Re: [PATCH v10 3/3] drm/mediatek: Implement OF graphs support for display paths

[AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>]

Il 07/10/24 08:57, CK Hu (胡俊光) ha scritto:
> Hi, Angelo:
> 
> On Fri, 2024-10-04 at 12:22 +0200, AngeloGioacchino Del Regno wrote:
>> Il 04/10/24 08:03, CK Hu (胡俊光) ha scritto:
>>> Hi, Angelo:
>>>
>>> On Tue, 2024-10-01 at 13:33 +0200, AngeloGioacchino Del Regno wrote:
>>>> Il 01/10/24 12:07, CK Hu (胡俊光) ha scritto:
>>>>> Hi, Angelo:
>>>>>
>>>>> On Tue, 2024-09-10 at 10:51 +0000, AngeloGioacchino Del Regno wrote:
>>>>>> It is impossible to add each and every possible DDP path combination
>>>>>> for each and every possible combination of SoC and board: right now,
>>>>>> this driver hardcodes configuration for 10 SoCs and this is going to
>>>>>> grow larger and larger, and with new hacks like the introduction of
>>>>>> mtk_drm_route which is anyway not enough for all final routes as the
>>>>>> DSI cannot be connected to MERGE if it's not a dual-DSI, or enabling
>>>>>> DSC preventively doesn't work if the display doesn't support it, or
>>>>>> others.
>>>>>>
>>>>>> Since practically all display IPs in MediaTek SoCs support being
>>>>>> interconnected with different instances of other, or the same, IPs
>>>>>> or with different IPs and in different combinations, the final DDP
>>>>>> pipeline is effectively a board specific configuration.
>>>>>>
>>>>>> Implement OF graphs support to the mediatek-drm drivers, allowing to
>>>>>> stop hardcoding the paths, and preventing this driver to get a huge
>>>>>> amount of arrays for each board and SoC combination, also paving the
>>>>>> way to share the same mtk_mmsys_driver_data between multiple SoCs,
>>>>>> making it more straightforward to add support for new chips.
>>>>>>
>>>>>> Reviewed-by: Alexandre Mergnat <amergnat@baylibre.com>
>>>>>> Tested-by: Alexandre Mergnat <amergnat@baylibre.com>
>>>>>> Acked-by: Sui Jingfeng <sui.jingfeng@linux.dev>
>>>>>> Tested-by: Michael Walle <mwalle@kernel.org> # on kontron-sbc-i1200
>>>>>> Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
>>>>>> ---
>>>>>
>>>>> [snip]
>>>>>
>>>>>> +
>>>>>> +bool mtk_ovl_adaptor_is_comp_present(struct device_node *node)
>>>>>> +{
>>>>>> +	enum mtk_ovl_adaptor_comp_type type;
>>>>>> +	int ret;
>>>>>> +
>>>>>> +	ret = ovl_adaptor_of_get_ddp_comp_type(node, &type);
>>>>>> +	if (ret)
>>>>>> +		return false;
>>>>>> +
>>>>>> +	if (type >= OVL_ADAPTOR_TYPE_NUM)
>>>>>> +		return false;
>>>>>> +
>>>>>> +	/*
>>>>>> +	 * ETHDR and Padding are used exclusively in OVL Adaptor: if this
>>>>>> +	 * component is not one of those, it's likely not an OVL Adaptor path.
>>>>>> +	 */
>>>>>
>>>>> I don't know your logic here.
>>>>> The OVL Adaptor pipeline is:
>>>>>
>>>>> mdp_rdma -> padding ---+      +-------+
>>>>>                         Merge -> |       |
>>>>> mdp_rdma -> padding ---+      |       |
>>>>>                                  |       |
>>>>> mdp_rdma -> padding ---+      |       |
>>>>>                         Merge -> |       |
>>>>> mdp_rdma -> padding ---+      |       |
>>>>>                                  | ETHDR |
>>>>> mdp_rdma -> padding ---+      |       |
>>>>>                         Merge -> |       |
>>>>> mdp_rdma -> padding ---+      |       |
>>>>>                                  |       |
>>>>> mdp_rdma -> padding ---+      |       |
>>>>>                         Merge -> |       |
>>>>> mdp_rdma -> padding ---+      +-------+
>>>>>
>>>>> So mdp_rdma and merge is not OVL Adaptor?
>>>>>
>>>>
>>>> Yes, and in device tree, you express that exactly like you just pictured.
>>>>
>>>> OVL Adaptor is treated like a software component internally, and manages
>>>> its own merge pipes exactly like before this commit.
>>>>
>>>> In case there will be any need to express OVL Adaptor as hardware component,
>>>> it will be possible to do so with no modification to the bindings.
>>>>
>>>>>
>>>>>> +	return type == OVL_ADAPTOR_TYPE_ETHDR || type == OVL_ADAPTOR_TYPE_PADDING;
>>>>>> +}
>>>>>> +
>>>>>>     
>>>>>
>>>>> [snip]
>>>>>
>>>>>> +
>>>>>> +/**
>>>>>> + * mtk_drm_of_ddp_path_build_one - Build a Display HW Pipeline for a CRTC Path
>>>>>> + * @dev:          The mediatek-drm device
>>>>>> + * @cpath:        CRTC Path relative to a VDO or MMSYS
>>>>>> + * @out_path:     Pointer to an array that will contain the new pipeline
>>>>>> + * @out_path_len: Number of entries in the pipeline array
>>>>>> + *
>>>>>> + * MediaTek SoCs can use different DDP hardware pipelines (or paths) depending
>>>>>> + * on the board-specific desired display configuration; this function walks
>>>>>> + * through all of the output endpoints starting from a VDO or MMSYS hardware
>>>>>> + * instance and builds the right pipeline as specified in device trees.
>>>>>> + *
>>>>>> + * Return:
>>>>>> + * * %0       - Display HW Pipeline successfully built and validated
>>>>>> + * * %-ENOENT - Display pipeline was not specified in device tree
>>>>>> + * * %-EINVAL - Display pipeline built but validation failed
>>>>>> + * * %-ENOMEM - Failure to allocate pipeline array to pass to the caller
>>>>>> + */
>>>>>> +static int mtk_drm_of_ddp_path_build_one(struct device *dev, enum mtk_crtc_path cpath,
>>>>>> +					 const unsigned int **out_path,
>>>>>> +					 unsigned int *out_path_len)
>>>>>> +{
>>>>>> +	struct device_node *next, *prev, *vdo = dev->parent->of_node;
>>>>>> +	unsigned int temp_path[DDP_COMPONENT_DRM_ID_MAX] = { 0 };
>>>>>> +	unsigned int *final_ddp_path;
>>>>>> +	unsigned short int idx = 0;
>>>>>> +	bool ovl_adaptor_comp_added = false;
>>>>>> +	int ret;
>>>>>> +
>>>>>> +	/* Get the first entry for the temp_path array */
>>>>>> +	ret = mtk_drm_of_get_ddp_ep_cid(vdo, 0, cpath, &next, &temp_path[idx]);
>>>>>> +	if (ret) {
>>>>>> +		if (next && temp_path[idx] == DDP_COMPONENT_DRM_OVL_ADAPTOR) {
>>>>>
>>>>> mdp_rdma would not be DDP_COMPONENT_DRM_OVL_ADAPTOR.
>>>>
>>>> This piece of code just avoids adding OVL_ADAPTOR more than once to the pipeline.
>>>>
>>>>>
>>>>>> +			dev_dbg(dev, "Adding OVL Adaptor for %pOF\n", next);
>>>>>> +			ovl_adaptor_comp_added = true;
>>>>>> +		} else {
>>>>>> +			if (next)
>>>>>> +				dev_err(dev, "Invalid component %pOF\n", next);
>>>>>> +			else
>>>>>> +				dev_err(dev, "Cannot find first endpoint for path %d\n", cpath);
>>>>>> +
>>>>>> +			return ret;
>>>>>> +		}
>>>>>> +	}
>>>>>> +	idx++;
>>>>>> +
>>>>>> +	/*
>>>>>> +	 * Walk through port outputs until we reach the last valid mediatek-drm component.
>>>>>> +	 * To be valid, this must end with an "invalid" component that is a display node.
>>>>>> +	 */
>>>>>> +	do {
>>>>>> +		prev = next;
>>>>>> +		ret = mtk_drm_of_get_ddp_ep_cid(next, 1, cpath, &next, &temp_path[idx]);
>>>>>> +		of_node_put(prev);
>>>>>> +		if (ret) {
>>>>>> +			of_node_put(next);
>>>>>> +			break;
>>>>>> +		}
>>>>>> +
>>>>>> +		/*
>>>>>> +		 * If this is an OVL adaptor exclusive component and one of those
>>>>>> +		 * was already added, don't add another instance of the generic
>>>>>> +		 * DDP_COMPONENT_OVL_ADAPTOR, as this is used only to decide whether
>>>>>> +		 * to probe that component master driver of which only one instance
>>>>>> +		 * is needed and possible.
>>>>>> +		 */
>>>>>> +		if (temp_path[idx] == DDP_COMPONENT_DRM_OVL_ADAPTOR) {
>>>>>
>>>>> merge would not be DDP_COMPONENT_DRM_OVL_ADAPTOR.
>>>>> Finally, the path would be:
>>>>>
>>>>> mdp_rdma -> ovl adaptor (padding) -> merge -> (ethdr is skipped here) ...
>>>>>
>>>>
>>>> Again, the path in the OF graph is expressed exactly like you said.
>>>
>>> I know the OF graph is expressed like I said.
>>> But I care about the path in driver should like this:
>>
>> Ok, now I understand your concern.
>>
>>>
>>> static const unsigned int mt8195_mtk_ddp_ext[] = {
>>>           DDP_COMPONENT_DRM_OVL_ADAPTOR,
>>>           DDP_COMPONENT_MERGE5,
>>>           DDP_COMPONENT_DP_INTF1,
>>> };
>>>
>>> In OF graph, the first component is mdp_rdma and mtk_ovl_adaptor_is_comp_present() would return false for mdp_rdma.
>>> So I think this would make mtk_drm_of_ddp_path_build_one() return error and the path is not created.
>>> If I'm wrong, please explain how this code would result in the path like mt8195_mtk_ddp_ext[].
>>>
>>
>> The MDP_RDMA usage in mtk_disp_ovl_adaptor is hardcoded: in function
>> mtk_ovl_adaptor_layer_config(), the rdma_l/r are always OVL_ADAPTOR_MDP_RDMAx,
>> then function mtk_ovl_adaptor_dma_dev_get(), always returns the MDP_RDMA0
>> component, same for mtk_ovl_adaptor_get_{num_formats,formats}() which always
>> call mtk_mdp_rdma_get_formats() for OVL_ADAPTOR_MDP_RDMA0.
>>
>> I have just rechecked how I expressed the path for MT8195 Tomato, where the
>> external display works with OF Graphs, and it was missing MDP_RDMA entirely.
>>
>> The path was ethdr -> merge -> dp_intf1 ... and it should be mdp_rdma -> (etc).
>>
>> Effectively, that is indeed wrong, as all of the steps must be expressed
>> inside of the graph.
>>
>> Since the OVL Adaptor's RDMA instances' compatible strings do *not* collide
>> with the others, as OVL Adaptor uses compatible mediatek,mt8195-vdo1-rdma,
>> and the regular one uses compatible mediatek,mt8195-disp-rdma, we can resolve
>> this issue by changing function mtk_ovl_adaptor_is_comp_present()
>>
>> from
>>
>> return type == OVL_ADAPTOR_TYPE_ETHDR || type == OVL_ADAPTOR_TYPE_PADDING;
>>
>> to
>>
>> return type == OVL_ADAPTOR_TYPE_ETHDR || type == OVL_ADAPTOR_TYPE_PADDING ||
>>          type == OVL_ADAPTOR_TYPE_MDP_RDMA;
>>
>> is that okay for you?
> 
> I just want the path to be like mt8195_mtk_ddp_ext[]. If so, I'm ok.
> 

Yes, that makes the path that you described to be exactly like
mt8195_mtk_ddp_ext[].

I will send a v11 later today.

Cheers,
Angelo

> Regards,
> CK
> 
>>
>>> If you does not test this with mt8195 external display path, maybe we could just drop the code about OVL adaptor with a TODO comment.
>>>
>>
>> And yes, as I said, external display paths were tested on 8195, actually both
>> on Kontron i1200 by Michael Walle and on MT8195 Tomato by myself.
>>
>> Thanks again,
>> Angelo
>>
>>> Regards,
>>> CK
>>>
>>>>
>>>> Regards,
>>>> Angelo
>>>>
>>>>> Regards,
>>>>> CK
>>>>>
>>>>>> +			if (!ovl_adaptor_comp_added)
>>>>>> +				ovl_adaptor_comp_added = true;
>>>>>> +			else
>>>>>> +				idx--;
>>>>>> +		}
>>>>>> +	} while (++idx < DDP_COMPONENT_DRM_ID_MAX);
>>>>>> +
>>>>>> +	/*
>>>>>> +	 * The device component might not be enabled: in that case, don't
>>>>>> +	 * check the last entry and just report that the device is missing.
>>>>>> +	 */
>>>>>> +	if (ret == -ENODEV)
>>>>>> +		return ret;
>>>>>> +
>>>>>> +	/* If the last entry is not a final display output, the configuration is wrong */
>>>>>> +	switch (temp_path[idx - 1]) {
>>>>>> +	case DDP_COMPONENT_DP_INTF0:
>>>>>> +	case DDP_COMPONENT_DP_INTF1:
>>>>>> +	case DDP_COMPONENT_DPI0:
>>>>>> +	case DDP_COMPONENT_DPI1:
>>>>>> +	case DDP_COMPONENT_DSI0:
>>>>>> +	case DDP_COMPONENT_DSI1:
>>>>>> +	case DDP_COMPONENT_DSI2:
>>>>>> +	case DDP_COMPONENT_DSI3:
>>>>>> +		break;
>>>>>> +	default:
>>>>>> +		dev_err(dev, "Invalid display hw pipeline. Last component: %d (ret=%d)\n",
>>>>>> +			temp_path[idx - 1], ret);
>>>>>> +		return -EINVAL;
>>>>>> +	}
>>>>>> +
>>>>>> +	final_ddp_path = devm_kmemdup(dev, temp_path, idx * sizeof(temp_path[0]), GFP_KERNEL);
>>>>>> +	if (!final_ddp_path)
>>>>>> +		return -ENOMEM;
>>>>>> +
>>>>>> +	dev_dbg(dev, "Display HW Pipeline built with %d components.\n", idx);
>>>>>> +
>>>>>> +	/* Pipeline built! */
>>>>>> +	*out_path = final_ddp_path;
>>>>>> +	*out_path_len = idx;
>>>>>> +
>>>>>> +	return 0;
>>>>>> +}
>>>>>> +
>>>>
>>>>
>>>>
>>
>>

-- 
AngeloGioacchino Del Regno
Senior Software Engineer

Collabora Ltd.
Platinum Building, St John's Innovation Park, Cambridge CB4 0DS, UK
Registered in England & Wales, no. 5513718