netdev - Re: [RFC PATCH net-next 1/4] net-sysfs: remove rtnl

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: <169770320930.433869.5743241833039124669@kwain>
Date: Thu, 19 Oct 2023 10:13:29 +0200
From: Antoine Tenart <atenart@...nel.org>
To: Greg KH <gregkh@...uxfoundation.org>
Cc: davem@...emloft.net, kuba@...nel.org, pabeni@...hat.com, edumazet@...gle.com, netdev@...r.kernel.org, mhocko@...e.com, stephen@...workplumber.org
Subject: Re: [RFC PATCH net-next 1/4] net-sysfs: remove rtnl_trylock from device attributes

Quoting Greg KH (2023-10-18 18:49:18)
> On Wed, Oct 18, 2023 at 05:47:43PM +0200, Antoine Tenart wrote:
> > +static inline struct kernfs_node *sysfs_rtnl_lock(struct kobject *kobj,
> > +                                               struct attribute *attr,
> > +                                               struct net_device *ndev)
> > +{
> > +     struct kernfs_node *kn;
> > +
> > +     /* First, we hold a reference to the net device we might use in the
> > +      * locking section as the unregistration path might run in parallel.
> > +      * This will ensure the net device won't be freed before we return.
> > +      */
> > +     dev_hold(ndev);
> > +     /* sysfs_break_active_protection was introduced to allow self-removal of
> > +      * devices and their associated sysfs files by bailing out of the
> > +      * sysfs/kernfs protection. We do this here to allow the unregistration
> > +      * path to complete in parallel. The following takes a reference on the
> > +      * kobject and the kernfs_node being accessed.
> > +      *
> > +      * This works because we hold a reference onto the net device and the
> > +      * unregistration path will wait for us eventually in netdev_run_todo
> > +      * (outside an rtnl lock section).
> > +      */
> > +     kn = sysfs_break_active_protection(kobj, attr);
> > +     WARN_ON_ONCE(!kn);
> 
> If this triggers, you will end up rebooting the machines that set
> panic-on-warn, do you mean to do that?  And note, the huge majority of
> Linux systems in the world have that enabled, so be careful.

Right. My understanding was this can not happen here and I added this
one as a "that should not happen and something is really wrong", as the
attribute should be valid until at least the call to
sysfs_break_active_protection.

Thanks,
Antoine