Re: [patch] clone_startup(), 2.5.31-A0

David Mosberger (davidm@napali.hpl.hp.com)
Tue, 13 Aug 2002 09:32:50 -0700


>>>>> On Tue, 13 Aug 2002 17:11:38 +0100, Christoph Hellwig <hch@infradead.org> said:

Chris> On Tue, Aug 13, 2002 at 10:09:24AM -0600, Erik Andersen
Chris> wrote:
>> > First the name souns horrible. What about spawn_thread or
>> create_thread > instead? it's not our good old clone and not a
>> lookalike, it's some > pthreadish monster..
>>
>> How about "clone2"?

Chris> Already used by ia64 for a hybrid between the good old clone
Chris> and the new monster :)

The original clone() system call was misdesigned. Even if you chose
to ignore ia64, clone() cannot be used by portable applications to
specify a stack (think "stack-growth direction").

clone() should have specified a stack memory *area* from the
beginning. (UNIX got this right from the beginning, see, e.g.,
sigaltstack()).

--david
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/