Re: [PATCH]: kill HOST_PTR_PRINTF

[Geoffrey Keating <geoffk at geoffk dot org>]

Gabriel Dos Reis <gdr@integrable-solutions.net> writes:

> "Kaveh R. Ghazi" <ghazi@caipclassic.rutgers.edu> writes:
> 
> |  > "Kaveh R. Ghazi" <ghazi@caipclassic.rutgers.edu> writes:
> |  > 
> |  > [...]
> |  > 
> |  > |	* aclocal.m4 (gcc_AC_FUNC_PRINTF_PTR): Delete.
> |  > |	* configure.ac: Don't call gcc_AC_FUNC_PRINTF_PTR.
> |  > 
> |  > Instead of deleting the test entirely, I would rather see a mandatory
> |  > check that halt the build process if "%p" is not supported, i.e. a
> |  > kind of assertion.
> |  > -- Gaby
> | 
> | IMHO, that's overkill.
> | 
> | Beyond the "-Werror would have barfed" rationale, I have more evidence
> | that %p works everywhere we host GCC.  If you look at the sources,
> | there are already a few uses of %p that have snuck in.  So if it was a
> | problem we would have known by now.
> | 
> | Do we assert for any other C90 features before using them? 
> 
> As I understand it we do require ISO C90 compiler, but not ISO C90
> library.  Given that, either
> 
>   (1) the lack of assertion of ISO C90 library function is a bug; or
>   (2) we move on assuming ISO C90 library.
> 
> If we don't do (2) then failing to assert is not a feature.

At least for this case, I think that it would be better to build the
compiler using %p, and possibly have tree dumping not work, than to
not build the compiler at all.  It's not like someone would get this
error message and install a new C99-compliant libc, so the error would
in practise not be very helpful.

The patch is OK.