From nobody Fri Dec 19 19:07:41 2025 Received: from out-174.mta1.migadu.com (out-174.mta1.migadu.com [95.215.58.174]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 144D628A1ED for ; Wed, 14 May 2025 18:42:29 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=95.215.58.174 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1747248152; cv=none; b=e5hHaLC+FcL9hEeGqJOcjbFtPrAYSuC3C0eY3vR0xhFT7wJEOVUmkuP9n43SSdg770TVOyNN+S5XIELDEA1E99qu57KvOo6eQEYxqZgqTf9PqN3K9WqSlyCoE69ZmyM62KuDu/EH+xjgcy63uQ0Ce9YylOoRIDr/JROm82vopgQ= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1747248152; c=relaxed/simple; bh=+2HNtNHSyHe09aiFKC4uxM3PAH6DOJWNjNKEmVXT27I=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=sgBeY6l1UHIiH+cPd0yC2Eh7ExfzQlHiunS+pp1m/AgtXj4xK3s5DfLnXIPOVNLfLCq6PeeNwRTNhn/YUWBSFxMDcF1rDI9y8/n9Pus4iDLajzm5enhXPAQuW76JXGJFpOGuwp2QILcGScN4xkR2244n2D9cgJ1wyJiXtVwUTng= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev; spf=pass smtp.mailfrom=linux.dev; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b=O82arQvG; arc=none smtp.client-ip=95.215.58.174 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=linux.dev Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=linux.dev Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linux.dev header.i=@linux.dev header.b="O82arQvG" X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1747248148; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=pnF7krW2jVoclxTmKq2Rm1VmAZ/F/mXL38we6iUw/xE=; b=O82arQvGzNDNLe5lnOtBcCE5uvdvACBYJ1PalJjlxxRDPRFgkRaG6x0Uchr66drqz47Jjr PXA6t11IzYBcrVXHGHNqjFWOsEV5uJE2n3iLvDJ3zn0qpesRQJXDy2quVFbtLTIlVna6ae XZZLZPbQIYDaDTrGf0FrrUl72yz+7IM= From: Shakeel Butt To: Andrew Morton Cc: Johannes Weiner , Michal Hocko , Roman Gushchin , Muchun Song , Vlastimil Babka , Alexei Starovoitov , Sebastian Andrzej Siewior , Harry Yoo , Yosry Ahmed , bpf@vger.kernel.org, linux-mm@kvack.org, cgroups@vger.kernel.org, linux-kernel@vger.kernel.org, Meta kernel team Subject: [PATCH v2 2/7] memcg: move preempt disable to callers of memcg_rstat_updated Date: Wed, 14 May 2025 11:41:53 -0700 Message-ID: <20250514184158.3471331-3-shakeel.butt@linux.dev> In-Reply-To: <20250514184158.3471331-1-shakeel.butt@linux.dev> References: <20250514184158.3471331-1-shakeel.butt@linux.dev> 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-Migadu-Flow: FLOW_OUT Content-Type: text/plain; charset="utf-8" Let's move the explicit preempt disable code to the callers of memcg_rstat_updated and also remove the memcg_stats_lock and related functions which ensures the callers of stats update functions have disabled preemption because now the stats update functions are explicitly disabling preemption. Signed-off-by: Shakeel Butt Acked-by: Vlastimil Babka --- mm/memcontrol.c | 74 +++++++++++++------------------------------------ 1 file changed, 19 insertions(+), 55 deletions(-) diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 2464a58fbf17..1750d86012f3 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -557,48 +557,22 @@ static u64 flush_last_time; =20 #define FLUSH_TIME (2UL*HZ) =20 -/* - * Accessors to ensure that preemption is disabled on PREEMPT_RT because i= t can - * not rely on this as part of an acquired spinlock_t lock. These function= s are - * never used in hardirq context on PREEMPT_RT and therefore disabling pre= emtion - * is sufficient. - */ -static void memcg_stats_lock(void) -{ - preempt_disable_nested(); - VM_WARN_ON_IRQS_ENABLED(); -} - -static void __memcg_stats_lock(void) -{ - preempt_disable_nested(); -} - -static void memcg_stats_unlock(void) -{ - preempt_enable_nested(); -} - - static bool memcg_vmstats_needs_flush(struct memcg_vmstats *vmstats) { return atomic64_read(&vmstats->stats_updates) > MEMCG_CHARGE_BATCH * num_online_cpus(); } =20 -static inline void memcg_rstat_updated(struct mem_cgroup *memcg, int val) +static inline void memcg_rstat_updated(struct mem_cgroup *memcg, int val, + int cpu) { struct memcg_vmstats_percpu __percpu *statc_pcpu; struct memcg_vmstats_percpu *statc; - int cpu; unsigned int stats_updates; =20 if (!val) return; =20 - /* Don't assume callers have preemption disabled. */ - cpu =3D get_cpu(); - cgroup_rstat_updated(memcg->css.cgroup, cpu); statc_pcpu =3D memcg->vmstats_percpu; for (; statc_pcpu; statc_pcpu =3D statc->parent_pcpu) { @@ -619,7 +593,6 @@ static inline void memcg_rstat_updated(struct mem_cgrou= p *memcg, int val) stats_updates =3D this_cpu_xchg(statc_pcpu->stats_updates, 0); atomic64_add(stats_updates, &statc->vmstats->stats_updates); } - put_cpu(); } =20 static void __mem_cgroup_flush_stats(struct mem_cgroup *memcg, bool force) @@ -717,6 +690,7 @@ void __mod_memcg_state(struct mem_cgroup *memcg, enum m= emcg_stat_item idx, int val) { int i =3D memcg_stats_index(idx); + int cpu; =20 if (mem_cgroup_disabled()) return; @@ -724,12 +698,14 @@ void __mod_memcg_state(struct mem_cgroup *memcg, enum= memcg_stat_item idx, if (WARN_ONCE(BAD_STAT_IDX(i), "%s: missing stat item %d\n", __func__, id= x)) return; =20 - memcg_stats_lock(); + cpu =3D get_cpu(); + __this_cpu_add(memcg->vmstats_percpu->state[i], val); val =3D memcg_state_val_in_pages(idx, val); - memcg_rstat_updated(memcg, val); + memcg_rstat_updated(memcg, val, cpu); trace_mod_memcg_state(memcg, idx, val); - memcg_stats_unlock(); + + put_cpu(); } =20 #ifdef CONFIG_MEMCG_V1 @@ -758,6 +734,7 @@ static void __mod_memcg_lruvec_state(struct lruvec *lru= vec, struct mem_cgroup_per_node *pn; struct mem_cgroup *memcg; int i =3D memcg_stats_index(idx); + int cpu; =20 if (WARN_ONCE(BAD_STAT_IDX(i), "%s: missing stat item %d\n", __func__, id= x)) return; @@ -765,24 +742,7 @@ static void __mod_memcg_lruvec_state(struct lruvec *lr= uvec, pn =3D container_of(lruvec, struct mem_cgroup_per_node, lruvec); memcg =3D pn->memcg; =20 - /* - * The caller from rmap relies on disabled preemption because they never - * update their counter from in-interrupt context. For these two - * counters we check that the update is never performed from an - * interrupt context while other caller need to have disabled interrupt. - */ - __memcg_stats_lock(); - if (IS_ENABLED(CONFIG_DEBUG_VM)) { - switch (idx) { - case NR_ANON_MAPPED: - case NR_FILE_MAPPED: - case NR_ANON_THPS: - WARN_ON_ONCE(!in_task()); - break; - default: - VM_WARN_ON_IRQS_ENABLED(); - } - } + cpu =3D get_cpu(); =20 /* Update memcg */ __this_cpu_add(memcg->vmstats_percpu->state[i], val); @@ -791,9 +751,10 @@ static void __mod_memcg_lruvec_state(struct lruvec *lr= uvec, __this_cpu_add(pn->lruvec_stats_percpu->state[i], val); =20 val =3D memcg_state_val_in_pages(idx, val); - memcg_rstat_updated(memcg, val); + memcg_rstat_updated(memcg, val, cpu); trace_mod_memcg_lruvec_state(memcg, idx, val); - memcg_stats_unlock(); + + put_cpu(); } =20 /** @@ -873,6 +834,7 @@ void __count_memcg_events(struct mem_cgroup *memcg, enu= m vm_event_item idx, unsigned long count) { int i =3D memcg_events_index(idx); + int cpu; =20 if (mem_cgroup_disabled()) return; @@ -880,11 +842,13 @@ void __count_memcg_events(struct mem_cgroup *memcg, e= num vm_event_item idx, if (WARN_ONCE(BAD_STAT_IDX(i), "%s: missing stat item %d\n", __func__, id= x)) return; =20 - memcg_stats_lock(); + cpu =3D get_cpu(); + __this_cpu_add(memcg->vmstats_percpu->events[i], count); - memcg_rstat_updated(memcg, count); + memcg_rstat_updated(memcg, count, cpu); trace_count_memcg_events(memcg, idx, count); - memcg_stats_unlock(); + + put_cpu(); } =20 unsigned long memcg_events(struct mem_cgroup *memcg, int event) --=20 2.47.1