Hi,
I'm using 6502bench SourceGen to disassemble a program that was compiled with an unknown APW C version.
I'm relying on CiderPress disassembler capability to look into APW's CLIB because, being a library, it has no relocation information and most disassemblers that I've tried won't deal with it. (do you have any suggestions on that topic?)
Well, I'm trying to match CLIB function signatures inside my program, and I need to find out the exact version of the CLIB that was used to build it. I've had a partial success with the CLIB from APW C 1.0B7.
Now I'm searching for APW C version 1.0B4, and possibly earlier versions like 1.0B1 that was released, too.
Also, any other beta versions that you might offer would be highly appreciated.

Thank you

On Monday, June 14, 2021 at 1:56:57 AM UTC-7, Enrico wrote:
> I'm relying on CiderPress disassembler capability to look into APW's CLIB because, being a library, it has no relocation information and most disassemblers that I've tried won't deal with it. (do you have any suggestions on that topic?)

To disassemble a IIgs load file (S16, EXE, PIF, etc), one needs to replicate the system loader behavior. To disassemble a IIgs OBJ or LIB file, one also needs to replicate linker behavior. The SourceGen OMF converter only acts like a loader.

One thing you might try is to write a simple program that references every function in the library, compile and link it into an EXE, and disassemble that. A simple table of function pointers ought to cause the linker to pull in the functions. Assuming that works, then depending on how everything gets laid out in the binary, creating it as an array of {pointer, "name"} structs might help keep things straight while poking around.

You can use APW DumpObj or ORCA DumpObj or MPW DumpObjIIgs to disassemble
CLIB. The DumpObjii do a decent job for intermediate OMF object files
(because they're not yet relocated).

-------
ProLine: kelvin@pro-kegs

Thank you for your suggestions, I'm going to try both and will let you know.

Hi Enrico,
files at http://www.brutaldeluxe.fr/apw/
Please share the disk image with the program you want to disassemble, I can help.
Cheers,
Antoine

Hello Antoine,

that page is giving me a 404

Le vendredi 18 juin 2021 à 21:27:37 UTC+2, Enrico a écrit :
> Hello Antoine,
>
> that page is giving me a 404

Correct link is :

http://www.brutaldeluxe.fr/public/apw/

Thank you so mutch Antoine :-()

Oops, yes, sorry, the link is http://www.brutaldeluxe.fr/public/apw/
av

Il giorno sabato 19 giugno 2021 alle 10:15:46 UTC+2 ntn.v...@gmail.com ha scritto:
> Oops, yes, sorry, the link is http://www.brutaldeluxe.fr/public/apw/
> av

Thank you, this is really useful. I can see a lot of updated versions in there. Unfortunately there's no earlier version than 1.0B7 for APW C. I'm still looking for that, if anyone can help-..............

I see two possible solutions;

http://www.byteworks.us/Byte_Works/Morgue_files/CPW.img
I think this is a small-c version.

and the disk labeled ; "0298 APW - C - Workdisk, Exsamples.po" from the brutal deluxe archive.

This is the earliest C compiler i have seen for the IIGS. Pre-dates APW-C. Not sure why it's labeled as APW-C.

The release notes mention it as the MegaMax C compiler. Was used as a ADTS pre-release system inside apple before APW-C.
Byteworks APW-C is a totally different system.

Hope this helps.

james

On Thursday, June 24, 2021 at 11:36:54 p.m. UTC-7, Enrico wrote:
> Thank you, this is really useful. I can see a lot of updated versions in there. Unfortunately there's no earlier version than 1.0B7 for APW C. I'm still looking for that, if anyone can help-..............

Taking a look at the files from Antoine,

D3_APWC1.0B7_19870414_ERR
&
0127 APW - C-compiler

Both have Version 1.0 of CLIB.
D3_APWC1.0B7_19870414_ERR, Has significant CLIB differences to 0127 APW - C-compiler CLIB.

Some the other CLIB's are as you mention. v1.0.b8. Some i have not been able to get a version # on.

Hope this helps.