linux-kernel - Re: [PATCH] soc: mediatek: mtk-mmsys: Add support for MT6795 Helio X10

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <1c0b5dfc-4eb0-aa9d-cf38-657124022ed5@collabora.com>
Date:   Mon, 16 May 2022 12:06:56 +0200
From:   AngeloGioacchino Del Regno 
        <angelogioacchino.delregno@...labora.com>
To:     CK Hu <ck.hu@...iatek.com>, matthias.bgg@...il.com
Cc:     linux-arm-kernel@...ts.infradead.org,
        linux-mediatek@...ts.infradead.org, linux-kernel@...r.kernel.org,
        konrad.dybcio@...ainline.org, marijn.suijten@...ainline.org,
        martin.botka@...ainline.org, ~postmarketos/upstreaming@...ts.sr.ht,
        phone-devel@...r.kernel.org, paul.bouchara@...ainline.org,
        kernel@...labora.com
Subject: Re: [PATCH] soc: mediatek: mtk-mmsys: Add support for MT6795 Helio
 X10

Il 16/05/22 11:50, CK Hu ha scritto:
> On Mon, 2022-05-16 at 10:44 +0200, AngeloGioacchino Del Regno wrote:
>> Il 16/05/22 07:43, CK Hu ha scritto:
>>> Hi, Angelo:
>>>
>>> On Fri, 2022-05-13 at 19:03 +0200, AngeloGioacchino Del Regno
>>> wrote:
>>>> Add MM support for the MT6795 SoC, using the mmsys default
>>>> routing
>>>> table.
>>>>
>>>> Signed-off-by: AngeloGioacchino Del Regno <
>>>> angelogioacchino.delregno@...labora.com>
>>>> ---
>>>>
>>>> In order for this patch to have any meaning and functionality, it
>>>> is
>>>> required to also have in-tree the [1] MT6795 clocks series, even
>>>> though that's not required to actually compile this code.
>>>>
>>>> [1]
>>>>
> https://urldefense.com/v3/__https://patchwork.kernel.org/project/linux-mediatek/list/?series=641493__;!!CTRNKA9wMg0ARbw!ywDb_Gj4oV0iYHK_PbtJSNejzsJGlFVjX-1O0G1mR0vZX_5VeMtlZJ_AS6K_Mg$
>>>>    
>>>>
>>>>    drivers/soc/mediatek/mtk-mmsys.c | 11 +++++++++++
>>>>    1 file changed, 11 insertions(+)
>>>>
>>>> diff --git a/drivers/soc/mediatek/mtk-mmsys.c
>>>> b/drivers/soc/mediatek/mtk-mmsys.c
>>>> index 06d8e83a2cb5..38b36b881a3b 100644
>>>> --- a/drivers/soc/mediatek/mtk-mmsys.c
>>>> +++ b/drivers/soc/mediatek/mtk-mmsys.c
>>>> @@ -57,6 +57,13 @@ static const struct mtk_mmsys_match_data
>>>> mt6779_mmsys_match_data = {
>>>>    	},
>>>>    };
>>>>    
>>>> +static const struct mtk_mmsys_driver_data
>>>> mt6795_mmsys_driver_data =
>>>> {
>>>> +	.clk_driver = "clk-mt6795-mm",
>>>> +	.routes = mmsys_default_routing_table,
>>>
>>> mt6795 should have its own routing table instead of default one.
>>> The
>>> default one is used for mt8173, mt2701, mt2712 which no one knows
>>> how
>>> to separate them. If you don't know the routing table of mt6795,
>>> just
>>> left this as NULL.
>>>
>>
>> Hello CK,
>>
>> I agree with you, in fact the first test that I've done was with
>> leaving
>> routes to NULL, but that did not work (I couldn't get the display to
>> work),
>> while assigning the default routing table resulted in having a
>> working
>> internal (DSI) display et al, which is why I have done that.
>>
>> I am sure that this may be a quirk that is required due to my
>> platform's
>> bootloader, but still, it won't work otherwise.
>>
>> If anyone from MediaTek wants to help with providing a better
>> solution for
>> that, you're welcome to - but otherwise, my proposal is to configure
>> MT6795
>> with that default routing table and eventually come back to this
>> later if
>> any issue with this gets discovered... I couldn't find any, though.
> 
> So you have a drm driver which support mt6795 and it would call
> mtk_mmsys_ddp_connect() to setup routing routing register. You could
> generate mt6795 routing table from default one by partial element that
> you use in mt6795. I think this table would be smaller than the default
> one.
> 

Okay, that makes a lot of sense: I'll try following this advice.

Regards,
Angelo