On Wed, Jul 30, 2025 at 07:55:09PM +0200, Toon Claes wrote: > Allow users of the commit-graph to explicitly prepare the commit-graph. > This can be useful when users want to start using bloom keys before > calling functions like prepare_revision_walk(). We'll use this exported > function in a subsequent commit. Hm. Ideally we wouldn't have to expose this low-level function and the commit-graph subsystem would know to handle this. We typically have patterns like this in our codebase: if (repo_find_commit_pos_in_graph(r, c, &graph_pos)) load_bloom_filter_from_graph(r->objects->commit_graph, filter, graph_pos); The call to `repo_find_commit_pos_in_graph()` knows to call `prepare_commit_graph()`, so no manual call to that function would be required. I haven't yet read the next commit though that adds the callsite. So let's read on. > diff --git a/commit-graph.h b/commit-graph.h > index 78ab7b875b..0f76681333 100644 > --- a/commit-graph.h > +++ b/commit-graph.h > @@ -131,6 +131,14 @@ struct repo_settings; > struct commit_graph *parse_commit_graph(struct repo_settings *s, > void *graph_map, size_t graph_size); > > +/* > + * Return 1 if commit_graph is non-NULL, and 0 otherwise. > + * > + * On the first invocation, this function attempts to load the commit > + * graph if the_repository is configured to have one. > + */ > +int prepare_commit_graph(struct repository *r); Let's fix the reference to `the_repository` while at it. Patrick
