netdev - Re: [PATCH net-next 3/6] eth: switch to netif_napi_add

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]

Date:   Wed, 11 May 2022 22:02:44 +0100
From:   Edward Cree <ecree.xilinx@...il.com>
To:     Jakub Kicinski <kuba@...nel.org>,
        Martin Habets <habetsm.xilinx@...il.com>
Cc:     davem@...emloft.net, netdev@...r.kernel.org, pabeni@...hat.com,
        edumazet@...gle.com
Subject: Re: [PATCH net-next 3/6] eth: switch to netif_napi_add_weight()

On 11/05/2022 20:45, Jakub Kicinski wrote:
> On Wed, 11 May 2022 18:57:53 +0100 Edward Cree wrote:
>> This isn't really a custom weight; napi_weight is initialised to
>>  64 and never changed, so probably we ought to be just using
>>  NAPI_POLL_WEIGHT here and end up on the non-_weight API.
>> Same goes for Falcon.
> 
> Ack, I wanted to be nice. I figured this must be a stub for a module
> param in your our of tree driver.

I mean, it *is*... but there's almost certainly a better way.  Configuring
 NAPI weight for tuning purposes (as opposed to some kind of correctness
 limitation in a driver) probably ought to be left to generic infrastructure
 rather than vendor-specific modparams in driver code.

> Should I send a patch to remove
> the non-const static napi_weight globals and switch back to non-_weight?

Yes please, unless Martin has any objections...?

-ed