Messages in this thread | | | Date | Thu, 2 Feb 2023 08:51:51 -0800 | From | Boqun Feng <> | Subject | Re: [RFC 2/5] rust: sync: Arc: Introduces ArcInner::count() |
| |
On Thu, Feb 02, 2023 at 05:17:55PM +0100, Greg KH wrote: [...] > > > > But there are correct ways to use a refcount, e.g. if you own > > > > `Arc` and `.count()` returns 1, then you know that you are the > > > > exclusive owner of the `Arc` and nobody else is going to touch it. > > > > > > But you should never know this, as it is not relevant. > > > > > > So no, please don't allow printing out of a reference count, that will > > > only cause problems and allow people to think it is safe to do so. > > > > > > > People already do it, even in *security* code, > > > > security/keys/keyring.c: > > > > int key_link(struct key *keyring, struct key *key) > > { > > ... > > kenter("{%d,%d}", keyring->serial, refcount_read(&keyring->usage)); > > ... > > } > > > > Should we fix that? > > Yes. But really, that's debugging code, it probably should all be > removed now. >
Well, there are also printings in proc_keys_show() and proc_key_users_show() in security/keys/proc.c, and I think it's hard to remove these since they are userspace related.
Anyway I realise I could have done a better job explaining what I'm doing here:
I want to read refcount in a later patch, which make Arc<T> implement Debug trait/interface, and that allows user to print Arc<T> for debug purposes, e.g.
pr_info!("{:#x?}", a); // a is an "Arc<T">
that's the only usage of the reading from refcount. I could open code an FFI call in that implementation, but I thought maybe I could add a helper function, hence the "count" function. And since "count" is a private function, so no one can use it outside this rust/kernel/sync/arc.rs file, therefore mis-usage by outsiders are impossible.
Maybe I made things confusing because I just learned the language and wanted to try out a few things which made things complicated (for review), hope the above explanation undo some of the confusion I created.
As I said, I'm open to remove the printing of the refcount, and if you and Peter think maybe it's OK to do that after the explanation above, I will improve the patch to make things clear ;-)
Regards, Boqun
> thanks, > > greg k-h
| |