netdev - Re: [PATCH iwl-net] ice: add missing napi deletion

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 for Android: free password hash cracker in your pocket

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

Message-ID: <ZJHgOXXHFjsOjlnA@boxer>
Date: Tue, 20 Jun 2023 19:22:01 +0200
From: Maciej Fijalkowski <maciej.fijalkowski@...el.com>
To: Jakub Kicinski <kuba@...nel.org>
CC: <intel-wired-lan@...ts.osuosl.org>, <netdev@...r.kernel.org>,
	<anthony.l.nguyen@...el.com>, <magnus.karlsson@...el.com>,
	<michal.swiatkowski@...el.com>
Subject: Re: [PATCH iwl-net] ice: add missing napi deletion

On Tue, Jun 20, 2023 at 09:53:35AM -0700, Jakub Kicinski wrote:
> On Tue, 20 Jun 2023 10:24:54 +0200 Maciej Fijalkowski wrote:
> > Error path from ice_probe() is missing ice_napi_del() calls, add it to
> > ice_deinit_eth() as ice_init_eth() was the one to add napi instances. It
> > is also a good habit to delete napis when removing driver and this also
> > addresses that. FWIW ice_napi_del() had no callsites.
> > 
> > Fixes: 6624e780a577 ("ice: split ice_vsi_setup into smaller functions")
> > Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@...el.com>
> 
> Is there user visible impact? I agree that it's a good habit, but
> since unregister cleans up NAPI instances automatically the patch
> is not necessarily a fix.

It's rather free_netdev() not unregistering per se, no? I sent this patch
as I found that cited commit didn't delete napis on ice_probe()'s error
path - I just saw that as a regression. 

But as you're saying when getting rid of netdev we actually do
netif_napi_del() - it seems redundant to do explicit napi delete on remove
path as it is supposed do free the netdev. Does it mean that many drivers
should be verified against that? Sorta tired so might be missing
something, pardon. If not, I'll send a v2 that just removes
ice_napi_del().