Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Message-ID: <4de89aeb-c127-908c-b403-f9cea128a0b0@igalia.com>
Date:   Thu, 20 Jan 2022 12:51:11 -0300
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.5.0
Subject: Re: [PATCH V3] panic: Move panic_print before kmsg dumpers
Content-Language: en-US
To:     Petr Mladek <pmladek@suse.com>
Cc:     Baoquan He <bhe@redhat.com>, linux-kernel@vger.kernel.org,
        akpm@linux-foundation.org, kernel@gpiccoli.net,
        senozhatsky@chromium.org, rostedt@goodmis.org,
        john.ogness@linutronix.de, feng.tang@intel.com,
        kexec@lists.infradead.org, dyoung@redhat.com,
        keescook@chromium.org, anton@enomsg.org, ccross@android.com,
        tony.luck@intel.com
References: <20220114183046.428796-1-gpiccoli@igalia.com>
 <20220119071318.GA4977@MiWiFi-R3L-srv> <YegytkfED+QI56Y8@alley>
 <94bb12a2-a788-cee6-7d4f-dc0ac581fb39@igalia.com> <YektvNyN6mAHv9jJ@alley>
From:   "Guilherme G. Piccoli" <gpiccoli@igalia.com>
In-Reply-To: <YektvNyN6mAHv9jJ@alley>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 20/01/2022 06:39, Petr Mladek wrote:
> [...]
> It makes perfect sense to disable the watchdogs during panic().
> For example, rcu_panic() just sets a variable:
> 
> static int rcu_panic(struct notifier_block *this, unsigned long ev, void *ptr)
> {
> 	rcu_cpu_stall_suppress = 1;
> 	return NOTIFY_DONE;
> }
> 
> It is quick and super-safe. It might make sense to implement similar
> thing for other watchdogs and do something like:
> 
> void panic_supress_watchdogs(void)
> {
> 	rcu_panic();
> 	softlockup_watchog_panic();
> 	wq_watchog_panic();
> }
> 
> It might be caller early in panic().
> 

For reference, I saw your great input about this subject in another
related thread, which I think we should mention here:

https://lore.kernel.org/lkml/Yel8WQiBn%2FHNQN83@alley/


> JFYI, there is an extension for the crash tool that might be able to read
> the trace log, see https://crash-utility.github.io/extensions.html
> 
> I haven't tested it myself yet.

Thanks, nice to know =)