Hello,
On Tue, Nov 29, 2011 at 08:02:53AM +0000, Chen Peter-B29397 wrote:
>
> > Peter Chen writes:
> > > The patch removes all the uses of cpu_is_mx(). Instead, it utilizes
> > > platform_device_id to distinguish the ehci differences among SoCs.
> > > It can be useful to imx ehci submission and device tree support later.
> > >
> > >
> > As the EHCI core is equivalent on all SoCs but only require different
> > clocks on each platform I think it's overkill to use platform_ids to
> > distinguish the SoCs. Indeed there is actually no need to make any
> > distinction in the driver, as you could simply provide the same set of
> > clocks on all platforms with some of them being dummy clocks on
> > certain platforms:
> > clock-imx25.c and clock-imx35.c:
> > _REGISTER_CLOCK("mxc-ehci.0", "usb", usbotg_clk)
> > _REGISTER_CLOCK("mxc-ehci.0", "system_bus", dummy_clk)
> > _REGISTER_CLOCK("mxc-ehci.0", "usb_phy", dummy_clk)
> > _REGISTER_CLOCK("mxc-ehci.1", "usb", usbotg_clk)
> > _REGISTER_CLOCK("mxc-ehci.1", "system_bus", dummy_clk)
> > _REGISTER_CLOCK("mxc-ehci.1", "usb_phy", dummy_clk)
> > _REGISTER_CLOCK("mxc-ehci.2", "usb", usbotg_clk)
> > _REGISTER_CLOCK("mxc-ehci.2", "system_bus", dummy_clk)
> > _REGISTER_CLOCK("mxc-ehci.2", "usb_phy", dummy_clk)
> > clock-imx27.c:
> > _REGISTER_CLOCK("mxc-ehci.0", "usb", usb_clk)
> > _REGISTER_CLOCK("mxc-ehci.0", "system_bus", usb_clk1)
> > _REGISTER_CLOCK("mxc-ehci.0", "usb_phy", dummy_clk)
> > _REGISTER_CLOCK("mxc-ehci.1", "usb", usb_clk)
> > _REGISTER_CLOCK("mxc-ehci.1", "system_bus", usb_clk1)
> > _REGISTER_CLOCK("mxc-ehci.1", "usb_phy", dummy_clk)
> > _REGISTER_CLOCK("mxc-ehci.2", "usb", usb_clk)
> > _REGISTER_CLOCK("mxc-ehci.2", "system_bus", usb_clk1)
> > _REGISTER_CLOCK("mxc-ehci.2", "usb_phy", dummy_clk)
> > clock-imx31.c:
> > _REGISTER_CLOCK("mxc-ehci.0", "usb", usb_clk1)
> > _REGISTER_CLOCK("mxc-ehci.0", "system_bus", usb_clk2)
> > _REGISTER_CLOCK("mxc-ehci.0", "usb_phy", dummy_clk)
> > _REGISTER_CLOCK("mxc-ehci.1", "usb", usb_clk1)
> > _REGISTER_CLOCK("mxc-ehci.1", "system_bus", usb_clk2)
> > _REGISTER_CLOCK("mxc-ehci.1", "usb_phy", dummy_clk)
> > _REGISTER_CLOCK("mxc-ehci.2", "usb", usb_clk1)
> > _REGISTER_CLOCK("mxc-ehci.2", "system_bus", usb_clk2)
> > _REGISTER_CLOCK("mxc-ehci.2", "usb_phy", dummy_clk)
> > clock-mx51-mx53.c:
> > _REGISTER_CLOCK("mxc-ehci.0", "usb", usboh3_clk)
> > _REGISTER_CLOCK("mxc-ehci.0", "system_bus", usb_ahb_clk)
> > _REGISTER_CLOCK("mxc-ehci.0", "usb_phy", usb_phy1_clk)
> > _REGISTER_CLOCK("mxc-ehci.1", "usb", usboh3_clk)
> > _REGISTER_CLOCK("mxc-ehci.1", "system_bus", usb_ahb_clk)
> > _REGISTER_CLOCK("mxc-ehci.1", "usb_phy", dummy_clk)
> > _REGISTER_CLOCK("mxc-ehci.2", "usb", usboh3_clk)
> > _REGISTER_CLOCK("mxc-ehci.2", "system_bus", usb_ahb_clk)
> > _REGISTER_CLOCK("mxc-ehci.2", "usb_phy", dummy_clk)
> >
> >
> > In the driver you would then do a clk_get() for those three clocks
> > independent of the platform.
> >
> The reason why I add this patch is not for clk things, it is for:
> 1. Coming device driver
> 2. Handle things difference between SoC's. not only clk, others
> like low power, wakeup, hsic, or other will comes in future.
Then are they really different per SoC (in a way not yet handled in the
driver)? If not, up to now these devices were named after the first SoC
introducing them, so i.MX25 has imx21-uart.0 and imx35-cspi.0.
Best regards
Uwe
--
Pengutronix e.K. | Uwe Kleine-König |
Industrial Linux Solutions | http://www.pengutronix.de/ |
--
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
