_setup(), like sizes[0] =
> > video->imagesize * 3. Then the issue rarely happens. For example, when I
> > do YUYV I see green, uninitialized data, at the bottom part of the
> > frame. If I do MJPG with smaller filled sizes, the transmission is fine.
> >
> > +-------------------------+
> > | |
> > | |
> > | |
> > | Good data |
> > | |
> > | |
> > | |
> > +-------------------------+
> > |xxxxxxxxxxxxxxxxxxxxxxxxx|
> > |xxxx Bad data xxxxxxxxx|
> > |xxxxxxxxxxxxxxxxxxxxxxxxx|
> > +-------------------------+
> >
>
> I did not stream with YUYV for some time. I will do that and try to
> reproduce the issues you describe.
>
> I also have an patch in the queue that will limit the sg support for
> devices with speed > HIGH_SPEED. Because of the overhead of the limited
> transfer payload of 1024*3 Bytes, it is possible that a simple memcpy
> will actually be fast enough. But for that patch I still have to make
> proper measurements. Btw. which USB speed are you transferring with?

This happens with both super speed and high speed setups. Another
developer also mentioned problems with sg support and had to disable it
https://lore.kernel.org/linux-usb/0de496bf-fbc6-a2a1-a967-9a0580a7b1eb@xxxxxxxxxxxxxxxx/T/#mf06aa48ec53ed4c174601f03f397bd296f09efe5
but didn't have problems with dwc2.

>
> Regards,
> Michael
>
> --
> Pengutronix e.K. | |
> Steuerwalder Str. 21 | http://www.pengutronix.de/ |
> 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
> Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |