Heiko, ? 2016?01?08? 17:46, Heiko Stuebner ??: > Hi Caesar, > > Am Freitag, 8. Januar 2016, 17:26:18 schrieb Caesar Wang: >> ? 2016?01?07? 18:05, Heiko Stuebner ??: >>> Am Donnerstag, 7. Januar 2016, 16:25:47 schrieb Caesar Wang: >>>> SCLk_I2S_OUT is the noc bus clock for i2s module, this clock is used by >>>> extra codecs. >>>> >>>> Due to it shouldn't belong to any driver, but we need it enabled, >>>> so just mark it as the CLK_IGNORE_UNUSED flag. >>> What makes you think it shouldn't belong to any driver? >>> >>> In most schematics I have, i2s_clkout is going to some clock-input of >>> the >>> audio codec - probably the MCLK input on your rt5616. And while the new >>> rt5616 driver does not seem to do clock handling, it should just handle >>> the clock using the normal APIs. >> You are right, that's seem has to find the root cause. >> I'm wonder that the codec/i2s is working in kernel v4.1 patch without it. >> >> In general, the MLCK will be handled in i2s driver if the cpu side is >> master. > looking at the rt5616 datasheet, that would be BCLK1 instead I think. This > one does differentiate between being master/slave. > > The master/sysclk (mclk) is also handled in other drivers in all cases. > (da7213.c, da7219.c, max98090.c, max98095.c, probably more) Saw the patch: https://patchwork.kernel.org/patch/5427131/ Add the follow that's working. --- a/arch/arm/boot/dts/rk3036.dtsi +++ b/arch/arm/boot/dts/rk3036.dtsi @@ -288,8 +288,8 @@ interrupts = <GIC_SPI 51 IRQ_TYPE_LEVEL_HIGH>; #address-cells = <1>; #size-cells = <0>; - clock-names = "i2s_clk", "i2s_hclk"; - clocks = <&cru SCLK_I2S>, <&cru HCLK_I2S>; + clock-names = "i2s_clk", "i2s_hclk", "i2s_clk_out"; + clocks = <&cru SCLK_I2S>, <&cru HCLK_I2S>, <&cru SCLK_I2S_OUT>; Okay, I believe Mark likes that works in codec driver. > > > Heiko > > _______________________________________________ > Linux-rockchip mailing list > Linux-rockchip at lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-rockchip -- Thanks, Caesar
