[openbeos] Re: question about 64-bit processors
- From: André Braga <meianoite@xxxxxxxxx>
- To: openbeos@xxxxxxxxxxxxx
- Date: Fri, 17 Sep 2004 04:28:40 -0300
On Fri, 17 Sep 2004 08:38:14 +0200, Marcus Overhagen <ml@xxxxxxxxxxxx> wrote:
> > BPointer meaningful_ptr_name1[B_POINTER_SIZE];
> > BPointer meaningful_ptr_name2[B_POINTER_SIZE];
>
> You don't seem to understand how pointers are represented by the compiler.
I should have probably have mentioned that BPointer would be a typedef
to char; this would allocate B_POINTER_SIZE bytes of memory. How you
would use those are up to you. Using this space to hold memory
addresses is perfectly possible, doable, and it's nothing short of a
classic technique (classic as in "used by unix IP socket
structures"-classic; namely, sin_zero[] in sockaddr_in, for example).
> > //I hope you'll tolerate seing code like this
> No we will not.
Come on! I was being sarcastic...
> You don't seem to understand how pointers are represented by the compiler.
> This doesn't make any sense!
Unless I completely missed it, that's how I imagined Ingo's suggestion
of wrapping those structures around typedefs so it could support
different architectures/pointer widths on compile time.
If I got anything wrong, please enlighten me. I *love* when people
correct me when I'm wrong, but I *hate* being told to shut up,
specially when I'm left ignorant WRT my mistakes.
--
"I invented the term Object-Oriented, and I can tell you I did not
have C++ in mind"
Alan Kay
Other related posts:
