Re: [PATCH v3 sched_ext/for-7.3 21/36] sched_ext: Maintain per-cpu effective cap copies for single-read checks

[Tejun Heo]

On Tue, 07 Jul 2026 00:38:20 +0000, sashiko-bot@xxxxxxxxxx wrote:
> - [High] Unconditional dereference of global CID routing tables in `scx_process_sync_ecaps()` leads to NULL pointer dereference on error paths and out-of-bounds reads during layout updates.

The routing-table reads sit behind the llist_empty() guard, and the
teardown/discard path doesn't call scx_process_sync_ecaps(), so there's
no window where it runs against a torn-down or mid-update layout. Same as
the v2 round. No change.

Thanks.

-- 
tejun