[PATCH v13 13/14] drm/mediatek: dpi: Add dp_intf support
Rex-BC Chen
rex-bc.chen at mediatek.com
Wed Jun 22 02:10:55 PDT 2022
On Tue, 2022-06-21 at 14:11 +0200, AngeloGioacchino Del Regno wrote:
> Il 21/06/22 13:37, Bo-Chen Chen ha scritto:
> > From: Guillaume Ranquet <granquet at baylibre.com>
> >
> > Dpintf is the displayport interface hardware unit. This unit is
> > similar
> > to dpi and can reuse most of the code.
> >
> > This patch adds support for mt8195-dpintf to this dpi driver. Main
> > differences are:
> > - 4 pixels for one round for dp_intf while dpi is 1 pixel for one
> > round.
> > Therefore, we add a new config "round_pixels" to control
> > setting of
>
> Okay, I like where this is going, but "round_pixels" explains itself
> like
> "the pixels are round" (which doesn't make sense).
>
> So... is this "4 pixels per iteration" (computation of 4 pixels per
> iteration),
> or "outputs 4 pixels at a time"?
> I can give you some ideas about a better name for this struct member,
> I think...
> ...something like "output_4pixel"?
>
> Alternatively, we can do something even cleaner here:
> u8 pixels_per_round;
>
> ...where we define it like...
>
> static const struct mtk_dpi_conf mt8195_dpintf_conf = {
> .pixels_per_round = 4,
> ..........
> }
>
> static const struct mtk_dpi_conf mt8192_conf = {
> .pixels_per_round = 1,
> .........
> }
>
> ... and then, in function mtk_dpi_set_display_mode(), we would have
> something like
>
> /*
>
> * Depending on the IP version, we may output a different
> amount
>
> * of pixels for each (choose: round/iteration): divide the
> clock
>
> * by this number and adjust the display porches accordingly.
>
> */
>
> vm.pixelclock = pll_rate / factor;
>
> vm.pixelclock /= dpi->conf->pixels_per_round;
>
>
>
> ......
>
>
>
> hsync.sync_width = vm.hsync_len / dpi->conf->pixels_per_round;
>
> hsync.back_porch = vm.hback_porch / dpi->conf-
> >pixels_per_round;
>
> hsync.front_porch = vm.hfront_porch / dpi->conf-
> >pixels_per_round;
>
> This would also eliminate any need to check if the IP has 4 pixels
> per round
> or if it has just 1, as in the latter case, we'd be dividing by 1
> (hence, we
> would not be dividing anything).
>
> Do you like this solution? :-)
>
> Cheers,
> Angelo
>
Hello Angelo,
Thanks for the advise!
After syncing with CK, I will use mothod of "pixels_per_round".
This soultion is pretty good and more clear for this config.
Thanks again!
BRs,
Bo-Chen
> > pixel clock and timing parameter for dp_intf.
> > - Input of dp_intf is two pixels per round, so we add a new
> > config
> > "input_2pixel" to control this.
> > - Some register contents differ slightly between the two
> > components. To
> > work around this I added register bits/masks with a DPINTF_
> > prefix
> > and use them where different.
> >
> > Based on a separate driver for dpintf created by
> > Jitao shi <jitao.shi at mediatek.com>.
> >
> > Signed-off-by: Markus Schneider-Pargmann <msp at baylibre.com>
> > Signed-off-by: Guillaume Ranquet <granquet at baylibre.com>
> > [Bo-Chen: Modify reviewers' comments.]
> > Signed-off-by: Bo-Chen Chen <rex-bc.chen at mediatek.com>
> > ---
> > drivers/gpu/drm/mediatek/mtk_dpi.c | 62
> > +++++++++++++++++++++
> > drivers/gpu/drm/mediatek/mtk_dpi_regs.h | 12 ++++
> > drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c | 4 ++
> > drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h | 1 +
> > drivers/gpu/drm/mediatek/mtk_drm_drv.c | 3 +
> > 5 files changed, 82 insertions(+)
> >
> > diff --git a/drivers/gpu/drm/mediatek/mtk_dpi.c
> > b/drivers/gpu/drm/mediatek/mtk_dpi.c
> > index ef7f828a4b1e..e562f0d55cc2 100644
> > --- a/drivers/gpu/drm/mediatek/mtk_dpi.c
> > +++ b/drivers/gpu/drm/mediatek/mtk_dpi.c
> > @@ -131,6 +131,11 @@ struct mtk_dpi_yc_limit {
> > * @color_fmt_trans_support: Enable color format transfer.
> > * @support_direct_pin: Dpi can directly connect pins, and enable
> > this config
> > * to do this.
> > + * @round_pixels: Dp_intf is 4 pixels for one round while dpi is
> > one pixel for
> > + * one round, so we need to enable this config for
> > dp_intf to do
> > + * corresponding settings.
> > + * @input_2pixel: Input pixel of dp_intf is 2 pixel per round, so
> > enable this
> > + * config to enable this feature.
> > * @dimension_mask: Mask used for HWIDTH, HPORCH, VSYNC_WIDTH and
> > VSYNC_PORCH
> > * (no shift).
> > * @hvsize_mask: Mask of HSIZE and VSIZE mask (no shift).
> > @@ -149,6 +154,8 @@ struct mtk_dpi_conf {
> > bool swap_input_support;
> > bool color_fmt_trans_support;
> > bool support_direct_pin;
> > + bool round_pixels;
> > + bool input_2pixel;
> > u32 dimension_mask;
> > u32 hvsize_mask;
> > u32 channel_swap_shift;
> > @@ -560,6 +567,14 @@ static int mtk_dpi_set_display_mode(struct
> > mtk_dpi *dpi,
> > pll_rate = clk_get_rate(dpi->tvd_clk);
> >
> > vm.pixelclock = pll_rate / factor;
> > +
> > + /*
> > + * For dp_intf, we need to divide clock by 4 because it's
> > + * 4 pixels for one round while dpi is 1 pixel for one round.
> > + */
> > + if (dpi->conf->round_pixels)
> > + vm.pixelclock /= 4;
> > +
> > if ((dpi->output_fmt == MEDIA_BUS_FMT_RGB888_2X12_LE) ||
> > (dpi->output_fmt == MEDIA_BUS_FMT_RGB888_2X12_BE))
> > clk_set_rate(dpi->pixel_clk, vm.pixelclock * 2);
> > @@ -581,6 +596,17 @@ static int mtk_dpi_set_display_mode(struct
> > mtk_dpi *dpi,
> > hsync.sync_width = vm.hsync_len;
> > hsync.back_porch = vm.hback_porch;
> > hsync.front_porch = vm.hfront_porch;
> > +
> > + /*
> > + * For dp_intf, we need to divide everything by 4 because it's
> > + * 4 pixels for one round while dpi is 1 pixel for one round.
> > + */
> > + if (dpi->conf->round_pixels) {
> > + hsync.sync_width = vm.hsync_len / 4;
> > + hsync.back_porch = vm.hback_porch / 4;
> > + hsync.front_porch = vm.hfront_porch / 4;
> > + }
> > +
> > hsync.shift_half_line = false;
> > vsync_lodd.sync_width = vm.vsync_len;
> > vsync_lodd.back_porch = vm.vback_porch;
> > @@ -629,6 +655,10 @@ static int mtk_dpi_set_display_mode(struct
> > mtk_dpi *dpi,
> > mtk_dpi_dual_edge(dpi);
> > mtk_dpi_config_disable_edge(dpi);
> > }
> > + if (dpi->conf->input_2pixel) {
> > + mtk_dpi_mask(dpi, DPI_CON, DPINTF_INPUT_2P_EN,
> > + DPINTF_INPUT_2P_EN);
> > + }
> > mtk_dpi_sw_reset(dpi, false);
> >
> > return 0;
> > @@ -869,6 +899,16 @@ static unsigned int
> > mt8183_calculate_factor(int clock)
> > return 2;
> > }
> >
> > +static unsigned int mt8195_dpintf_calculate_factor(int clock)
> > +{
> > + if (clock < 70000)
> > + return 4;
> > + else if (clock < 200000)
> > + return 2;
> > + else
> > + return 1;
> > +}
> > +
> > static const u32 mt8173_output_fmts[] = {
> > MEDIA_BUS_FMT_RGB888_1X24,
> > };
> > @@ -878,6 +918,11 @@ static const u32 mt8183_output_fmts[] = {
> > MEDIA_BUS_FMT_RGB888_2X12_BE,
> > };
> >
> > +static const u32 mt8195_output_fmts[] = {
> > + MEDIA_BUS_FMT_RGB888_1X24,
> > + MEDIA_BUS_FMT_YUYV8_1X16,
> > +};
> > +
> > static const struct mtk_dpi_conf mt8173_conf = {
> > .cal_factor = mt8173_calculate_factor,
> > .reg_h_fre_con = 0xe0,
> > @@ -943,6 +988,20 @@ static const struct mtk_dpi_conf mt8192_conf =
> > {
> > .csc_enable_bit = CSC_ENABLE,
> > };
> >
> > +static const struct mtk_dpi_conf mt8195_dpintf_conf = {
> > + .cal_factor = mt8195_dpintf_calculate_factor,
> > + .max_clock_khz = 600000,
> > + .output_fmts = mt8195_output_fmts,
> > + .num_output_fmts = ARRAY_SIZE(mt8195_output_fmts),
> > + .round_pixels = true,
> > + .input_2pixel = true,
> > + .dimension_mask = DPINTF_HPW_MASK,
> > + .hvsize_mask = DPINTF_HSIZE_MASK,
> > + .channel_swap_shift = DPINTF_CH_SWAP,
> > + .yuv422_en_bit = DPINTF_YUV422_EN,
> > + .csc_enable_bit = DPINTF_CSC_ENABLE,
> > +};
> > +
> > static int mtk_dpi_probe(struct platform_device *pdev)
> > {
> > struct device *dev = &pdev->dev;
> > @@ -1065,6 +1124,9 @@ static const struct of_device_id
> > mtk_dpi_of_ids[] = {
> > { .compatible = "mediatek,mt8192-dpi",
> > .data = &mt8192_conf,
> > },
> > + { .compatible = "mediatek,mt8195-dp-intf",
> > + .data = &mt8195_dpintf_conf,
> > + },
> > { },
> > };
> > MODULE_DEVICE_TABLE(of, mtk_dpi_of_ids);
More information about the Linux-mediatek
mailing list