Needed_changes_to_compile_with_gcc_13.2.1

[Brumi-2021]

Hi , to not forget the current investigation, (and thanks specially to @bernd-herzog , @u-foka , and @NotherNgineer ...)

Please find attached the changes that I need to apply , for been able to compile with the new version gcc 13.2.1 arm-none-eabi.
Till now I was using the version gcc 10.3.1, but recently pentoo linux updated it to the new gcc 13.2.1

Notes :
1-) Clearly , using the compiler gcc 13.2.1 (same as with the previous gcc 10.3.1) , those upgraded gcc versions produced with the same source code , a slighly bigger binary size , than the binary produced with the current official upstream 9.2.1.
(therefore in my compile test, following @NotherNgineer advices , I usually delete temporary some application to reduce the binary size , in ui_navigation.cpp )
Ex .
// {"sstvtx", "SSTV", TX, ui::Color::green(), &bitmap_icon_sstv, new ViewFactory()},
//{"touchtune", "TouchTune", TX, ui::Color::green(), &bitmap_icon_touchtunes, new ViewFactory()},

2-) I could compile all from vscode , using embeded CMake tool , (same as I was doing with 10.3.1 , thanks to @bernd-herzog help )
Now no problem, even it gives me that msg :
"[build] WARNING: Compiler version mismatch, please use the official compiler version 9.2.1 when sharing builds! Current compiler version: 13.2.1"

3-) Once reflashed the new compiled binary (created with gcc. 13.2.1) ,
I could confirm that using our Portapack , we can generate DEBUG_DUMP_XXX.TXT ,
and confirm with it , that the generated bin , comes from gcc a real 13.2.1 (just a double check of the used compiler version)

4-) Recently, I was getting some GURU mediation errors , just at power up with my binary compiled with gcc 13.2.1,
but all those problems have been dissappeared after re-sync my cloned github repo with latest today's next .
(I still do not know why , but now it works well , no more GURU errors)
https://github.com/user-attachments/assets/7bad6661-5196-4710-bc0c-6da2d782aa24

5-) Regarding the event_m0.cpp file changes , those two changes are not related to this gcc compatibility issue,
but trying to solve the above (4) , I added two lines , to delete two more warnings. And now all seems ok ,
But I will need your approval about them .
I could delete from 28 compile warnings (actual) --> 26 compile warnings (after this PR) .
Now all remaining 26 warnings , seems to be related to USB

6-) The applied changes in the file : /opt/portapack-mayhem/firmware/tools/make_spi_image.py

Are not strictly necessary .But due to the point (1) , initially I was getting so many uncompleted compilations with exit code(1) , and it was so annoying and I could not test the full compile and link process.
Then , in my set up investigation , I decided to comment those two below lines ,
It is true that now when excedding 1MB binary flash size , we will not get the usual error message :
Example,

But anyway , we still could always check that other output line .
And if the room margin is negative (-) , means that we are exceding the 1MB flash size. (we do not have positive (+) room margin)

Post-note about (6) ,
based on above revision comments, I have re-activated that Binary limit size protection check with a second commit inside that PR, (as it was originally)

---

There is no rush to be merged that PR .
But soon or later gcc version 9.2.1 might be obsolete ...
But before merging it , we need to be completely sure that this PR is fully compatible with current upstream 9.2.1 compilation , without breaking anything . (Hopefully it should be compatible and transparent) .

Waiting for your testing support and comments .

@u-foka , in the past , I know that you were also using gcc version 12,x.x...or even 13.x.x ,
so maybe you can also have a chance to test those changes ...

Thanks in advance for your test , comment reviews , and feedbacks
Cheers.

[NotherNgineer]

I fully agree with the goal to enable alternative gcc versions for Mayhem testing.

[gullradriel]

Nice ! When will you undraft it ?

[u-foka]

Looks promising, great work thanks! :)

[Brumi-2021]

I resolved both previous conversations .
I added one new commit , changing back the Limit flash size protection of 1MB .

So far , I am not deleting the double compile options specs. (I know that it was already added in yesterday @NotherNgineer's PR : Fix "standalone pacman" compile errors with gcc >9.2.1 #2223) .
This PR has exactly same change , hopefully it will not create any conflict . If not , we will amend it .

[NotherNgineer]

lgtm

[u-foka]

Looks good to me thanks!

[gullradriel]

Merged,
Many thanks to all ;-)