netdev - Re: [net-next 01/12] net/mlx5e: Add UDP GSO support

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-next>] [day] [month] [year] [list]

Date:   Sun, 1 Jul 2018 21:45:18 -0400
From:   Willem de Bruijn <willemdebruijn.kernel@...il.com>
To:     Alexander Duyck <alexander.duyck@...il.com>
Cc:     borisp@...lanox.com, David Miller <davem@...emloft.net>,
        Network Development <netdev@...r.kernel.org>,
        Saeed Mahameed <saeedm@...lanox.com>, ogerlitz@...lanox.com,
        yossiku@...lanox.com
Subject: Re: [net-next 01/12] net/mlx5e: Add UDP GSO support

>> I've noticed that we could get cleaner code in our driver if we remove
>> these two lines from net/ipv4/udp_offload.c:
>> if (skb_is_gso(segs))
>>               mss *= skb_shinfo(segs)->gso_segs;
>>
>> I think that this is correct in case of GSO_PARTIAL segmentation for the
>> following reasons:
>> 1. After this change the UDP payload field is consistent with the IP
>> header payload length field. Currently, IPv4 length is 1500 and UDP
>> total length is the full unsegmented length.

How does this simplify the driver? Does it currently have to
change the udph->length field to the mss on the wire, because the
device only splits + replicates the headers + computes the csum?

> I don’t recall that the UDP header length field will match the IP length
> field. I had intentionally left it at the original value used to compute
> the UDP header checksum. That way you could just adjust it by
> cancelling out the length from the partial checksum.

>> 2. AFAIU, in GSO_PARTIAL no tunnel headers should be modified except the
>> IP ID field, including the UDP length field.
>> What do you think?
>
>
> For outer headers this is correct. For inner headers this is not. The inner
> UDP header will need to have the length updated and the checksum
> recomputed as a part of the segmentation.