Gosse Wijbenga wrote:



There was a similar report last October with AIX 5.2; see <http://mail.nessus.org/mailman/htdig/nessus/2004-October/msg00140.html>. To my knowledge, it was never resolved, although I was hoping the person would be able to narrow the problem down with truss. Perhaps you can try.


Georg,

Yes, I saw that post. It looks very similar.
At this moment I'm having a little trouble compiling libnasl:

exec.c: In function `execute_nasl_script':
exec.c:1827: warning: assignment discards qualifiers from pointer target type

No idea. In my exec.c there is an explicit cast:

     truc = (lex_ctxt*)ctx.tree;

And truc is defined previously as

lex_ctxt* truc = NULL;

   So I don't see why this should spout a warning.


and

nasl_func.c: In function `insert_nasl_func':
nasl_func.c:100: warning: passing arg 4 of `qsort' from incompatible pointer type


Maybe you've got a hint on how to handle these???

No idea either. In that function qsort is called as:

    qsort(pf->args_names, pf->nb_named_args,
            sizeof(pf->args_names[0]), strcmp);

AIX defines qsort as:

void qsort (Base, NumberOfElements, Size, ComparisonPointer)
void * Base;
size_t  NumberOfElements,  Size;
int  (*ComparisonPointer)(const void*, const void*);

[See http://publib16.boulder.ibm.com/pseries/en_US/libs/basetrf2/qsort.htm#a1819299]

Which is slightly different from glibc's qsort:
[/usr/include/stdlib.h]

typedef int (*__compar_fn_t) (__const void *, __const void *);
(...)
extern void qsort (void *__base, size_t __nmemb, size_t __size,
                   __compar_fn_t __compar);

But not incompatible, strcmp() should fit the definition quite OK. Both in AIX:

int  strcmp ( String1,  String2)
const char *String1, *String2;

(http://publib16.boulder.ibm.com/doc_link/en_US/a_doc_lib/libs/basetrf2/strcmp.htm#e65d10e933mich)

and in glibc's:

[/usr/include/string.h]

extern int strcmp (__const char *__s1, __const char *__s2)

Maybe you want to review your compiler's headres and see how qsort/strcmp is defined there to see if you spot the problem.

Regards

Javier
_______________________________________________
Nessus mailing list
[email protected]
http://mail.nessus.org/mailman/listinfo/nessus

Reply via email to