On Wed, Oct 29, 2008 at 2:53 PM, Stefan Behnel <[EMAIL PROTECTED]> wrote:
>
> I'm fine with calling InitThread() when "with nogil" or "with gil" is used
> anywhere in the source.

OK

> And I think it should be left to the users to handle
> all other cases themselves.

Here I'm not following you. What other cases are you talking about?

> Also note that UtilityCode() accepts a keyword "init", so you can just put the
> call to InitThreads() in there and it will do the right thing. I think the
> "proto" can be dropped completely in this case.

Yes, I reviewed the code you recently pushed. But the "init" (if I got
it right) seems to be intended for initializing global variables. So I
was not sure if using "init" was strictly correct.

Moreover, if you look carefully at my path, you will find that the
"proto" stuff starts with "ifndef __PYX_FORCE_INIT_THREADS", just in
case uses want to pass this flag to the C compiler a
-D__PYX_FORCE_INIT_THREADS=1/0 to force/skip the hardwired call. For
example, if you suspect that Py2.4.1 could have the same problem, then
you could try do pass -D__PYX_FORCE_INIT_THREADS=1 and see what
happens.

But if you think the above stuff is just a nonsense, and we should
just uncondionally make the PyEval_InitThread() call in the Py2.3
case, just let me know and I'll write that like you want.





-- 
Lisandro Dalcín
---------------
Centro Internacional de Métodos Computacionales en Ingeniería (CIMEC)
Instituto de Desarrollo Tecnológico para la Industria Química (INTEC)
Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET)
PTLC - Güemes 3450, (3000) Santa Fe, Argentina
Tel/Fax: +54-(0)342-451.1594
_______________________________________________
Cython-dev mailing list
[email protected]
http://codespeak.net/mailman/listinfo/cython-dev

Reply via email to