Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <CAMdYzYpF4FNTBPZsEFeWRuEwSies36QM_As8osPWZSr2q-viEA@mail.gmail.com>
 <87y1y48spg.fsf@jogness.linutronix.de>
In-Reply-To: <87y1y48spg.fsf@jogness.linutronix.de>
From:   Peter Geis <pgwipeout@gmail.com>
Date:   Fri, 10 Jun 2022 11:34:47 -0400
Message-ID: <CAMdYzYr-Wo713Y4qjboTpoK6GcrYfKCfRJAEizwXw6-=dymVzg@mail.gmail.com>
Subject: Re: [BUG] Threaded printk breaks early debugging
To:     John Ogness <john.ogness@linutronix.de>
Cc:     Petr Mladek <pmladek@suse.com>,
        Sergey Senozhatsky <senozhatsky@chromium.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        "open list:ARM/Rockchip SoC..." <linux-rockchip@lists.infradead.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Fri, Jun 10, 2022 at 11:05 AM John Ogness <john.ogness@linutronix.de> wrote:
>
> Hi Peter,
>
> On 2022-06-10, Peter Geis <pgwipeout@gmail.com> wrote:
> > However I've run into an issue debugging early boot issues. Anything
> > that causes the kernel threading system to die (for example here, a
> > NPE) causes the boot console to halt before it outputs the error.
>
> A null pointer exception should trigger a panic, which will cause printk
> to enter direct mode for the oops. It should be just as effective as
> before introducing the printing kthreads.

This might be a side effect of the fact that this is on a low powered
arm64 board. I noticed with threading enabled during large bursts the
console drops an excessive amount of messages. It's especially
apparent during the handover from earlycon to the normal console.

>
> Can you please provide a test case or patch that will trigger the
> problem you refer to?

Unfortunately no, unless you have a Rockchip rk356x device with both
DSI and HDMI. We're working on adding DSI support to the new video
output driver, and when both are enabled it tosses a NPE.

>
> > Would it be possible to have both a kconfig option and a runtime
> > option to toggle printk_fallback_preferred_direct()? Currently I've
> > had to force it at all times while debugging.
>
> It would be trivial to provide a boot parameter for this behavior, but I
> think we first need to see a real problem and also see if we can fix
> that problem.
>
> John Ogness