Re: [PATCH 8/8] arm64: dts: rockchip: RockPro64: enable wifi module at sdio0

From: Soeren Moch
Date: Tue Dec 10 2019 - 05:16:16 EST

On 10.12.19 10:14, Kalle Valo wrote:
> Heiko StÃbner <heiko@xxxxxxxxx> writes:
>> Hi Soeren,
>> Am Dienstag, 10. Dezember 2019, 00:29:21 CET schrieb Soeren Moch:
>>> On 10.12.19 00:08, Heiko StÃbner wrote:
>>>> Am Montag, 9. Dezember 2019, 23:38:22 CET schrieb Soeren Moch:
>>>>> RockPro64 supports an Ampak AP6359SA based wifi/bt combo module.
>>>>> The BCM4359/9 wifi controller in this module is connected to sdio0,
>>>>> enable this interface.
>>>>> Signed-off-by: Soeren Moch <smoch@xxxxxx>
>>>>> ---
>>>>> Not sure where to place exactly the sdio0 node in the dts because
>>>>> existing sd nodes are not sorted alphabetically.
>>>>> This last patch in this brcmfmac patch series probably should be picked
>>>>> up by Heiko independently of the rest of this series. It was sent together
>>>>> to show how this brcmfmac extension for 4359-sdio support with RSDB is
>>>>> used and tested.
>>>> node placement looks good so I can apply it, just a general questions
>>>> I only got patch 8/8 are patches 1-7 relevant for this one and what are they?
>>> Patches 1-7 are the patches to support the BCM4359 chipset with SDIO
>>> interface in the linux brcmfmac net-wireless driver, see [1].
>>> So this patch series has 2 parts:
>>> patches 1-7: add support for the wifi chipset in the wireless driver,
>>> this has to go through net-wireless
>>> patch 8: enable the wifi module with this chipset on RockPro64, this patch
>> Thanks for the clarification :-) .
>> As patch 8 "only" does the core sdio node, it doesn't really depend on the
>> earlier ones and you can submit any uart-hooks for bluetooth once the
>> other patches land I guess.
>>> If this was confusing, what would be the ideal way to post such series?
>> I think every maintainer has some slightly different perspective on this,
>> but personally I like getting the whole series to follow the discussion but
>> also to just see when the driver-side changes get merged, as the dts-parts
>> need to wait for that in a lot of cases.
> FWIW I prefer the same as Heiko. If I don't see all the patches in the
> patchset I start worrying if patchwork lost them, or something, and then
> it takes more time from me to investigate what happened. So I strongly
> recommend sending the whole series to everyone as it saves time.
Thanks for your explanation.
I will keep this in mind for future submissions.