[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH] allow -Werror on IA-64 with GPT enabled
From: |
Andrew Clausen |
Subject: |
Re: [PATCH] allow -Werror on IA-64 with GPT enabled |
Date: |
Sat, 6 Oct 2001 11:38:37 +1000 |
User-agent: |
Mutt/1.2.5i |
On Fri, Oct 05, 2001 at 08:09:12PM -0500, address@hidden wrote:
> > Shouldn't we be checking for something like #if sizeof(int)
> > == 8 instead?
>
> sizeof(long), but that didn't work for me. gcc errored out on such a
> consruct.
In #ifdef's? You can still autoconfuse it...
> The simple:
> touch foo.c
> gcc -dM -E foo.c
> spits out all the #defines set by gcc for you. __LP64__ is the best of
> those. BITS_PER_LONG is what I'd use in the kernel, but it isn't set from
> user space. __LP64__ accurately describes what we're looking for. If longs
> are 64 bits, print with %lx. Otherwise, print with %llx. Works fine until
> we get to >64 bit computing, which will be necessary within our lifetime.
> :-)
sizeof(long) isn't interesting. The idea is: if int's are too small, then
use longs (and assume they are big enough). If you type-cast to the bigger
size, then it should do automatic type conversion, right?
In fact, can't you just typecast everything to long long, and not worry about
#ifdef's at all?
Andrew