On 3/5/21 5:31 PM, Eric Blake wrote:
As promised in an earlier patch, get rid of the temporary nbdkit_backend_reopen, and replace it with the finer-grained combination of nbdkit_context_get_backend, nbdkit_context_next_open, nbdkit_context_next_close, and nbdkit_context_set_next. Update the retry filter to use the new API. Change the (temporary) preprocessor guard from NBDKIT_RETRY_FILTER to NBDKIT_TYPESAFE (will be dropped later once more filters are converted). Drop the now-unused backend_reopen() function. Expose .prepare and .finalize callbacks into struct nbdkit_next_ops for use during manual context manipulation.
Insufficient testing on my part:
+++ b/docs/nbdkit-filter.pod @@ -132,25 +132,33 @@ C<nbdkit_next_after_fork>, C<nbdkit_next_preconnect>, C<nbdkit_next_list_exports>, C<nbdkit_next_default_export>, C<nbdkit_next_open>) and a structure called C<struct nbdkit_next_ops>. These abstract the next plugin or filter in the chain. There is also
+an opaque pointer C<backend>, C<context> or C<nxdata> which must be +passed along when calling these functions. The value of C<backend> is +stable between C<.after_fork>, C<.preconnect>, C<.list_exports>, and +C<.default_export>, and can also be obtained by using +C<nbdkit_context_get_backend> on the C<context> parameter to C<.open>.
This promise makes sense,
@@ -468,31 +563,21 @@ C<nbdkit_is_tls>. The filter should generally call C<next> as its first step, to allocate from the plugin outwards, so that C<.close> running from the outer filter to the plugin will be in reverse. Skipping a call to -C<next> is acceptable if the filter will not access C<next_ops> -during any of the remaining callbacks reached on the same connection. +C<next> is acceptable if the filter will not access C<next_ops> during +any of the remaining callbacks reached on the same connection. The +C<next> function is provided for convenience; the same functionality +can be obtained manually (other than error checking) by using the +following: + + nbdkit_context_set_next (context, nbdkit_next_context_open + (nbdkit_context_get_backend (context), readonly, exportname));
as does this documentation,
+++ b/server/filters.c @@ -263,12 +263,14 @@ filter_default_export (struct backend *b, int readonly, int is_tls) static int next_open (struct context *c, int readonly, const char *exportname) { - struct backend *b = c->b; - struct context *c_next = backend_open (b->next, readonly, exportname); + struct backend *b = nbdkit_context_get_backend (c); + struct context *c_next = nbdkit_next_context_open (b, readonly, exportname);
and this even tries to mirror the documentation,
+ struct context *old;
if (c_next == NULL) return -1; - c->c_next = c_next; + old = nbdkit_context_set_next (c, c_next); + assert (old == NULL); return 0; }
@@ -694,12 +696,35 @@ filter_register (struct backend *next, size_t index, const char *filename, return (struct backend *) f; }
-int -nbdkit_backend_reopen (struct context *c, int readonly, - const char *exportname, struct context **nxdata) +struct backend * +nbdkit_context_get_backend (struct context *c) { - int r = backend_reopen (c, readonly, exportname); + assert (c); + return c->b;
but this is returning the wrong context (that of the current filter, rather than of the backend)
+} + +struct context * +nbdkit_next_context_open (struct backend *b, + int readonly, const char *exportname) +{ + assert (b); + return backend_open (b->next, readonly, exportname);
and this is compensating by calling backend_open() on the wrong context (it should be calling it on b, rather than on b->next). I'm fixing the bug and enhancing test-layers-filter.c to catch the problem. -- Eric Blake, Principal Software Engineer Red Hat, Inc. +1-919-301-3226 Virtualization: qemu.org | libvirt.org