From mboxrd@z Thu Jan 1 00:00:00 1970 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 bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id A5DF0C5DF94 for ; Fri, 21 Aug 2026 16:40:25 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender:List-Subscribe:List-Help :List-Post:List-Archive:List-Unsubscribe:List-Id:MIME-Version: Content-Transfer-Encoding:Content-Type:References:In-Reply-To:Date:Cc:To:From :Subject:Message-ID:Reply-To:Content-ID:Content-Description:Resent-Date: Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:List-Owner; bh=6LdCN8NzDzrZlxlx10gtZoWk+7I60n+3MMYtlcuqxbk=; b=C7m7SiQ2w4MHfShtXR5OIfbsFO ejDnknmQO3pOfiT+GjhdzQucrekreh4xrHx+fpmyVy98KDemsanPfe575iAUeiAIgzX1qSEO/8vI6 c8sBD3oj4Z4Yvc1UaCapcIfLQXmXmiYY68HkuaIF2jZsWe1CiXvG8ZWlt71SicNcRa2scPe3NKuG6 hxOrRPzrZgN9zXrZd8Ki7skq4QyjIgPz4PypKbdBY4DHV3dAJs2KY6iCDB/hQym1VLPuyl0k/X63/ xtySn9/TI1tNPPEV+PA/UkPnISyxU/Cv8X4BbcrEKDktEVCTGDSn8k3Ktw4PkpoaXqbcMWb6hbceb /jPzB8Hw==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.99.1 #2 (Red Hat Linux)) id 1wxSHz-0000000Dq3I-13DQ; Fri, 21 Aug 2026 16:40:19 +0000 Received: from mta-65-225.siemens.flowmailer.net ([185.136.65.225]) by bombadil.infradead.org with esmtps (Exim 4.99.1 #2 (Red Hat Linux)) id 1wxSHw-0000000Dq2x-20Wv for linux-arm-kernel@lists.infradead.org; Fri, 21 Aug 2026 16:40:17 +0000 Received: by mta-65-225.siemens.flowmailer.net with ESMTPSA id 20260821164012d42138873900020780 for ; Fri, 21 Aug 2026 18:40:12 +0200 DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; s=fm2; d=siemens.com; i=florian.bezdeka@siemens.com; h=Date:From:Subject:To:Message-ID:MIME-Version:Content-Type:Content-Transfer-Encoding:Cc:References:In-Reply-To; bh=6LdCN8NzDzrZlxlx10gtZoWk+7I60n+3MMYtlcuqxbk=; b=bMcd9dKvvOlWyh3WwT71xtjxR+P/Lc0WkdnLkmgoZtQfXT4UPhRHxsq1XchYkVkAebI4xA PI7gglJrp0TkXd8L8Ms1d3XNYYefC5qiSp1gCpRSTZ5bz9xzhqkmJNQiut4M30bneWJR9kS6 yGLHzXxrYNvm+e6X0z1t10ppT5UQcwD8Mgp/SLs8DYfIdYICRYG6IxHas0cRajHaO1eqRPZX XsrebHGFTcs82AYj1Xhu34N/V8k20u0TUq7i2JOAVawUwtZmvsne3J5YXOkvXFN5VsIcoGqE 8arcNlOglmxEvDdM3ZyaXdCONkpQWlcT20E6CVj0qeL24mWQQ3FQIXJw==; Message-ID: Subject: Re: [PATCH RFC 3/3] net: stmmac: Migrate IRQ balancing to cpumask_local_spread() From: Florian Bezdeka To: Yury Norov Cc: Maxime Chevallier , Andrew Lunn , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Paolo Abeni , Maxime Coquelin , Alexandre Torgue , Yury Norov , Rasmus Villemoes , Andrew Morton , Sebastian Andrzej Siewior , Clark Williams , Steven Rostedt , Thomas Gleixner , Jan Kiszka , netdev@vger.kernel.org, linux-stm32@st-md-mailman.stormreply.com, linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org, linux-rt-devel@lists.linux.dev Date: Fri, 21 Aug 2026 18:40:11 +0200 In-Reply-To: References: <20260819-flo-net-7-2-make-stmmac-default-affinity-aware-v1-0-3f79a99cadaf@siemens.com> <20260819-flo-net-7-2-make-stmmac-default-affinity-aware-v1-3-3f79a99cadaf@siemens.com> Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-Flowmailer-Platform: Siemens Feedback-ID: 519:519-68982:519-21489:flowmailer X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.9.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20260821_094016_689434_D1F42E1B X-CRM114-Status: GOOD ( 25.34 ) X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+linux-arm-kernel=archiver.kernel.org@lists.infradead.org Hi Yury, On Wed, 2026-08-19 at 14:45 -0400, Yury Norov wrote: > On Wed, Aug 19, 2026 at 04:30:32PM +0200, Florian Bezdeka wrote: > > The previous balancing mechanism was based on num_online_cpus(), which > > is a problem for systems cpu-isolating workloads. IRQs were targeting > > CPUs that were isolated for those sensitive workloads. > >=20 > > With a migration to cpumask_local_spread() we > > - get NUMA locality > > - honor the default SMP affinity mask, which avoids targeting > > isolated CPUs. > >=20 > > This also aligns with the pattern used by most network drivers dealing > > with IRQ affinities / affinity hints. > >=20 > > Signed-off-by: Florian Bezdeka > > --- > > drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 21 +++++++++++++++= ++---- > > 1 file changed, 17 insertions(+), 4 deletions(-) > >=20 > > diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/driver= s/net/ethernet/stmicro/stmmac/stmmac_main.c > > index a71f0df263785dd8badc45292ca3067ab33bda05..949ced7e46d2814b57c6bd8= 6b4886ac3bf33996c 100644 > > --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c > > +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c > > @@ -3826,11 +3826,16 @@ static int stmmac_request_irq_multi_msi(struct = net_device *dev) > > struct stmmac_priv *priv =3D netdev_priv(dev); > > struct stmmac_msi *msi =3D priv->msi; > > enum request_irq_err irq_err; > > + cpumask_var_t affinity; > > int irq_idx =3D 0; > > char *int_name; > > + int node; > > int ret; > > int i; > > =20 > > + if (!zalloc_cpumask_var(&affinity, GFP_KERNEL)) > > + return -ENOMEM; > > + > > /* For common interrupt */ > > int_name =3D msi->int_name_mac; > > sprintf(int_name, "%s:%s", dev->name, "mac"); > > @@ -3916,6 +3921,7 @@ static int stmmac_request_irq_multi_msi(struct ne= t_device *dev) > > } > > =20 > > /* Request Rx MSI irq */ > > + node =3D dev_to_node(&priv->dev->dev); > > for (i =3D 0; i < priv->plat->rx_queues_to_use; i++) { > > if (i >=3D MTL_MAX_RX_QUEUES) > > break; > > @@ -3935,8 +3941,10 @@ static int stmmac_request_irq_multi_msi(struct n= et_device *dev) > > irq_idx =3D i; > > goto irq_error; > > } > > - irq_set_affinity_hint(msi->rx_irq[i], > > - cpumask_of(i % num_online_cpus())); > > + > > + cpumask_clear(affinity); > > + cpumask_set_cpu(cpumask_local_spread(i, node), affinity); > > + irq_set_affinity_and_hint(msi->rx_irq[i], affinity); >=20 > If you want to spread more than one IRQ, you'd better convert your > loop into for_each_numa_hop_mask(). That way you don't need to > introduce new function. See the comment on top of > cpumask_local_spread(): >=20 I was knowingly ignoring this "optimization". As already noted the patch 1 thread: I'm currently trying to demonstrate the shortcomings, optimizations will follow once we agree on a plan / vision.