[Bug other/46633] [meta-bug] frontends use BITS_PER_UNIT when they mean TYPE_PRECISION (char_type_node)

["joseph at codesourcery dot com" <gcc-bugzilla at gcc dot gnu dot org>]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46633

--- Comment #1 from joseph at codesourcery dot com <joseph at codesourcery dot com> 2010-11-24 15:13:38 UTC ---
On Wed, 24 Nov 2010, amylaar at gcc dot gnu.org wrote:

> E.g. consider this code from c-family/c-common.c:fix_string_type :
> 
>   else if (TREE_TYPE (value) == char16_array_type_node)
>     {
>       nchars = length / (TYPE_PRECISION (char16_type_node) / BITS_PER_UNIT);
>       e_type = char16_type_node;
> 
> On a bit-addressed architecture, you would have BITS_PER_UNIT == 1, but
> probably TYPE_PRECISION (char_type_node) == 8.

String representations when either BITS_PER_UNIT or TYPE_PRECISION 
(char_type_node) doesn't match host char have their own problems.  
However, my inclination is that TREE_STRING_LENGTH counts in units of 
BITS_PER_UNIT, which means the above logic is correct and the problem is 
that the "char" case should involve (TYPE_PRECISION (char_type_node) / 
BITS_PER_UNIT).  <http://gcc.gnu.org/ml/gcc/2003-06/msg01159.html> gives 
some thoughts of mine on these issues about representation of strings.