Hi Raphaël Poggi, On 09/14/2014 04:12 AM, Raphaël Poggi wrote:
+static struct clk_lookup periph_clocks_lookups[] = { >+ CLKDEV_CON_DEV_ID("macb_clk", "macb0", &macb0_clk), >+ CLKDEV_CON_DEV_ID("mci_clk", "atmel_mci0", &mmc0_clk), >+ CLKDEV_CON_DEV_ID("mci_clk", "atmel_mci1", &mmc1_clk), >+ CLKDEV_CON_DEV_ID("spi_clk", "atmel_spi0", &spi0_clk), >+ CLKDEV_DEV_ID("at91sam9x5-gpio0", &pioA_clk), >+ CLKDEV_DEV_ID("at91sam9x5-gpio1", &pioB_clk), >+ CLKDEV_DEV_ID("at91sam9x5-gpio2", &pioC_clk), >+ CLKDEV_DEV_ID("at91sam9x5-gpio3", &pioD_clk), >+ CLKDEV_DEV_ID("at91sam9x5-gpio4", &pioE_clk),Why not using clkdev_add_physbase for the gpio clocks ? This allow to use these clocks in device tree and non device context.
Thanks for remind this. I will fix it in v2.
+#else >+void at91_add_device_i2c(short i2c_id, struct i2c_board_info *devices, >+ int nr_devices) {} >+#endifMaybe we can use the new atmel i2c driver, no ?
Yes, we can. However, I think use the GPIO i2c first, after the i2c fix has been applied, then switch to use atmel i2c driver. In this way, it won't block the sama5d4ek board support go into mainline.
Btw, if the i2c fix has been applied before I send the v2 patch, I will switch to use atmel i2c driver.
Best Regards, Bo Shen _______________________________________________ barebox mailing list barebox@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/barebox
