[PATCH] drm/fsl-dcu: fix endian issue when using clk_register_divider
Meng Yi
meng.yi at nxp.com
Mon Sep 5 08:46:56 UTC 2016
> Subject: [PATCH] drm/fsl-dcu: fix endian issue when using clk_register_divider
>
> Since using clk_register_divider to setup the pixel clock, regmap is no longer
> used. Regmap did take care of DCU using different endianness. Check
> endianness using the device-tree property "big-endian" to determine the
> location of DIV_RATIO.
>
> Cc: stable at vger.kernel.org
> Fixes: 2d701449bce1 ("drm/fsl-dcu: use common clock framework for pixel
> clock divider")
> Reported-by: Meng Yi <meng.yi at nxp.com>
> Signed-off-by: Stefan Agner <stefan at agner.ch>
> ---
> drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c | 7 ++++++-
> 1 file changed, 6 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c b/drivers/gpu/drm/fsl-
> dcu/fsl_dcu_drm_drv.c
> index 7882387..8dd042e 100644
> --- a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c
> +++ b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_drv.c
> @@ -330,6 +330,7 @@ static int fsl_dcu_drm_probe(struct platform_device
> *pdev)
> const char *pix_clk_in_name;
> const struct of_device_id *id;
> int ret;
> + u8 div_ratio_shift = 0;
>
> fsl_dev = devm_kzalloc(dev, sizeof(*fsl_dev), GFP_KERNEL);
> if (!fsl_dev)
> @@ -382,11 +383,15 @@ static int fsl_dcu_drm_probe(struct platform_device
> *pdev)
> pix_clk_in = fsl_dev->clk;
> }
>
> + if (of_property_read_bool(dev->of_node, "big-endian"))
> + div_ratio_shift = 24;
> +
> +
> pix_clk_in_name = __clk_get_name(pix_clk_in);
> snprintf(pix_clk_name, sizeof(pix_clk_name), "%s_pix",
> pix_clk_in_name);
> fsl_dev->pix_clk = clk_register_divider(dev, pix_clk_name,
> pix_clk_in_name, 0, base + DCU_DIV_RATIO,
> - 0, 8, CLK_DIVIDER_ROUND_CLOSEST, NULL);
> + div_ratio_shift, 8, CLK_DIVIDER_ROUND_CLOSEST,
> NULL);
> if (IS_ERR(fsl_dev->pix_clk)) {
> dev_err(dev, "failed to register pix clk\n");
> ret = PTR_ERR(fsl_dev->pix_clk);
> --
> 2.9.0
Tested-by: Meng Yi <meng.yi at nxp.com>
On LS1021A-TWR board.
Meng
More information about the dri-devel
mailing list