linux-kernel - Re: [PATCH] mfd: qcom-pm8xxx: fix OF populate on driver rebind

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] [day] [month] [year] [list]

Message-ID: <aWYPRryYNYcl8IjI@hovoldconsulting.com>
Date: Tue, 13 Jan 2026 10:24:22 +0100
From: Johan Hovold <johan@...nel.org>
To: Lee Jones <lee@...nel.org>
Cc: linux-arm-msm@...r.kernel.org, linux-kernel@...r.kernel.org,
	stable@...r.kernel.org
Subject: Re: [PATCH] mfd: qcom-pm8xxx: fix OF populate on driver rebind

On Fri, Jan 09, 2026 at 03:27:38PM +0000, Lee Jones wrote:
> On Fri, 19 Dec 2025, Johan Hovold wrote:
> 
> > Since commit c6e126de43e7 ("of: Keep track of populated platform
> > devices") child devices will not be created by of_platform_populate()
> > if the devices had previously been deregistered individually so that the
> > OF_POPULATED flag is still set in the corresponding OF nodes.
> > 
> > Switch to using of_platform_depopulate() instead of open coding so that
> > the child devices are created if the driver is rebound.
> > 
> > Fixes: c6e126de43e7 ("of: Keep track of populated platform devices")
> > Cc: stable@...r.kernel.org	# 3.16
> > Signed-off-by: Johan Hovold <johan@...nel.org>

> > -static int pm8xxx_remove_child(struct device *dev, void *unused)
> > -{
> > -	platform_device_unregister(to_platform_device(dev));
> > -	return 0;
> > -}
> > -
> >  static void pm8xxx_remove(struct platform_device *pdev)
> >  {
> >  	struct pm_irq_chip *chip = platform_get_drvdata(pdev);
> >  
> > -	device_for_each_child(&pdev->dev, NULL, pm8xxx_remove_child);
> > +	of_platform_depopulate(&pdev->dev);
> >  	irq_domain_remove(chip->irqdomain);
> 
> Have you explored devm_of_platform_populate()?

Yeah, but mixing devres and explicit release risks introducing bugs. And
here we want to make sure the children are deregistered before freeing
the irqdomain.

Johan