Re: [PATCH v4 04/11] drm/bridge: synopsys: dw-hdmi: add bus format negociation

From: Neil Armstrong
Date: Mon Mar 02 2020 - 03:08:26 EST

Next message: Thomas Gleixner: "Re: [patch 4/8] x86/entry: Move irq tracing on syscall entry to C-code"
Previous message: Uwe Kleine-König: "Re: [PATCH V2] thermal: imx_thermal: Use __maybe_unused instead of CONFIG_PM_SLEEP"
In reply to: Neil Armstrong: "Re: [PATCH v4 04/11] drm/bridge: synopsys: dw-hdmi: add bus format negociation"
Next in thread: Laurent Pinchart: "Re: [PATCH v4 04/11] drm/bridge: synopsys: dw-hdmi: add bus format negociation"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On 29/02/2020 11:09, Jonas Karlman wrote:
> Hi Jernej,
>
> On 2020-02-29 08:42, Jernej Åkrabec wrote:
>> Hi Neil!
>>
>> Dne Äetrtek, 06. februar 2020 ob 20:18:27 CET je Neil Armstrong napisal(a):
>>> Add the atomic_get_output_bus_fmts, atomic_get_input_bus_fmts to negociate
>>> the possible output and input formats for the current mode and monitor,
>>> and use the negotiated formats in a basic atomic_check callback.
>>>
>>> Signed-off-by: Neil Armstrong <narmstrong@xxxxxxxxxxxx>
>>> ---
>>> drivers/gpu/drm/bridge/synopsys/dw-hdmi.c | 272 +++++++++++++++++++++-
>>> 1 file changed, 268 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
>>> b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c index
>>> fec4a4bcd1fe..15048ad694bc 100644
>>> --- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
>>> +++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
>>> @@ -2095,11 +2095,10 @@ static int dw_hdmi_setup(struct dw_hdmi *hdmi,
>>> struct drm_display_mode *mode)
>>> hdmi->hdmi_data.video_mode.mpixelrepetitionoutput = 0;
>>> hdmi->hdmi_data.video_mode.mpixelrepetitioninput = 0;
>>>
>>> - /* TOFIX: Get input format from plat data or fallback to RGB888 */
>>> if (hdmi->plat_data->input_bus_format)
>>> hdmi->hdmi_data.enc_in_bus_format =
>>> hdmi->plat_data->input_bus_format;
>>> - else
>>> + else if (hdmi->hdmi_data.enc_in_bus_format == MEDIA_BUS_FMT_FIXED)
>>> hdmi->hdmi_data.enc_in_bus_format =
>> MEDIA_BUS_FMT_RGB888_1X24;
>>>
>>> /* TOFIX: Get input encoding from plat data or fallback to none */
>>> @@ -2109,8 +2108,8 @@ static int dw_hdmi_setup(struct dw_hdmi *hdmi, struct
>>> drm_display_mode *mode) else
>>> hdmi->hdmi_data.enc_in_encoding =
>> V4L2_YCBCR_ENC_DEFAULT;
>>>
>>> - /* TOFIX: Default to RGB888 output format */
>>> - hdmi->hdmi_data.enc_out_bus_format = MEDIA_BUS_FMT_RGB888_1X24;
>>> + if (hdmi->hdmi_data.enc_out_bus_format == MEDIA_BUS_FMT_FIXED)
>>> + hdmi->hdmi_data.enc_out_bus_format =
>> MEDIA_BUS_FMT_RGB888_1X24;
>>>
>>> hdmi->hdmi_data.pix_repet_factor = 0;
>>> hdmi->hdmi_data.hdcp_enable = 0;
>>> @@ -2388,6 +2387,267 @@ static const struct drm_connector_helper_funcs
>>> dw_hdmi_connector_helper_funcs = .atomic_check =
>>> dw_hdmi_connector_atomic_check,
>>> };
>>>
>>> +/*
>>> + * Possible output formats :
>>> + * - MEDIA_BUS_FMT_UYYVYY16_0_5X48,
>>> + * - MEDIA_BUS_FMT_UYYVYY12_0_5X36,
>>> + * - MEDIA_BUS_FMT_UYYVYY10_0_5X30,
>>> + * - MEDIA_BUS_FMT_UYYVYY8_0_5X24,
>>> + * - MEDIA_BUS_FMT_YUV16_1X48,
>>> + * - MEDIA_BUS_FMT_RGB161616_1X48,
>>> + * - MEDIA_BUS_FMT_UYVY12_1X24,
>>> + * - MEDIA_BUS_FMT_YUV12_1X36,
>>> + * - MEDIA_BUS_FMT_RGB121212_1X36,
>>> + * - MEDIA_BUS_FMT_UYVY10_1X20,
>>> + * - MEDIA_BUS_FMT_YUV10_1X30,
>>> + * - MEDIA_BUS_FMT_RGB101010_1X30,
>>> + * - MEDIA_BUS_FMT_UYVY8_1X16,
>>> + * - MEDIA_BUS_FMT_YUV8_1X24,
>>> + * - MEDIA_BUS_FMT_RGB888_1X24,
>>> + */
>>> +
>>> +/* Can return a maximum of 12 possible output formats for a mode/connector
>>> */ +#define MAX_OUTPUT_SEL_FORMATS 12
>>> +
>>> +static u32 *dw_hdmi_bridge_atomic_get_output_bus_fmts(struct drm_bridge
>>> *bridge, + struct
>> drm_bridge_state *bridge_state,
>>> + struct drm_crtc_state
>> *crtc_state,
>>> + struct
>> drm_connector_state *conn_state,
>>> + unsigned int
>> *num_output_fmts)
>>> +{
>>> + struct drm_connector *conn = conn_state->connector;
>>> + struct drm_display_info *info = &conn->display_info;
>>> + struct drm_display_mode *mode = &crtc_state->mode;
>>> + u8 max_bpc = conn_state->max_requested_bpc;
>>> + bool is_hdmi2_sink = info->hdmi.scdc.supported ||
>>> + (info->color_formats &
>> DRM_COLOR_FORMAT_YCRCB420);
>>> + u32 *output_fmts;
>>> + int i = 0;
>>> +
>>> + *num_output_fmts = 0;
>>> +
>>> + output_fmts = kcalloc(MAX_OUTPUT_SEL_FORMATS, sizeof(*output_fmts),
>>> + GFP_KERNEL);
>>> + if (!output_fmts)
>>> + return NULL;
>>> +
>>> + /*
>>> + * If the current mode enforces 4:2:0, force the output but format
>>> + * to 4:2:0 and do not add the YUV422/444/RGB formats
>>> + */
>>> + if (conn->ycbcr_420_allowed &&
>>> + (drm_mode_is_420_only(info, mode) ||
>>> + ())) {
>>> +
>>> + /* Order bus formats from 16bit to 8bit if supported */
>>> + if (max_bpc >= 16 && info->bpc == 16 &&
>>> + (info->hdmi.y420_dc_modes &
>> DRM_EDID_YCBCR420_DC_48))
>>> + output_fmts[i++] =
>> MEDIA_BUS_FMT_UYYVYY16_0_5X48;
>>> +
>>> + if (max_bpc >= 12 && info->bpc >= 12 &&
>>> + (info->hdmi.y420_dc_modes &
>> DRM_EDID_YCBCR420_DC_36))
>>> + output_fmts[i++] =
>> MEDIA_BUS_FMT_UYYVYY12_0_5X36;
>>> +
>>> + if (max_bpc >= 10 && info->bpc >= 10 &&
>>> + (info->hdmi.y420_dc_modes &
>> DRM_EDID_YCBCR420_DC_30))
>>> + output_fmts[i++] =
>> MEDIA_BUS_FMT_UYYVYY10_0_5X30;
>>> +
>>> + /* Default 8bit fallback */
>>> + output_fmts[i++] = MEDIA_BUS_FMT_UYYVYY8_0_5X24;
>>> +
>>> + *num_output_fmts = i;
>>> +
>>> + return output_fmts;
>>
>> Driver shouldn't return just yet for case "is_hdmi2_sink &&
>> drm_mode_is_420_also(info, mode)", because monitor/TV also supports YCbCr
>> 4:4:4 in that case. IMO YCbCr 4:4:4 should be even prefered. What do you
>> think?
>
> I think we need to have some way for controller driver and userspace to control what
> hdmi output format gets selected. I know for a fact that some Samsung TV have issues
> with 444 YCbCr modes at 4k 50/60hz but have no problems with 420 modes.
> The Samsung TV edid lie and/or the TV is not fully following HDMI specs.

In this case this must be handled in the EDID quirks, by disabling SCDC & co for this TV.

>
> From a personal and mediaplayer userspace perspective I would like to prefer
> 420/444 YCbCr mode as soon as any yuv drm plane is active and rgb 444 anytime else.
>
> On Rockchip SoCs the display controller cannot output yuv422 to dw-hdmi block,
> the optimal output format selection in such case should put yuv422 last.

I think 4:2:2 should be moved after 4:4:4 indeed, but on another side, 10bit cannot be
done in 4k60 4:4:4, so only 4:2:2 is possible when the tv is not 4:2:0 capable.

>
> Maybe dw-hdmi can call a dw-hdmi glue driver callback to get the preferred output format order?

This seems complex to handle...

>
> On a side note but related issue, the dw-hdmi format negotiation code should
> probably also filter modes on tmds rate, something like [1].
> It is needed to filter out deep color modes that is not supported by the sink or hdmi spec.

Indeed.

>
> [1] https://github.com/Kwiboo/linux-rockchip/commit/fc3df6903384e764ab6ac59879c489cbef55fcbe
>
> Best regards,
> Jonas

Neil

>
>>
>> Best regards,
>> Jernej
>>
>>> + }
>>> +
>>> + /*
>>> + * Order bus formats from 16bit to 8bit and from YUV422 to RGB
>>> + * if supported. In any case the default RGB888 format is added
>>> + */
>>> +
>>> + if (max_bpc >= 16 && info->bpc == 16) {
>>> + if (info->color_formats & DRM_COLOR_FORMAT_YCRCB444)
>>> + output_fmts[i++] = MEDIA_BUS_FMT_YUV16_1X48;
>>> +
>>> + output_fmts[i++] = MEDIA_BUS_FMT_RGB161616_1X48;
>>> + }
>>> +
>>> + if (max_bpc >= 12 && info->bpc >= 12) {
>>> + if (info->color_formats & DRM_COLOR_FORMAT_YCRCB422)
>>> + output_fmts[i++] = MEDIA_BUS_FMT_UYVY12_1X24;
>>> +
>>> + if (info->color_formats & DRM_COLOR_FORMAT_YCRCB444)
>>> + output_fmts[i++] = MEDIA_BUS_FMT_YUV12_1X36;
>>> +
>>> + output_fmts[i++] = MEDIA_BUS_FMT_RGB121212_1X36;
>>> + }
>>> +
>>> + if (max_bpc >= 10 && info->bpc >= 10) {
>>> + if (info->color_formats & DRM_COLOR_FORMAT_YCRCB422)
>>> + output_fmts[i++] = MEDIA_BUS_FMT_UYVY10_1X20;
>>> +
>>> + if (info->color_formats & DRM_COLOR_FORMAT_YCRCB444)
>>> + output_fmts[i++] = MEDIA_BUS_FMT_YUV10_1X30;
>>> +
>>> + output_fmts[i++] = MEDIA_BUS_FMT_RGB101010_1X30;
>>> + }
>>> +
>>> + if (info->color_formats & DRM_COLOR_FORMAT_YCRCB422)
>>> + output_fmts[i++] = MEDIA_BUS_FMT_UYVY8_1X16;
>>> +
>>> + if (info->color_formats & DRM_COLOR_FORMAT_YCRCB444)
>>> + output_fmts[i++] = MEDIA_BUS_FMT_YUV8_1X24;
>>> +
>>> + /* Default 8bit RGB fallback */
>>> + output_fmts[i++] = MEDIA_BUS_FMT_RGB888_1X24;
>>> +
>>> + *num_output_fmts = i;
>>> +
>>> + return output_fmts;
>>> +}
>>> +
>>> +/*
>>> + * Possible input formats :
>>> + * - MEDIA_BUS_FMT_RGB888_1X24
>>> + * - MEDIA_BUS_FMT_YUV8_1X24
>>> + * - MEDIA_BUS_FMT_UYVY8_1X16
>>> + * - MEDIA_BUS_FMT_UYYVYY8_0_5X24
>>> + * - MEDIA_BUS_FMT_RGB101010_1X30
>>> + * - MEDIA_BUS_FMT_YUV10_1X30
>>> + * - MEDIA_BUS_FMT_UYVY10_1X20
>>> + * - MEDIA_BUS_FMT_UYYVYY10_0_5X30
>>> + * - MEDIA_BUS_FMT_RGB121212_1X36
>>> + * - MEDIA_BUS_FMT_YUV12_1X36
>>> + * - MEDIA_BUS_FMT_UYVY12_1X24
>>> + * - MEDIA_BUS_FMT_UYYVYY12_0_5X36
>>> + * - MEDIA_BUS_FMT_RGB161616_1X48
>>> + * - MEDIA_BUS_FMT_YUV16_1X48
>>> + * - MEDIA_BUS_FMT_UYYVYY16_0_5X48
>>> + */
>>> +
>>> +/* Can return a maximum of 4 possible input formats for an output format */
>>> +#define MAX_INPUT_SEL_FORMATS 4
>>> +
>>> +static u32 *dw_hdmi_bridge_atomic_get_input_bus_fmts(struct drm_bridge
>>> *bridge, + struct
>> drm_bridge_state *bridge_state,
>>> + struct drm_crtc_state
>> *crtc_state,
>>> + struct
>> drm_connector_state *conn_state,
>>> + u32 output_fmt,
>>> + unsigned int
>> *num_input_fmts)
>>> +{
>>> + u32 *input_fmts;
>>> + int i = 0;
>>> +
>>> + *num_input_fmts = 0;
>>> +
>>> + input_fmts = kcalloc(MAX_INPUT_SEL_FORMATS, sizeof(*input_fmts),
>>> + GFP_KERNEL);
>>> + if (!input_fmts)
>>> + return NULL;
>>> +
>>> + switch (output_fmt) {
>>> + /* 8bit */
>>> + case MEDIA_BUS_FMT_RGB888_1X24:
>>> + input_fmts[i++] = MEDIA_BUS_FMT_RGB888_1X24;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_YUV8_1X24;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_UYVY8_1X16;
>>> + break;
>>> + case MEDIA_BUS_FMT_YUV8_1X24:
>>> + input_fmts[i++] = MEDIA_BUS_FMT_YUV8_1X24;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_UYVY8_1X16;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_RGB888_1X24;
>>> + break;
>>> + case MEDIA_BUS_FMT_UYVY8_1X16:
>>> + input_fmts[i++] = MEDIA_BUS_FMT_UYVY8_1X16;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_YUV8_1X24;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_RGB888_1X24;
>>> + break;
>>> +
>>> + /* 10bit */
>>> + case MEDIA_BUS_FMT_RGB101010_1X30:
>>> + input_fmts[i++] = MEDIA_BUS_FMT_RGB101010_1X30;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_YUV10_1X30;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_UYVY10_1X20;
>>> + break;
>>> + case MEDIA_BUS_FMT_YUV10_1X30:
>>> + input_fmts[i++] = MEDIA_BUS_FMT_YUV10_1X30;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_UYVY10_1X20;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_RGB101010_1X30;
>>> + break;
>>> + case MEDIA_BUS_FMT_UYVY10_1X20:
>>> + input_fmts[i++] = MEDIA_BUS_FMT_UYVY10_1X20;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_YUV10_1X30;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_RGB101010_1X30;
>>> + break;
>>> +
>>> + /* 12bit */
>>> + case MEDIA_BUS_FMT_RGB121212_1X36:
>>> + input_fmts[i++] = MEDIA_BUS_FMT_RGB121212_1X36;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_YUV12_1X36;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_UYVY12_1X24;
>>> + break;
>>> + case MEDIA_BUS_FMT_YUV12_1X36:
>>> + input_fmts[i++] = MEDIA_BUS_FMT_YUV12_1X36;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_UYVY12_1X24;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_RGB121212_1X36;
>>> + break;
>>> + case MEDIA_BUS_FMT_UYVY12_1X24:
>>> + input_fmts[i++] = MEDIA_BUS_FMT_UYVY12_1X24;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_YUV12_1X36;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_RGB121212_1X36;
>>> + break;
>>> +
>>> + /* 16bit */
>>> + case MEDIA_BUS_FMT_RGB161616_1X48:
>>> + input_fmts[i++] = MEDIA_BUS_FMT_RGB161616_1X48;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_YUV16_1X48;
>>> + break;
>>> + case MEDIA_BUS_FMT_YUV16_1X48:
>>> + input_fmts[i++] = MEDIA_BUS_FMT_YUV16_1X48;
>>> + input_fmts[i++] = MEDIA_BUS_FMT_RGB161616_1X48;
>>> + break;
>>> +
>>> + /* 420 */
>>> + case MEDIA_BUS_FMT_UYYVYY8_0_5X24:
>>> + case MEDIA_BUS_FMT_UYYVYY10_0_5X30:
>>> + case MEDIA_BUS_FMT_UYYVYY12_0_5X36:
>>> + case MEDIA_BUS_FMT_UYYVYY16_0_5X48:
>>> + input_fmts[i++] = output_fmt;
>>> + break;
>>> + }
>>> +
>>> + *num_input_fmts = i;
>>> +
>>> + if (*num_input_fmts == 0) {
>>> + kfree(input_fmts);
>>> + input_fmts = NULL;
>>> + }
>>> +
>>> + return input_fmts;
>>> +}
>>> +
>>> +static int dw_hdmi_bridge_atomic_check(struct drm_bridge *bridge,
>>> + struct drm_bridge_state
>> *bridge_state,
>>> + struct drm_crtc_state
>> *crtc_state,
>>> + struct drm_connector_state
>> *conn_state)
>>> +{
>>> + struct dw_hdmi *hdmi = bridge->driver_private;
>>> +
>>> + dev_dbg(hdmi->dev, "selected output format %x\n",
>>> + bridge_state->output_bus_cfg.format);
>>> +
>>> + hdmi->hdmi_data.enc_out_bus_format =
>>> + bridge_state->output_bus_cfg.format;
>>> +
>>> + dev_dbg(hdmi->dev, "selected input format %x\n",
>>> + bridge_state->input_bus_cfg.format);
>>> +
>>> + hdmi->hdmi_data.enc_in_bus_format =
>>> + bridge_state->input_bus_cfg.format;
>>> +
>>> + return 0;
>>> +}
>>> +
>>> static int dw_hdmi_bridge_attach(struct drm_bridge *bridge)
>>> {
>>> struct dw_hdmi *hdmi = bridge->driver_private;
>>> @@ -2499,6 +2759,9 @@ static const struct drm_bridge_funcs
>>> dw_hdmi_bridge_funcs = { .atomic_reset = drm_atomic_helper_bridge_reset,
>>> .attach = dw_hdmi_bridge_attach,
>>> .detach = dw_hdmi_bridge_detach,
>>> + .atomic_check = dw_hdmi_bridge_atomic_check,
>>> + .atomic_get_output_bus_fmts =
>> dw_hdmi_bridge_atomic_get_output_bus_fmts,
>>> + .atomic_get_input_bus_fmts =
>> dw_hdmi_bridge_atomic_get_input_bus_fmts,
>>> .enable = dw_hdmi_bridge_enable,
>>> .disable = dw_hdmi_bridge_disable,
>>> .mode_set = dw_hdmi_bridge_mode_set,
>>> @@ -2963,6 +3226,7 @@ __dw_hdmi_probe(struct platform_device *pdev,
>>>
>>> hdmi->bridge.driver_private = hdmi;
>>> hdmi->bridge.funcs = &dw_hdmi_bridge_funcs;
>>> +
>>> #ifdef CONFIG_OF
>>> hdmi->bridge.of_node = pdev->dev.of_node;
>>> #endif
>>
>>
>>
>>

Next message: Thomas Gleixner: "Re: [patch 4/8] x86/entry: Move irq tracing on syscall entry to C-code"
Previous message: Uwe Kleine-König: "Re: [PATCH V2] thermal: imx_thermal: Use __maybe_unused instead of CONFIG_PM_SLEEP"
In reply to: Neil Armstrong: "Re: [PATCH v4 04/11] drm/bridge: synopsys: dw-hdmi: add bus format negociation"
Next in thread: Laurent Pinchart: "Re: [PATCH v4 04/11] drm/bridge: synopsys: dw-hdmi: add bus format negociation"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]