Re: [PATCH v3 1/3] cgroup/cpuset: return PERR_NOCPUS in remote_partition_enable()

[Hui Peng <benquike@gmail.com>]

On Mon, Sep 21, 2026 at 03:08:55PM +0800, Ridong Chen wrote:
> What is going on here?
>
> Please do not reply with a new series. And I think the changes should be listed
> in the cover letter. Otherwise, it will get very confusing.
>
> ...
>
> I cannot apply this patch. It seems quite different from v2.

Hi Ridong,

Sorry for the confusion and the noise. Two mistakes happened when I
prepared v3 after splitting patch 2/2 per Guopeng's feedback:

1. I mistakenly passed --in-reply-to pointing to the v2 thread and
   omitted the 0/3 cover letter.
2. Patch 1/3 was accidentally overwritten with a broken diff instead of
   the remote_partition_enable() hunk from v2 1/2:

@@ -1591,10 +1591,11 @@ static int remote_partition_enable(struct cpuset *cs, int new_prs,
 	 * above it or remote partition root underneath it is not allowed.
 	 */
 	compute_excpus(cs, tmp->new_cpus);
-	WARN_ON_ONCE(cpumask_intersects(tmp->new_cpus, subpartitions_cpus));
 	if (!cpumask_intersects(tmp->new_cpus, cpu_active_mask) ||
 	    cpumask_subset(top_cpuset.effective_cpus, tmp->new_cpus))
 		return PERR_INVCPUS;
+	if (cpumask_intersects(tmp->new_cpus, subpartitions_cpus))
+		return PERR_NOCPUS;

Would it be OK if I post a clean, standalone [PATCH v4 0/3..3/3] series
(as a new thread with a 0/3 cover letter listing the changelog and with
the v2 1/2 diff restored in patch 1/3)?

Best regards,
Hui