netdev - Re: [PATCH net V2] Crypto/chcr: Registering cxgb4 to xfrmdev

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: <f1f8ed22-9821-08cb-995c-7500355f2680@chelsio.com>
Date:   Mon, 27 Jul 2020 15:29:50 +0530
From:   Ayush Sawal <ayush.sawal@...lsio.com>
To:     Steffen Klassert <steffen.klassert@...unet.com>,
        Herbert Xu <herbert@...dor.apana.org.au>
Cc:     ayush.sawal@...cdesigners.com, David Miller <davem@...emloft.net>,
        netdev@...r.kernel.org, secdev@...lsio.com, lkp@...el.com
Subject: Re: [PATCH net V2] Crypto/chcr: Registering cxgb4 to xfrmdev_ops


On 7/27/2020 2:46 PM, Steffen Klassert wrote:
> On Sat, Jul 25, 2020 at 04:20:34PM +1000, Herbert Xu wrote:
>> On Fri, Jul 24, 2020 at 05:01:08PM -0700, David Miller wrote:
>>> Please start submitting chcr patches to the crypto subsystem, where it
>>> belongs, instead of the networking GIT trees.
>> Hi Dave:
>>
>> I think this patch belongs to the networking tree.  The reason is
>> that it's related to xfrm offload which has nothing to do with the
>> Crypto API.
> Hm, I think some of this code is just misplaced under drivers/crypto.
> All functions in 'drivers/crypto/chelsio/chcr_ipsec.c' implement
> networking (IPsec). So it should be under drivers/net, then it
> can be merged via the net or net-next tree as usual for network
> drivers.

Ok,
We have started the work towards shifting inline ipsec, nic tls & chtls 
code present in drivers/crypto/chelsio/ into 
drivers/net/ethernet/chelsio/. Only co-processor code will exist in the 
drivers/crypto/chelsio directory.

Thanks,
Ayush

>> Do xfrm offload drivers usually go through the networking tree or
>> would it be better directed through the xfrm tree?
> The drivers go through the networking trees, and I think it should
> stay like this. Otherwise we would create needless merge conflicts.
>