[PATCH v10 00/18] drm: Add Samsung MIPI DSIM bridge

[Alexander Stein]

Hi Marek,

Am Freitag, 16. Dezember 2022, 13:58:13 CET schrieb Marek Vasut:
> On 12/16/22 13:41, Alexander Stein wrote:
> > Hi,
> 
> Hi,
> 
> > thanks for the new update.
> > 
> > Am Mittwoch, 14. Dezember 2022, 13:58:49 CET schrieb Jagan Teki:
> >> This series supports common bridge support for Samsung MIPI DSIM
> >> which is used in Exynos and i.MX8MM SoC's.
> >> 
> >> The final bridge supports both the Exynos and i.MX8M Mini/Nano/Plus.
> >> 
> >> Patch 0001 - 0004: adding devm_drm_of_dsi_get_bridge
> >> 
> >> Patch 0005 - 0006: optional PHY, PMS_P offset
> >> 
> >> Patch 0007       : introduce hw_type
> >> 
> >> Patch 0008	 : fixing host init
> >> 
> >> Patch 0009	 : atomic_check
> >> 
> >> Patch 0010	 : input_bus_flags
> >> 
> >> Patch 0011	 : atomic_get_input_bus_fmts
> >> 
> >> Patch 0012 - 0013: component vs bridge
> >> 
> >> Patch 0014	 : DSIM bridge
> >> 
> >> Patch 0015 - 0016: i.MX8M Mini/Nano
> >> 
> >> Patch 0017 - 0018: i.MX8M Plus
> > 
> > Tested using LVDS bridges and works on:
> > * imx8mp-tqma8mpql-mba8mpxl (i.MX8M Plus)
> > * imx8mm-tqma8mqml-mba8mx (i.MX8M Mini)
> > * imx8mn-tqma8mqnl-mba8mx (i.MX8M Nano)
> > 
> > But I noticed that tqma8mqml and tqma8mqnl only have a stable output if I
> > specify 'data-lanes = <0 1 2>' (not <0 1 2 3>!) on the DSI-LVDS-Bridge (TI
> > SN65DSI83) input endpoint, e.g.
> > 
> > dsi_lvds_bridge: bridge at 2d {
> > 
> > 	compatible = "ti,sn65dsi83";
> > 	reg = <0x2d>;
> > 	enable-gpios = <&gpio_delays 0 130000 0>;
> > 	vcc-supply = <&reg_sn65dsi83_1v8>;
> > 	status = "okay";
> > 	ports {
> > 	
> > 		#address-cells = <1>;
> > 		#size-cells = <0>;
> > 		port at 0 {
> > 		
> > 			reg = <0>;
> > 			lvds_bridge_in: endpoint {
> > 			
> > 				data-lanes = <0 1 2>;
> > 				remote-endpoint = <&mipi_dsi_out>;
> > 			
> > 			};
> > 		
> > 		};
> > 		port at 2 {
> > 		
> > 			reg = <2>;
> > 			lvds_bridge_out: endpoint {
> > 			
> > 				remote-endpoint = <&panel_in_lvds0>;
> > 			
> > 			};
> > 		
> > 		};
> > 	
> > 	};
> > 
> > };
> > 
> > I have no idea if this is on the DSIM side or the DSI-LVDS bridge side.
> > Register 0xe5 on sn65dsi83 shows that CHA_LPP_ERR and CHA_SOT_BIT_ERR
> > error
> > bits are set, indicating there are low level protocol and SoT leader
> > sequence bit errors. I can't find any significant difference in sn65dsi83
> > driver compared to downstream-kernel, so I would assume something in DSIM
> > driver is different.
> 
> Make sure the HS clock rate expected by the DSI83 bridge chip and the
> one generated by the DSIM bridge from PMS PLL match. There is currently
> no negotiation of this frequency and both sides divine it on their own.
> 
> See sn65dsi83_get_dsi_range() and samsung_dsim_set_pll().

Oh, nice, thanks for the pointer. When setting
> samsung,burst-clock-frequency = <668250000>;
in imx8mm.dtsi
I get a non-flickering display using 4 lanes. Although admittedly this is just 
random guessing. I'm not sure which clock exactly has to be in the range 
CHA_DSI_CLK_RANGE is configured to. With 4 lanes SN65DSI84 is configured for 
205-210 MHz (0x29), while I get these PLL PMS settings on DSIM:
> samsung-dsim 32e10000.dsi: PLL freq 668250000, (p 4, m 99, s 0)
> samsung-dsim 32e10000.dsi: hs_clk = 668250000, byte_clk = 83531250, esc_clk 
= 16706250

AFAICS DSIM bridge is configurung hs_clk, byte_clk and esc_clk just from DT 
properties, while SN65DSI84 is using display mode and number of lanes.

Is it expected that the DSIM PLL frequencies are set in DT for a specific 
bridge/display setup?

Best regards,
Alexander