[U-Boot] [PATCH V3 2/2] arm: mx5: Add support for DENX M53EVK

[Benoît Thébaudeau]

Dear Marek Vasut,

On Friday, May 3, 2013 9:56:40 PM, Marek Vasut wrote:
> Dear Benoît Thébaudeau,
> 
> > Dear Marek Vasut,
> > 
> > On Friday, May 3, 2013 4:59:13 PM, Benoît Thébaudeau wrote:
> > > Dear Marek Vasut,
> > > 
> > > On Thursday, April 25, 2013 10:16:03 PM, Marek Vasut wrote:
> > > > Add basic support for the DENX M53EVK board. Currently supported is:
> > > > MMC (incl. booting)
> > > > NAND (incl. booting)
> > > > Ethernet, I2C, USB, SATA, RTC.
> > > > 
> > > > Signed-off-by: Marek Vasut <marex at denx.de>
> > > 
> > > [...]
> > > 
> > > > diff --git a/board/denx/m53evk/m53evk.c b/board/denx/m53evk/m53evk.c
> > > > new file mode 100644
> > > > index 0000000..3289f28
> > > > --- /dev/null
> > > > +++ b/board/denx/m53evk/m53evk.c
> > > > @@ -0,0 +1,408 @@
> > > 
> > > [...]
> > > 
> > > > +#ifdef CONFIG_USB_EHCI_MX5
> > > > +int board_ehci_hcd_init(int port)
> > > > +{
> > > > +	if (port == 0) {
> > > > +		/* USB OTG PWRON */
> > > > +		mxc_request_iomux(MX53_PIN_GPIO_4, IOMUX_CONFIG_ALT1);
> > > > +		mxc_iomux_set_pad(MX53_PIN_GPIO_4,
> > > > +				PAD_CTL_PKE_ENABLE |
> > > > +				PAD_CTL_100K_PD |
> > > > +				PAD_CTL_DRV_HIGH
> > > > +				);
> > > > +		gpio_direction_output(IOMUX_TO_GPIO(MX53_PIN_GPIO_4), 0);
> > > > +
> > > > +		/* USB OTG Over Current */
> > > > +		mxc_request_iomux(MX53_PIN_GPIO_18, IOMUX_CONFIG_ALT1);
> > > > +		mxc_iomux_set_input(MX53_USBOH3_IPP_IND_OTG_OC_SELECT_INPUT, 1);
> > > 
> > > It looks like there is something wrong here: The last line connects
> > > OTG.OC to EIM_D21.ALT6, but the line before sets GPIO_18 as GPIO7[13]
> > > instead of setting
> > > EIM_D21 as OTG.OC.
> > 
> > According to http://www.denx-cs.de/doku/?q=m53pinout , USB_OTG_OC is on
> > ball D7 (i.e. GPIO_18), and ball V3 (i.e. EIM_D21) is not on the SoM
> > connector.
> > 
> > So can you clarify what's going on with this signal? Is it handled by
> > software as a GPIO, and EIM_D21 is just tied high or low on the SoM to
> > make the hardware controller not complain about a false OC detection?
> 
> Yes, the OTG OC is handled by GPIO.

So should "mxc_iomux_set_input(MX53_USBOH3_IPP_IND_OTG_OC_SELECT_INPUT, 1);" be
dropped or not?

Best regards,
Benoît