"-march=native" possibly being misled on Debian i386?

[Toby Douglass]

On Sun, Jun 11, 2017 at 4:43 PM, Toby Douglass
<toby.douglass@minodes.com> wrote:
> And in fact;
>
> minnow@minnow:~$ gcc -mtune=native -dM -E - < /dev/null | grep 86
> #define __i586 1
> #define __DBL_MAX__ ((double)1.79769313486231570815e+308L)
> #define __i386 1
> #define __i586__ 1
> #define __i386__ 1
> #define i386 1
>
> The docs say;
>
> "While picking a specific cpu-type will schedule things appropriately
> for that particular chip, the compiler will not generate any code that
> does not run on the i386 without the -march=cpu-type option being
> used."
>
> But in fact if I understand the output of the command Ruoyao provided,
> tune *is* setting arch (whilst not setting tune =-) so this compiler
> in this specific case ends up setting arch correctly.

This is rubbish.  All that's happening is that in the absence of
march, the default arch is being used.