On 5/17/19 12:16 PM, Richard W.M. Jones wrote:
On Wed, May 15, 2019 at 10:57:58PM -0500, Eric Blake wrote:
> +static int
> +can_cache_wrapper (void *h)
> +{
> + CAMLparam0 ();
> + CAMLlocal1 (rv);
> +
> + caml_leave_blocking_section ();
> +
> + rv = caml_callback_exn (can_cache_fn, *(value *) h);
> + if (Is_exception_result (rv)) {
> + nbdkit_error ("%s", caml_format_exception (Extract_exception (rv)));
> + caml_enter_blocking_section ();
> + CAMLreturnT (int, -1);
> + }
> +
> + caml_enter_blocking_section ();
> + CAMLreturnT (int, Int_val (rv));
The not very obvious implicit assumption here is that the order of the
Cache* flags in the OCaml code is the same as the numbering of the
NBDKIT_CACHE_* flags in the C code, which it is, so we should be good.
The fua code did similarly, so I wasn't too worried (copy and paste is
nice...)
--
Eric Blake, Principal Software Engineer
Red Hat, Inc. +1-919-301-3226
Virtualization:
qemu.org |
libvirt.org