[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <51ab2b78-a409-4201-97a0-30e976571df8@redhat.com>
Date: Thu, 24 Jul 2025 21:56:07 -0400
From: Waiman Long <llong@...hat.com>
To: Thomas Gleixner <tglx@...utronix.de>, Waiman Long <llong@...hat.com>,
Pavel Machek <pavel@....cz>
Cc: kernel list <linux-kernel@...r.kernel.org>, mingo@...hat.com,
bp@...en8.de, dave.hansen@...ux.intel.com, x86@...nel.org, hpa@...or.com,
peterz@...radead.org, will@...nel.org, miriam.rachel.korenblit@...el.com,
linux-wireless@...r.kernel.org, Petr Mladek <pmladek@...e.com>,
John Ogness <jogness@...utronix.de>
Subject: Re: locking problems in iwlwifi? was Re: 6.16-rcX: crashing way too
often on thinkpad X220
On 7/24/25 6:21 PM, Thomas Gleixner wrote:
> On Thu, Jul 24 2025 at 15:22, Waiman Long wrote:
>> On 7/24/25 3:07 PM, Pavel Machek wrote:
>>>> The lockdep warning just means that console_owner_lock is acquired both in
>>>> softirq context and in task context with interrupt enabled. That can leads
>>>> to deadlock. So the remedy is to always take console_owner_lock with
>>>> interrupt disabled, i.e. with
>>>> raw_spin_lock_irqsave/raw_spin_lock_irqrestore.
>>> So you suspect problem is in the printk code?
>> This lockdep problem is likely unrelated to the IRQ flag problem
>> reported in the original email as it shouldn't affect the IRQ flag.
> Wrong you are ... :)
>
Sorry for coming to a conclusion early without looking more deeply into
the printk code. I do admit that printk() isn't an area that I know
enough about.
Will look harder next time.
Thanks,
Longman
Powered by blists - more mailing lists