[Nasm-devel] Include path for incbin

[Frank K. <fbk...@at...>]

Alexei A. Frounze wrote:

[fbk]
>>One thing I can see that will have to go is the "//" style comments. (I
>>told ya we were quaint!). That's something I think I can fix myself :)
> 
> Weird. Include a ... either syntax policy or something or a beautifier to
> the NASM package then :)

Okay. Write one up :)

>>More troubling is the fact that if I do "nasm -f bin -o incbin.com
>>incbin.asm" - no "-I" switch at all - Nasm segfaults! This is with a
>>djgpp build of Nasm. I haven't tried other builds, or tried to track it
>>down much. I'll look for it - just wanted to let you know I *did* find a
>>problem.
> 
> Damn! I lost one pointer check and I got away with it under w2k, though it
> appeared in DOS7.
> This is how it should be done:
> 
> char**
> pp_get_include_path_ptr (char **pPrevPath)
> {
> #define GetMemberOffset(StructType,MemberName)\
>   ((size_t)&((StructType*)0)->MemberName)
>     IncPath *i;
> 
>     if (pPrevPath == NULL)
>     {
>         if (ipath != NULL)
>             return &ipath->path;
>         else
>             return NULL;
>     }
> 
>     i = (IncPath*) ((char*)pPrevPath - GetMemberOffset(IncPath,path));
>     i = i->next;
>     if (i != NULL)
>         return &i->path;
>     else
>         return NULL;
> #undef GetMemberOffset
> }

Yep! That seems to get it. Thanks, Alexei!

Anyone else got comments on this? Are Alexei and I the only ones who 
think this is a good idea? Anyone think it's a bad idea?

> P.S. Sorry, it was unintentional. I shouldn't be doing programming at night
> time, it's not working good that way :)

Can you program in the daytime??? Must try that. :)

Best,
Frank