Am Freitag, dem 25.02.2022 um 10:17 -0600 schrieb Adam Ford:
> When attempting to communicate with the ISI, the system hangs.
> Through trial and error, it appears removing the two root clocks
> fixes the issues. Since IMX8MN_DISPBLK_PD_ISI enables the
> pgc_dispmix, and pgc_dispmix enables those same root clocks, it should be
> safe.
>
This doesn't look right. Why would this make a difference?
The only difference is that those clocks may be disabled at the end of
the power-up routine when they are assigned to the blk-ctrl domain. If
ISI communication hangs while those clocks are disabled, then that is
more of a hint for a missing clock enable on the ISI driver side than
an issue on the blk-ctrl side.
Regards,
Lucas
> Fixes: 7f511d514e8c ("soc: imx: imx8m-blk-ctrl: add i.MX8MN DISP blk-ctrl")
> Signed-off-by: Adam Ford <[email protected]>
>
> diff --git a/drivers/soc/imx/imx8m-blk-ctrl.c b/drivers/soc/imx/imx8m-blk-ctrl.c
> index 122f9c884b38..cacf379aef80 100644
> --- a/drivers/soc/imx/imx8m-blk-ctrl.c
> +++ b/drivers/soc/imx/imx8m-blk-ctrl.c
> @@ -574,9 +574,8 @@ static const struct imx8m_blk_ctrl_domain_data imx8mn_disp_blk_ctl_domain_data[]
> },
> [IMX8MN_DISPBLK_PD_ISI] = {
> .name = "dispblk-isi",
> - .clk_names = (const char *[]){ "disp_axi", "disp_apb", "disp_axi_root",
> - "disp_apb_root"},
> - .num_clks = 4,
> + .clk_names = (const char *[]){ "disp_axi", "disp_apb" },
> + .num_clks = 2,
> .gpc_name = "isi",
> .rst_mask = BIT(6) | BIT(7),
> .clk_mask = BIT(6) | BIT(7),