On 12/23/25 9:35 PM, Alex G. wrote:
> On Friday, December 19, 2025 7:29:07 AM CST Konrad Dybcio wrote:
>> On 12/19/25 5:34 AM, Alexandru Gagniuc wrote:
>>> IPQ8074, IPQ6018, and IPQ9574 support an m3 firmware image in addtion
>>> to the q6 firmware. The firmware releases from qcom provide both q6
>>> and m3 firmware for these SoCs. Support loading the m3 firmware image.
>>>
>>> Signed-off-by: Alexandru Gagniuc <[email protected]>
>>> ---
[...]
>>> +static int q6v5_wcss_load_aux(struct q6v5_wcss *wcss, const char
>>> *fw_name)
>>> +{
>>> + const struct firmware *extra_fw;
>>> + int ret;
>>> +
>>> + dev_info(wcss->dev, "loading additional firmware image %s\n",
> fw_name);
Your email client is messing up the reply context - if it happens to
be Thunderbird, set:
mailnews.wraplength = 0
mailnews.send_plaintext_flowed = false
in the config
>>
>> I don't think this log line is useful beyond development
>
> Remoteproc driver prints the main (q6) fimrware name, so I thought it would
> be
> prudent to print the names of any additional firmwares:
>
> remoteproc remoteproc0: Booting fw image IPQ9574/q6_fw.mdt, size 8140
>
>>> +
>>> + ret = request_firmware(&extra_fw, fw_name, wcss->dev);
>>> + if (ret)
>>> + return 0;
>>
>> return ret, perhaps? Unless you want to say that "it's fine if the M3 image
>> is missing, particularly not to impose any new requirements on existing
>> setups". But you haven't spelt that out explicitly.
>
> I intended to not abort when aux firmware is missing. Maybe the better way to
> handle this is to check for "-ENOENT" in the caller instead of return 0 here.
>
>> You also haven't provided an explanation as to why the firmware should be
>> loaded. Is it necessary for some functionality? Is it that case on the
>> newly-supported IPQ9574?
>
> I don't have a good answer. I reasoned that since the qcom provides it [1],
> the M3 firmware would need to be loaded. I haven't done much testing without
> it.
Well, could you please try?
IIRC it was strictly necessary for ATH1xk-on-PCIe so I'm assuming it's going
to be a necessity here as well
Konrad
