Re: [PATCH v2 3/6] arm64: dts: qcom: r0q: add touchscreen support

[Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>]

On 10/27/25 7:32 PM, Eric Gonçalves wrote:
> 
> 
> On October 20, 2025 9:18:18 AM GMT-03:00, Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> wrote:
>> On 10/17/25 8:22 PM, Eric Gonçalves wrote:
>>>
>>>
>>> On October 17, 2025 5:20:43 AM GMT-03:00, Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> wrote:
>>>> On 10/16/25 10:08 PM, Ghatto wrote:
>>>>> On Thu, Oct 16, 2025 at 6:06 AM Konrad Dybcio
>>>>> <konrad.dybcio@oss.qualcomm.com> wrote:
>>>>>>
>>>>>>
>>>>>>
>>>>>> On 10/14/25 9:04 PM, Ghatto wrote:
>>>>>>> On Tue, Oct 14, 2025 at 11:18 AM Konrad Dybcio
>>>>>>> <konrad.dybcio@oss.qualcomm.com> wrote:
>>>>>>>>
>>>>>>>> On 10/14/25 5:10 PM, Ghatto wrote:
>>>>>>>>> On Tue, Oct 14, 2025 at 7:01 AM Konrad Dybcio
>>>>>>>>> <konrad.dybcio@oss.qualcomm.com> wrote:
>>>>>>>>>>
>>>>>>>>>> On 10/14/25 6:41 AM, Eric Gonçalves wrote:
>>>>>>>>>>> Enable the ST-Microelectronics FTS2BA61Y touchscreen. This patch
>>>>>>>>>>> depends on "Input: add support for the STM FTS2BA61Y touchscreen".
>>>>>>>>>>
>>>>>>>>>> The second sentence doesn't really make sense to be included in
>>>>>>>>>> the git log
>>>>>>>>> I'll keep it to the cover letter then
>>>>>>>>>>
>>>>>>>>>>> The device has an issue where SPI 8 (the bus which the touchscreen is
>>>>>>>>>>> connected to) is not working properly right now, so
>>>>>>>>>>> spi-gpio is used instead.
>>>>>>>>>>
>>>>>>>>>> Some Samsung devices used to use spi/i2c-gpio intentionally, also
>>>>>>>>>> on downstream. I'm assuming this isn't the case for r0q.
>>>>>>>>> It isn't, the device uses fts2ba61y on the spi8 bus - I hosted the
>>>>>>>>> DT at https://github.com/ghatt-o/ss_experiments/blob/main/r0q.dts if you
>>>>>>>>> want to take a look.
>>>>>>>>>>
>>>>>>>>>> Did you enable gpi_dma1, qupv3_id_1 before spi8, when testing
>>>>>>>>> The driver probes, but it fails to recognize the touchscreen device
>>>>>>>>
>>>>>>>> Could you post a complete dmesg and the precise DT diff you used?
>>>>>>> https://pastebin.com/QkYa8nMp (android dmesg) mainline dmesg doesn't have
>>>>>>
>>>>>> The link has expired 🙁
>>>>> https://pastebin.com/s4abJS9M shouldn't expire now!
>>>>
>>>> And yet it did!
>>>>
>>>> Feel free to just attach it to your message.. the list may bounce
>>>> it, but it will still reach the To/Cc recipients
>>> Attached the dmesg.txt file.
>>
>> I'm interested in what happens early when the SPI hosts are programmed,
>> whether there's any additional debug messages.
>>
>> If your log buffer is overrunning, you can add log_buf_len=8M to your
>> cmdline (which will make the buffer 8MiB long)
> Is it that big of a deal though? spi-gpio
> works just fine for touchscreen.

It's like using CPU rendering even though your computer has a GPU in it..

Plus we don't expect the upstream driver to have such odd issues, so I'd
really like to get to the bottom of it

Konrad