From: sashiko-bot@kernel.org
To: "Tejun Heo" <tj@kernel.org>
Cc: sched-ext@lists.linux.dev
Subject: Re: [PATCH v2 sched_ext/for-7.3 15/36] sched_ext: Add scx_cmask_ref for validated arena cmask access
Date: Mon, 06 Jul 2026 01:58:39 +0000 [thread overview]
Message-ID: <20260706015840.5CD561F000E9@smtp.kernel.org> (raw)
In-Reply-To: <20260706014058.439853-16-tj@kernel.org>
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Non-atomic Read-Modify-Write in `scx_cmask_ref_or` and `scx_cmask_ref_copy` clobbers concurrent BPF updates to adjacent bits.
--
commit 1141e7d64b0f45ca499fc3944be4ed4e896ff3b8
Author: Tejun Heo <tj@kernel.org>
sched_ext: Add scx_cmask_ref for validated arena cmask access
This commit introduces scx_cmask_ref to safely validate and access BPF
arena memory used for struct scx_cmask. It takes a snapshot of the base
and nr_cids to protect against concurrent BPF mutations, and provides bounded
shard read and writeback functions like scx_cmask_ref_or() and
scx_cmask_ref_copy().
> diff --git a/kernel/sched/ext/cid.c b/kernel/sched/ext/cid.c
> index 6a22bead4c601..42fc8348b3b4c 100644
> --- a/kernel/sched/ext/cid.c
> +++ b/kernel/sched/ext/cid.c
[ ... ]
> @@ -669,6 +675,12 @@ static __always_inline bool cmask_word_op2(u64 *av, const u64 *bp, u64 mask,
> return (READ_ONCE(*bp) & ~READ_ONCE(*av)) & mask;
> case CMASK_OP2_INTERSECTS:
> return (READ_ONCE(*av) & READ_ONCE(*bp)) & mask;
> + case CMASK_OP2_REF_OR:
> + WRITE_ONCE(*av, READ_ONCE(*av) | (READ_ONCE(*bp) & mask));
> + return false;
> + case CMASK_OP2_REF_COPY:
> + WRITE_ONCE(*av, (READ_ONCE(*av) & ~mask) | (READ_ONCE(*bp) & mask));
[Severity: High]
Does this non-atomic read-modify-write clobber concurrent BPF updates to
adjacent bits?
Since this code uses READ_ONCE() and WRITE_ONCE() but performs a
read-modify-write sequence, if a BPF program concurrently updates other bits
in the same 64-bit word (for example, using atomic bit operations), could
those updates be lost when this kernel thread writes back the modified word?
Would a cmpxchg loop or a proper atomic bitwise operation be necessary here
to prevent data loss under concurrency?
> + return false;
> }
> unreachable();
> }
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260706014058.439853-1-tj@kernel.org?part=15
next prev parent reply other threads:[~2026-07-06 1:58 UTC|newest]
Thread overview: 73+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-06 1:40 [PATCHSET v2 sched_ext/for-7.3] sched_ext: Capability-based CPU delegation for sub-schedulers Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 01/36] sched_ext: Add SCX_CALL_CID_OP_TASK() for cid-form op dispatch Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 02/36] sched_ext: Rename extra_enq_flags to remote_activate_enq_flags Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 03/36] sched_ext: Fix premature ops->priv publication in scx_alloc_and_add_sched() Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 04/36] tools/sched_ext: scx - Fix cmask_subset(), cmask_equal() and cmask_weight() Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 05/36] sched_ext: Use READ_ONCE/WRITE_ONCE in cmask word ops and drop _RACY variants Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 06/36] tools/sched_ext: scx_qmap - Use bare u64/u32/s32 integer types Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 07/36] sched_ext: Reject direct slice and dsq_vtime writes for cid-form schedulers Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 08/36] sched_ext: Make scx_bpf_kick_cid() return void Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 09/36] sched_ext: Make the kick machinery per-sched Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 10/36] sched_ext: Add ops.init_cids() to finalize the cid layout before init Tejun Heo
2026-07-06 2:08 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 11/36] sched_ext: Add CID sharding Tejun Heo
2026-07-06 1:53 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 12/36] sched_ext: Add shard boundaries to scx_bpf_cid_override() Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 13/36] sched_ext: Defer scx_sched kobj sysfs add into the enable workfns Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 14/36] sched_ext: Add per-shard scx_sched storage scaffolding Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 15/36] sched_ext: Add scx_cmask_ref for validated arena cmask access Tejun Heo
2026-07-06 1:58 ` sashiko-bot [this message]
2026-07-06 23:40 ` Tejun Heo
2026-07-06 9:04 ` Andrea Righi
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 16/36] sched_ext: Build the set_cmask scratch from trusted geometry Tejun Heo
2026-07-06 1:59 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 17/36] sched_ext: RCU-protect the sub-sched tree's children/sibling lists Tejun Heo
2026-07-06 2:08 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 18/36] sched_ext: Add scx_skip_subtree_pre() Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 19/36] sched_ext: Add per-shard cap delegation for sub-schedulers Tejun Heo
2026-07-06 2:13 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 20/36] sched_ext: Add coalescing sub_caps_updated() notifier " Tejun Heo
2026-07-06 2:06 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 21/36] sched_ext: Maintain per-cpu effective cap copies for single-read checks Tejun Heo
2026-07-06 2:07 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 22/36] sched_ext: Add sub_ecaps_updated() effective-cap change notifier Tejun Heo
2026-07-06 2:09 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 23/36] sched_ext: Generalize local-DSQ handling to rq-owned DSQs Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 24/36] sched_ext: Add reject DSQ for cap-rejected dispatches Tejun Heo
2026-07-06 2:08 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 25/36] sched_ext: Add the SCX_CAP_ENQ_IMMED cap Tejun Heo
2026-07-06 2:10 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 26/36] sched_ext: Assign a unique id to each scheduler instance Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 27/36] sched_ext: Route task slice writes through set_task_slice() Tejun Heo
2026-07-06 2:05 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 9:10 ` Andrea Righi
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 28/36] sched_ext: Tie cpu occupancy to SCX_CAP_BASE through the task slice Tejun Heo
2026-07-06 2:07 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 29/36] sched_ext: Add the SCX_CAP_ENQ cap Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 30/36] sched_ext: Gate kicks on SCX_CAP_BASE and preemption on SCX_CAP_PREEMPT Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 31/36] sched_ext: Authorize remote-move inserts against the placing scheduler Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 32/36] sched_ext: Route ops.update_idle() to sub-schedulers and re-notify owed scheds Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 33/36] sched_ext: Replay ecaps notifications suppressed by bypass Tejun Heo
2026-07-06 2:11 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 34/36] sched_ext: Add scx_bpf_sub_kill() to evict a child sub-scheduler Tejun Heo
2026-07-06 2:12 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 35/36] tools/sched_ext: scx_qmap - Expand hierarchical sub-scheduling Tejun Heo
2026-07-06 2:10 ` sashiko-bot
2026-07-06 23:40 ` Tejun Heo
2026-07-06 1:40 ` [PATCH v2 sched_ext/for-7.3 36/36] tools/sched_ext: scx_qmap - Add sub-sched cap fault injection Tejun Heo
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=20260706015840.5CD561F000E9@smtp.kernel.org \
--to=sashiko-bot@kernel.org \
--cc=sashiko-reviews@lists.linux.dev \
--cc=sched-ext@lists.linux.dev \
--cc=tj@kernel.org \
/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.