linux-kernel - Re: [PATCH net v3 1/3] netpoll: fix incorrect refcount handling causing incorrect cleanup

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: <v6yjo7yfimoof3mws4ymzrb5lkqd2t3cv4d4zzmh2ljux6rna2@zyj2tn53gep6>
Date: Wed, 10 Sep 2025 11:50:15 -0700
From: Breno Leitao <leitao@...ian.org>
To: Jakub Kicinski <kuba@...nel.org>
Cc: Calvin Owens <calvin@...nvd.org>, Andrew Lunn <andrew+netdev@...n.ch>, 
	"David S. Miller" <davem@...emloft.net>, Eric Dumazet <edumazet@...gle.com>, 
	Paolo Abeni <pabeni@...hat.com>, Shuah Khan <shuah@...nel.org>, Simon Horman <horms@...nel.org>, 
	david decotigny <decot@...glers.com>, linux-kernel@...r.kernel.org, netdev@...r.kernel.org, 
	linux-kselftest@...r.kernel.org, asantostc@...il.com, efault@....de, kernel-team@...a.com, 
	stable@...r.kernel.org, jv@...sburgh.net
Subject: Re: [PATCH net v3 1/3] netpoll: fix incorrect refcount handling
 causing incorrect cleanup

On Wed, Sep 10, 2025 at 10:58:58AM -0700, Jakub Kicinski wrote:
> On Wed, 10 Sep 2025 07:12:03 -0700 Breno Leitao wrote:
> > On Tue, Sep 09, 2025 at 04:16:25PM -0700, Jakub Kicinski wrote:
> > > On Tue, 9 Sep 2025 13:17:27 -0700 Breno Leitao wrote:  
> > > > On Mon, Sep 08, 2025 at 06:29:58PM -0700, Jakub Kicinski wrote:  
> > > > > On Mon, 8 Sep 2025 13:47:24 -0700 Calvin Owens wrote:    
> > > > > > I wonder if there might be a demon lurking in bonding+netpoll that this
> > > > > > was papering over? Not a reason not to fix the leaks IMO, I'm just
> > > > > > curious, I don't want to spend time on it if you already did :)    
> > > > > 
> > > > > +1, I also feel like it'd be good to have some bonding tests in place
> > > > > when we're removing a hack added specifically for bonding.    
> > > > 
> > > > Do you prefer to have a separated bonding selftest, or, is it better to
> > > > add some bond operations in the torture selftest?  
> > > 
> > > Normal test is preferable, given the flakiness rate and patch volume
> > > I'm a bit scared of randomized testing as part of CI.  
> > 
> > Ok, I will create a selftest to cover the netpoll part of bonding, as
> > soon as my understanding is good enough. I don't think it will be quick,
> > but, it is on my hi-pri todo list.
> > 
> > Do you want to have the selftest done before merging this patch, or, can
> > they go in parallel?
> 
> I said "it'd be good to have some bonding tests in place when we're
> removing a hack added specifically for bonding."
> "In place" means part of CI when we're merging this fix.
> Please read emails more carefully.

Apologies for the misunderstanding, It was unclear that the bonding
selftest should come before the fix. Thanks for the clarification.

I am planning to create a selftest similar to the original reported to cause
the issue[1], where I create a bond device with two netdevsim, and bind
netconsole to it.

[1] https://lore.kernel.org/lkml/96b940137a50e5c387687bb4f57de8b0435a653f.1404857349.git.decot@googlers.com/