From nobody Fri Oct 17 10:31:35 2025 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 246E3C4332F for ; Wed, 19 Oct 2022 12:04:59 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232071AbiJSME4 (ORCPT ); Wed, 19 Oct 2022 08:04:56 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:35840 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S229687AbiJSMDf (ORCPT ); Wed, 19 Oct 2022 08:03:35 -0400 Received: from ams.source.kernel.org (ams.source.kernel.org [145.40.68.75]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 53AD550199; Wed, 19 Oct 2022 04:39:47 -0700 (PDT) Received: from smtp.kernel.org (relay.kernel.org [52.25.139.140]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by ams.source.kernel.org (Postfix) with ESMTPS id 38CDEB82391; Wed, 19 Oct 2022 08:51:45 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 70733C433C1; Wed, 19 Oct 2022 08:51:43 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=linuxfoundation.org; s=korg; t=1666169503; bh=nUoYxlS4P3sUkSZzW3n7UycxsTcJ6Of6LllWPr/hLhI=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=ggoYda00wrGE0upzqdwYOJUwwCgthxXevOOu+DTPXEOOrWCpyDocYi0L+MDsVpshT kf4/9NCWIVdbHtBHyeslsheGRboo89CC4FRHW/qPWuiByc1g3mk7PnkJICV18SynsW UYeiBaLB9JL1c9TG6CJRJEz/azAdA8IoQY4+8k1U= From: Greg Kroah-Hartman To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman , stable@vger.kernel.org, Florian Westphal , Antoine Tenart , Sasha Levin Subject: [PATCH 6.0 300/862] netfilter: conntrack: fix the gc rescheduling delay Date: Wed, 19 Oct 2022 10:26:27 +0200 Message-Id: <20221019083303.262065457@linuxfoundation.org> X-Mailer: git-send-email 2.38.0 In-Reply-To: <20221019083249.951566199@linuxfoundation.org> References: <20221019083249.951566199@linuxfoundation.org> User-Agent: quilt/0.67 MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" From: Antoine Tenart [ Upstream commit 95eabdd207024312876d0ebed90b4c977e050e85 ] Commit 2cfadb761d3d ("netfilter: conntrack: revisit gc autotuning") changed the eviction rescheduling to the use average expiry of scanned entries (within 1-60s) by doing: for (...) { expires =3D clamp(nf_ct_expires(tmp), ...); next_run +=3D expires; next_run /=3D 2; } The issue is the above will make the average ('next_run' here) more dependent on the last expiration values than the firsts (for sets > 2). Depending on the expiration values used to compute the average, the result can be quite different than what's expected. To fix this we can do the following: for (...) { expires =3D clamp(nf_ct_expires(tmp), ...); next_run +=3D (expires - next_run) / ++count; } Fixes: 2cfadb761d3d ("netfilter: conntrack: revisit gc autotuning") Cc: Florian Westphal Signed-off-by: Antoine Tenart Signed-off-by: Florian Westphal Signed-off-by: Sasha Levin --- net/netfilter/nf_conntrack_core.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack= _core.c index 1357a2729a4b..2e6d5f1e6d63 100644 --- a/net/netfilter/nf_conntrack_core.c +++ b/net/netfilter/nf_conntrack_core.c @@ -67,6 +67,7 @@ struct conntrack_gc_work { struct delayed_work dwork; u32 next_bucket; u32 avg_timeout; + u32 count; u32 start_time; bool exiting; bool early_drop; @@ -1466,6 +1467,7 @@ static void gc_worker(struct work_struct *work) unsigned int expired_count =3D 0; unsigned long next_run; s32 delta_time; + long count; =20 gc_work =3D container_of(work, struct conntrack_gc_work, dwork.work); =20 @@ -1475,10 +1477,12 @@ static void gc_worker(struct work_struct *work) =20 if (i =3D=3D 0) { gc_work->avg_timeout =3D GC_SCAN_INTERVAL_INIT; + gc_work->count =3D 1; gc_work->start_time =3D start_time; } =20 next_run =3D gc_work->avg_timeout; + count =3D gc_work->count; =20 end_time =3D start_time + GC_SCAN_MAX_DURATION; =20 @@ -1498,8 +1502,8 @@ static void gc_worker(struct work_struct *work) =20 hlist_nulls_for_each_entry_rcu(h, n, &ct_hash[i], hnnode) { struct nf_conntrack_net *cnet; - unsigned long expires; struct net *net; + long expires; =20 tmp =3D nf_ct_tuplehash_to_ctrack(h); =20 @@ -1513,6 +1517,7 @@ static void gc_worker(struct work_struct *work) =20 gc_work->next_bucket =3D i; gc_work->avg_timeout =3D next_run; + gc_work->count =3D count; =20 delta_time =3D nfct_time_stamp - gc_work->start_time; =20 @@ -1528,8 +1533,8 @@ static void gc_worker(struct work_struct *work) } =20 expires =3D clamp(nf_ct_expires(tmp), GC_SCAN_INTERVAL_MIN, GC_SCAN_INT= ERVAL_CLAMP); + expires =3D (expires - (long)next_run) / ++count; next_run +=3D expires; - next_run /=3D 2u; =20 if (nf_conntrack_max95 =3D=3D 0 || gc_worker_skip_ct(tmp)) continue; @@ -1570,6 +1575,7 @@ static void gc_worker(struct work_struct *work) delta_time =3D nfct_time_stamp - end_time; if (delta_time > 0 && i < hashsz) { gc_work->avg_timeout =3D next_run; + gc_work->count =3D count; gc_work->next_bucket =3D i; next_run =3D 0; goto early_exit; --=20 2.35.1