From: Tejun Heo <tj@kernel.org>
To: Juntong Deng <juntong.deng@outlook.com>
Cc: ast@kernel.org, daniel@iogearbox.net, john.fastabend@gmail.com,
andrii@kernel.org, martin.lau@linux.dev, eddyz87@gmail.com,
song@kernel.org, yonghong.song@linux.dev, kpsingh@kernel.org,
sdf@fomichev.me, haoluo@google.com, jolsa@kernel.org,
memxor@gmail.com, void@manifault.com, arighi@nvidia.com,
changwoo@igalia.com, bpf@vger.kernel.org,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH sched_ext/for-6.15 v3 3/5] sched_ext: Add scx_kfunc_ids_ops_context_sensitive for unified filtering of context-sensitive SCX kfuncs
Date: Thu, 27 Feb 2025 11:32:38 -1000 [thread overview]
Message-ID: <Z8DZ9pqlWim8EIwk@slm.duckdns.org> (raw)
In-Reply-To: <AM6PR03MB5080C1F0E0F10BCE67101F6F99CD2@AM6PR03MB5080.eurprd03.prod.outlook.com>
Hello,
On Thu, Feb 27, 2025 at 09:23:20PM +0000, Juntong Deng wrote:
> > > + if (prog->type == BPF_PROG_TYPE_STRUCT_OPS &&
> > > + prog->aux->st_ops != &bpf_sched_ext_ops)
> > > + return 0;
> >
> > Why can't other struct_ops progs call scx_kfunc_ids_unlocked kfuncs?
> >
>
> Return 0 means allowed. So kfuncs in scx_kfunc_ids_unlocked can be
> called by other struct_ops programs.
Hmm... would that mean a non-sched_ext bpf prog would be able to call e.g.
scx_bpf_dsq_insert()?
> > > + /* prog->type == BPF_PROG_TYPE_STRUCT_OPS && prog->aux->st_ops == &bpf_sched_ext_ops*/
> > > +
> > > + moff = prog->aux->attach_st_ops_member_off;
> > > + flags = scx_ops_context_flags[SCX_MOFF_IDX(moff)];
> > > +
> > > + if ((flags & SCX_OPS_KF_UNLOCKED) &&
> > > + btf_id_set8_contains(&scx_kfunc_ids_unlocked, kfunc_id))
> > > + return 0;
> >
> > Wouldn't this disallow e.g. ops.dispatch() from calling scx_dsq_move()?
> >
>
> No, because
>
> > > [SCX_OP_IDX(dispatch)] = SCX_OPS_KF_DISPATCH | SCX_OPS_KF_ENQUEUE,
>
> Therefore, kfuncs (scx_bpf_dsq_move_*) in scx_kfunc_ids_dispatch can be
> called in the dispatch context.
I see, scx_dsq_move_*() are in both groups, so it should be fine. I'm not
fully sure the groupings are the actually implemented filtering are in sync.
They are intended to be but the grouping didn't really matter in the
previous implementation. So, they need to be carefully audited.
> > Have you tested that the before and after behaviors match?
>
> I tested the programs in tools/testing/selftests/sched_ext and
> tools/sched_ext and all worked fine.
>
> If there are other cases that are not covered, we may need to add new
> test cases.
Right, the coverage there isn't perfect. Testing all conditions would be too
much but it'd be nice to have a test case which at least confirms that all
allowed cases verify successfully.
Thanks.
--
tejun
next prev parent reply other threads:[~2025-02-27 21:32 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-02-26 19:24 [PATCH sched_ext/for-6.15 v3 0/5] bpf, sched_ext: Make kfunc filters support struct_ops context to reduce runtime overhead Juntong Deng
2025-02-26 19:28 ` [PATCH sched_ext/for-6.15 v3 1/5] bpf: Add struct_ops context information to struct bpf_prog_aux Juntong Deng
2025-02-26 19:28 ` [PATCH sched_ext/for-6.15 v3 2/5] sched_ext: Declare context-sensitive kfunc groups that can be used by different SCX operations Juntong Deng
2025-02-27 19:19 ` Tejun Heo
2025-02-28 21:57 ` Tejun Heo
2025-03-03 21:12 ` Juntong Deng
2025-03-04 18:00 ` Tejun Heo
2025-03-07 0:07 ` Juntong Deng
2025-02-26 19:28 ` [PATCH sched_ext/for-6.15 v3 3/5] sched_ext: Add scx_kfunc_ids_ops_context_sensitive for unified filtering of context-sensitive SCX kfuncs Juntong Deng
2025-02-27 20:25 ` Tejun Heo
2025-02-27 21:23 ` Juntong Deng
2025-02-27 21:32 ` Tejun Heo [this message]
2025-02-28 2:34 ` Alexei Starovoitov
2025-02-28 21:31 ` Tejun Heo
2025-02-28 23:06 ` Alexei Starovoitov
2025-02-28 18:42 ` Juntong Deng
2025-02-28 21:20 ` Tejun Heo
2025-02-28 2:38 ` Alexei Starovoitov
2025-02-28 21:14 ` Tejun Heo
2025-02-26 19:28 ` [PATCH sched_ext/for-6.15 v3 4/5] sched_ext: Removed mask-based runtime restrictions on calling kfuncs in different contexts Juntong Deng
2025-02-27 16:24 ` Andrea Righi
2025-02-27 16:49 ` Juntong Deng
2025-02-26 19:28 ` [PATCH sched_ext/for-6.15 v3 5/5] selftests/sched_ext: Update enq_select_cpu_fails to adapt to struct_ops context filter Juntong Deng
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=Z8DZ9pqlWim8EIwk@slm.duckdns.org \
--to=tj@kernel.org \
--cc=andrii@kernel.org \
--cc=arighi@nvidia.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=changwoo@igalia.com \
--cc=daniel@iogearbox.net \
--cc=eddyz87@gmail.com \
--cc=haoluo@google.com \
--cc=john.fastabend@gmail.com \
--cc=jolsa@kernel.org \
--cc=juntong.deng@outlook.com \
--cc=kpsingh@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=martin.lau@linux.dev \
--cc=memxor@gmail.com \
--cc=sdf@fomichev.me \
--cc=song@kernel.org \
--cc=void@manifault.com \
--cc=yonghong.song@linux.dev \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.