This is the mail archive of the
mailing list for the GCC project.
Re: [RFH] Getting rid of WCHAR_TYPE_SIZE
- From: mike stump <mrs at windriver dot com>
- To: gcc at gcc dot gnu dot org, neil at daikokuya dot demon dot co dot uk
- Cc: dewar at gnat dot com, kenner at vlsi1 dot ultra dot nyu dot edu
- Date: Wed, 8 May 2002 14:59:02 -0700 (PDT)
- Subject: Re: [RFH] Getting rid of WCHAR_TYPE_SIZE
- References: <20020508201330.GA449@daikokuya.demon.co.uk>
> Date: Wed, 8 May 2002 21:13:30 +0100
> To: email@example.com
> Cc: Richard Kenner <firstname.lastname@example.org>, Robert Dewar <email@example.com>
> From: Neil Booth <firstname.lastname@example.org>
> Successfully removing WCHAR_UNSIGNED tempted me to look at
> getting rid of WCHAR_TYPE_SIZE like Mark suggested.
> It turns out there is only one user in the whole compiler,
> and it's in Ada (huh?). In targtyps.c, we have:
> get_target_wchar_t_size ()
> /* We never want wide chacters less than "short" in Ada. */
> return MAX (SHORT_TYPE_SIZE, WCHAR_TYPE_SIZE);
I suspect this is is due to:
kankakee bash grep 'WCHAR_TYPE_SIZE 8' * */*
m68k/vxm68k.h:#define WCHAR_TYPE_SIZE 8
sparc/vxsparc.h:#define WCHAR_TYPE_SIZE 8
sparc/vxsparc64.h:#define WCHAR_TYPE_SIZE 8
:-( These are wrong, were wrong, and will always be wrong. The better
solution, is to fix that code, and remove the stupid hack around in
Zack will probably be submitting work soon to fix these.
#define WCHAR_TYPE_SIZE 16
is solution we've adopted for these files.