Received-SPF: pass (google.com: domain of linux-kernel+bounces-49069-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
Sender: Tejun Heo <htejun@gmail.com>
Date: Thu, 1 Feb 2024 15:39:05 -1000
From: Tejun Heo <tj@kernel.org>
To: Helge Deller <deller@gmx.de>
Cc: Helge Deller <deller@kernel.org>,
	Lai Jiangshan <jiangshanlai@gmail.com>,
	linux-kernel@vger.kernel.org, linux-parisc@vger.kernel.org
Subject: Re: [PATCH][RFC] workqueue: Fix kernel panic on CPU hot-unplug
Message-ID: <ZbxHuS6vOc0MB7-R@slm.duckdns.org>
References: <ZbqfMR_mVLaSCj4Q@carbonx1>
 <ZbrJq3X63hIMkbl2@slm.duckdns.org>
 <8a266076-b3dc-4a39-aac4-089e2ef77da3@gmx.de>
 <ZbvM1V8CYNeds14r@slm.duckdns.org>
 <e6916a78-7872-442d-922d-31ea3920da4f@gmx.de>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <e6916a78-7872-442d-922d-31ea3920da4f@gmx.de>

Hello,

On Thu, Feb 01, 2024 at 06:56:20PM +0100, Helge Deller wrote:
> root@debian:~# drgn --main-symbols -s ./vmlinux ./wq_dump.py 2>&1 | tee L
> Affinity Scopes
> ===============
> wq_unbound_cpumask=0000ffff
> 
> CPU
>   nr_pods  16
>   pod_cpus [0]=00000001 [1]=00000002 [2]=00000004 [3]=00000008 [4]=00000010 [5]=00000020 [6]=00000040 [7]=00000080 [8]=00000100 [9]=00000200 [10]=00000400 [11]=00000800 [12]=00001000 [13]=00002000 [14]=00004000 [15]=00008000
>   pod_node [0]=0 [1]=0 [2]=0 [3]=0 [4]=0 [5]=0 [6]=0 [7]=0 [8]=0 [9]=0 [10]=0 [11]=0 [12]=0 [13]=0 [14]=0 [15]=0
>   cpu_pod  [0]=0 [1]=1

wq_unbound_cpumask is saying there are 16 possible cpus but
for_each_possible_cpu() iteration is only giving two. Can you please apply
the following patch and post the boot dmesg? Thanks.

diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index ffb625db9771..d3fa2bea4d75 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -7146,6 +7146,9 @@ void __init workqueue_init_early(void)
 	BUG_ON(!alloc_cpumask_var(&wq_requested_unbound_cpumask, GFP_KERNEL));
 	BUG_ON(!zalloc_cpumask_var(&wq_isolated_cpumask, GFP_KERNEL));
 
+	printk("XXX workqueue_init_early: possible_cpus=%*pb\n",
+	       cpumask_pr_args(cpu_possible_mask));
+
 	cpumask_copy(wq_unbound_cpumask, cpu_possible_mask);
 	restrict_unbound_cpumask("HK_TYPE_WQ", housekeeping_cpumask(HK_TYPE_WQ));
 	restrict_unbound_cpumask("HK_TYPE_DOMAIN", housekeeping_cpumask(HK_TYPE_DOMAIN));
@@ -7290,6 +7293,9 @@ void __init workqueue_init(void)
 	struct worker_pool *pool;
 	int cpu, bkt;
 
+	printk("XXX workqueue_init: possible_cpus=%*pb\n",
+	       cpumask_pr_args(cpu_possible_mask));
+
 	wq_cpu_intensive_thresh_init();
 
 	mutex_lock(&wq_pool_mutex);
@@ -7401,6 +7407,9 @@ void __init workqueue_init_topology(void)
 	struct workqueue_struct *wq;
 	int cpu;
 
+	printk("XXX workqueue_init_topology: possible_cpus=%*pb\n",
+	       cpumask_pr_args(cpu_possible_mask));
+
 	init_pod_type(&wq_pod_types[WQ_AFFN_CPU], cpus_dont_share);
 	init_pod_type(&wq_pod_types[WQ_AFFN_SMT], cpus_share_smt);
 	init_pod_type(&wq_pod_types[WQ_AFFN_CACHE], cpus_share_cache);