Hi Maxime,
Thanks for your comments,
> Subject: [EXT] Re: [PATCH v15 4/8] drm: bridge: Cadence: Add MHDP8501
> DP/HDMI driver
>
> Hi,
>
> On Wed, Mar 06, 2024 at 11:16:21AM +0100, Alexander Stein wrote:
> > +static int cdns_mhdp8501_read_hpd(struct cdns_mhdp8501_device
> *mhdp)
> > +{
> > + u8 status;
> > + int ret;
> > +
> > + mutex_lock(&mhdp->mbox_mutex);
> > +
> > + ret = cdns_mhdp_mailbox_send(&mhdp->base,
> MB_MODULE_ID_GENERAL,
> > + GENERAL_GET_HPD_STATE, 0, NULL);
> > + if (ret)
> > + goto err_get_hpd;
> > +
> > + ret = cdns_mhdp_mailbox_recv_header(&mhdp->base,
> MB_MODULE_ID_GENERAL,
> > + GENERAL_GET_HPD_STATE,
> > + sizeof(status));
> > + if (ret)
> > + goto err_get_hpd;
> > +
> > + ret = cdns_mhdp_mailbox_recv_data(&mhdp->base, &status,
> sizeof(status));
> > + if (ret)
> > + goto err_get_hpd;
> > +
> > + mutex_unlock(&mhdp->mbox_mutex);
> > +
> > + return status;
> > +
> > +err_get_hpd:
> > + dev_err(mhdp->dev, "read hpd failed: %d\n", ret);
> > + mutex_unlock(&mhdp->mbox_mutex);
> > +
> > + return ret;
> > +}
> > +
> > +enum drm_connector_status cdns_mhdp8501_detect(struct
> > +cdns_mhdp8501_device *mhdp) {
> > + u8 hpd = 0xf;
> > +
> > + hpd = cdns_mhdp8501_read_hpd(mhdp);
> > + if (hpd == 1)
> > + return connector_status_connected;
> > + else if (hpd == 0)
> > + return connector_status_disconnected;
> > +
> > + dev_warn(mhdp->dev, "Unknown cable status, hdp=%u\n", hpd);
> > + return connector_status_unknown;
> > +}
> > +
> > +static void hotplug_work_func(struct work_struct *work) {
> > + struct cdns_mhdp8501_device *mhdp = container_of(work,
> > + struct cdns_mhdp8501_device,
> > + hotplug_work.work);
> > + enum drm_connector_status status = cdns_mhdp8501_detect(mhdp);
> > +
> > + drm_bridge_hpd_notify(&mhdp->bridge, status);
> > +
> > + if (status == connector_status_connected) {
> > + /* Cable connected */
> > + DRM_INFO("HDMI/DP Cable Plug In\n");
> > + enable_irq(mhdp->irq[IRQ_OUT]);
> > + } else if (status == connector_status_disconnected) {
> > + /* Cable Disconnected */
> > + DRM_INFO("HDMI/DP Cable Plug Out\n");
> > + enable_irq(mhdp->irq[IRQ_IN]);
> > + }
> > +}
> > +
> > +static irqreturn_t cdns_mhdp8501_irq_thread(int irq, void *data) {
> > + struct cdns_mhdp8501_device *mhdp = data;
> > +
> > + disable_irq_nosync(irq);
> > +
> > + mod_delayed_work(system_wq, &mhdp->hotplug_work,
> > + msecs_to_jiffies(HOTPLUG_DEBOUNCE_MS));
> > +
> > + return IRQ_HANDLED;
> > +}
>
> AFAICT from the rest of the driver, you support HDMI modes with a character
> rate > 340Mchar/s, and thus you need to enable the scrambler.
>
> If you unplug/replug a monitor with the scrambler enabled though, and if
> there's no userspace component to react to the userspace event, the code you
> have here won't enable the scrambler again.
>
> You can test that by using modetest with a 4k@60Hz mode or something, and
> then disconnecting / reconnecting the monitor.
>
> We addressed it in vc4 in commit 6bed2ea3cb38 ("drm/vc4: hdmi: Reset link
> on hotplug").
>
> Arguably, the whole handling of the HDMI scrambling setup should be turned
> into a helper, and I wanted to extend the work I've been doing around the
> HDMI infra to support the scrambler setup once it landed.
>
Yes, for userspace components that do not handle HPD events (such as modetest),
if they are connected to a 4K display and enable scramble then the cable is unplugged/plugged, HDMI will not work.
However, this is a userspace component limitation. fbdev and weston could work well for this case.
The patch for vc4 in commit 6bed2ea3cb38 ("drm/vc4: hdmi: Reset link on hotplug")
assumes unplug/replug the same monitor as stated in its commit log.
It does not support the case where unplug/plug to different displays.
For example, if the cable is unplugged from a 4K monitor and then plugged into a 1080p monitor,
4K video mode will be output to the 1080p monitor because this userspace component cannot respond to the monitor change.
Therefore, for userspace components that do not handle HPD events (such as modetest),
this patch can only partially solve the limitation, but it is not a perfect solution.
If helper functions are used to restore the HDMI scramble bit after hotplug, I would like to use it.
> > +static int cdns_hdmi_mode_config(struct cdns_mhdp8501_device *mhdp,
> > + struct drm_display_mode *mode,
> > + struct video_info *video_info)
> > +{
> > + int ret;
> > + u32 val;
> > + u32 vsync_lines = mode->vsync_end - mode->vsync_start;
> > + u32 eof_lines = mode->vsync_start - mode->vdisplay;
> > + u32 sof_lines = mode->vtotal - mode->vsync_end;
> > + u32 hblank = mode->htotal - mode->hdisplay;
> > + u32 hactive = mode->hdisplay;
> > + u32 vblank = mode->vtotal - mode->vdisplay;
> > + u32 vactive = mode->vdisplay;
> > + u32 hfront = mode->hsync_start - mode->hdisplay;
> > + u32 hback = mode->htotal - mode->hsync_end;
> > + u32 vfront = eof_lines;
> > + u32 hsync = hblank - hfront - hback;
> > + u32 vsync = vsync_lines;
> > + u32 vback = sof_lines;
> > + u32 v_h_polarity = ((mode->flags & DRM_MODE_FLAG_NHSYNC) ? 0 : 1)
> +
> > + ((mode->flags & DRM_MODE_FLAG_NVSYNC) ? 0 : 2);
> > +
> > + ret = cdns_mhdp_reg_write(&mhdp->base, SCHEDULER_H_SIZE, (hactive
> << 16) + hblank);
> > + if (ret < 0)
> > + return ret;
> > +
> > + ret = cdns_mhdp_reg_write(&mhdp->base, SCHEDULER_V_SIZE, (vactive
> << 16) + vblank);
> > + if (ret < 0)
> > + return ret;
> > +
> > + ret = cdns_mhdp_reg_write(&mhdp->base,
> HDTX_SIGNAL_FRONT_WIDTH, (vfront << 16) + hfront);
> > + if (ret < 0)
> > + return ret;
> > +
> > + ret = cdns_mhdp_reg_write(&mhdp->base, HDTX_SIGNAL_SYNC_WIDTH,
> (vsync << 16) + hsync);
> > + if (ret < 0)
> > + return ret;
> > +
> > + ret = cdns_mhdp_reg_write(&mhdp->base, HDTX_SIGNAL_BACK_WIDTH,
> (vback << 16) + hback);
> > + if (ret < 0)
> > + return ret;
> > +
> > + ret = cdns_mhdp_reg_write(&mhdp->base, HSYNC2VSYNC_POL_CTRL,
> v_h_polarity);
> > + if (ret < 0)
> > + return ret;
> > +
> > + /* Reset Data Enable */
> > + cdns_mhdp_reg_read(&mhdp->base, HDTX_CONTROLLER, &val);
> > + val &= ~F_DATA_EN(1);
> > + ret = cdns_mhdp_reg_write(&mhdp->base, HDTX_CONTROLLER, val);
> > + if (ret < 0)
> > + return ret;
> > +
> > + /* Set bpc */
> > + val &= ~F_VIF_DATA_WIDTH(3);
> > + switch (video_info->bpc) {
> > + case 10:
> > + val |= F_VIF_DATA_WIDTH(1);
> > + break;
> > + case 12:
> > + val |= F_VIF_DATA_WIDTH(2);
> > + break;
> > + case 16:
> > + val |= F_VIF_DATA_WIDTH(3);
> > + break;
> > + case 8:
> > + default:
> > + val |= F_VIF_DATA_WIDTH(0);
> > + break;
> > + }
> > +
> > + /* select color encoding */
> > + val &= ~F_HDMI_ENCODING(3);
> > + switch (video_info->color_fmt) {
> > + case DRM_COLOR_FORMAT_YCBCR444:
> > + val |= F_HDMI_ENCODING(2);
> > + break;
> > + case DRM_COLOR_FORMAT_YCBCR422:
> > + val |= F_HDMI_ENCODING(1);
> > + break;
> > + case DRM_COLOR_FORMAT_YCBCR420:
> > + val |= F_HDMI_ENCODING(3);
> > + break;
> > + case DRM_COLOR_FORMAT_RGB444:
> > + default:
> > + val |= F_HDMI_ENCODING(0);
> > + break;
> > + }
>
> It looks like you're only using RGB444?
The limitation will be removed in the next version, RGB444, YUV444, YUV422 and YUV420 could be supported by HDMI driver.
>
> > +static int cdns_hdmi_avi_info_set(struct cdns_mhdp8501_device *mhdp,
> > + struct drm_display_mode *mode)
> > +{
> > + struct hdmi_avi_infoframe frame;
> > + struct drm_connector_state *conn_state = mhdp->curr_conn->state;
> > + struct drm_display_mode *adj_mode;
> > + enum hdmi_quantization_range qr;
> > + u8 buf[32];
> > + int ret;
> > +
> > + /* Initialise info frame from DRM mode */
> > + drm_hdmi_avi_infoframe_from_display_mode(&frame,
> mhdp->curr_conn,
> > +mode);
> > +
> > + frame.colorspace =
> cdns_hdmi_colorspace(mhdp->video_info.color_fmt);
> > +
> > + drm_hdmi_avi_infoframe_colorimetry(&frame, conn_state);
> > +
> > + adj_mode = &mhdp->bridge.encoder->crtc->state->adjusted_mode;
> > +
> > + qr = drm_default_rgb_quant_range(adj_mode);
> > +
> > + drm_hdmi_avi_infoframe_quant_range(&frame, mhdp->curr_conn,
> > + adj_mode, qr);
> > +
> > + ret = hdmi_avi_infoframe_pack(&frame, buf + 1, sizeof(buf) - 1);
> > + if (ret < 0) {
> > + dev_err(mhdp->dev, "failed to pack AVI infoframe: %d\n", ret);
> > + return -1;
> > + }
> > +
> > + buf[0] = 0;
> > + cdns_hdmi_infoframe_set(mhdp, 0, sizeof(buf), buf,
> > +HDMI_INFOFRAME_TYPE_AVI);
> > +
> > + return 0;
> > +}
> > +
> > +static void cdns_hdmi_vendor_info_set(struct cdns_mhdp8501_device
> *mhdp,
> > + struct drm_display_mode *mode) {
> > + struct hdmi_vendor_infoframe frame;
> > + u8 buf[32];
> > + int ret;
> > +
> > + /* Initialise vendor frame from DRM mode */
> > + ret = drm_hdmi_vendor_infoframe_from_display_mode(&frame,
> mhdp->curr_conn, mode);
> > + if (ret < 0) {
> > + dev_info(mhdp->dev, "No vendor infoframe\n");
> > + return;
> > + }
> > +
> > + ret = hdmi_vendor_infoframe_pack(&frame, buf + 1, sizeof(buf) - 1);
> > + if (ret < 0) {
> > + dev_warn(mhdp->dev, "Unable to pack vendor infoframe: %d\n",
> ret);
> > + return;
> > + }
> > +
> > + buf[0] = 0;
> > + cdns_hdmi_infoframe_set(mhdp, 3, sizeof(buf), buf,
> > +HDMI_INFOFRAME_TYPE_VENDOR); }
> > +
> > +static void cdns_hdmi_drm_info_set(struct cdns_mhdp8501_device
> *mhdp)
> > +{
> > + struct drm_connector_state *conn_state;
> > + struct hdmi_drm_infoframe frame;
> > + u8 buf[32];
> > + int ret;
> > +
> > + conn_state = mhdp->curr_conn->state;
> > +
> > + if (!conn_state->hdr_output_metadata)
> > + return;
> > +
> > + ret = drm_hdmi_infoframe_set_hdr_metadata(&frame, conn_state);
> > + if (ret < 0) {
> > + dev_dbg(mhdp->dev, "couldn't set HDR metadata in infoframe\n");
> > + return;
> > + }
> > +
> > + ret = hdmi_drm_infoframe_pack(&frame, buf + 1, sizeof(buf) - 1);
> > + if (ret < 0) {
> > + dev_dbg(mhdp->dev, "couldn't pack HDR infoframe\n");
> > + return;
> > + }
> > +
> > + buf[0] = 0;
> > + cdns_hdmi_infoframe_set(mhdp, 3, sizeof(buf), buf,
> > +HDMI_INFOFRAME_TYPE_DRM); }
>
> This is another thing that was supposed to be handled by my current series
> here:
> https://lore.kernel.org/r/20240222-kms-hdmi-connector-state-v7-0-8f4af575f
> ce2@xxxxxxxxxx
I will use the hdmi helper function hdmi_clear_infoframe() and hdmi_write_infoframe()
from Dmitry's patch in the next version.
>
> In particular, you're missing the SPD infoframes here.
>
I will add SPD support in the next version.
> > +static void cdns_hdmi_mode_set(struct cdns_mhdp8501_device *mhdp) {
>
> You should use atomic_enable here
Actually, this mode_set() function is called by atomic_enable(),
I will merge the function into atomic_enable() in the next version.
>
> > + struct drm_display_mode *mode = &mhdp->mode;
> > + union phy_configure_opts phy_cfg;
> > + int ret;
> > +
> > + /* video mode check */
> > + if (mode->clock == 0 || mode->hdisplay == 0 || mode->vdisplay == 0)
> > + return;
> > +
> > + cdns_hdmi_lanes_config(mhdp);
> > +
> > + phy_cfg.hdmi.pixel_clk_rate = mode->clock;
> > + phy_cfg.hdmi.bpc = mhdp->video_info.bpc;
> > + phy_cfg.hdmi.color_space =
> > +cdns_hdmi_colorspace(mhdp->video_info.color_fmt);
> > +
> > + /* Mailbox protect for HDMI PHY access */
> > + mutex_lock(&mhdp->mbox_mutex);
> > + ret = phy_configure(mhdp->phy, &phy_cfg);
> > + mutex_unlock(&mhdp->mbox_mutex);
> > + if (ret) {
> > + dev_err(mhdp->dev, "%s: phy_configure() failed: %d\n",
> > + __func__, ret);
> > + return;
> > + }
> > +
> > + cdns_hdmi_sink_config(mhdp);
> > +
> > + ret = cdns_hdmi_ctrl_init(mhdp, mhdp->hdmi.hdmi_type);
> > + if (ret < 0) {
> > + dev_err(mhdp->dev, "%s, ret = %d\n", __func__, ret);
> > + return;
> > + }
> > +
> > + /* Config GCP */
> > + if (mhdp->video_info.bpc == 8)
> > + cdns_hdmi_disable_gcp(mhdp);
> > + else
> > + cdns_hdmi_enable_gcp(mhdp);
> > +
> > + ret = cdns_hdmi_avi_info_set(mhdp, mode);
> > + if (ret < 0) {
> > + dev_err(mhdp->dev, "%s ret = %d\n", __func__, ret);
> > + return;
> > + }
> > +
> > + /* vendor info frame is enabled only for HDMI1.4 4K mode */
> > + cdns_hdmi_vendor_info_set(mhdp, mode);
> > +
> > + cdns_hdmi_drm_info_set(mhdp);
> > +
> > + ret = cdns_hdmi_mode_config(mhdp, mode, &mhdp->video_info);
> > + if (ret < 0) {
> > + dev_err(mhdp->dev, "CDN_API_HDMITX_SetVic_blocking ret
> = %d\n", ret);
> > + return;
> > + }
> > +}
> > +
> > +static int cdns_hdmi_bridge_attach(struct drm_bridge *bridge,
> > + enum drm_bridge_attach_flags flags) {
> > + struct cdns_mhdp8501_device *mhdp = bridge->driver_private;
> > +
> > + if (!(flags & DRM_BRIDGE_ATTACH_NO_CONNECTOR)) {
> > + dev_err(mhdp->dev, "do not support creating a drm_connector\n");
> > + return -EINVAL;
> > + }
> > +
> > + return 0;
> > +}
> > +
> > +static enum drm_mode_status
> > +cdns_hdmi_bridge_mode_valid(struct drm_bridge *bridge,
> > + const struct drm_display_info *info,
> > + const struct drm_display_mode *mode) {
> > + struct cdns_mhdp8501_device *mhdp = bridge->driver_private;
> > + enum drm_mode_status mode_status = MODE_OK;
> > + union phy_configure_opts phy_cfg;
> > + int ret;
> > +
> > + /* We don't support double-clocked and Interlaced modes */
> > + if (mode->flags & DRM_MODE_FLAG_DBLCLK ||
> > + mode->flags & DRM_MODE_FLAG_INTERLACE)
> > + return MODE_BAD;
> > +
> > + /* MAX support pixel clock rate 594MHz */
> > + if (mode->clock > 594000)
> > + return MODE_CLOCK_HIGH;
> > +
> > + if (mode->hdisplay > 3840)
> > + return MODE_BAD_HVALUE;
> > +
> > + if (mode->vdisplay > 2160)
> > + return MODE_BAD_VVALUE;
> > +
> > + /* Check modes supported by PHY */
> > + phy_cfg.hdmi.pixel_clk_rate = mode->clock;
> > +
> > + /* Mailbox protect for HDMI PHY access */
> > + mutex_lock(&mhdp->mbox_mutex);
> > + ret = phy_validate(mhdp->phy, PHY_MODE_HDMI, 0, &phy_cfg);
> > + mutex_unlock(&mhdp->mbox_mutex);
> > + if (ret < 0)
> > + return MODE_CLOCK_RANGE;
> > +
> > + return mode_status;
> > +}
>
> mode_valid is only called when the userspace asks for the current modes
> available on a connector, but not when the userspace programs a mode.
> This is atomic_check's job, so you probably want to have a similar
> atomic_check here.
>
atomic_check() and hdmi_tmds_char_rate_valid() will be added in the next version.
> > +static bool
> > +cdns_hdmi_bridge_mode_fixup(struct drm_bridge *bridge,
> > + const struct drm_display_mode *mode,
> > + struct drm_display_mode *adjusted_mode) {
> > + struct cdns_mhdp8501_device *mhdp = bridge->driver_private;
> > + struct video_info *video = &mhdp->video_info;
> > +
> > + /* The only currently supported format */
> > + video->bpc = 8;
> > + video->color_fmt = DRM_COLOR_FORMAT_RGB444;
> > +
> > + return true;
> > +}
>
> mode_fixup is deprecated. I guess you wanted to use reset here.
> Generally speaking, the bpc and color_format should be part of your state
> anyway.
>
mode_fixup() will be removed in the next version.
> I know Dmitry was interested in plugging my HDMI series into the bridge
> infrastructure. It would benefit your driver quite a lot I think, so you probably
> want to sync up with him.
>
Thanks, the driver will use the new HDMI helper functions that Dmitry added in the patch
'make use of the HDMI connector infrastructure'.
B.R
Sandor
> Maxime
