On Sun, 29 Apr 2012 20:30:47 -0600
Eduardo Silva <[email protected]> wrote:

> We need to redesign that part:
> 
> http://bugs.monkey-project.com/ticket/144
> 
> On Sun, Apr 29, 2012 at 8:16 PM, Eduardo Silva <[email protected]> wrote:
> > Hi,
> >
> > what does -march=i486 for the case when the atomic functions are not
> > available ?, is the compiler providing some workaround ?
> >
> > regards,

486 is the first CPU that added the necessary instructions. If GCC is 
configured to produce code for 386 it can't use those, and errors out.

The GCC support was added in 4.1, I suppose there are indeed some GCC 3 
toolchains for not too well supported devices.

I agree that ideally the counter should be lockless, but please commit this 
patch in the meantime - until that redesign, we should error out in configure 
and not in the compile.

- Lauri
_______________________________________________
Monkey mailing list
[email protected]
http://lists.monkey-project.com/listinfo/monkey

Reply via email to