Re: gcc13 compilation error on Void Linux

[Greg Troxel <gdt%lexort.com@localhost>]

Aleksey Cheusov <vle%gmx.net@localhost> writes:

>> Aleksey Cheusov <vle%gmx.net@localhost> writes:
>
>  >>> LONG_MIN is in limits.h, so this smells like a failure to include the
>  >>> right headers.
>  >>
>  >> Please have a look at email I replied to. There you'll find a required
>  >> information and detailed analysis.
>
>> Why isn't it a better approach to just patch the configure test to
>> have -O2?
>
> As for me fixing a problem by changing environment is always better than
> patching sources, especially patching of autogenerated files consisting
> of 1000s of lines.

Sure, but patching the configure test only changes that, and doesn't
change how the rest of the package is built.

>> The patch's comment is nowhere near good enough.
>
> How about the following?
>
>    # The following line fixes a build failure on Linuxes with gcc with
>    # -DFORTIFY_SOURCE enabled by default. On such systems libiberty's
>    # ./configure script fails to find limits.h header file without -O
>    # flag passed to gcc.
>
>> and the patch lacks a URL to an upstream bugtracker entry.
>
> I'll register a bug to gcc.

That will help a lot.

>> That remains true.  I still don't know in what component you think the
>> bug lies.
>
> gcc13's libiberty component.

Any other opinions?