On 02/18/2014 03:20 PM, Stefan Agner wrote: > Am 2014-02-18 20:27, schrieb Stephen Warren: >> On 02/16/2014 12:50 PM, Stefan Agner wrote: >>> +#if defined(CONFIG_TEGRA20) || defined(CONFIG_TEGRA30) >>> +/* USB2_IF_ULPI_TIMING_CTRL_0 */ >>> +#define ULPI_OUTPUT_PINMUX_BYP (1 << 10) >>> +#define ULPI_CLKOUT_PINMUX_BYP (1 << 11) >>> + >>> +/* USB2_IF_ULPI_TIMING_CTRL_1 */ >>> +#define ULPI_DATA_TRIMMER_LOAD (1 << 0) >>> +#define ULPI_DATA_TRIMMER_SEL(x) (((x) & 0x7) << 1) >>> +#define ULPI_STPDIRNXT_TRIMMER_LOAD (1 << 16) >>> +#define ULPI_STPDIRNXT_TRIMMER_SEL(x) (((x) & 0x7) << 17) >>> +#define ULPI_DIR_TRIMMER_LOAD (1 << 24) >>> +#define ULPI_DIR_TRIMMER_SEL(x) (((x) & 0x7) << 25) >>> +#endif >> >> Are those defines really Tegra20/30-only, or were they simply omitted >> from the Tegra114 header because they aren't used by the driver? >> >> Actually, if these aren't used, is it even worth including this text in >> the header? Perhaps delete unused text in a separate patch before this >> one, so that this patch simply moves text between files? > > I don't have access to a Tegra114 TRM right now, but I can it be that > the ULPI PHY is not in place there? If its not used in the driver, I > guess we can remove those defines completely...
Looking further, all those values are used, but by a piece of code that's #ifdef CONFIG_USB_ULPI, and that define is only enabled on a couple of Tegra20 boards. Looking at the TRMs, both the Tegra20/30 TRMs mention the register that contains these fields (but doesn't document the register:-/), but neither the Tegra114/124 TRMs mention the same registers, so perhaps those values are Tegra20/30-only, so the ifdef you have is fine. _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
