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: <20230410092634.4870-1-David.Rau.opensource@dm.renesas.com> <d9dac02a-328d-485e-9aff-158852eeb990@sirena.org.uk>
In-Reply-To: <d9dac02a-328d-485e-9aff-158852eeb990@sirena.org.uk>
From:   Guenter Roeck <groeck@google.com>
Date:   Tue, 11 Apr 2023 07:28:06 -0700
Message-ID: <CABXOdTfGD_fyi+itsr9Nd834Pb6+4Ej8M9SM5AN=uhXQ1K4DiA@mail.gmail.com>
Subject: Re: [PATCH] ASoC: da7219: Improve the relability of AAD IRQ process
To:     Mark Brown <broonie@kernel.org>
Cc:     David Rau <David.Rau.opensource@dm.renesas.com>,
        support.opensource@diasemi.com, lgirdwood@gmail.com,
        perex@perex.cz, tiwai@suse.com, alsa-devel@alsa-project.org,
        linux-kernel@vger.kernel.org, bailideng@google.com
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, Apr 11, 2023 at 4:32=E2=80=AFAM Mark Brown <broonie@kernel.org> wro=
te:
>
> On Mon, Apr 10, 2023 at 09:26:34AM +0000, David Rau wrote:
>
> > This commit improves the control of ground switches in AAD IRQ
>
> In what way does it do this - what was previously unrelabile and how
> does this change address that?

One very specific problem is that da7219_aad_handle_gnd_switch_time()
is currently called after interrupts were enabled. As a result, the
delay time is not initialized if there is an interrupt before the
initialization. This results in a negative value passed to msleep().
Since the parameter to msleep() is unsigned, this causes it to sleep
forever which in turn causes a substantial number of hung task crashes
in ChromeOS. Plus, of course, the code doesn't really do anything in
this situation.

A secondary problem may be that calling msleep() with a potentially
large sleep time on a system worker isn't really a good idea, but I
didn't explore the impact further.

Guenter