Re: [PATCH] drm/bridge/synopsys: dsi: use adjusted_mode in mode_set

[Laurent Pinchart]

Hi Philippe,

On Saturday, 3 February 2018 00:41:26 EET Philippe CORNU wrote:
> On 01/29/2018 11:40 AM, Laurent Pinchart wrote:
> > On Monday, 29 January 2018 12:17:37 EET Philippe CORNU wrote:
> >> On 01/29/2018 10:46 AM, Laurent Pinchart wrote:
> >>> On Thursday, 25 January 2018 17:55:04 EET Philippe Cornu wrote:
> >>>> The "adjusted_mode" clock value (ie the real pixel clock) is more
> >>>> accurate than "mode" clock value (ie the panel/bridge requested
> >>>> clock value). It offers a better preciseness for timing
> >>>> computations and allows to reduce the extra dsi bandwidth in
> >>>> burst mode (from ~20% to ~10-12%, hw platform dependant).
> >>>>
> >>>> Signed-off-by: Philippe Cornu <philippe.cornu@xxxxxx>
> >>>
> >>> The adjusted mode is documented as
> >>>
> >>>   /**
> >>>    * @adjusted_mode:
> >>>    *
> >>>    * Internal display timings which can be used by the driver to handle
> >>>    * differences between the mode requested by userspace in @mode and
> >>>    what
> >>>    * is actually programmed into the hardware. It is purely driver
> >>>    * implementation defined what exactly this adjusted mode means.
> >>>    Usually
> >>>    * it is used to store the hardware display timings used between the
> >>>    * CRTC and encoder blocks.
> >>>    */
> >>>
> >>> This is easy to handle when the CRTC and encoder are controlled by the
> >>> same driver, as the field is "implementation defined" by a single
> >>> driver. However, when using bridges, there are two drivers involved, and
> >>> they must both agree to meaningfully use the adjusted mode. I can't see
> >>> how to do so without standardizing the meaning of the adjusted mode
> >>> field.
> >>
> >> This is exactly the reason why my first implementation used the dsi
> >> bridge "optional pixel clock" instead of the adjusted_mode (see [1])
> >>
> >> But after digging more into the drm source code, I think using
> >> adjusted_mode instead of the pixel clock here brings more advantages
> >> because:
> >> * adjusted_mode is an argument of bridge mode_set() probably for being
> >> used in any manner, maybe like this :)
> >> * if the bridge "user" (crtc or a master bridge drivers) does not need
> >> to modify its adjusted_mode then mode & adjusted_mode mode_set()
> >> arguments will have the same values so "no consequence" for the bridge.
> >> * if the bridge "user" (crtc or master bridge drivers) needs to adjust
> >> any value of the mode then this adjustment is available for the bridge.
> > 
> > Remember that there can be multiple chained bridges, and a single
> > adjusted mode field.
> > 
> >> * rockchip crtc updates a part of the mode (the clock), stm is doing the
> >> same (see [2]) but any future "user" of the dw_mipi_dsi bridge can
> >> adjust something else (blankings...) and the dw_mipi_dsi bridge will be
> >> then aware of...
> >>
> >> But maybe it is a wrong usage of the "adjusted_mode offer"...
> > 
> > I don't disagree that there's a need for using adjusted values, but I
> > believe we need to create a clear API to do so. Using the adjust_mode
> > field as-is when it's clearly documented as being implementation-defined
> > is asking for trouble. 
> 
> Laurent, do you think we can use "adjusted mode" here in this small 
> patch as the actual 2 "users" of this bridge (rockchip & stm) use both 
> "adjusted mode" in their crtc?

The Synopsys DSI driver might only be used by two display controller drivers 
today, it is nonetheless a standard bridge driver that should not make any 
assumption of a particular use of the adjusted_mode in particular display 
controller drivers. To use the adjusted_mode value in bridge drivers we need 
to standardize its usage, otherwise we'll end up with incompatibilities 
between bridge drivers and display controller drivers.

> >> [1] https://patchwork.freedesktop.org/patch/200240/
> >> [2] https://patchwork.freedesktop.org/patch/200720/
> >>
> >>> Daniel, what's your opinion on this ?
> 
> Daniel, any opinion on the adjusted_mode usage?
> 
> >>>> ---
> >>>> Note: This patch replaces "drm/bridge/synopsys: dsi: add optional
> >>>> pixel clock"
> >>>>
> >>>> drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c | 12 ++++++------
> >>>> 1 file changed, 6 insertions(+), 6 deletions(-)
> >>>>
> >>>> diff --git a/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c
> >>>> b/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c index
> >>>> ed8af32f8e52..b926b62e9e33 100644
> >>>> --- a/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c
> >>>> +++ b/drivers/gpu/drm/bridge/synopsys/dw-mipi-dsi.c
> >>>> @@ -707,20 +707,20 @@ static void dw_mipi_dsi_bridge_mode_set(struct
> >>>> drm_bridge *bridge,
> >>>>
> >>>>    	clk_prepare_enable(dsi->pclk);
> >>>>
> >>>> -	ret = phy_ops->get_lane_mbps(priv_data, mode, dsi->mode_flags,
> >>>> +	ret = phy_ops->get_lane_mbps(priv_data, adjusted_mode,
> >>>> dsi->mode_flags,
> >>>>    				     dsi->lanes, dsi->format, &dsi->lane_mbps);
> >>>>    if (ret)
> >>>>    	DRM_DEBUG_DRIVER("Phy get_lane_mbps() failed\n");
> >>>>
> >>>>    pm_runtime_get_sync(dsi->dev);
> >>>>    dw_mipi_dsi_init(dsi);
> >>>>
> >>>> -	dw_mipi_dsi_dpi_config(dsi, mode);
> >>>> +	dw_mipi_dsi_dpi_config(dsi, adjusted_mode);
> >>>>    dw_mipi_dsi_packet_handler_config(dsi);
> >>>>    dw_mipi_dsi_video_mode_config(dsi);
> >>>> -	dw_mipi_dsi_video_packet_config(dsi, mode);
> >>>> +	dw_mipi_dsi_video_packet_config(dsi, adjusted_mode);
> >>>>    dw_mipi_dsi_command_mode_config(dsi);
> >>>> -	dw_mipi_dsi_line_timer_config(dsi, mode);
> >>>> -	dw_mipi_dsi_vertical_timing_config(dsi, mode);
> >>>> +	dw_mipi_dsi_line_timer_config(dsi, adjusted_mode);
> >>>> +	dw_mipi_dsi_vertical_timing_config(dsi, adjusted_mode);
> >>>>    dw_mipi_dsi_dphy_init(dsi);
> >>>>    dw_mipi_dsi_dphy_timing_config(dsi);
> >>>>
> >>>> @@ -734,7 +734,7 @@ static void dw_mipi_dsi_bridge_mode_set(struct
> >>>> drm_bridge *bridge,
> >>>>
> >>>>    dw_mipi_dsi_dphy_enable(dsi);
> >>>>
> >>>> -	dw_mipi_dsi_wait_for_two_frames(mode);
> >>>> +	dw_mipi_dsi_wait_for_two_frames(adjusted_mode);
> >>>>
> >>>>    /* Switch to cmd mode for panel-bridge pre_enable & panel prepare
> >>>>    */
> >>>>    dw_mipi_dsi_set_mode(dsi, 0);

-- 
Regards,

Laurent Pinchart