Re: [PATCH 6/6] eventfs: clean up dentry ops and add revalidate function

[Linus Torvalds]

On Tue, 30 Jan 2024 at 18:46, Al Viro <viro@xxxxxxxxxxxxxxxxxx> wrote:
>
> What's to stop ->d_revalidate() from being called in parallel with
> __dentry_kill()?

Oh, you're right.

For some reason I thought we did the d_release() _after_ the RCU grace
period, but we don't.

Why don't we, btw? It would be so much better if we did the
d_release() from __d_free(). We have all that smarts in fs/dcache.c to
decide if we need to RCU-delay it or not, and then we don't let
filesystems use it.

I assume the reason is that some 'd_delete' cases might want to sleep
etc. Still, for things like this that just want to release memory, it
would be *much* better to have d_release called when the dentry is
actually released.

Hmm. Not very many d_delete cases, but I did see a couple that
definitely want process context (dma_buf_release goes to things that
do vfree() etc).

So I guess the "make low-level filesystems do their own kfree_rcu() is
what we're doing.

In this case it's as simple as doing that

-       kfree(ei);
+       kfree_rcu(ei, rcu);

and we'd just make the rcu entry a union with something that isn't
that 'is_freed' field so that it doesn't take more space.

             Linus