Re: [PATCH 07/11] ASoC: SOF: core: Skip firmware test for undefined fw_name

[Venkata Prasad Potturu]

On 12/14/23 17:27, Péter Ujfalusi wrote:
> On 14/12/2023 12:58, Venkata Prasad Potturu wrote:
> > On 12/14/23 16:18, Péter Ujfalusi wrote:
> > Thanks for your time Peter!
> > > On 09/12/2023 22:53, Cristian Ciocaltea wrote:
> > > > Some SOF drivers like AMD ACP do not always rely on a single static
> > > > firmware file, but may require multiple files having their names
> > > > dynamically computed on probe time, e.g. based on chip name.
> > > I see, AMD vangogh needs two binary files to be loaded sometimes, it
> > > is not using the default name as it hints via the sof_dev_desc
> > > ("sof-vangogh.ri").
> > >
> > > The constructed names for the two files are just using different pattern:
> > > sof-%PLAT%.ri
> > > vs
> > > sof-%PLAT%-code.bin
> > > sof-%PLAT%-data.bin
> > >
> > > iow, instead of the combined .ri file which includes the code and data
> > > segment it has 'raw' bin files and cannot use the core for loading the
> > > firmware.
> > >
> > > What is the reason for this? an .ri file can have two 'modules' one to
> > > be written to IRAM the other to DRAM.
> > > sof_ipc3_load_fw_to_dsp()
> > For AMD Vangogh platform devices signed firmware image is required, so
> > split .ri image into code and data images.
> >
> > Only Code.bin will be signed and loaded into corresponding IRAM location.
> This is not different than what the Intel .ri files are made of. The
> module which is to be loaded to IRAM is signed code the module which
> goes to DRAM is not signed.
> The loader itself is not looking into the sections of the .ri image, it
> just parses the header and copies them where they belong.
>
> if the issue is name collision then you could try to put the signed
> firmware file under 'signed' folder (fw_path_postfix) of the platform
> like Intel does with the community signed ones?

We have a limitation that code image can't be signed during compilation.

So splitting the .ri image into code and data bin and sign the code bin
and load into IRAM.

> It would be great if somehow we can handle all of these in core, have
> shared code and familiar prints among vendors, platforms..
>
> Fwiw, I'm planning the path, filename creation to be moved to core for
> the current platforms, but it implies that they do use single firmware file.
> struct sof_dev_desc would only have two strings:
> vendor - AMD / iMX / Intel / Mediatek
> platform - tgl, vaggogh, etc
>
> I need to adjust it based on what I have learned today about vangogh.