[RFA] s/char/void/g in floatformat.[ch]

DJ Delorie dj@redhat.com
Mon Aug 15 14:52:00 GMT 2005


> The fact that the floatformat_xxx functions use `char *' for byte
> buffers used to store floating-point numbers doesn't help with GCC
> 4's `unsigned char' vs. '(signed) char' "madness".  GDB now uses
> `gdb_byte' instead of '(unsigned) char', but we obviously cannot use
> that in libiberty.  Therefore I propose we take the easy way out and
> use `void *' for the byte buffers.  That should remove the need for
> nasty casts in all code that uses the floatformat stuff.
> 
> Mark

Ok.

> P.S. I still think this code should be removed from libiberty/ and
> moved back into gdb/.  Or does newlib still use it?

opcodes uses it for m68k.



More information about the Binutils mailing list