Re: [RFA] Darwin/x86 port (v4 - part 1/4: machoread.c)

[Tristan Gingold <gingold@adacore.com>]

On Nov 18, 2008, at 2:31 AM, Stan Shebs wrote:

> Tristan Gingold wrote:
>>
>> typedef struct oso_el
>> {
> Something about what this struct is for, please?

Comments added.

> And while you and I know that "oso" is short for "other source" and  
> refers to the collection of debug info from .o files, it's new to  
> everybody else. :-)

I thought it means "object source".  Anyway, I added comments.

> (Getting function bounds set correctly is going to be loads of fun,  
> ahem, I had to look at Apple's hairy code for this several months  
> ago...)

Yes, there is work to do in this area!

>> DEF_VEC_O(oso_el);
>> static VEC(oso_el) *oso_vector;
> The vector macros should follow the "space before paren" just as  
> everybody else does, although I note that the doc in vec.h confuses  
> things by not following the general rule.

I added a space.

>> #define DSYM_SUFFIX ".dSYM/Contents/Resources/DWARF/"
> Again, need to say what dSYM files are, and maybe mention that the  
> yucky wired-in string is guaranteed to have that form (or at least  
> that we're doing the same thing as Apple's GDB).

Done.

> In comments, we should prefer to say "Mach-O" rather than "macho" or  
> "mach-o", since that is the proper name of the file format.

I think I fixed all.  'macho' is still used in functions name.

> With these doc and formatting changes, machoread.c is OK to go into  
> the trunk.

Thanks.

> (Ideally we will migrate this file to the list of generic files, but  
> we will need to check that it compiles on other hosts first.)

Tristan.