在 2025/9/16 15:30, Eric Dumazet 写道:

On Mon, Sep 15, 2025 at 11:47 PM [off-list ref] wrote:

quoted

From: Xuanqiang Luo <redacted>

After replacing R/W locks with RCU in commit 3ab5aee7fe84 ("net: Convert
TCP & DCCP hash tables to use RCU / hlist_nulls"), a race window emerged
during the switch from reqsk/sk to sk/tw.

Now that both timewait sock (tw) and full sock (sk) reside on the same
ehash chain, it is appropriate to introduce hlist_nulls replace
operations, to eliminate the race conditions caused by this window.

---
Changes:
   v2:
     * Patch 1
         * Use WRITE_ONCE() to initialize old->pprev.
     * Patch 2&3
         * Optimize sk hashed check. Thanks Kuni for pointing it out!

   v1: https://lore.kernel.org/all/20250915070308.111816-1-xuanqiang.luo@linux.dev/ (local)

Note : I think you sent an earlier version, you should have added a
link to the discussion,
and past feedback/suggestions.

Lack of credit is a bit annoying frankly.

I will take a look at your series, thanks.

This patch's solution isn't very related to previous ones, so I didn't
include prior discussions.

But to help others get the full picture, I'll share past links and
earlier discussions related to this type of issue in the next version
or a reply. Sorry for any oversight.

Thanks
Xuanqiang.

quoted

Xuanqiang Luo (3):
   rculist: Add __hlist_nulls_replace_rcu() and
     hlist_nulls_replace_init_rcu()
   inet: Avoid ehash lookup race in inet_ehash_insert()
   inet: Avoid ehash lookup race in inet_twsk_hashdance_schedule()

  include/linux/rculist_nulls.h | 61 +++++++++++++++++++++++++++++++++++
  include/net/sock.h            | 23 +++++++++++++
  net/ipv4/inet_hashtables.c    |  4 ++-
  net/ipv4/inet_timewait_sock.c | 15 ++++-----
  4 files changed, 93 insertions(+), 10 deletions(-)

--
2.25.1