[Gc] Problems with exceptions in OSX
hans.boehm at hp.com
Wed Sep 9 20:00:06 PDT 2009
I'll generate a revised patch with the one line page and either check it in or post it. I can only test on 10.2/PowerPC unfortunately.
> -----Original Message-----
> From: gc-bounces at napali.hpl.hp.com
> [mailto:gc-bounces at napali.hpl.hp.com] On Behalf Of Ivan Maidanski
> Sent: Wednesday, September 09, 2009 1:39 PM
> To: gc at napali.hpl.hp.com
> Subject: Re: [Gc] Problems with exceptions in OSX
> "Boehm, Hans" <hans.boehm at hp.com> wrote:
> > Ivan just reminded me of this patch, also his diff104_cvs.
> > I was trying to understand it and the original code, and
> failed at both. Since I'm not a MacOS programmer, I'm
> probably missing something basic.
> > Could someone explain the role of thread_state here, in
> either version of the code? As far as I can tell, it gets
> set to NULL, and is then passed to thread_get_state, which
> presumably is intended to somehow set it, so that it can then
> be used by thread_set_state or the like. Since I think this
> is compiled as C code, and there are thus no reference
> parameters, it's unclear to me how thread_get_state can
> possibly have any impact on a null pointer value. Can
> someone enlighten me?
> > I realize that this is largely orthogonal to the patch ...
> > Thanks.
> > Hans
> I think it's a typo - "thread_state_t" should be, in fact,
> changed to "thread_state_data_t" (which is an array type so
> always passed by reference in C), and it shouldn't be
> initialized to null.
> For more, eg., see (as a reference) the following code:
> Gc mailing list
> Gc at linux.hpl.hp.com
More information about the Gc