public inbox for netdev@vger.kernel.org
 help / color / mirror / Atom feed
From: Willem de Bruijn <willemdebruijn.kernel@gmail.com>
To: Alice Mikityanska <alice.kernel@fastmail.im>,
	 Daniel Borkmann <daniel@iogearbox.net>,
	 "David S. Miller" <davem@davemloft.net>,
	 Eric Dumazet <edumazet@google.com>,
	 Jakub Kicinski <kuba@kernel.org>,
	 Paolo Abeni <pabeni@redhat.com>,
	 Xin Long <lucien.xin@gmail.com>,
	 Willem de Bruijn <willemdebruijn.kernel@gmail.com>,
	 David Ahern <dsahern@kernel.org>,
	 Nikolay Aleksandrov <razor@blackwall.org>
Cc: Shuah Khan <shuah@kernel.org>,
	 Stanislav Fomichev <stfomichev@gmail.com>,
	 Andrew Lunn <andrew+netdev@lunn.ch>,
	 Simon Horman <horms@kernel.org>,
	 Florian Westphal <fw@strlen.de>,
	 netdev@vger.kernel.org,  Alice Mikityanska <alice@isovalent.com>,
	 Gal Pressman <gal@nvidia.com>
Subject: Re: [PATCH net-next v2 02/12] udp: gso: Simplify handling length in GSO_PARTIAL
Date: Fri, 06 Mar 2026 15:55:21 -0500	[thread overview]
Message-ID: <willemdebruijn.kernel.346c48b49836b@gmail.com> (raw)
In-Reply-To: <20260226201600.222044-3-alice.kernel@fastmail.im>

Alice Mikityanska wrote:
> From: Alice Mikityanska <alice@isovalent.com>
> 
> Taking further the idea of commit b10b446ce7ad ("udp: gso: Use single
> MSS length in UDP header for GSO_PARTIAL"), simplify the implementation
> and fix the checksum (apparently ignored by hardware anyway).
> 
> The mentioned commit started using msslen for uh->len, but still uses
> newlen to adjust uh->check. If the formula for check is fixed, newlen is
> assigned but never used before the loop, and newlen is overwritten after
> the loop. This makes msslen not really necessary, as we can reuse
> newlen, if we don't adjust mss before. 

That's a big if. Why is the udp length now set to the segment length,
and not to the GSO packet length, as before?

> The adjustment of mss can be
> simply dropped, because mss is not used anywhere else below.
> 
> This brings us back to one variable, drops an unneeded arithmetic for
> mss, and fixes the UDP checksum.
> 
> Signed-off-by: Alice Mikityanska <alice@isovalent.com>
> Cc: Gal Pressman <gal@nvidia.com>
> ---
>  net/ipv4/udp_offload.c | 13 ++-----------
>  1 file changed, 2 insertions(+), 11 deletions(-)
> 
> diff --git a/net/ipv4/udp_offload.c b/net/ipv4/udp_offload.c
> index 6b1654c1ad4a..e831234326c4 100644
> --- a/net/ipv4/udp_offload.c
> +++ b/net/ipv4/udp_offload.c
> @@ -483,11 +483,11 @@ struct sk_buff *__udp_gso_segment(struct sk_buff *gso_skb,
>  	struct sock *sk = gso_skb->sk;
>  	unsigned int sum_truesize = 0;
>  	struct sk_buff *segs, *seg;
> -	__be16 newlen, msslen;
>  	struct udphdr *uh;
>  	unsigned int mss;
>  	bool copy_dtor;
>  	__sum16 check;
> +	__be16 newlen;
>  	int ret = 0;
>  
>  	mss = skb_shinfo(gso_skb)->gso_size;
> @@ -556,15 +556,6 @@ struct sk_buff *__udp_gso_segment(struct sk_buff *gso_skb,
>  		return segs;
>  	}
>  
> -	msslen = htons(sizeof(*uh) + mss);
> -
> -	/* GSO partial and frag_list segmentation only requires splitting
> -	 * the frame into an MSS multiple and possibly a remainder, both
> -	 * cases return a GSO skb. So update the mss now.
> -	 */
> -	if (skb_is_gso(segs))
> -		mss *= skb_shinfo(segs)->gso_segs;
> -
>  	seg = segs;
>  	uh = udp_hdr(seg);
>  
> @@ -587,7 +578,7 @@ struct sk_buff *__udp_gso_segment(struct sk_buff *gso_skb,
>  		if (!seg->next)
>  			break;
>  
> -		uh->len = msslen;
> +		uh->len = newlen;
>  		uh->check = check;
>  
>  		if (seg->ip_summed == CHECKSUM_PARTIAL)
> -- 
> 2.52.0
> 



  reply	other threads:[~2026-03-06 20:55 UTC|newest]

Thread overview: 25+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-02-26 20:15 [PATCH net-next v2 00/12] BIG TCP for UDP tunnels Alice Mikityanska
2026-02-26 20:15 ` [PATCH net-next v2 01/12] net/sched: act_csum: don't mangle UDP tunnel GSO packets Alice Mikityanska
2026-02-26 20:15 ` [PATCH net-next v2 02/12] udp: gso: Simplify handling length in GSO_PARTIAL Alice Mikityanska
2026-03-06 20:55   ` Willem de Bruijn [this message]
2026-03-06 22:19     ` Alice Mikityanska
2026-03-07 23:23       ` Willem de Bruijn
2026-03-07 23:34         ` Alice Mikityanska
2026-03-07 23:53           ` Willem de Bruijn
2026-02-26 20:15 ` [PATCH net-next v2 03/12] geneve: Fix off-by-one comparing with GRO_LEGACY_MAX_SIZE Alice Mikityanska
2026-02-26 20:20   ` Alice Mikityanska
2026-02-26 20:15 ` [PATCH net-next v2 04/12] net: Use helpers to get/set UDP len tree-wide Alice Mikityanska
2026-02-26 20:19   ` Alice Mikityanska
2026-02-26 20:15 ` [PATCH net-next v2 05/12] net: Enable BIG TCP with partial GSO Alice Mikityanska
2026-02-26 20:15 ` [PATCH net-next v2 06/12] udp: Support gro_ipv4_max_size > 65536 Alice Mikityanska
2026-03-06 21:24   ` Willem de Bruijn
2026-03-06 21:31     ` Willem de Bruijn
2026-02-26 20:15 ` [PATCH net-next v2 07/12] udp: Support BIG TCP GSO packets where they can occur Alice Mikityanska
2026-02-26 20:15 ` [PATCH net-next v2 08/12] udp: Validate UDP length in udp_gro_receive Alice Mikityanska
2026-02-26 20:15 ` [PATCH net-next v2 09/12] udp: Set length in UDP header to 0 for big GSO packets Alice Mikityanska
2026-02-26 20:15 ` [PATCH net-next v2 10/12] vxlan: Enable BIG TCP packets Alice Mikityanska
2026-02-26 20:15 ` [PATCH net-next v2 11/12] geneve: " Alice Mikityanska
2026-02-26 20:16 ` [PATCH net-next v2 12/12] selftests: net: Add a test for BIG TCP in UDP tunnels Alice Mikityanska
2026-02-27  1:30   ` Jakub Kicinski
2026-02-27  9:35     ` Alice Mikityanska
2026-02-27 18:17 ` [syzbot ci] Re: BIG TCP for " syzbot ci

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=willemdebruijn.kernel.346c48b49836b@gmail.com \
    --to=willemdebruijn.kernel@gmail.com \
    --cc=alice.kernel@fastmail.im \
    --cc=alice@isovalent.com \
    --cc=andrew+netdev@lunn.ch \
    --cc=daniel@iogearbox.net \
    --cc=davem@davemloft.net \
    --cc=dsahern@kernel.org \
    --cc=edumazet@google.com \
    --cc=fw@strlen.de \
    --cc=gal@nvidia.com \
    --cc=horms@kernel.org \
    --cc=kuba@kernel.org \
    --cc=lucien.xin@gmail.com \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=razor@blackwall.org \
    --cc=shuah@kernel.org \
    --cc=stfomichev@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox