Re: [PATCH V5 09/12] usb: xhci: Add NVIDIA Tegra xHCI host-controller driver

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 




Hi,

On Mon, Nov 17, 2014 at 03:51:29PM -0800, Andrew Bresticker wrote:
> >> +static const struct {
> >> +     const char *name;
> >> +     unsigned int num;
> >> +} tegra_xhci_phy_types[] = {
> >> +     {
> >> +             .name = "usb3",
> >> +             .num = TEGRA_XUSB_USB3_PHYS,
> >> +     }, {
> >> +             .name = "utmi",
> >> +             .num = TEGRA_XUSB_UTMI_PHYS,
> >> +     }, {
> >> +             .name = "hsic",
> >> +             .num = TEGRA_XUSB_HSIC_PHYS,
> >> +     },
> >
> > these should be using the generic PHY subsystem (drivers/phy).
> 
> This driver does use the generic PHY subsystem, see patch 7 in this series :).

true, I got misled by the fact that you added your phy provider to your
pinctrl. Odd.

> >> +static int tegra_xhci_load_firmware(struct tegra_xhci_hcd *tegra)
> >> +{
> >> +     struct device *dev = tegra->dev;
> >> +     struct tegra_xhci_fw_cfgtbl *cfg_tbl;
> >> +     struct tm fw_tm;
> >> +     u32 val, code_tag_blocks, code_size_blocks;
> >> +     u64 fw_base;
> >> +     time_t fw_time;
> >> +     unsigned long timeout;
> >> +
> >> +     if (csb_readl(tegra, XUSB_CSB_MP_ILOAD_BASE_LO) != 0) {
> >> +             dev_info(dev, "Firmware already loaded, Falcon state 0x%x\n",
> >> +                      csb_readl(tegra, XUSB_FALC_CPUCTL));
> >> +             return 0;
> >> +     }
> >
> > won't this prevent rmmod && insmod from succeeding ? Maybe not, as you
> > return 0 anyway, but if the firmware side craps out, it might be useful
> > to reload the firmware and reset that block. I wonder if this prevents
> > that.
> 
> rmmod && insmod does work with the above since, as you mentioned, we
> return successfully if the controller is already running.  Attempting
> to re-load the firmware if the controller is already running won't
> work - it requires a full power-gate/ungate sequence of the
> controller.  Unfortunately, that sequence is rather complicated and I
> haven't had a chance to implement it yet.  it will hopefully be coming
> later, though.

Alright, could you add a comment on the source code stating that ?
Perhaps a REVISIT note or FIXME ?

cheers

-- 
balbi

Attachment: signature.asc
Description: Digital signature


[Index of Archives]     [Device Tree Compilter]     [Device Tree Spec]     [Linux Driver Backports]     [Video for Linux]     [Linux USB Devel]     [Linux PCI Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Yosemite Backpacking]
  Powered by Linux