From nobody Mon Feb 9 08:54:39 2026 Received: from mail-yw1-f202.google.com (mail-yw1-f202.google.com [209.85.128.202]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 72DF1153BE5 for ; Wed, 27 Mar 2024 21:31:38 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.128.202 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1711575100; cv=none; b=V2Jzqb9ZmExJfSfJuP5WDZVXshkvG2jJkuOgJYWnd9c1Ybl5xX56i1rfqGIIsO2/yg+r4V2R3/BfBQGRRybhVq31qczMe+MAN8lgnnaLTNYvEdAGmqf8yp9t4H8gLdYt0Dr+xE+0bOG/IHt8VOXvKXXsQHoWN4fKXUD58l7JUYg= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1711575100; c=relaxed/simple; bh=bBIE5Ow/pYwgtpsi8Q9h2nMAHvXb3FBjBYB5fON9v+o=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=O5wiPpt2zQid3j8VlPMzW6xIpC5uQrixGHn3r74gMYtsxqsWZEJQdN5i2dYI7vlwV12UEk3nWzv8i8I+esRJ4eSPIVyxd7V8PixygDbq9JBP7Wf7o+tsT+ZAtXV64MsRfDWtCialH8skcAs3j3UcruD+TDzMssVkqT9cdRnl76o= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--yuanchu.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=wPEWaDx5; arc=none smtp.client-ip=209.85.128.202 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=flex--yuanchu.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="wPEWaDx5" Received: by mail-yw1-f202.google.com with SMTP id 00721157ae682-60a54004e9fso5812827b3.3 for ; Wed, 27 Mar 2024 14:31:38 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1711575097; x=1712179897; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=xaCDmpnOXna8KuNDUqa6ynYFsa6ZfJ9LVAHJUw2Aofg=; b=wPEWaDx5HTkYByOae+MOT/UVhBJ84y2h9ndPFVFYN/t7E3lZlwxr6f8Zkk8cGaabvf 2iRZ0EjXD7q0I+Eb7S3/TCCVMM8AZwsqvD+6V7ulSGgq5nEPRt9Ys0UHcxAQiGM6punv U8gkabHKRFn7zSYiLDOtvgM4yYxj5eAZ4KITpIwlaKfv5YOItTtS57b1ieuxrvSCRAI5 79EPa8+i9BEpKAHRiqn7f5stprOM/FptmIpY9fKCrUx3gWXRN+CBEgcLs/yrBkIkiT+6 yA0WVTfTakDWsGZZA7ZwvvGYZMdMA1ARb1X3vCl+IWyqf//nLWGH3TPkvk4XU5prYHeU Hf/Q== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1711575097; x=1712179897; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=xaCDmpnOXna8KuNDUqa6ynYFsa6ZfJ9LVAHJUw2Aofg=; b=JcgvCE4/KTxEVAAxycv7deNBJQsC8+5tHdm8CpqHumsiD55sh5qIqB+lNs20ae0qEj 5j1WWp4/dH2kTH+DE7KeYkgIpu80vpKoqPSlKr6T1eSiX+28DiWnQmo1Nlubtne+J09P LrhGKrx8vz5yJzMzU5TrY6R8FYUav4pQPGy0J4eZmlHlGWa7OLWy/G35FNpEH2lN3bt5 +YQyHlCPFbhyNd7FsH7kc+lzcjvp3ZmCMZ9lft3tAKCXuW+2kAf2key2jiMxD+sIo+nn FRi27liyFCBT5MbzvfztpHENJjcC7r2SUEVTRdOLAfL3lOM0Eo6yGphKKokhwKe549DV TV8Q== X-Forwarded-Encrypted: i=1; AJvYcCUx03KfmFM7/xMGpwBvLoMcp8VLmgEwgrgG4nQ3QfwmYjLH+yjrBq1t/b1LEnRQ/1PoDQpFi41OJz5I+zCOYjqBmwmLPxjpGGaFR7vx X-Gm-Message-State: AOJu0Yw6Myuuyc4ETI0/GhY+WtjBwyirKi9at2Fo3ru7BDlZamk980sC +X6pWjM6yX65+mY7xSeFAWFSRvJUyeIt2p1kQ1lNSM6pkpRrtz28EtgmakrBmBUEw37ohetJfr2 xBUobQQ== X-Google-Smtp-Source: AGHT+IHPhGifJWkaGVpVQkyYgUvbZ7gXC3zKKNE+5jad1F4KE6vLxnBUAJz0fdIi3IF4p9BgtNJpeDc4GOj3 X-Received: from yuanchu-desktop.svl.corp.google.com ([2620:15c:2a3:200:6df3:ef42:a58e:a6b1]) (user=yuanchu job=sendgmr) by 2002:a05:690c:b9a:b0:610:dc1b:8f39 with SMTP id ck26-20020a05690c0b9a00b00610dc1b8f39mr216073ywb.0.1711575097511; Wed, 27 Mar 2024 14:31:37 -0700 (PDT) Date: Wed, 27 Mar 2024 14:31:04 -0700 In-Reply-To: <20240327213108.2384666-1-yuanchu@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20240327213108.2384666-1-yuanchu@google.com> X-Mailer: git-send-email 2.44.0.396.g6e790dbe36-goog Message-ID: <20240327213108.2384666-6-yuanchu@google.com> Subject: [RFC PATCH v3 5/8] mm: extend working set reporting to memcgs From: Yuanchu Xie To: David Hildenbrand , "Aneesh Kumar K.V" , Khalid Aziz , Henry Huang , Yu Zhao , Dan Williams , Gregory Price , Huang Ying Cc: Wei Xu , David Rientjes , Greg Kroah-Hartman , "Rafael J. Wysocki" , Andrew Morton , Johannes Weiner , Michal Hocko , Roman Gushchin , Muchun Song , Shuah Khan , Yosry Ahmed , Matthew Wilcox , Sudarshan Rajagopalan , Kairui Song , "Michael S. Tsirkin" , Vasily Averin , Nhat Pham , Miaohe Lin , Qi Zheng , Abel Wu , "Vishal Moola (Oracle)" , Kefeng Wang , Yuanchu Xie , linux-kernel@vger.kernel.org, linux-mm@kvack.org, cgroups@vger.kernel.org, linux-kselftest@vger.kernel.org Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Break down the system-wide working set reporting into per-memcg reports, which aggregages its children hierarchically. The per-node working set reporting histograms and refresh/report threshold files are presented as memcg files, showing a report containing all the nodes. Memcg interface: /sys/fs/cgroup/.../memory.workingset.page_age The memcg equivalent of the sysfs workingset page age histogram, breaks down the workingset of this memcg and its children into page age intervals. Each node is prefixed with a node header and a newline. Non-proactive direct reclaim on this memcg can also wake up userspace agents that are waiting on this file. e.g. N0 1000 anon=3D0 file=3D0 2000 anon=3D0 file=3D0 3000 anon=3D0 file=3D0 4000 anon=3D0 file=3D0 5000 anon=3D0 file=3D0 18446744073709551615 anon=3D0 file=3D0 /sys/fs/cgroup/.../memory.workingset.page_age_intervals Configures the intervals for the page age histogram. This file operates on a per-node basis, allowing for different intervals for each node. e.g. echo N0=3D1000,2000,3000,4000,5000 > memory.workingset.page_age_intervals /sys/fs/cgroup/.../memory.workingset.refresh_interval The memcg equivalent of the sysfs refresh interval. A per-node number of how much time a page age histogram is valid for, in milliseconds. e.g. echo N0=3D2000 > memory.workingset.refresh_interval /sys/fs/cgroup/.../memory.workingset.report_threshold The memcg equivalent of the sysfs report threshold. A per-node number of how often userspace agent waiting on the page age histogram can be woken up, in milliseconds. e.g. echo N0=3D1000 > memory.workingset.report_threshold Signed-off-by: Yuanchu Xie --- include/linux/memcontrol.h | 5 + include/linux/workingset_report.h | 6 +- mm/internal.h | 2 + mm/memcontrol.c | 267 +++++++++++++++++++++++++++++- mm/workingset_report.c | 10 +- 5 files changed, 286 insertions(+), 4 deletions(-) diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h index 20ff87f8e001..7d7bc0928961 100644 --- a/include/linux/memcontrol.h +++ b/include/linux/memcontrol.h @@ -335,6 +335,11 @@ struct mem_cgroup { struct lru_gen_mm_list mm_list; #endif =20 +#ifdef CONFIG_WORKINGSET_REPORT + /* memory.workingset.page_age file */ + struct cgroup_file workingset_page_age_file; +#endif + struct mem_cgroup_per_node *nodeinfo[]; }; =20 diff --git a/include/linux/workingset_report.h b/include/linux/workingset_r= eport.h index 589d240d6251..502542c812b3 100644 --- a/include/linux/workingset_report.h +++ b/include/linux/workingset_report.h @@ -9,6 +9,7 @@ struct mem_cgroup; struct pglist_data; struct node; struct lruvec; +struct cgroup_file; =20 #ifdef CONFIG_WORKINGSET_REPORT =20 @@ -38,7 +39,10 @@ struct wsr_state { unsigned long report_threshold; unsigned long refresh_interval; =20 - struct kernfs_node *page_age_sys_file; + union { + struct kernfs_node *page_age_sys_file; + struct cgroup_file *page_age_cgroup_file; + }; =20 /* breakdown of workingset by page age */ struct mutex page_age_lock; diff --git a/mm/internal.h b/mm/internal.h index 36480c7ac0dd..3730c8399ad4 100644 --- a/mm/internal.h +++ b/mm/internal.h @@ -212,6 +212,8 @@ extern void reclaim_throttle(pg_data_t *pgdat, enum vms= can_throttle_state reason void notify_workingset(struct mem_cgroup *memcg, struct pglist_data *pgdat= ); /* Requires wsr->page_age_lock held */ void wsr_refresh_scan(struct lruvec *lruvec, unsigned long refresh_interva= l); +int workingset_report_intervals_parse(char *src, + struct wsr_report_bins *bins); #else static inline void notify_workingset(struct mem_cgroup *memcg, struct pglist_data *pgdat) diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 2f07141de16c..75bda5f7994d 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c @@ -7005,6 +7005,245 @@ static ssize_t memory_reclaim(struct kernfs_open_fi= le *of, char *buf, return nbytes; } =20 +#ifdef CONFIG_WORKINGSET_REPORT +static int memory_ws_page_age_intervals_show(struct seq_file *m, void *v) +{ + int nid; + struct mem_cgroup *memcg =3D mem_cgroup_from_seq(m); + + for_each_node_state(nid, N_MEMORY) { + struct wsr_state *wsr; + struct wsr_page_age_histo *page_age; + int i, nr_bins; + + wsr =3D &mem_cgroup_lruvec(memcg, NODE_DATA(nid))->wsr; + mutex_lock(&wsr->page_age_lock); + page_age =3D wsr->page_age; + if (!page_age) + goto no_page_age; + seq_printf(m, "N%d=3D", nid); + nr_bins =3D page_age->bins.nr_bins; + for (i =3D 0; i < nr_bins; ++i) { + struct wsr_report_bin *bin =3D + &page_age->bins.bins[i]; + + seq_printf(m, "%u", jiffies_to_msecs(bin->idle_age)); + if (i + 1 < nr_bins) + seq_putc(m, ','); + } + seq_putc(m, ' '); +no_page_age: + mutex_unlock(&wsr->page_age_lock); + } + seq_putc(m, '\n'); + + return 0; +} + +static ssize_t memory_wsr_interval_parse(struct kernfs_open_file *of, char= *buf, + size_t nbytes, unsigned int *nid_out, + struct wsr_report_bins *bins) +{ + char *node, *intervals; + unsigned int nid; + int err; + + buf =3D strstrip(buf); + intervals =3D buf; + node =3D strsep(&intervals, "=3D"); + + if (*node !=3D 'N') + return -EINVAL; + + err =3D kstrtouint(node + 1, 0, &nid); + if (err) + return err; + + if (nid >=3D nr_node_ids || !node_state(nid, N_MEMORY)) + return -EINVAL; + + err =3D workingset_report_intervals_parse(intervals, bins); + if (err < 0) + return err; + + *nid_out =3D nid; + return err; +} + +static ssize_t memory_ws_page_age_intervals_write(struct kernfs_open_file = *of, + char *buf, size_t nbytes, + loff_t off) +{ + unsigned int nid; + int err; + struct wsr_page_age_histo *page_age =3D NULL, *old; + struct wsr_state *wsr; + struct mem_cgroup *memcg =3D mem_cgroup_from_css(of_css(of)); + + page_age =3D + kzalloc(sizeof(struct wsr_page_age_histo), GFP_KERNEL_ACCOUNT); + + if (!page_age) { + err =3D -ENOMEM; + goto failed; + } + + err =3D memory_wsr_interval_parse(of, buf, nbytes, &nid, &page_age->bins); + if (err < 0) + goto failed; + + if (err =3D=3D 0) { + kfree(page_age); + page_age =3D NULL; + } + + wsr =3D &mem_cgroup_lruvec(memcg, NODE_DATA(nid))->wsr; + mutex_lock(&wsr->page_age_lock); + old =3D xchg(&wsr->page_age, page_age); + mutex_unlock(&wsr->page_age_lock); + kfree(old); + return nbytes; +failed: + kfree(page_age); + return err; +} + +static int memory_ws_refresh_interval_show(struct seq_file *m, void *v) +{ + int nid; + struct mem_cgroup *memcg =3D mem_cgroup_from_seq(m); + + for_each_node_state(nid, N_MEMORY) { + struct wsr_state *wsr =3D + &mem_cgroup_lruvec(memcg, NODE_DATA(nid))->wsr; + + seq_printf(m, "N%d=3D%u ", nid, + jiffies_to_msecs(READ_ONCE(wsr->refresh_interval))); + } + seq_putc(m, '\n'); + + return 0; +} + +static ssize_t memory_wsr_threshold_parse(char *buf, size_t nbytes, + unsigned int *nid_out, + unsigned int *msecs) +{ + char *node, *threshold; + unsigned int nid; + int err; + + buf =3D strstrip(buf); + threshold =3D buf; + node =3D strsep(&threshold, "=3D"); + + if (*node !=3D 'N') + return -EINVAL; + + err =3D kstrtouint(node + 1, 0, &nid); + if (err) + return err; + + if (nid >=3D nr_node_ids || !node_state(nid, N_MEMORY)) + return -EINVAL; + + err =3D kstrtouint(threshold, 0, msecs); + if (err) + return err; + + *nid_out =3D nid; + + return nbytes; +} + +static ssize_t memory_ws_refresh_interval_write(struct kernfs_open_file *o= f, + char *buf, size_t nbytes, + loff_t off) +{ + unsigned int nid, msecs; + struct wsr_state *wsr; + struct mem_cgroup *memcg =3D mem_cgroup_from_css(of_css(of)); + ssize_t ret =3D memory_wsr_threshold_parse(buf, nbytes, &nid, &msecs); + + if (ret < 0) + return ret; + + wsr =3D &mem_cgroup_lruvec(memcg, NODE_DATA(nid))->wsr; + WRITE_ONCE(wsr->refresh_interval, msecs_to_jiffies(msecs)); + return ret; +} + +static int memory_ws_report_threshold_show(struct seq_file *m, void *v) +{ + int nid; + struct mem_cgroup *memcg =3D mem_cgroup_from_seq(m); + + for_each_node_state(nid, N_MEMORY) { + struct wsr_state *wsr =3D + &mem_cgroup_lruvec(memcg, NODE_DATA(nid))->wsr; + + seq_printf(m, "N%d=3D%u ", nid, + jiffies_to_msecs(READ_ONCE(wsr->report_threshold))); + } + seq_putc(m, '\n'); + + return 0; +} + +static ssize_t memory_ws_report_threshold_write(struct kernfs_open_file *o= f, + char *buf, size_t nbytes, + loff_t off) +{ + unsigned int nid, msecs; + struct wsr_state *wsr; + struct mem_cgroup *memcg =3D mem_cgroup_from_css(of_css(of)); + ssize_t ret =3D memory_wsr_threshold_parse(buf, nbytes, &nid, &msecs); + + if (ret < 0) + return ret; + + wsr =3D &mem_cgroup_lruvec(memcg, NODE_DATA(nid))->wsr; + WRITE_ONCE(wsr->report_threshold, msecs_to_jiffies(msecs)); + return ret; +} + +static int memory_ws_page_age_show(struct seq_file *m, void *v) +{ + int nid; + struct mem_cgroup *memcg =3D mem_cgroup_from_seq(m); + + for_each_node_state(nid, N_MEMORY) { + struct wsr_state *wsr =3D + &mem_cgroup_lruvec(memcg, NODE_DATA(nid))->wsr; + struct wsr_report_bin *bin; + + if (!READ_ONCE(wsr->page_age)) + continue; + + wsr_refresh_report(wsr, memcg, NODE_DATA(nid)); + mutex_lock(&wsr->page_age_lock); + if (!wsr->page_age) + goto unlock; + seq_printf(m, "N%d\n", nid); + for (bin =3D wsr->page_age->bins.bins; + bin->idle_age !=3D WORKINGSET_INTERVAL_MAX; bin++) + seq_printf(m, "%u anon=3D%lu file=3D%lu\n", + jiffies_to_msecs(bin->idle_age), + bin->nr_pages[0] * PAGE_SIZE, + bin->nr_pages[1] * PAGE_SIZE); + + seq_printf(m, "%lu anon=3D%lu file=3D%lu\n", WORKINGSET_INTERVAL_MAX, + bin->nr_pages[0] * PAGE_SIZE, + bin->nr_pages[1] * PAGE_SIZE); + +unlock: + mutex_unlock(&wsr->page_age_lock); + } + + return 0; +} +#endif + static struct cftype memory_files[] =3D { { .name =3D "current", @@ -7073,7 +7312,33 @@ static struct cftype memory_files[] =3D { .flags =3D CFTYPE_NS_DELEGATABLE, .write =3D memory_reclaim, }, - { } /* terminate */ +#ifdef CONFIG_WORKINGSET_REPORT + { + .name =3D "workingset.page_age_intervals", + .flags =3D CFTYPE_NOT_ON_ROOT | CFTYPE_NS_DELEGATABLE, + .seq_show =3D memory_ws_page_age_intervals_show, + .write =3D memory_ws_page_age_intervals_write, + }, + { + .name =3D "workingset.refresh_interval", + .flags =3D CFTYPE_NOT_ON_ROOT | CFTYPE_NS_DELEGATABLE, + .seq_show =3D memory_ws_refresh_interval_show, + .write =3D memory_ws_refresh_interval_write, + }, + { + .name =3D "workingset.report_threshold", + .flags =3D CFTYPE_NOT_ON_ROOT | CFTYPE_NS_DELEGATABLE, + .seq_show =3D memory_ws_report_threshold_show, + .write =3D memory_ws_report_threshold_write, + }, + { + .name =3D "workingset.page_age", + .flags =3D CFTYPE_NOT_ON_ROOT | CFTYPE_NS_DELEGATABLE, + .file_offset =3D offsetof(struct mem_cgroup, workingset_page_age_file), + .seq_show =3D memory_ws_page_age_show, + }, +#endif + {} /* terminate */ }; =20 struct cgroup_subsys memory_cgrp_subsys =3D { diff --git a/mm/workingset_report.c b/mm/workingset_report.c index 3ed3b0e8f8ad..b00ffbfebcab 100644 --- a/mm/workingset_report.c +++ b/mm/workingset_report.c @@ -20,9 +20,12 @@ void wsr_init(struct lruvec *lruvec) { struct wsr_state *wsr =3D &lruvec->wsr; + struct mem_cgroup *memcg =3D lruvec_memcg(lruvec); =20 memset(wsr, 0, sizeof(*wsr)); mutex_init(&wsr->page_age_lock); + if (memcg && !mem_cgroup_is_root(memcg)) + wsr->page_age_cgroup_file =3D &memcg->workingset_page_age_file; } =20 void wsr_destroy(struct lruvec *lruvec) @@ -34,7 +37,7 @@ void wsr_destroy(struct lruvec *lruvec) memset(wsr, 0, sizeof(*wsr)); } =20 -static int workingset_report_intervals_parse(char *src, +int workingset_report_intervals_parse(char *src, struct wsr_report_bins *bins) { int err =3D 0, i =3D 0; @@ -490,5 +493,8 @@ void notify_workingset(struct mem_cgroup *memcg, struct= pglist_data *pgdat) { struct wsr_state *wsr =3D &mem_cgroup_lruvec(memcg, pgdat)->wsr; =20 - kernfs_notify(wsr->page_age_sys_file); + if (mem_cgroup_is_root(memcg)) + kernfs_notify(wsr->page_age_sys_file); + else + cgroup_file_notify(wsr->page_age_cgroup_file); } --=20 2.44.0.396.g6e790dbe36-goog