Very nice. In case you have trouble finding it, the line goes in asmx.c at line 532 as part of the AsmInit() function.
One caveat - it doesn't accept gas syntax... or apparently old Hitachi syntax... as a matter of fact, I'm having trouble finding which syntax it's using. Maybe you can clarify that.
int_m.src - hitachi assembler format (most SEGA's libs for 32X and Saturn are in this format) https://pastebin.com/1tuKMYHw
In the case of GAS source file:
[*]ASMX uses a ; as a mark of commentary instead of !, /* and */.
[*]align and constant lenght specifying pseudo-ops, including long doesn't require dot as a prefix in ASMX.
[*]The ASMX equivalent of .section may be the SEG/RSEG/SEG.U pseudo-op (I just never use a rom division by sections before)
[*]I also doubt that there is a something similar to subs division by global and local in ASMX
Thanks. Knowing the syntax asmx uses for SH will allow us 32X/Saturn/DC folks to modify our code to work for cases where we want to use asmx instead of gas or shasm.
Chilly Willy wrote: Mon Apr 07, 2025 5:06 pm
Thanks. Knowing the syntax asmx uses for SH will allow us 32X/Saturn/DC folks to modify our code to work for cases where we want to use asmx instead of gas or shasm.
Now I'm trying to add the PowerPC - the heart of Sega Model 3 arcade hardware - architecture support (this arch assembly module is rewritten by the go ARM module, so the source file for a while still have many "tails" of it), but there is a problem with o, ., a and l postfixes for some commands - compiled ASMX exe file perfectly understand these commands without of postfixes, but the same commands with them for this time gives an error. Where can be the root of problem?
The first block of quoted code correctly checks strings. The second quoted code is the problem code that is trying to pass integers to a string function.
Chilly Willy wrote: Tue Jun 10, 2025 9:52 pm
The first block of quoted code correctly checks strings. The second quoted code is the problem code that is trying to pass integers to a string function.
I replaced the code from second quote with a code from first, and all the same getting a same error. Here is a test source for the each ASM command translation checking.
That wasn't the only place with that error. I was just pointing out the first instance and that anything like it would need to be fixed. I'll go through it and find the others and check against the test code.
Chilly Willy wrote: Thu Jun 12, 2025 8:45 pm
That wasn't the only place with that error. I was just pointing out the first instance and that anything like it would need to be fixed. I'll go through it and find the others and check against the test code.