This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Problem with egcs? Not sure.


Alex Buell <alex.buell@tahallah.demon.co.uk> writes:

> http://www.rpi.edu/~payned/xwpe/xwpe-1.5.12a.tgz

> finally downloaded egcs-1.1b, and built it. 


> gcc -g -O2  -DCC_HAS_PROTOS=1 -DGPM=1 -DRANDLIB=1 -DNOPRINTER=1 -DNOVSWTCH=1  -DLIBRARY_DIR=\"/usr/lib/xwpe\" -c we_wind.c
> we_wind.c: In function `e_schr_lst_wsv':
> we_wind.c:1290: called object is not a function
> we_wind.c:1290: called object is not a function
> we_wind.c:1290: warning: unreachable code at beginning of switch statement
> we_wind.c:1290: called object is not a function
> make: *** [we_wind.o] Error 1

> int e_schr_lst_wsv(char *str, int xa, int ya, int n, int strlen, int ft,
>   int fz, struct dirfile **df, FENSTER *f)
> {
[snip]
>           strcpy(str, (*df)->name[num]);  /* <-- ERROR HAPPENS HERE */

Did you notice what the strcpy() expands to?  Check line 10469 of
we_wind.i: you'll find a switch statement, as well as some calls to
strlen().  Unfortunately, strlen is not a function in the current
scope: the function argument named `strlen' hides the standard
function strlen().  I believe this is well-defined, and glibc should
call __strlen or such instead, but I'm no C language lawyer.

Maybe you should report this as a bug to the maintainers of glibc...
Unless some of them is listening to this discussion :-)

Anyway, I'd recommend xwpe people to rename the function argument;
it's bad practice to redefine well-known names :-)

-- 
Alexandre Oliva
mailto:oliva@dcc.unicamp.br mailto:aoliva@acm.org
http://www.dcc.unicamp.br/~oliva
Universidade Estadual de Campinas, SP, Brasil



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]