[U-Boot] [PATCH v2 02/10] usb: omap5: Add glue logic to support DM for USB host

[Simon Glass]

Hi Jean-Jacques,

On 14 March 2018 at 10:18, Jean-Jacques Hiblot <jjhiblot at ti.com> wrote:
> The omap5 uses the dwc3. The dwc3 supports the driver model but it requires
> some glue logic to load the the driver.
>
> Signed-off-by: Jean-Jacques Hiblot <jjhiblot at ti.com>
> ---
>
> Changes in v2: None
>
>  drivers/usb/host/Kconfig          | 10 +++++++++
>  drivers/usb/host/Makefile         |  1 +
>  drivers/usb/host/dwc3-omap-glue.c | 47 +++++++++++++++++++++++++++++++++++++++
>  3 files changed, 58 insertions(+)
>  create mode 100644 drivers/usb/host/dwc3-omap-glue.c
>
> diff --git a/drivers/usb/host/Kconfig b/drivers/usb/host/Kconfig
> index 90b2f78..405ab5f 100644
> --- a/drivers/usb/host/Kconfig
> +++ b/drivers/usb/host/Kconfig
> @@ -79,6 +79,16 @@ config USB_XHCI_DRA7XX_INDEX
>           Select the DRA7XX xHCI USB index.
>           Current supported values: 0, 1.
>
> +config USB_DM_XHCI_OMAP
> +       bool "Support for OMAP family on-chip xHCI USB controller (DM version)"
> +       depends on DM_USB
> +       depends on ARCH_OMAP2PLUS
> +       default y if DRA7XX
> +       help
> +         Enables support for the on-chip xHCI controller on TI OMAP family SoCs
> +         using the Driver Model.
> +         This driver provides the glue logic to probe the generic dwc3 driver.
> +
>  config USB_XHCI_FSL
>         bool "Support for NXP Layerscape on-chip xHCI USB controller"
>         default y if ARCH_LS1021A || FSL_LSCH3 || FSL_LSCH2
> diff --git a/drivers/usb/host/Makefile b/drivers/usb/host/Makefile
> index 7f9ba24..416d0eb 100644
> --- a/drivers/usb/host/Makefile
> +++ b/drivers/usb/host/Makefile
> @@ -60,6 +60,7 @@ obj-$(CONFIG_USB_XHCI_OMAP) += xhci-omap.o
>  obj-$(CONFIG_USB_XHCI_PCI) += xhci-pci.o
>  obj-$(CONFIG_USB_XHCI_RCAR) += xhci-rcar.o
>  obj-$(CONFIG_USB_XHCI_STI) += dwc3-sti-glue.o
> +obj-$(CONFIG_USB_DM_XHCI_OMAP) += dwc3-omap-glue.o
>
>  # designware
>  obj-$(CONFIG_USB_DWC2) += dwc2.o
> diff --git a/drivers/usb/host/dwc3-omap-glue.c b/drivers/usb/host/dwc3-omap-glue.c
> new file mode 100644
> index 0000000..1110fb6
> --- /dev/null
> +++ b/drivers/usb/host/dwc3-omap-glue.c
> @@ -0,0 +1,47 @@
> +/*
> + * OMAP5 family DWC3 specific Glue layer
> + *
> + * Copyright (c) 2017
> + * Jean-Jacques Hiblot <jjhiblot at ti.com>
> + * based on dwc3-sti-glue
> + *
> + * SPDX-License-Identifier:    GPL-2.0+
> + */
> +
> +#include <common.h>
> +#include <dm.h>
> +#include <errno.h>
> +
> +DECLARE_GLOBAL_DATA_PTR;
> +
> +static int omap5_dwc3_glue_bind(struct udevice *dev)
> +{
> +       int dwc3_node;
> +
> +       /* check if one subnode is present */
> +       dwc3_node = fdt_first_subnode(gd->fdt_blob, dev_of_offset(dev));

Can this use the livetree interface? E.g. here it is dev_read_first_subnode()

> +       if (dwc3_node <= 0) {
> +               printf("Can't find subnode for %s\n", dev->name);
> +               return -ENODEV;
> +       }
> +       /* check if the subnode compatible string is the dwc3 one*/
> +       if (fdt_node_check_compatible(gd->fdt_blob, dwc3_node,
> +                                     "snps,dwc3") != 0) {

There is no

> +               printf("Can't find dwc3 subnode for %s\n", dev->name);

debug()

> +               return -ENODEV;

Can you use -ENOENT or something else here? At present -ENODEV means
there is no device, which in fact we have a device (@dev). This can be
confusing since there are places in DM core where -ENODEV signals that
there was no device, and not to worry about it.

> +       }
> +
> +       return dm_scan_fdt_dev(dev);
> +}
> +
> +static const struct udevice_id omap5_dwc3_glue_ids[] = {
> +       { .compatible = "ti,dwc3" },
> +       { }
> +};
> +
> +U_BOOT_DRIVER(dwc3_omap5_glue) = {
> +       .name = "dwc3_omap5_glue",
> +       .id = UCLASS_MISC,
> +       .of_match = omap5_dwc3_glue_ids,
> +       .bind = omap5_dwc3_glue_bind,
> +};
> --
> 2.7.4
>

Regards,
Simon