Re: s-enum-types-h Fail on Darwin

[Owen Taylor <otaylor redhat com>]

Jeshua Lacock <gnome-devel openosx com> writes:

> Hello,
> 
> I am trying to build atk-0.13, with glib 1.3.15, autoconf 1.6, and
> libtool 1.4.2 installed on Mac OS X (10.1.3) and it fails trying to
> create the "atk-enum-types.h" and "atk-enum-types.c" files.
> 
> Here is the error:
> 
> Making all in atk
> ( cd . && glib-mkenums \
>                  --fhead "#ifndef __ATK_ENUM_TYPES_H__\n#define
> __ATK_ENUM_TYPES_H__\n\n#include <glib-object.h>\n\nG_BEGIN_DECLS\n" \
>                  --fprod "/* enumerations from \"@filename \" */\n" \
>                  --vhead "GType @enum_name _get_type (void);\n#define
> ATK_TYPE_ ENUMSHORT@ (@enum_name _get_type())\n"         \
>                  --ftail "G_END_DECLS\n\n#endif /*
> __ATK_ENUM_TYPES_H__ */" \
>          atk.h atkaction.h atkcomponent.h atkdocument.h
> atkeditabletext.h atkgobjectaccessible.h atkhyperlink.h atkhypertext.h
> atknoopobject.h atknoopobjectfactory.h atkobject.h atkobjectfactory.h
> atkimage.h atkregistry.h atkrelation.h atkrelationset.h atkselection.h
> atkstate.h atkstateset.h atkstreamablecontent.h atktable.h atktext.h
> atkutil.h atkvalue.h ) > tmp-atk-enum-types.h \
> && (cmp -s tmp-atk-enum-types.h ./atk-enum-types.h || cp
> tmp-atk-enum-types.h ./atk-enum-types.h ) \
> && rm -f tmp-atk-enum-types.h \
> && echo timestamp > s-enum-types-h
> make[1]: *** [s-enum-types-h] Error 139
> make: *** [all-recursive] Error 1

Well, atk-0.13 and glib-1.3.15 are pretty old. GTK+-2.0.0 and ATK-1.0.0
were released almost a month ago :-)

I think your particular problem is probably something that was fixed
recently:

Mon Mar 25 17:51:05 2002  Owen Taylor  <otaylor redhat com>

	* glib-mkenums.in (parse_entries): Fix various bugs and
	excessive stack usage that crept in the conversion from
	gtk-mkenums. (#74431)

This isn't in Glib-2.0.0, but it is in the 2.0.1 tarballs that I put
up on ftp.gtk.org yesterday. (They haven't been announced officially
yet, but are final.)

If you follow the bug reference and the mailing list references from
there, you'll find lots of details. The essence though is that
glib-mkenums was using lots and lots of stack space, which caused
problems operating systems with fixed sized stacks.

Regards,
                                        Owen