Re: [PATCH 0/2] Demangler crash handler

[Doug Evans <xdje42@gmail.com>]

On Fri, May 9, 2014 at 8:33 AM, Gary Benson <gbenson@redhat.com> wrote:
> [...]
> +  if (crash_signal != 0)
> +    {
> +      static int warning_printed = 0;
> +
> +      if (!warning_printed)
> +       {
> +         warning ("internal error: demangler failed with signal %d\n"
> +                  "Unable to demangle '%s'\n"
> +                  "This is a bug, "
> +                  "please report it to the GDB maintainers.",
> +                  crash_signal, name);
> +
> +         warning_printed = 1;
> +       }
> +
> +      result = NULL;
> +    }
> +
> +  return result;

Hi.

Applying "Consistency Is Good" to this patch,
I wonder if we should do something similar to what we do for internal errors.
I'm not sure I would use the same flag (grep for
internal_problem_modes and friends in utils.c), but
OTOH I wouldn't want a proliferation of options for controlling each
particular kind of "crash".

What do you think?