Summary: | gnustep-base's configure doesn't complains about no ffcall, while there is | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Artem Baguinski <artm> |
Component: | [OLD] Library | Assignee: | Gentoo Linux bug wranglers <bug-wranglers> |
Status: | RESOLVED INVALID | ||
Severity: | normal | ||
Priority: | High | ||
Version: | 1.4_rc4 | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- |
Description
Artem Baguinski
2003-12-26 05:39:57 UTC
oh, the problem is here: checking "for forwarding callback in runtime"... no which is caused by the following bit of configure.ac failing: AC_MSG_CHECKING("for forwarding callback in runtime") AC_EGREP_HEADER(__objc_msg_forward, objc/objc-api.h, AC_MSG_RESULT(yes), AC_MSG_RESULT(no); enable_ffcall=no) i don't know the reason it fails but my objc/objc-api.h DOES have __objc_msg_forward (gcc 3.3.2) some additional investigation shows that c preprocessor, invoked by the AC_EGREP_HEADER, makes all extern declarations disappear. that results in egrep not finding the function in question. oh, no, i was wrong, cpp doesn't make ALL extern declarations disappear, but only some... this is the cpp output corresponding to the place where "missing" variable is declared... could it be that variable declarations are removed? ---------------------------------------------------------------- void objc_free (void *mem); # 412 "/usr/local/include/objc/objc-api.h" 3 extern void *(*_objc_malloc) (size_t); extern void *(*_objc_atomic_malloc) (size_t); extern void *(*_objc_valloc) (size_t); extern void *(*_objc_realloc) (void *, size_t); extern void *(*_objc_calloc) (size_t, size_t); extern void (*_objc_free) (void *); Method_t class_get_class_method (MetaClass class, SEL aSel); ---------------------------------------------------------------- dear developers, the problem was with my system [some other software i use have installed /usr/local/inlude/objc/* which wasn't quite correct i figured]. |