Re: stdbool.h with AIX C-compiler

[Heinrich Mislik]

On 15 May 2008 at 20:39, Eric Blake wrote:

> Heinrich Mislik <Heinrich.Mislik <at> univie.ac.at> writes:
> 
> Hello Heinrich, and thanks for the report.
> 
> > here is a problem with stdbool.h from gnulib:
> >     pbool(256);
> 
> Inherently non-portable when attempted with a C89 compiler.  And
> Gnulib's 
> <stdbool.h> replacement documents it as one of the limitations of
> using the 
> replacement <stdbool.h>:
> 
> |   Limitations of this substitute, when used in a C89
> environment:
> |
> |       - <stdbool.h> must be #included before the '_Bool' type can
> be used.
> |
> |       - You cannot assume that _Bool is a typedef; it might be a
> macro.
> |
> |       - Bit-fields of type 'bool' are not supported.  Portable
> code
> |         should use 'unsigned int foo : 1;' rather than 'bool foo :
> 1;'.
> |
> |       - In C99, casts and automatic conversions to '_Bool' or
> 'bool' are
> |         performed in such a way that every nonzero value gets
> converted
> |         to 'true', and zero gets converted to 'false'.  This
> doesn't work
> |         with this substitute.  With this substitute, only the
> values 0 and 1
> |         give the expected result when converted to _Bool' or
> 'bool'.
> 
> > I came across this, when the -iregex option of find didn't work 
> > (behaves like -regex). The above test-program reflects the
> situation 
> > in line 778 of regcomp.c:
> > 
> >  err = re_string_construct (&regexp, pattern, length,
> preg-translate,
> >                               syntax & RE_ICASE, dfa);
> 
> Therefore the bug is in regcomp.c, for violating the limits of
> <stdbool.h> 
> usage.  The proper fix is not in the stdbool module (since there is
> no way to 
> make C89 behave the way C99 requires bool to act), but in the regex
> module, to 
> explicitly compare the result of & against 0 in order to create a
> true 0/1 
> value regardless of what underlying type we are really using.

After reading the comments in stdbool.h I came to the same conclusion 
and to the same fix. After that -iregex was completly broken. I found 
one second place, that needs the same change:

Line 703 in regexec.c is now:

  err = re_string_allocate (&mctx.input, string, length,
        dfa->nodes_len + 1,
    preg->translate, (preg->syntax & RE_ICASE) != 0, dfa);

Now -iregex seems to work OK. But as you mentioned, there may be more 
gotchas around.


Regarding your other mail:

Attached find the relevant parts from config.log. I don't now, wether 
the compiler is in error here. And reporting bugs for AIX costs a lot 
of time, which I don't have right now.

Cheers

Heinrich
-- 
Heinrich Mislik
Zentraler Informatikdienst der Universitaet Wien
A-1010 Wien, Universitaetsstrasse 7
Tel.: (+43 1) 4277-14056, Fax: (+43 1) 4277-9140


The following section of this message contains a file attachment
prepared for transmission using the Internet MIME message format.
If you are using Pegasus Mail, or any other MIME-compliant system,
you should be able to save it or view it from within your mailer.
If you cannot, please ask your system administrator for assistance.

   ---- File information -----------
     File:  config.log.bool
     Date:  16 May 2008, 13:04
     Size:  5577 bytes.
     Type:  Unknown

config.log.bool
Description: Binary data