Re: [PATCH v20 4/8] fork: Add shadow stack support to clone3()

On Fri, Sep 05, 2025 at 05:21:59PM +0200, Christian Brauner wrote:

On Tue, Sep 02, 2025 at 11:21:48AM +0100, Mark Brown wrote:

quoted

+		.shadow_stack_token	= args.shadow_stack_token,

I'm not sure why this has to be named "shadow_stack_token" I think
that's just confusing and we should just call it "shadow_stack" and be
done with it. It's also a bit long of a field name imho.

I'm not hugely attached to the name, if you want to rename that's
perfectly fine by me.  My thinking was that there's a potential
confusion with it being a pointer to the base of the shadow stack by
comparison with the existing "stack" but I do agree that the resulting
name is quite long and if someone does actually get confused they should
discover the problem fairly rapidly in testing.  ss_token would shorter
but the abbreviation is less clear, whatever name you prefer is fine by
me.

I have a kernel-6.18.clone3 branch
https://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs.git/log/?h=kernel-6.18.clone3
because there's another cross-arch cleanup that cleans up copy_thread(),
copy_sighand(), and copy_process() and - surprisingly - also adds
clone3() support for nios2...

Anyway, if you just want me to slap it on top of that branch then I can
simply rename while applying so no need to resend in that case.

That would be amazing, I'm totally happy with you doing that.  If I do
need to rebase and resend let me know.

It's probably worth mentioning that the RISC-V shadow stack support was
getting near to being merged, if that ends up being OK for this release
(it's not been posted yet though) there might be some cross tree merge
needed or something.