Yo Hal!

On Thu, 30 Oct 2025 23:01:39 -0700
Hal Murray <[email protected]> wrote:

> [email protected] said:
> > This seems fixed, but it's not really a compiler bug, but a
> > compiler that was unable to prove that a reference was safe, and
> > Gary already committed the usual approach to be more explicit about
> > checks and help the reasoning.  
> 
> I think it's a compiler bug.  No compiler can prove a whole program
> is safe for any serious program.  It's warnings should be restricted
> to when it can prove that it is unsafe.

Coverity does amazingly well at whole program analysis.

> Gary's fix is shutting up the compiler but not fixing the bug.

Oh?  I see no bug in gpsd.  You mean a bug in your CC?

>  That
> code was doing something important enough to have code for it.  With
> Gary's fix, nothing happens to fix the nasty case.

I've looked at that code a lot.  Seems solid to me.  What do you not like?

RGDS
GARY
---------------------------------------------------------------------------
Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        [email protected]  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can't measure it, you can't improve it." - Lord Kelvin

Attachment: pgpvwtknki_V2.pgp
Description: OpenPGP digital signature

Reply via email to