> Problem #1.- Continuous reset of the USB controller
> To check if the hardware is fine, I've run the BSP FSL-2.6.31_09.12.01
> from Freescale and I can load any gadget driver with positive results.
>
> The problem appears on new kernels when fsl_udc_core is compiled with
> VERBOSE and DEBUG defined. I've tried 3.2.2, 3.3.2 and 3.4-rc3 kernels,
> all of them with the same results; if I load -any- gadget driver and
> plug the USB cord into my Linux Host, I get the following error
> messages, continuously, non stopping:
> ------------------------------------------------------------------------
> --- [fsl_udc_irq] reset int
> [reset_irq] Controller reset
> [dr_controller_setup] vir[qh_base] is ffdf9000 phy[qh_base] is
> 0x838c4000 reg is 0x838c4000
> [fsl_udc_irq] reset int
> [reset_irq] Controller reset
> [dr_controller_setup] vir[qh_base] is ffdf9000 phy[qh_base] is
> 0x838c4000 reg is 0x838c4000
> ..........
> ..........
> ------------------------------------------------------------------------
> ---
> If fsl_udc_core is compiled with VERBOSE -OR- DEBUG undefined, I can
> load any gadget driver with no "reset" problem.
>
PORTSCX_PORT_RESET is cleared by hardware automatically when the reset
is completed, if you have inserted too many debug message after reset
occurs, the PORTSCX_PORT_RESET will be 0.
The code at upstream may have some problems, you may not need below code:
1833 } else {
1834 VDBG("Controller reset");
1835 /* initialize usb hw reg except for regs for EP, not
1836 * touch usbintr reg */
1837 dr_controller_setup(udc);
1838
1839 /* Reset all internal used Queues */
1840 reset_queues(udc);
1841
1842 ep0_setup(udc);
1843
1844 /* Enable DR IRQ reg, Set Run bit, change udc state */
1845 dr_controller_run(udc);
1846 udc->usb_state = USB_STATE_ATTACHED;
1847 }
>
>
> Problem #2.- g_webcam enumeration
> The enumeration starts as soon as the USB cord is plugged into my Linux
> Host, but with no user space support (the device /dev/video0 hasn't been
> opened).
>
> With the USB cord unplugged I've traced the call chain when g_webcam is
> modprobed down to "fsl_pullup". This function is only called once and,
> as expected, "is_on" parameter is 0 so the pullup should disabled, but
> when I plug the USB cord ..... enumeration starts :(
>
Some logic about controlling controller run/stop is incomplete at current
i.mx framework (lack of vbus interrupt to trigger fsl_vbus_session)
Temp solution for let your thing work (only for g_webcam):
1. Does not set USB_CMD_RUN_STOP at dr_controller_run
2. Change fsl_pullup:
1234 if (can_pullup(udc))
-->
1234 if (is_on)
> Any advice is really appreciated !!!
>
> TIA for your help,
> Gonzalo.
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-usb" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
