[Bug gdb/31331] Wenum-constexpr-conversion should be fixed, soon treated as a hard error

["simon.marchi at polymtl dot ca" <sourceware-bugzilla@sourceware.org>]

https://sourceware.org/bugzilla/show_bug.cgi?id=31331

--- Comment #17 from Simon Marchi <simon.marchi at polymtl dot ca> ---
(In reply to Carlos Galvez from comment #11)
> So I applied this change:
> 
> $ git diff gdbsupport/
> diff --git a/gdbsupport/enum-flags.h b/gdbsupport/enum-flags.h
> index 50780043477..212a51396e8 100644
> --- a/gdbsupport/enum-flags.h
> +++ b/gdbsupport/enum-flags.h
> @@ -94,7 +94,7 @@ struct enum_underlying_type
>    DIAGNOSTIC_PUSH
>    DIAGNOSTIC_IGNORE_ENUM_CONSTEXPR_CONVERSION
>    typedef typename
> -    integer_for_size<sizeof (T), static_cast<bool>(T (-1) < T (0))>::type
> +    integer_for_size<sizeof (T), std::is_signed<typename
> std::underlying_type<T>::type>::value>::type
>      type;
>    DIAGNOSTIC_POP
>  };
> 
> 
> And leads to the above-attached diff in testsuite/gdb.sum. I'm not sure what
> to make of it, but it does seem to reduce the number of unexpected failures?
> 
> < # of expected passes		104586
> < # of unexpected failures	685
> ---
> > # of expected passes		104601
> > # of unexpected failures	680
> 108988c108993
> < # of known failures		99
> ---
> > # of known failures		97

For the record, if we choose to go with underlying_type, I think we can get rid
of this integer_for_size thing altogether and use std::underlying_type to
define enum_flags::underlying_type directly.

I pinged Pedro Alves, original author of this code, he'll take a look soon.

-- 
You are receiving this mail because:
You are on the CC list for the bug.