On Sun, Jul 7, 2024 at 3:45 AM Michael Trimarchi
<mich...@amarulasolutions.com> wrote:
>
> The osc_24m is the clock-output-name and not the one that
> is used as internal name reference from the strcmp. The clock
> that use osc_24m, will not be able to reparent it as they should.
> We need anyway register the osc_24m clock fixed factor in the clock
> tree.
>
> Fixes: 710c4ffb890 ("clk: imx: clk-imx8mn add gpmi nand clocks")
> Fixes: 85b1c11989c ("clk: imx: Add ECSPI to iMX8MN")
> Cc: Marek Vasut <ma...@denx.de>
> Signed-off-by: Michael Trimarchi <mich...@amarulasolutions.com>
> ---
> drivers/clk/imx/clk-imx8mn.c | 15 +++++++++++----
> 1 file changed, 11 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/clk/imx/clk-imx8mn.c b/drivers/clk/imx/clk-imx8mn.c
> index ed9e16d7c1..bfd1677520 100644
> --- a/drivers/clk/imx/clk-imx8mn.c
> +++ b/drivers/clk/imx/clk-imx8mn.c
> @@ -57,15 +57,15 @@ static const char *imx8mn_usdhc2_sels[] =
> {"clock-osc-24m", "sys_pll1_400m", "sy
> "sys_pll3_out", "sys_pll1_266m",
> "audio_pll2_out", "sys_pll1_100m", };
>
> #if CONFIG_IS_ENABLED(DM_SPI)
> -static const char *imx8mn_ecspi1_sels[] = {"osc_24m", "sys_pll2_200m",
> "sys_pll1_40m",
> +static const char *imx8mn_ecspi1_sels[] = {"clock-osc-24m", "sys_pll2_200m",
> "sys_pll1_40m",
> "sys_pll1_160m", "sys_pll1_800m",
> "sys_pll3_out",
> "sys_pll2_250m", "audio_pll2_out",
> };
>
> -static const char *imx8mn_ecspi2_sels[] = {"osc_24m", "sys_pll2_200m",
> "sys_pll1_40m",
> +static const char *imx8mn_ecspi2_sels[] = {"clock-osc-24m", "sys_pll2_200m",
> "sys_pll1_40m",
> "sys_pll1_160m", "sys_pll1_800m",
> "sys_pll3_out",
> "sys_pll2_250m", "audio_pll2_out",
> };
>
> -static const char *imx8mn_ecspi3_sels[] = {"osc_24m", "sys_pll2_200m",
> "sys_pll1_40m",
> +static const char *imx8mn_ecspi3_sels[] = {"clock-osc-24m", "sys_pll2_200m",
> "sys_pll1_40m",
> "sys_pll1_160m", "sys_pll1_800m",
> "sys_pll3_out",
> "sys_pll2_250m", "audio_pll2_out",
> };
> #endif
> @@ -105,7 +105,7 @@ static const char *imx8mn_usdhc3_sels[] =
> {"clock-osc-24m", "sys_pll1_400m", "sy
> static const char *imx8mn_qspi_sels[] = {"clock-osc-24m", "sys_pll1_400m",
> "sys_pll2_333m", "sys_pll2_500m",
> "audio_pll2_out", "sys_pll1_266m",
> "sys_pll3_out", "sys_pll1_100m", };
>
> -static const char * const imx8mn_nand_sels[] = {"osc_24m", "sys_pll2_500m",
> "audio_pll1_out",
> +static const char * const imx8mn_nand_sels[] = {"clock-osc-24m",
> "sys_pll2_500m", "audio_pll1_out",
> "sys_pll1_400m",
> "audio_pll2_out", "sys_pll3_out",
> "sys_pll2_250m",
> "video_pll_out", };
>
> @@ -119,7 +119,9 @@ static const char * const imx8mn_usb_phy_sels[] =
> {"clock-osc-24m", "sys_pll1_10
>
> static int imx8mn_clk_probe(struct udevice *dev)
> {
> + struct clk osc_24m_clk;
> void __iomem *base;
> + int ret;
>
> base = (void *)ANATOP_BASE_ADDR;
>
> @@ -238,6 +240,11 @@ static int imx8mn_clk_probe(struct udevice *dev)
> clk_dm(IMX8MN_SYS_PLL2_1000M,
> imx_clk_fixed_factor("sys_pll2_1000m", "sys_pll2_out", 1, 1));
>
I know it's late, but I didn't get around to testing my Nano board for
a while. Sorry for the delayed feedback...
> + ret = clk_get_by_name(dev, "osc_24m", &osc_24m_clk);
> + if (ret)
> + return ret;
> + clk_dm(IMX8MN_CLK_24M, dev_get_clk_ptr(osc_24m_clk.dev));
> +
These four lines appear to have introduced a regression on the
imx8mn-beacon board. In the SPL phase, I get an error message
indicating it cannot find the i2c clk, then access to the PMIC fails.
If I remove these four lines, the error message disappears, and the
PMIC is happy again.
I have confirmed that CLK, and SPL_CLK are both defined. I checked
the u-boot-spl.dtb and it shows clock-osc-24m with clock-output-names
= "osc_24m"
I have also tried to mark the 24m clock as with bootph-pre-ram;:
&{/clock-osc-24m} {
bootph-pre-ram;
};
Unfortunately, nothing is helping, and I am not sure what else to try.
clk_fixed_rate.o is built into my SPL, so I don't think it's a config
issue. I need to the PMIC to increase the voltage since we run the
Nano in overdrive mode to run the LPDDR4 at the highest speed.
Might anyone have any suggestions?
adam
> base = dev_read_addr_ptr(dev);
> if (!base)
> return -EINVAL;
> --
> 2.43.0
>
