Re: [rfc] epoll interface change and glibc bits ...

[Jamie Lokier <lk@tantalophile.demon.co.uk>]

Mark Mielke wrote:
> It looks fine to me for as long as we can guarantee that sizeof(void *)
> will be less than or equal to sizeof(__uint64_t) (relatively safe).

It is fine, for epoll itself, even if sizeof(void *) > sizeof(__uint64_t).

Conversion functions for legacy 32-bit code running on a 128-bit chip
will be more complex, but epoll is the _least_ of problems in that case...

> I still prefer 'userdata' over 'obj', but the name of thing is not very
> important to me.

The AIO subsystem has a very similar cookie, which is simply declared
as `__u64 aio_data' in user-supplied requests, and returned in the
responses.  It's converted to a field called `ki_user_data' in the
kernel.

Just a few reference points...  I like `user_data' myself.

> I'm not sure if this is wise or not, but an 'fd' member might be
> useful as well:
> [...]
> For applications that wish to store fd's (probably common due to
> poll() roots), this would help them avoid casting magic as well. Also,
> it allows for 64 bit file descriptors if that ever became necessary.

That is a good idea, if we go for the union.

-- Jamie