Netdev List
 help / color / mirror / Atom feed
From: Yizhou Zhao <zhaoyz24@mails.tsinghua.edu.cn>
To: David Ahern <dsahern@kernel.org>,
	Ido Schimmel <idosch@nvidia.com>,
	Simon Horman <horms@verge.net.au>, Julian Anastasov <ja@ssi.bg>,
	"David S. Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Pablo Neira Ayuso <pablo@netfilter.org>,
	Florian Westphal <fw@strlen.de>, Phil Sutter <phil@nwl.cc>
Cc: Yizhou Zhao <zhaoyz24@mails.tsinghua.edu.cn>,
	netdev@vger.kernel.org, lvs-devel@vger.kernel.org,
	linux-kernel@vger.kernel.org, netfilter-devel@vger.kernel.org,
	coreteam@netfilter.org, stable@vger.kernel.org,
	Yuxiang Yang <yangyx22@mails.tsinghua.edu.cn>,
	Ao Wang <wangao@seu.edu.cn>, Xuewei Feng <fengxw06@126.com>,
	Qi Li <qli01@tsinghua.edu.cn>, Ke Xu <xuke@tsinghua.edu.cn>
Subject: [PATCH nf v3 0/2] ipvs: fix destination overload state updates
Date: Mon, 13 Jul 2026 17:48:00 +0800	[thread overview]
Message-ID: <cover.1783931964.git.zhaoyz24@mails.tsinghua.edu.cn> (raw)

IPVS updates a destination's overload status from connection accounting
and destination configuration paths, while schedulers read it from packet
processing paths.

Patch 1, authored by Julian, updates the overload state whenever a
destination's connection thresholds change. It also limits the upper
threshold to the range that can be compared safely with the connection
counter.

Patch 2 moves overload state out of dest->flags, which it previously
shared with the independent availability state. It uses a separate bitset
and bitops so updates to the two states cannot clobber each other. KCSAN
reports the original race between __ip_vs_update_dest() and
ip_vs_sh_schedule(), as well as between ip_vs_bind_dest() and the SH
scheduler.

The series keeps reader-side synchronization lightweight. test_bit() does
not provide a fresh cross-field snapshot, so schedulers may still observe
stale destination state as they could before this change.

Changes in v3:
- Add Julian's fix to properly refresh OVERLOAD on destination edit and 
reject upper thresholds above INT_MAX before comparing them with the signed
connection counter as patch 1/2.
- Keep AVAILABLE in dest->flags and move OVERLOAD to a separate bitset.
- Link to v2: https://lore.kernel.org/netfilter-devel/20260708060454.20534-1-zhaoyz24@mails.tsinghua.edu.cn/

Julian Anastasov (1):
  ipvs: properly update the overload flag on dest edit

Yizhou Zhao (1):
  ipvs: use bitops for destination overload state

 include/net/ip_vs.h              | 10 ++++++++
 include/uapi/linux/ip_vs.h       |  6 -----
 net/netfilter/ipvs/ip_vs_conn.c  | 44 ++++++++++++++++++++++----------
 net/netfilter/ipvs/ip_vs_ctl.c   | 26 +++++++++++++------
 net/netfilter/ipvs/ip_vs_dh.c    |  4 +--
 net/netfilter/ipvs/ip_vs_fo.c    |  2 +-
 net/netfilter/ipvs/ip_vs_lblc.c  |  4 +--
 net/netfilter/ipvs/ip_vs_lblcr.c |  8 +++---
 net/netfilter/ipvs/ip_vs_lc.c    |  2 +-
 net/netfilter/ipvs/ip_vs_mh.c    |  2 +-
 net/netfilter/ipvs/ip_vs_nq.c    |  2 +-
 net/netfilter/ipvs/ip_vs_ovf.c   |  2 +-
 net/netfilter/ipvs/ip_vs_rr.c    |  2 +-
 net/netfilter/ipvs/ip_vs_sed.c   |  4 +--
 net/netfilter/ipvs/ip_vs_sh.c    |  2 +-
 net/netfilter/ipvs/ip_vs_twos.c  |  4 +--
 net/netfilter/ipvs/ip_vs_wlc.c   |  4 +--
 net/netfilter/ipvs/ip_vs_wrr.c   |  2 +-
 18 files changed, 81 insertions(+), 49 deletions(-)


base-commit: 2c7c88a412aa6d09cd04b414211b4ef8553b5309
-- 
2.34.1


             reply	other threads:[~2026-07-13  9:49 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-07-13  9:48 Yizhou Zhao [this message]
2026-07-13  9:48 ` [PATCH nf v3 1/2] ipvs: properly update the overload flag on dest edit Yizhou Zhao
2026-07-13  9:48 ` [PATCH nf v3 2/2] ipvs: use bitops for destination overload state Yizhou Zhao
2026-07-14 11:16   ` Julian Anastasov
2026-07-14 14:22     ` Yizhou Zhao

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=cover.1783931964.git.zhaoyz24@mails.tsinghua.edu.cn \
    --to=zhaoyz24@mails.tsinghua.edu.cn \
    --cc=coreteam@netfilter.org \
    --cc=davem@davemloft.net \
    --cc=dsahern@kernel.org \
    --cc=edumazet@google.com \
    --cc=fengxw06@126.com \
    --cc=fw@strlen.de \
    --cc=horms@verge.net.au \
    --cc=idosch@nvidia.com \
    --cc=ja@ssi.bg \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lvs-devel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=netfilter-devel@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=pablo@netfilter.org \
    --cc=phil@nwl.cc \
    --cc=qli01@tsinghua.edu.cn \
    --cc=stable@vger.kernel.org \
    --cc=wangao@seu.edu.cn \
    --cc=xuke@tsinghua.edu.cn \
    --cc=yangyx22@mails.tsinghua.edu.cn \
    /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