netdev - Re: [PATCH net] bnxt_en: bring back rtnl_lock() in bnxt_fw_reset

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: <aCKdCwjxSLcfw27k@mini-arch>
Date: Mon, 12 May 2025 18:14:51 -0700
From: Stanislav Fomichev <stfomichev@...il.com>
To: Jakub Kicinski <kuba@...nel.org>
Cc: Michael Chan <michael.chan@...adcom.com>, davem@...emloft.net,
	netdev@...r.kernel.org, edumazet@...gle.com, pabeni@...hat.com,
	andrew+netdev@...n.ch, pavan.chebbi@...adcom.com,
	andrew.gospodarek@...adcom.com, sdf@...ichev.me,
	Kalesh AP <kalesh-anakkur.purayil@...adcom.com>
Subject: Re: [PATCH net] bnxt_en: bring back rtnl_lock() in
 bnxt_fw_reset_task()

On 05/12, Jakub Kicinski wrote:
> On Mon, 12 May 2025 16:43:12 -0700 Stanislav Fomichev wrote:
> > On 05/12, Michael Chan wrote:
> > > On Mon, May 12, 2025 at 7:20 AM Stanislav Fomichev <stfomichev@...il.com> wrote:  
> > > > Will the following work instead? netdev_ops_assert_locked should take
> > > > care of asserting either ops lock or rtnl lock depending on the device
> > > > properties.  
> > > 
> > > It works for netif_set_real_num_tx_queues() but I also need to replace
> > > the ASSERT_RTNL() with netdev_ops_assert_locked(dev) in
> > > __udp_tunnel_nic_reset_ntf().  
> > 
> > Sounds good!
> 
> Mm... To me it sounds concerning. UDP tunnel port tracking doesn't have
> any locks, it depends on RTNL. Are y'all sure we can just drop the
> ASSERT_RTNL() and nothing will blow up? Or did I misunderstand?
> 
> I'd go with Michael's patch for net and revisit in net-next if you're
> filling bold.

Good point. But in this case, we need to cover more? bnxt_open from
bnxt_resume and the callers of bnxt_reset?