From nobody Fri Oct 3 13:32:30 2025 Received: from m16.mail.163.com (m16.mail.163.com [220.197.31.4]) by smtp.subspace.kernel.org (Postfix) with ESMTP id 0C5542561A2; Sun, 31 Aug 2025 10:09:15 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=220.197.31.4 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1756634958; cv=none; b=Y5a+xw225j+a7CNjIbdt8c6qu3nbeWs2VoNepCOd/GC+b9TBvXiPNM/47dN1KU3S8aYbApIIYSTMCa4E83HFwHSrFt+a4fJl6cUVvZPG+wGtOaDOKcqy4oeMPZfzDoJ7tOvjCweoERQHYbftpiWs95fVFb0SFbQL+n0tffri2vQ= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1756634958; c=relaxed/simple; bh=MXRPEyMnVIGdQ6y5qQCmjy/Ms0VWdqWsceL1Lo9x8ko=; h=From:To:Cc:Subject:Date:Message-Id:In-Reply-To:References: MIME-Version; b=TrMrpbAyvpjA8s4AJbZ+ZR+qKV2RIX9UcienHCqyUDa/BCKzoJFwa9xX/M9pe/hH3jb2MuRMwP/dD418zHa+EfFu0cgP2oJyLW0Np5VCErvLAQ3i7zyWBh3KrBe7fbjd1uVCdRGvHwtVEOuZ+jSqtAYLEanuak1yucEzm2vf89k= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=163.com; spf=pass smtp.mailfrom=163.com; dkim=pass (1024-bit key) header.d=163.com header.i=@163.com header.b=JB/sXVpk; arc=none smtp.client-ip=220.197.31.4 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=163.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=163.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=163.com header.i=@163.com header.b="JB/sXVpk" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=163.com; s=s110527; h=From:To:Subject:Date:Message-Id:MIME-Version; bh=bl Yk/xMFukiqktu/30G9TbZdILw5g1q0KR+Rl8F8jJw=; b=JB/sXVpkhIBib7lN7b a0Rm+sTwRcKG3rCzq7QQBq6eSJMjoaQ/pQerJm2p6zg5TfvT4bZRxyaYLonMIn+y +j8lgJFwfb9oy0w1P1YMW1/3kEFcfy4DbXpYaEkskCgYxhUku0vGn1Yrw2cGpvL9 yGONqSk10rnJM9Q1uAOMUScOw= Received: from zhaoxin-MS-7E12.. (unknown []) by gzga-smtp-mtada-g1-1 (Coremail) with SMTP id _____wBH7U4YH7RoR1JrFA--.61212S3; Sun, 31 Aug 2025 18:08:35 +0800 (CST) From: Xin Zhao To: willemdebruijn.kernel@gmail.com, edumazet@google.com, ferenc@fejes.dev Cc: davem@davemloft.net, kuba@kernel.org, pabeni@redhat.com, horms@kernel.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Xin Zhao Subject: [PATCH net-next v10 1/2] net: af_packet: remove last_kactive_blk_num field Date: Sun, 31 Aug 2025 18:08:21 +0800 Message-Id: <20250831100822.1238795-2-jackzxcui1989@163.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20250831100822.1238795-1-jackzxcui1989@163.com> References: <20250831100822.1238795-1-jackzxcui1989@163.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: _____wBH7U4YH7RoR1JrFA--.61212S3 X-Coremail-Antispam: 1Uf129KBjvJXoWxCr1fXFW8tr1rur4kWFWfKrg_yoWrGr43pF W3A343Jw4DGr4agw4xZwnrZr15Ww4rWFy8G3sxJw4fA3s3Aryava9F9Fy7WFyrJrZ3Zay7 Xrs5t345Cr4kJrJanT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x0z_F4iUUUUUU= X-CM-SenderInfo: pmdfy650fxxiqzyzqiywtou0bp/1tbiRxG6Cmi0Dcz5bQAAsv Content-Type: text/plain; charset="utf-8" kactive_blk_num (K) is incremented on block close. last_kactive_blk_num (L) is set to match K on block open and each timer. So the only time that they differ is if a block is closed in tpacket_rcv and no new block could be opened. So the origin check L=3D=3DK in timer callback only skip the case 'no new b= lock to open'. If we remove L=3D=3DK check, it will make prb_curr_blk_in_use che= ck earlier, which will not cause any side effect. Signed-off-by: Xin Zhao Reviewed-by: Jason Xing Reviewed-by: Willem de Bruijn Suggested-by: Willem de Bruijn --- net/packet/af_packet.c | 60 ++++++++++++++++++++---------------------- net/packet/internal.h | 6 ----- 2 files changed, 28 insertions(+), 38 deletions(-) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index a7017d7f0..d4eb4a4fe 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -669,7 +669,6 @@ static void init_prb_bdqc(struct packet_sock *po, p1->knum_blocks =3D req_u->req3.tp_block_nr; p1->hdrlen =3D po->tp_hdrlen; p1->version =3D po->tp_version; - p1->last_kactive_blk_num =3D 0; po->stats.stats3.tp_freeze_q_cnt =3D 0; if (req_u->req3.tp_retire_blk_tov) p1->retire_blk_tov =3D req_u->req3.tp_retire_blk_tov; @@ -693,7 +692,6 @@ static void _prb_refresh_rx_retire_blk_timer(struct tpa= cket_kbdq_core *pkc) { mod_timer(&pkc->retire_blk_timer, jiffies + pkc->tov_in_jiffies); - pkc->last_kactive_blk_num =3D pkc->kactive_blk_num; } =20 /* @@ -750,38 +748,36 @@ static void prb_retire_rx_blk_timer_expired(struct ti= mer_list *t) write_unlock(&pkc->blk_fill_in_prog_lock); } =20 - if (pkc->last_kactive_blk_num =3D=3D pkc->kactive_blk_num) { - if (!frozen) { - if (!BLOCK_NUM_PKTS(pbd)) { - /* An empty block. Just refresh the timer. */ - goto refresh_timer; - } - prb_retire_current_block(pkc, po, TP_STATUS_BLK_TMO); - if (!prb_dispatch_next_block(pkc, po)) - goto refresh_timer; - else - goto out; + if (!frozen) { + if (!BLOCK_NUM_PKTS(pbd)) { + /* An empty block. Just refresh the timer. */ + goto refresh_timer; + } + prb_retire_current_block(pkc, po, TP_STATUS_BLK_TMO); + if (!prb_dispatch_next_block(pkc, po)) + goto refresh_timer; + else + goto out; + } else { + /* Case 1. Queue was frozen because user-space was + * lagging behind. + */ + if (prb_curr_blk_in_use(pbd)) { + /* + * Ok, user-space is still behind. + * So just refresh the timer. + */ + goto refresh_timer; } else { - /* Case 1. Queue was frozen because user-space was - * lagging behind. + /* Case 2. queue was frozen,user-space caught up, + * now the link went idle && the timer fired. + * We don't have a block to close.So we open this + * block and restart the timer. + * opening a block thaws the queue,restarts timer + * Thawing/timer-refresh is a side effect. */ - if (prb_curr_blk_in_use(pbd)) { - /* - * Ok, user-space is still behind. - * So just refresh the timer. - */ - goto refresh_timer; - } else { - /* Case 2. queue was frozen,user-space caught up, - * now the link went idle && the timer fired. - * We don't have a block to close.So we open this - * block and restart the timer. - * opening a block thaws the queue,restarts timer - * Thawing/timer-refresh is a side effect. - */ - prb_open_block(pkc, pbd); - goto out; - } + prb_open_block(pkc, pbd); + goto out; } } =20 diff --git a/net/packet/internal.h b/net/packet/internal.h index 1e743d031..d367b9f93 100644 --- a/net/packet/internal.h +++ b/net/packet/internal.h @@ -24,12 +24,6 @@ struct tpacket_kbdq_core { unsigned short kactive_blk_num; unsigned short blk_sizeof_priv; =20 - /* last_kactive_blk_num: - * trick to see if user-space has caught up - * in order to avoid refreshing timer when every single pkt arrives. - */ - unsigned short last_kactive_blk_num; - char *pkblk_start; char *pkblk_end; int kblk_size; --=20 2.34.1 From nobody Fri Oct 3 13:32:30 2025 Received: from m16.mail.163.com (m16.mail.163.com [220.197.31.5]) by smtp.subspace.kernel.org (Postfix) with ESMTP id CDABA25333F; Sun, 31 Aug 2025 10:09:11 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=220.197.31.5 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1756634954; cv=none; b=hu3y0qgXtiDQepo09N46ms2UDL8vmuQ/AzBT99elBKTcTRClKFhTe216HPAHL90udkSEhNMVcBanrY2TGVTxLk+ZlnhXEtFmaXN2xcxVFKpu5j3sUmXB+m5L3bxW1vNKXBu6cFgRfKBNg+oyH11dCpKcOqYmsCMOL1WrdBviAjU= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1756634954; c=relaxed/simple; bh=7qmMCrSvCXWBzVLmm00fIfxsMF1iK91xEagZCY5Y4J8=; h=From:To:Cc:Subject:Date:Message-Id:In-Reply-To:References: MIME-Version; b=hY7AE2Gypdcz+wcV/UFYtZR4Rj2Jwvtrmzg1+6JZwyeQCjXlfVWOMfJpPHZpzmMhgOO6qrX0hzCKgVfPVq/LYN1o45oN5bBFXoCuc1l6/dvJdhOVrNlG9h/qTAlomwNVscU65xmfymbWL75oA2mbLcPhQI2hzbcWFIKxdkcwV5M= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=163.com; spf=pass smtp.mailfrom=163.com; dkim=pass (1024-bit key) header.d=163.com header.i=@163.com header.b=Nibyk53m; arc=none smtp.client-ip=220.197.31.5 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=163.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=163.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=163.com header.i=@163.com header.b="Nibyk53m" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=163.com; s=s110527; h=From:To:Subject:Date:Message-Id:MIME-Version; bh=1q 7nt1GgkLAOlvi8QWZdSXDrk74zprU57mw7aZ4nq+8=; b=Nibyk53mILxrfsU6Kr 5oXQGwa6c36O3fpq/mKTkxgqnEAXOvOvYkzbbWOKfCspZkB5JjuZGGj2BfqnMzJ9 cbzuUIajDSEF+yUGD2NN9IrS4Ad/dnkDloa5HRB3JuMZhEQ5ZJMttxLREd37ydlM eZSc83z69h4ANFEB1mykLZAOA= Received: from zhaoxin-MS-7E12.. (unknown []) by gzga-smtp-mtada-g1-1 (Coremail) with SMTP id _____wBH7U4YH7RoR1JrFA--.61212S4; Sun, 31 Aug 2025 18:08:37 +0800 (CST) From: Xin Zhao To: willemdebruijn.kernel@gmail.com, edumazet@google.com, ferenc@fejes.dev Cc: davem@davemloft.net, kuba@kernel.org, pabeni@redhat.com, horms@kernel.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Xin Zhao Subject: [PATCH net-next v10 2/2] net: af_packet: Use hrtimer to do the retire operation Date: Sun, 31 Aug 2025 18:08:22 +0800 Message-Id: <20250831100822.1238795-3-jackzxcui1989@163.com> X-Mailer: git-send-email 2.34.1 In-Reply-To: <20250831100822.1238795-1-jackzxcui1989@163.com> References: <20250831100822.1238795-1-jackzxcui1989@163.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: _____wBH7U4YH7RoR1JrFA--.61212S4 X-Coremail-Antispam: 1Uf129KBjvJXoW3Wry5Kr1fAF45JFy5Ar15CFg_yoW3tF1Dpa y5W34xGwsrXF4Ygw48Jrn7ZFyFgw1DAry5GrZ3Xwsayas3try3tF1j9FyYgFWftFZ2vw47 JrsYqrZ8Cr1DXrJanT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x0zRIJmUUUUUU= X-CM-SenderInfo: pmdfy650fxxiqzyzqiywtou0bp/1tbioxy6Cmi0EKHNJgAAsJ Content-Type: text/plain; charset="utf-8" In a system with high real-time requirements, the timeout mechanism of ordinary timers with jiffies granularity is insufficient to meet the demands for real-time performance. Meanwhile, the optimization of CPU usage with af_packet is quite significant. Use hrtimer instead of timer to help compensate for the shortcomings in real-time performance. In HZ=3D100 or HZ=3D250 system, the update of TP_STATUS_USER is not real-ti= me enough, with fluctuations reaching over 8ms (on a system with HZ=3D250). This is unacceptable in some high real-time systems that require timely processing of network packets. By replacing it with hrtimer, if a timeout of 2ms is set, the update of TP_STATUS_USER can be stabilized to within 3 ms. Signed-off-by: Xin Zhao Reviewed-by: Willem de Bruijn --- Changes in v8: - Simplify the logic related to setting timeout. Changes in v7: - Only update the hrtimer expire time within the hrtimer callback. Changes in v1: - Do not add another config for the current changes. --- net/packet/af_packet.c | 79 +++++++++--------------------------------- net/packet/diag.c | 2 +- net/packet/internal.h | 6 ++-- 3 files changed, 20 insertions(+), 67 deletions(-) diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c index d4eb4a4fe..3e3bb4216 100644 --- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -203,8 +203,7 @@ static void prb_retire_current_block(struct tpacket_kbd= q_core *, static int prb_queue_frozen(struct tpacket_kbdq_core *); static void prb_open_block(struct tpacket_kbdq_core *, struct tpacket_block_desc *); -static void prb_retire_rx_blk_timer_expired(struct timer_list *); -static void _prb_refresh_rx_retire_blk_timer(struct tpacket_kbdq_core *); +static enum hrtimer_restart prb_retire_rx_blk_timer_expired(struct hrtimer= *); static void prb_fill_rxhash(struct tpacket_kbdq_core *, struct tpacket3_hd= r *); static void prb_clear_rxhash(struct tpacket_kbdq_core *, struct tpacket3_hdr *); @@ -579,33 +578,13 @@ static __be16 vlan_get_protocol_dgram(const struct sk= _buff *skb) return proto; } =20 -static void prb_del_retire_blk_timer(struct tpacket_kbdq_core *pkc) -{ - timer_delete_sync(&pkc->retire_blk_timer); -} - static void prb_shutdown_retire_blk_timer(struct packet_sock *po, struct sk_buff_head *rb_queue) { struct tpacket_kbdq_core *pkc; =20 pkc =3D GET_PBDQC_FROM_RB(&po->rx_ring); - - spin_lock_bh(&rb_queue->lock); - pkc->delete_blk_timer =3D 1; - spin_unlock_bh(&rb_queue->lock); - - prb_del_retire_blk_timer(pkc); -} - -static void prb_setup_retire_blk_timer(struct packet_sock *po) -{ - struct tpacket_kbdq_core *pkc; - - pkc =3D GET_PBDQC_FROM_RB(&po->rx_ring); - timer_setup(&pkc->retire_blk_timer, prb_retire_rx_blk_timer_expired, - 0); - pkc->retire_blk_timer.expires =3D jiffies; + hrtimer_cancel(&pkc->retire_blk_timer); } =20 static int prb_calc_retire_blk_tmo(struct packet_sock *po, @@ -671,29 +650,22 @@ static void init_prb_bdqc(struct packet_sock *po, p1->version =3D po->tp_version; po->stats.stats3.tp_freeze_q_cnt =3D 0; if (req_u->req3.tp_retire_blk_tov) - p1->retire_blk_tov =3D req_u->req3.tp_retire_blk_tov; + p1->interval_ktime =3D ms_to_ktime(req_u->req3.tp_retire_blk_tov); else - p1->retire_blk_tov =3D prb_calc_retire_blk_tmo(po, - req_u->req3.tp_block_size); - p1->tov_in_jiffies =3D msecs_to_jiffies(p1->retire_blk_tov); + p1->interval_ktime =3D ms_to_ktime(prb_calc_retire_blk_tmo(po, + req_u->req3.tp_block_size)); p1->blk_sizeof_priv =3D req_u->req3.tp_sizeof_priv; rwlock_init(&p1->blk_fill_in_prog_lock); =20 p1->max_frame_len =3D p1->kblk_size - BLK_PLUS_PRIV(p1->blk_sizeof_priv); prb_init_ft_ops(p1, req_u); - prb_setup_retire_blk_timer(po); + hrtimer_setup(&p1->retire_blk_timer, prb_retire_rx_blk_timer_expired, + CLOCK_MONOTONIC, HRTIMER_MODE_REL_SOFT); + hrtimer_start(&p1->retire_blk_timer, p1->interval_ktime, + HRTIMER_MODE_REL_SOFT); prb_open_block(p1, pbd); } =20 -/* Do NOT update the last_blk_num first. - * Assumes sk_buff_head lock is held. - */ -static void _prb_refresh_rx_retire_blk_timer(struct tpacket_kbdq_core *pkc) -{ - mod_timer(&pkc->retire_blk_timer, - jiffies + pkc->tov_in_jiffies); -} - /* * Timer logic: * 1) We refresh the timer only when we open a block. @@ -717,7 +689,7 @@ static void _prb_refresh_rx_retire_blk_timer(struct tpa= cket_kbdq_core *pkc) * prb_calc_retire_blk_tmo() calculates the tmo. * */ -static void prb_retire_rx_blk_timer_expired(struct timer_list *t) +static enum hrtimer_restart prb_retire_rx_blk_timer_expired(struct hrtimer= *t) { struct packet_sock *po =3D timer_container_of(po, t, rx_ring.prb_bdqc.retire_blk_timer); @@ -730,9 +702,6 @@ static void prb_retire_rx_blk_timer_expired(struct time= r_list *t) frozen =3D prb_queue_frozen(pkc); pbd =3D GET_CURR_PBLOCK_DESC_FROM_CORE(pkc); =20 - if (unlikely(pkc->delete_blk_timer)) - goto out; - /* We only need to plug the race when the block is partially filled. * tpacket_rcv: * lock(); increment BLOCK_NUM_PKTS; unlock() @@ -749,26 +718,16 @@ static void prb_retire_rx_blk_timer_expired(struct ti= mer_list *t) } =20 if (!frozen) { - if (!BLOCK_NUM_PKTS(pbd)) { - /* An empty block. Just refresh the timer. */ - goto refresh_timer; + if (BLOCK_NUM_PKTS(pbd)) { + /* Not an empty block. Need retire the block. */ + prb_retire_current_block(pkc, po, TP_STATUS_BLK_TMO); + prb_dispatch_next_block(pkc, po); } - prb_retire_current_block(pkc, po, TP_STATUS_BLK_TMO); - if (!prb_dispatch_next_block(pkc, po)) - goto refresh_timer; - else - goto out; } else { /* Case 1. Queue was frozen because user-space was * lagging behind. */ - if (prb_curr_blk_in_use(pbd)) { - /* - * Ok, user-space is still behind. - * So just refresh the timer. - */ - goto refresh_timer; - } else { + if (!prb_curr_blk_in_use(pbd)) { /* Case 2. queue was frozen,user-space caught up, * now the link went idle && the timer fired. * We don't have a block to close.So we open this @@ -777,15 +736,12 @@ static void prb_retire_rx_blk_timer_expired(struct ti= mer_list *t) * Thawing/timer-refresh is a side effect. */ prb_open_block(pkc, pbd); - goto out; } } =20 -refresh_timer: - _prb_refresh_rx_retire_blk_timer(pkc); - -out: + hrtimer_forward_now(&pkc->retire_blk_timer, pkc->interval_ktime); spin_unlock(&po->sk.sk_receive_queue.lock); + return HRTIMER_RESTART; } =20 static void prb_flush_block(struct tpacket_kbdq_core *pkc1, @@ -917,7 +873,6 @@ static void prb_open_block(struct tpacket_kbdq_core *pk= c1, pkc1->pkblk_end =3D pkc1->pkblk_start + pkc1->kblk_size; =20 prb_thaw_queue(pkc1); - _prb_refresh_rx_retire_blk_timer(pkc1); =20 smp_wmb(); } diff --git a/net/packet/diag.c b/net/packet/diag.c index 6ce1dcc28..c8f43e0c1 100644 --- a/net/packet/diag.c +++ b/net/packet/diag.c @@ -83,7 +83,7 @@ static int pdiag_put_ring(struct packet_ring_buffer *ring= , int ver, int nl_type, pdr.pdr_frame_nr =3D ring->frame_max + 1; =20 if (ver > TPACKET_V2) { - pdr.pdr_retire_tmo =3D ring->prb_bdqc.retire_blk_tov; + pdr.pdr_retire_tmo =3D ktime_to_ms(ring->prb_bdqc.interval_ktime); pdr.pdr_sizeof_priv =3D ring->prb_bdqc.blk_sizeof_priv; pdr.pdr_features =3D ring->prb_bdqc.feature_req_word; } else { diff --git a/net/packet/internal.h b/net/packet/internal.h index d367b9f93..f8cfd9213 100644 --- a/net/packet/internal.h +++ b/net/packet/internal.h @@ -20,7 +20,6 @@ struct tpacket_kbdq_core { unsigned int feature_req_word; unsigned int hdrlen; unsigned char reset_pending_on_curr_blk; - unsigned char delete_blk_timer; unsigned short kactive_blk_num; unsigned short blk_sizeof_priv; =20 @@ -39,12 +38,11 @@ struct tpacket_kbdq_core { /* Default is set to 8ms */ #define DEFAULT_PRB_RETIRE_TOV (8) =20 - unsigned short retire_blk_tov; + ktime_t interval_ktime; unsigned short version; - unsigned long tov_in_jiffies; =20 /* timer to retire an outstanding block */ - struct timer_list retire_blk_timer; + struct hrtimer retire_blk_timer; }; =20 struct pgv { --=20 2.34.1