[PATCH] rust: lsm: replace context+len with lsm_context

[Paul Moore]

On Fri, Nov 1, 2024 at 1:24 PM Alice Ryhl <aliceryhl at google.com> wrote:
> On Fri, Nov 1, 2024 at 6:11 PM Paul Moore <paul at paul-moore.com> wrote:
> > On Fri, Nov 1, 2024 at 1:04 PM Miguel Ojeda
> > <miguel.ojeda.sandonis at gmail.com> wrote:
> > > On Fri, Nov 1, 2024 at 5:56 PM Paul Moore <paul at paul-moore.com> wrote:
> > > >
> > > > Thanks Alice.  Would you like me to pull this in via the LSM tree with
> > > > the associated LSM changes, or would you prefer to do this some other
> > > > way?
> > > >
> > > > I'm going to merge this into lsm/dev for now so that we fix the issue
> > > > in linux-next, but I'm happy to drop it or do something else, let me
> > > > know.
> > >
> > > Christian has the VFS side, and both are needed for this -- do you
> > > mean you will cross-merge vfs' branch too?
> >
> > I think our last emails crossed paths.  I'm not going to merge this
> > via the LSM tree as we don't have the Rust security.c helpers.
> > Ideally it would have been better to have the Rust LSM/security
> > helpers in the LSM tree for reasons like this, but it looks like it's
> > too late for that now.
>
> If Christian is okay with rewriting the vfs.rust.file tree, we can
> drop commit 94d356c0335f ("rust: security: add abstraction for
> secctx") from there and I'll update it and send it for inclusion in
> the LSM tree instead. I'll need to drop the piece that ties together
> `struct cred` and `secctx` from the patch, but I can follow up with a
> small patch for that for the 6.14 merge window.

I can only guess at what Chrisitian wants to do, but my guess is that
he isn't going to be very excited about rewriting a VFS tree at this
stage ... which is very understandable as far as I'm concerned.

I wouldn't worry too much about this right now, I'm going to plan on
holding Casey's patchset in a staging area until after the upcoming
merge window.

-- 
paul-moore.com