netdev - Re: [PATCH iwl-net 1/2] idpf: preserve IRQ affinity settings across resets

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: <96df3bc7-85ab-4e21-a26d-3785874454a8@intel.com>
Date: Mon, 2 Dec 2024 06:03:45 -0700
From: Ahmed Zaki <ahmed.zaki@...el.com>
To: Jakub Kicinski <kuba@...nel.org>
CC: <intel-wired-lan@...ts.osuosl.org>, <netdev@...r.kernel.org>
Subject: Re: [PATCH iwl-net 1/2] idpf: preserve IRQ affinity settings across
 resets



On 2024-11-11 7:53 p.m., Jakub Kicinski wrote:
> On Fri,  8 Nov 2024 17:12:05 -0700 Ahmed Zaki wrote:
>> From: Sudheer Mogilappagari <sudheer.mogilappagari@...el.com>
>>
>> Currently the IRQ affinity settings are getting lost when interface
>> goes through a soft reset (due to MTU configuration, changing number
>> of queues etc). Use irq_set_affinity_notifier() callbacks to keep
>> the IRQ affinity info in sync between driver and kernel.
> 
> Could you try doing this in the core? Store the mask in napi_struct
> if it has IRQ associated with it?
> 
> Barely any drivers get this right.

The napi structs are allocated/freed with open/close ndos. I don't think 
we should expect the user to re-set CPU affinity after link down/up.

netdev->_rx perhaps?