| Message ID | 534CF8A3.3010404@gmail.com |
|---|---|
| State | New |
| Headers | show |
On Tue, Apr 15, 2014 at 10:15:15AM +0100, Iain Paton wrote: > I'm working on a board using the imx6solo that implements i2c4 and while > building the devicetree it became clear that the devicetree fragment for > i2c4 in imx6dl.dtsi it's missing the necessary clock. > > Investigating further turned up that the ecspi5 clock in CCM_CCGR1 on the > quad/dual is replaced with the i2c4 clock on duallite/solo. > > Would the following patch be sufficient to allow correct use of this clock? Yes, Iain. It looks correct to me. Care to send a formal patch for it? Shawn > > I'm not familiar enough with the clock framework to be sure I'm not missing > something here, but have tested this patch on the board and found it to work. > > The board is element14's RIoTboard http://www.riotboard.org and I have dts > mostly ready to submit for it pending resolving this and a couple of other > minor things, and assuming Eric doesn't beat me to it. > > Iain > > diff --git a/arch/arm/mach-imx/clk-imx6q.c b/arch/arm/mach-imx/clk-imx6q.c > index b0e7f9d..2961b16 100644 > --- a/arch/arm/mach-imx/clk-imx6q.c > +++ b/arch/arm/mach-imx/clk-imx6q.c > @@ -352,7 +352,13 @@ static void __init imx6q_clocks_init(struct device_node *ccm_node) > clk[ecspi2] = imx_clk_gate2("ecspi2", "ecspi_root", base + 0x6c, 2); > clk[ecspi3] = imx_clk_gate2("ecspi3", "ecspi_root", base + 0x6c, 4); > clk[ecspi4] = imx_clk_gate2("ecspi4", "ecspi_root", base + 0x6c, 6); > - clk[ecspi5] = imx_clk_gate2("ecspi5", "ecspi_root", base + 0x6c, 8); > + if (cpu_is_imx6dl()) > + /* > + * ecspi5 is replaced with i2c4 on imx6dl & imx6s > + */ > + clk[ecspi5] = imx_clk_gate2("i2c4", "ipg_per", base + 0x6c, 8); > + else > + clk[ecspi5] = imx_clk_gate2("ecspi5", "ecspi_root", base + 0x6c, 8); > clk[enet] = imx_clk_gate2("enet", "ipg", base + 0x6c, 10); > clk[esai] = imx_clk_gate2("esai", "esai_podf", base + 0x6c, 16); > clk[gpt_ipg] = imx_clk_gate2("gpt_ipg", "ipg", base + 0x6c, 20); > >
On 16/04/14 08:47, Shawn Guo wrote: > On Tue, Apr 15, 2014 at 10:15:15AM +0100, Iain Paton wrote: >> I'm working on a board using the imx6solo that implements i2c4 and while >> building the devicetree it became clear that the devicetree fragment for >> i2c4 in imx6dl.dtsi it's missing the necessary clock. >> >> Investigating further turned up that the ecspi5 clock in CCM_CCGR1 on the >> quad/dual is replaced with the i2c4 clock on duallite/solo. >> >> Would the following patch be sufficient to allow correct use of this clock? > > Yes, Iain. It looks correct to me. Care to send a formal patch for it? Sure, it's on it's way.
diff --git a/arch/arm/mach-imx/clk-imx6q.c b/arch/arm/mach-imx/clk-imx6q.c index b0e7f9d..2961b16 100644 --- a/arch/arm/mach-imx/clk-imx6q.c +++ b/arch/arm/mach-imx/clk-imx6q.c @@ -352,7 +352,13 @@ static void __init imx6q_clocks_init(struct device_node *ccm_node) clk[ecspi2] = imx_clk_gate2("ecspi2", "ecspi_root", base + 0x6c, 2); clk[ecspi3] = imx_clk_gate2("ecspi3", "ecspi_root", base + 0x6c, 4); clk[ecspi4] = imx_clk_gate2("ecspi4", "ecspi_root", base + 0x6c, 6); - clk[ecspi5] = imx_clk_gate2("ecspi5", "ecspi_root", base + 0x6c, 8); + if (cpu_is_imx6dl()) + /* + * ecspi5 is replaced with i2c4 on imx6dl & imx6s + */ + clk[ecspi5] = imx_clk_gate2("i2c4", "ipg_per", base + 0x6c, 8); + else + clk[ecspi5] = imx_clk_gate2("ecspi5", "ecspi_root", base + 0x6c, 8); clk[enet] = imx_clk_gate2("enet", "ipg", base + 0x6c, 10); clk[esai] = imx_clk_gate2("esai", "esai_podf", base + 0x6c, 16); clk[gpt_ipg] = imx_clk_gate2("gpt_ipg", "ipg", base + 0x6c, 20);