Received-SPF: pass (google.com: domain of linux-kernel+bounces-48557-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Sender: Tejun Heo <htejun@gmail.com>
Date: Thu, 1 Feb 2024 06:54:45 -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: <ZbvM1V8CYNeds14r@slm.duckdns.org>
References: <ZbqfMR_mVLaSCj4Q@carbonx1>
 <ZbrJq3X63hIMkbl2@slm.duckdns.org>
 <8a266076-b3dc-4a39-aac4-089e2ef77da3@gmx.de>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <8a266076-b3dc-4a39-aac4-089e2ef77da3@gmx.de>

Hello, Helge.

On Thu, Feb 01, 2024 at 05:41:10PM +0100, Helge Deller wrote:
> > Hmm... I have a hard time imagining a scenario where some CPUs don't have
> > pwq installed on wq->cpu_pwq. Can you please run `drgn
> > tools/workqueue/wq_dump.py` before triggering the hotplug event and paste
> > the output along with full dmesg?
> 
> I'm not sure if parisc is already fully supported with that tool, or
> if I'm doing something wrong:
> 
> root@debian:~# uname -a
> Linux debian 6.8.0-rc1-32bit+ #1292 SMP PREEMPT Thu Feb  1 11:31:38 CET 2024 parisc GNU/Linux
> 
> root@debian:~# drgn --main-symbols -s ./vmlinux ./wq_dump.py
> Traceback (most recent call last):
>   File "/usr/bin/drgn", line 33, in <module>
>     sys.exit(load_entry_point('drgn==0.0.25', 'console_scripts', 'drgn')())
>              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>   File "/usr/lib/python3/dist-packages/drgn/cli.py", line 301, in _main
>     runpy.run_path(script, init_globals={"prog": prog}, run_name="__main__")
>   File "<frozen runpy>", line 291, in run_path
>   File "<frozen runpy>", line 98, in _run_module_code
>   File "<frozen runpy>", line 88, in _run_code
>   File "./wq_dump.py", line 78, in <module>
>     worker_pool_idr         = prog['worker_pool_idr']
>                               ~~~~^^^^^^^^^^^^^^^^^^^
> KeyError: 'worker_pool_idr'

Does the kernel have CONFIG_DEBUG_INFO enabled? If you can look up
worker_pool_idr in gdb, drgn should be able to do the same.

> Maybe you have an idea? I'll check further, but otherwise it's probably
> easier for me to add some printk() to the kernel function wq_update_pod()
> and send that info?

Can you first try with drgn? The script dumps all the config info, so it's
likely easier to view that way. If that doesn't work out, I can write up a
debug patch.

Thanks.

-- 
tejun