Re: USB lockups on BeagleBone/AM335x

[Felipe Balbi <balbi@ti.com>]

[-- Attachment #1: Type: text/plain, Size: 3525 bytes --]

On Fri, Feb 28, 2014 at 09:34:27AM +1300, Chris Kimber wrote:
> Hi,
> 
> On 22/02/2014 4:33 a.m., Felipe Balbi wrote:
> >On Fri, Feb 21, 2014 at 03:27:09PM +1300, Chris Kimber wrote:
> >>Hey
> >>
> >>On 21/02/2014 2:14 p.m., Felipe Balbi wrote:
> >>>Hi,
> >>>
> >>>(please avoid top-posting)
> >>>
> >>>On Fri, Feb 21, 2014 at 12:11:40AM +0000, Chris Kimber wrote:
> >>>>Hey,
> >>>>
> >>>>Thanks for the response.
> >>>>
> >>>>I've disabled the DMA (CONFIG_MUSB_PIO_ONLY=y) but the problem still
> >>>>persists (for both USB sticks & USB serial ports).
> >>>hmm, that's weird.
> >>>
> >>>>Now it looks like dsps_interrupt() never fires and causes the hang
> >>>>up...
> >>>>
> >>>>[   94.865635] tty ttyUSB0: serial_write - 11 byte(s)
> >>>>[   94.865656] cp210x ttyUSB0: usb_serial_generic_write_start - length = 11, data = 54 45 53 54 49 4e 47 20 34 32 0a
> >>>>[   94.865680] musb-hdrc musb-hdrc.1.auto: qh ce461a00 periodic slot 10
> >>>>[   94.865700] musb-hdrc musb-hdrc.1.auto: qh ce461a00 urb ce481e80 dev2 ep1out-bulk, hw_ep 10, ce43db00/11
> >>>>[   94.865721] musb-hdrc musb-hdrc.1.auto: --> hw10 urb ce481e80 spd2 dev2 ep1out h_addr00 h_port00 bytes 11
> >>>>[   94.865740] musb-hdrc musb-hdrc.1.auto: TX ep10 fifo d0832c48 count 11 buf ce43db00
> >>>>[   94.865755] musb-hdrc musb-hdrc.1.auto: Start TX10 pio
> >>>>[   94.865792] musb-hdrc musb-hdrc.1.auto: usbintr (0) epintr(400)
> >>>>[   94.865810] musb-hdrc musb-hdrc.1.auto: ** IRQ host usb0000 tx0400 rx0000
> >>>>[   94.865826] musb-hdrc musb-hdrc.1.auto: OUT/TX10 end, csr 2100
> >>>>[   94.865866] musb-hdrc musb-hdrc.1.auto: complete ce481e80 usb_serial_generic_write_bulk_callback+0x0/0xd4 [usbserial] (0), dev2 ep1out, 11/11
> >>>so the first one completed just fine.
> >>>
> >>>>[   94.865971] tty ttyUSB0: serial_write - 11 byte(s)
> >>>>[   94.865991] cp210x ttyUSB0: usb_serial_generic_write_start - length = 11, data = 54 45 53 54 49 4e 47 20 34 33 0a
> >>>>[   94.866015] musb-hdrc musb-hdrc.1.auto: qh ce461a00 periodic slot 10
> >>>>[   94.866035] musb-hdrc musb-hdrc.1.auto: qh ce461a00 urb ce481e80 dev2 ep1out-bulk, hw_ep 10, ce43db00/11
> >>>>[   94.866055] musb-hdrc musb-hdrc.1.auto: --> hw10 urb ce481e80 spd2 dev2 ep1out h_addr00 h_port00 bytes 11
> >>>>[   94.866075] musb-hdrc musb-hdrc.1.auto: TX ep10 fifo d0832c48 count 11 buf ce43db00
> >>>>[   94.866089] musb-hdrc musb-hdrc.1.auto: Start TX10 pio
> >>>can you dump TXCSR, DEVCTL and IRQ registers when this happens ?
> >>>
> >>I added the IRQ registers to the debugfs interface..
> >>
> >>IntrTX      : 0000
> >>IntrRX      : 0000
> >>IntrTXE     : ffff
> >>IntrRXE     : fffe
> >>IntrUSB     : 08
> >>IntrUSBE    : f7
> >>TxCSRp      : 0000
> >>DevCtl      : 5d
> >>
> >>Oddly the VBUSERR bit is set, if I disconnect the USB cable then it
> >>goes to 0. I just tried plugging in via powered hub and that doesn't
> >>change the status.
> >that might be the issue you're having, then. for whatever reason VBUS
> >level fluctuates and that confuses MUSB.
> >
> >I'll see if there's anyway to work around that.
> >
> 
> I've finally managed to find some time to check the old 3.2 kernel's
> Interrupt registers and have found that the VBUSERR bit is still set,
> so this might be a red herring.
> 
> I've ordered a BeagleBone Black just to make sure this isn't just
> related to the AM335x Silicon v1 problems you mentioned before.
> 
> The search continues.

alright, keep us posted ;-)

-- 
balbi

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]