IMHO if you want avoid specific binaries for specific cpus, or you want avoid dinamic runtime code paths for specific cpus, maybe the best practice is to use following gcc switches to generate code

for general OS4 use, considering also Amiga classic 1200/4000 accellerators
-mcpu=604e -mtune=604e

or use standard base code with specific tuning for superscalar cpus and their caches
-mcpu=604e -mtune=604e/7400/8540/etc..

With a focus on Amiga systems, PowerPC ISA is the same for all PowerPC family, there are very few instructions different and all them are not so much revelant for size/speed code.
Maybe the most revelant is the ISEL instruction to avoid jumps, but it works only on e5500/440 cpu (Power ISA v2.06)

For G4 specific code of course you can use
-mcpu=7400 -mtune=7400 -maltivec -mabi=altivec

but in this case your code will run only on G4, if altivec code is generated.

One switch I suggest to use always is
-fomit-frame-pointer
to free one register and use it for calculations

Last, I use always -O3 to highly optimize code and -O2 if I want to read generated assembler against C code

@MigthyMax

Your work should be tested by a1222 owners and if confirmed it's working ok than could be adopted to build a1222 specific binaries.
IMHO it should be ok also for building standard os4 binaries.
Thanks a lot for your efforts!

So someone could try to compile and assemble SPE code with native GCC 4 and link objects with GCC 11.
If problem rely on linker, and if it's a text file it could be easily fixed with a diff between GCC4 and GCC11 linker cfg text files

How can we get a fixed version of GCC to avoid all these workarounds/hacks?
Is there a guy able to fix it for a native version of GCC?

@sailor

..try to declare them as APTR (void *) and casting vars to double using pointer (*) and address (&) operators.
You shoud not got penality in benchmarks.

IMHO best debugging for "sunday sometimes" programmers is massive use of "printf" in #ifdef #endif blocks.
So you can have easily a debug or a release build modifing only a single parameter.

If you want look inside generated asm code with GCC you have to use "-S" and "-mregnames" mixed with others parameters too like "-O1" "-O2" "-O3" and others.
In this way you can appreciate also compiler's optimizing patterns and strategies, it works also for Altivec and SPE code.

Make a good debugger for AmigaOS is not trivial.

As always you did a great job!
Thanks.

I'm interested in a modern opensource text editor for OS4.
Above basic features it needs to support Amiga flavors programming languages indentation, black theme and possibility to build macro to run compilers and other stuff and become a sort of IDE just like Flowstudio on MOS.
..Something native AmigaOS lightweight and fast, not porting from other systems usually too much slow and full of bugs.
If some developer is interested I can get involved and support in my spare time.

When you open a library you can always check against a specific version and than change path until right version is found, if ever..
Anyway probably in AmigaOs you can't load two or more libraries with same name, Is it right?

I'd like to use Codebench due it's confortable IDE but on my system is really slow and too much prone to crash.
I'm still waiting for a new free/developer/commercial release.

As code editor I'd like to use LiteXL but I have to admit I have not understand how to use it.. I'd like to have menus to load/save files and all other stuff found in other common text editors.
LiteXL sadly it's not comfortable for my use, maybe I have to give it a second try after reading docs (?!?)

..So I use Annotate to write code and IMHO until now it's by far the best editor for Amigaos4.
I wonder how to bypass a requester asking for datatype.library v44 everytime I use a custom screen..

To compile code I just use GCC commandline or makefile in a shell. Anyway I'm always looking for valid alternatives..

On MOS we have FLowstudio and it's years ahead any OS4 software and for me it's the best solution possible.

@joerg

I'm quite sure avoiding all exceptions handler of FPU code, soft floats should be significantly faster than hard floats.

Could we have a new updated SDK with full SPE support, compiler and libs?
What about GCC8 for SPE?

IMHO we need first a native OS4 GCC SDK full support for SPE code.
As already said GCC8 should be last one able to generate SPE code. Later all common libs like newlib/clib/math libs/SDL libs should have a native SPE version.
Once it will be reality some programmers shoul consider at least to recompile their projects to be full compatible with A1222,
Than other new projects, programs and games, could include an A1222 version in an easy way.

Double post, removed

"Sound Blaster Live!" sounds very well, no problem at all with os4 drivers

Someone has to release a new SDK with full support for A1222 to let new apps, and rebuild of the existing ones, to take advantage of new vector unit, reduce slow fpu emulation and avoid crashes/gurus/grim reapers.
I hope also to have full integration of afx clib in next release.

I’ll try to swap from annotate to litexl to code next time, thanks for update!

@kas1e

It could be really useful to improve QEMU PegII board, there's really everything on this PDF.

Nice!

Every time I read that some company is making money by developing Amiga hardware I can't help but laugh out loud. Every time I read that a developer believes he can make money with Amiga software development I can't help but laugh out loud. Some things are done out of passion and that's it, if you have other expectations it means that more than 30 years of Amiga history have taught us nothing in this regard. I am Italian like the people responsible for Acube but this is not what makes me take a stand in their favor, in fact we don't know each other. I am absolutely certain that Acube inserted the driver not to make money but only to support users. Anyone who doesn't understand this is probably because they don't want to understand it, so the question is closed.