linux-kernel - Re: [PATCH v4 4/8] mfd: mc13xxx: Use devm_mfd_add_devices and devm_regmap_add_irq

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <alpine.DEB.2.21.2509180710190.9711@blala.de>
Date: Thu, 18 Sep 2025 07:25:05 +0000 (UTC)
From: Alexander Kurz <akurz@...la.de>
To: Dmitry Torokhov <dmitry.torokhov@...il.com>
cc: Lee Jones <lee@...nel.org>, Rob Herring <robh@...nel.org>,
        Krzysztof Kozlowski <krzk+dt@...nel.org>,
        Conor Dooley <conor+dt@...nel.org>,
        Dzmitry Sankouski <dsankouski@...il.com>,
        "Dr. David Alan Gilbert" <linux@...blig.org>,
        Heiko Stuebner <heiko@...ech.de>,
        Uwe Kleine-König <u.kleine-koenig@...libre.com>,
        devicetree@...r.kernel.org, linux-input@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v4 4/8] mfd: mc13xxx: Use devm_mfd_add_devices and
 devm_regmap_add_irq_chip


Hi Dimitry
On Tue, 16 Sep 2025, Dmitry Torokhov wrote:

> Hi Alexander,
> 
> On Sun, Sep 14, 2025 at 07:37:19PM +0000, Alexander Kurz wrote:
> > Use devm_mfd_add_devices() for adding MFD child devices and
> > devm_regmap_add_irq_chip() for IRQ chip registration.
> > 
> > This reduces the amount of required cleanup.
> > 
> > Signed-off-by: Alexander Kurz <akurz@...la.de>
> > ---
> >  drivers/mfd/mc13xxx-core.c | 9 ++++-----
> >  1 file changed, 4 insertions(+), 5 deletions(-)
> > 
> > diff --git a/drivers/mfd/mc13xxx-core.c b/drivers/mfd/mc13xxx-core.c
> > index 920797b806ce..091c9171b2b7 100644
> > --- a/drivers/mfd/mc13xxx-core.c
> > +++ b/drivers/mfd/mc13xxx-core.c
> > @@ -381,7 +381,7 @@ static int mc13xxx_add_subdevice_pdata(struct mc13xxx *mc13xxx,
> >  	if (!cell.name)
> >  		return -ENOMEM;
> >  
> > -	return mfd_add_devices(mc13xxx->dev, -1, &cell, 1, NULL, 0,
> > +	return devm_mfd_add_devices(mc13xxx->dev, -1, &cell, 1, NULL, 0,
> >  			       regmap_irq_get_domain(mc13xxx->irq_data));
> >  }
> >  
> > @@ -455,8 +455,9 @@ int mc13xxx_common_init(struct device *dev)
> >  	mc13xxx->irq_chip.irqs = mc13xxx->irqs;
> >  	mc13xxx->irq_chip.num_irqs = ARRAY_SIZE(mc13xxx->irqs);
> >  
> > -	ret = regmap_add_irq_chip(mc13xxx->regmap, mc13xxx->irq, IRQF_ONESHOT,
> > -				  0, &mc13xxx->irq_chip, &mc13xxx->irq_data);
> > +	ret = devm_regmap_add_irq_chip(dev, mc13xxx->regmap, mc13xxx->irq,
> > +				       IRQF_ONESHOT, 0, &mc13xxx->irq_chip,
> > +				       &mc13xxx->irq_data);
> >  	if (ret)
> >  		return ret;
> >  
> > @@ -502,8 +503,6 @@ void mc13xxx_common_exit(struct device *dev)
> >  {
> >  	struct mc13xxx *mc13xxx = dev_get_drvdata(dev);
> >  
> > -	mfd_remove_devices(dev);
> > -	regmap_del_irq_chip(mc13xxx->irq, mc13xxx->irq_data);
> >  	mutex_destroy(&mc13xxx->lock);
> 
> This causes the mutex be destroyed while the sub-devices are still
> present. The power button will try to call mc13xxx_lock() and
> mc13xxx_unlock() and of mutex debugging is enabled you'll get errors.
Thanks for noting this, actually I have introduced this change
in v4 of this series.

> I'd remove mutex_destroy() as well (and transitively get rid of 
> mc13xxx_common_exit()) and then look into getting rid of  mc13xxx_lock()
> and mc13xxx_unlock() because, as I mentioned in another email, they are
> IMO not needed.
I would prefer not to extend the scope of this series even further
and just drop this patch for v5.

There are still more issues todo with mc13xxx, e.g. mc13xxx-led does
not work since commit 78efa53e715e ("leds: Init leds class earlier").
Cleaning up potentially legacy mutex is just one more topic on this list. 
> 
> But this version of the patch is broken as far as I can tell.
> 
> Thanks.
> 
> -- 
> Dmitry
> 
Thanks, Alexander