All of lore.kernel.org
 help / color / mirror / Atom feed
From: Eric Dumazet <eric.dumazet@gmail.com>
To: "David S . Miller" <davem@davemloft.net>,
	Jakub Kicinski <kuba@kernel.org>
Cc: netdev <netdev@vger.kernel.org>,
	Eric Dumazet <edumazet@google.com>,
	Eric Dumazet <eric.dumazet@gmail.com>
Subject: [PATCH net-next 1/2] gro: optimize skb_gro_postpull_rcsum()
Date: Wed, 24 Nov 2021 12:24:45 -0800	[thread overview]
Message-ID: <20211124202446.2917972-2-eric.dumazet@gmail.com> (raw)
In-Reply-To: <20211124202446.2917972-1-eric.dumazet@gmail.com>

From: Eric Dumazet <edumazet@google.com>

We can leverage third argument to csum_partial():

  X = csum_sub(X, csum_partial(start, len, 0));

  -->

  X = csum_add(X, ~csum_partial(start, len, 0));

  -->

  X = ~csum_partial(start, len, ~X);

This removes one add/adc pair and its dependency against the carry flag.

Signed-off-by: Eric Dumazet <edumazet@google.com>
---
 include/net/gro.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/include/net/gro.h b/include/net/gro.h
index 9c22a010369cb89f9511d78cc322be56170d7b20..b1139fca7c435ca0c353c4ed17628dd7f3df4401 100644
--- a/include/net/gro.h
+++ b/include/net/gro.h
@@ -173,8 +173,8 @@ static inline void skb_gro_postpull_rcsum(struct sk_buff *skb,
 					const void *start, unsigned int len)
 {
 	if (NAPI_GRO_CB(skb)->csum_valid)
-		NAPI_GRO_CB(skb)->csum = csum_sub(NAPI_GRO_CB(skb)->csum,
-						  csum_partial(start, len, 0));
+		NAPI_GRO_CB(skb)->csum = ~csum_partial(start, len,
+						       ~NAPI_GRO_CB(skb)->csum);
 }
 
 /* GRO checksum functions. These are logical equivalents of the normal
-- 
2.34.0.rc2.393.gf8c9666880-goog


  reply	other threads:[~2021-11-24 20:24 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-11-24 20:24 [PATCH net-next 0/2] net: small csum optimizations Eric Dumazet
2021-11-24 20:24 ` Eric Dumazet [this message]
2021-11-24 20:24 ` [PATCH net-next 2/2] net: optimize skb_postpull_rcsum() Eric Dumazet
2021-11-25  9:41   ` David Laight
2021-11-25 13:32     ` Eric Dumazet
2021-11-25 14:29       ` David Laight
2021-12-02 13:10   ` Vladimir Oltean
2021-12-02 14:51     ` Eric Dumazet
2021-12-02 16:29       ` Vladimir Oltean
2021-12-02 19:32         ` Eric Dumazet
2021-12-02 20:37           ` Eric Dumazet
2021-12-02 21:07             ` Vladimir Oltean
2021-12-02 20:40           ` Vladimir Oltean
2021-12-02 20:58             ` Vladimir Oltean
2021-12-02 20:58             ` David Laight
2021-12-02 21:40               ` Vladimir Oltean
2021-12-03 14:51                 ` David Laight
2021-12-03 14:57                 ` David Laight
2021-12-03 16:14                   ` Vladimir Oltean
2021-12-03 16:30                     ` Eric Dumazet
2021-12-03 16:47                       ` David Laight
2021-12-03 16:58                         ` Eric Dumazet
2021-12-03 17:41                           ` David Laight
2021-12-02 15:06     ` David Laight
2021-12-02 15:22       ` Eric Dumazet
2021-11-26  5:20 ` [PATCH net-next 0/2] net: small csum optimizations patchwork-bot+netdevbpf

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=20211124202446.2917972-2-eric.dumazet@gmail.com \
    --to=eric.dumazet@gmail.com \
    --cc=davem@davemloft.net \
    --cc=edumazet@google.com \
    --cc=kuba@kernel.org \
    --cc=netdev@vger.kernel.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.