From nobody Mon Feb 9 11:09:47 2026 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (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 919E21F4705 for ; Fri, 31 Jan 2025 22:29:26 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1738362566; cv=none; b=FtWczE2hOjThagOVQFxzAnQ6dcTJXjakNUuXRA5pjj164FYkNXyeeasgB+TQ83kIioV/0z2YGutcxVlghTb9GQpZ7XEwS39PPNx4wddLE+qBRtIVXdqHKsp1ViaNPmTjnRcdrhBRiI3JXg28PTazfX6Z/x6FFXIBVmwxPMR07F4= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1738362566; c=relaxed/simple; bh=Bym4XLs9ayGHpB6cy9Lr9srUq23S23rW9fACtmSOnvA=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=WPx7qi7CTTohZWUcKtgwCrwFdfSw7Q50TSZTXgi3MrP2kHhnCyIDUT1N4pzEuVob+LydoYljulw7rRvdvTGrxOCZJXdOfusA37HYsbrJ9UPFX67j7tadnYJbHkILMOSCpxskfPz7UIPZZbtmvTw5XjdSHhnS5hRcClO+xmdezUA= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=uS4FYg+b; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="uS4FYg+b" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 01208C4CEE4; Fri, 31 Jan 2025 22:29:25 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1738362566; bh=Bym4XLs9ayGHpB6cy9Lr9srUq23S23rW9fACtmSOnvA=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=uS4FYg+bf32AOV5PPLHwyOyi624/ySwj5D8KI+Cxj3vzz2KEAekzjBDAmGnwtp0Js FKXTPpiUeNH4kt4fkghKuBKfXt7d7WPiN+bgRL15HZtRh80CYprHKGw8Pu3sIufx0D ZdJRcxh1N8CLypfj42NeojHE0uua7kwhmPKJJmm8mFqv55DmrJ8LawZAH7gzghsJF3 Ucum7dwqq5WE+QnVRxnJT8r4rPY+PZaXCmwkIF8FYjTOTgkDb7J5GZVhUjYpVMo2Eb wa1KA/c7XeHBgoynilRFWe4cpT6uNbJBzOtUmFjRJpIKMJHHbcKGUTYmIZkmUFUlcz /35qATlGW2w+A== From: Jaegeuk Kim To: linux-kernel@vger.kernel.org, linux-f2fs-devel@lists.sourceforge.net Cc: Jaegeuk Kim Subject: [PATCH 2/2] f2fs: add a sysfs entry to request donate file-backed pages Date: Fri, 31 Jan 2025 22:27:57 +0000 Message-ID: <20250131222914.1634961-3-jaegeuk@kernel.org> X-Mailer: git-send-email 2.48.1.362.g079036d154-goog In-Reply-To: <20250131222914.1634961-1-jaegeuk@kernel.org> References: <20250131222914.1634961-1-jaegeuk@kernel.org> 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 Content-Type: text/plain; charset="utf-8" 1. ioctl(fd1, F2FS_IOC_DONATE_RANGE, {0,3}); 2. ioctl(fd2, F2FS_IOC_DONATE_RANGE, {1,2}); 3. ioctl(fd3, F2FS_IOC_DONATE_RANGE, {3,1}); 4. echo 1024 > /sys/fs/f2fs/tuning/reclaim_caches_kb This gives a way to reclaim file-backed pages by iterating all f2fs mounts = until reclaiming 1MB page cache ranges, registered by #1, #2, and #3. 5. cat /sys/fs/f2fs/tuning/reclaim_caches_kb -> gives total number of registered file ranges. Signed-off-by: Jaegeuk Kim Reviewed-by: Chao Yu --- Documentation/ABI/testing/sysfs-fs-f2fs | 7 ++ fs/f2fs/f2fs.h | 2 + fs/f2fs/shrinker.c | 90 +++++++++++++++++++++++++ fs/f2fs/sysfs.c | 63 +++++++++++++++++ 4 files changed, 162 insertions(+) diff --git a/Documentation/ABI/testing/sysfs-fs-f2fs b/Documentation/ABI/te= sting/sysfs-fs-f2fs index 3e1630c70d8a..81deae2af84d 100644 --- a/Documentation/ABI/testing/sysfs-fs-f2fs +++ b/Documentation/ABI/testing/sysfs-fs-f2fs @@ -828,3 +828,10 @@ Date: November 2024 Contact: "Chao Yu" Description: It controls max read extent count for per-inode, the value of= threshold is 10240 by default. + +What: /sys/fs/f2fs/tuning/reclaim_caches_kb +Date: February 2025 +Contact: "Jaegeuk Kim" +Description: It reclaims the given KBs of file-backed pages registered by + ioctl(F2FS_IOC_DONATE_RANGE). + For example, writing N tries to drop N KBs spaces in LRU. diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 805585a7d2b6..bd0d8138b71d 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -4241,6 +4241,8 @@ unsigned long f2fs_shrink_count(struct shrinker *shri= nk, struct shrink_control *sc); unsigned long f2fs_shrink_scan(struct shrinker *shrink, struct shrink_control *sc); +unsigned int f2fs_donate_files(void); +void f2fs_reclaim_caches(unsigned int reclaim_caches_kb); void f2fs_join_shrinker(struct f2fs_sb_info *sbi); void f2fs_leave_shrinker(struct f2fs_sb_info *sbi); =20 diff --git a/fs/f2fs/shrinker.c b/fs/f2fs/shrinker.c index 83d6fb97dcae..45efff635d8e 100644 --- a/fs/f2fs/shrinker.c +++ b/fs/f2fs/shrinker.c @@ -130,6 +130,96 @@ unsigned long f2fs_shrink_scan(struct shrinker *shrink, return freed; } =20 +unsigned int f2fs_donate_files(void) +{ + struct f2fs_sb_info *sbi; + struct list_head *p; + unsigned int donate_files =3D 0; + + spin_lock(&f2fs_list_lock); + p =3D f2fs_list.next; + while (p !=3D &f2fs_list) { + sbi =3D list_entry(p, struct f2fs_sb_info, s_list); + + /* stop f2fs_put_super */ + if (!mutex_trylock(&sbi->umount_mutex)) { + p =3D p->next; + continue; + } + spin_unlock(&f2fs_list_lock); + + donate_files +=3D sbi->donate_files; + + spin_lock(&f2fs_list_lock); + p =3D p->next; + mutex_unlock(&sbi->umount_mutex); + } + spin_unlock(&f2fs_list_lock); + + return donate_files; +} + +static unsigned int do_reclaim_caches(struct f2fs_sb_info *sbi, + unsigned int reclaim_caches_kb) +{ + struct inode *inode; + struct f2fs_inode_info *fi; + unsigned int nfiles =3D sbi->donate_files; + pgoff_t npages =3D reclaim_caches_kb >> (PAGE_SHIFT - 10); + + while (npages && nfiles--) { + pgoff_t len; + + spin_lock(&sbi->inode_lock[DONATE_INODE]); + if (list_empty(&sbi->inode_list[DONATE_INODE])) { + spin_unlock(&sbi->inode_lock[DONATE_INODE]); + break; + } + fi =3D list_first_entry(&sbi->inode_list[DONATE_INODE], + struct f2fs_inode_info, gdonate_list); + list_move_tail(&fi->gdonate_list, &sbi->inode_list[DONATE_INODE]); + inode =3D igrab(&fi->vfs_inode); + spin_unlock(&sbi->inode_lock[DONATE_INODE]); + + if (!inode) + continue; + + len =3D fi->donate_end - fi->donate_start + 1; + npages =3D npages < len ? 0 : npages - len; + invalidate_inode_pages2_range(inode->i_mapping, + fi->donate_start, fi->donate_end); + iput(inode); + cond_resched(); + } + return npages << (PAGE_SHIFT - 10); +} + +void f2fs_reclaim_caches(unsigned int reclaim_caches_kb) +{ + struct f2fs_sb_info *sbi; + struct list_head *p; + + spin_lock(&f2fs_list_lock); + p =3D f2fs_list.next; + while (p !=3D &f2fs_list && reclaim_caches_kb) { + sbi =3D list_entry(p, struct f2fs_sb_info, s_list); + + /* stop f2fs_put_super */ + if (!mutex_trylock(&sbi->umount_mutex)) { + p =3D p->next; + continue; + } + spin_unlock(&f2fs_list_lock); + + reclaim_caches_kb =3D do_reclaim_caches(sbi, reclaim_caches_kb); + + spin_lock(&f2fs_list_lock); + p =3D p->next; + mutex_unlock(&sbi->umount_mutex); + } + spin_unlock(&f2fs_list_lock); +} + void f2fs_join_shrinker(struct f2fs_sb_info *sbi) { spin_lock(&f2fs_list_lock); diff --git a/fs/f2fs/sysfs.c b/fs/f2fs/sysfs.c index 4bd7b17a20c8..579226a05a69 100644 --- a/fs/f2fs/sysfs.c +++ b/fs/f2fs/sysfs.c @@ -916,6 +916,39 @@ static struct f2fs_base_attr f2fs_base_attr_##_name = =3D { \ .show =3D f2fs_feature_show, \ } =20 +static ssize_t f2fs_tune_show(struct f2fs_base_attr *a, char *buf) +{ + unsigned int res; + + if (!strcmp(a->attr.name, "reclaim_caches_kb")) + res =3D f2fs_donate_files(); + + return sysfs_emit(buf, "%u\n", res); +} + +static ssize_t f2fs_tune_store(struct f2fs_base_attr *a, + const char *buf, size_t count) +{ + unsigned long t; + int ret; + + ret =3D kstrtoul(skip_spaces(buf), 0, &t); + if (ret) + return ret; + + if (!strcmp(a->attr.name, "reclaim_caches_kb")) + f2fs_reclaim_caches(t); + + return ret ? ret : count; +} + +#define F2FS_TUNE_RW_ATTR(_name) \ +static struct f2fs_base_attr f2fs_base_attr_##_name =3D { \ + .attr =3D {.name =3D __stringify(_name), .mode =3D 0644 }, \ + .show =3D f2fs_tune_show, \ + .store =3D f2fs_tune_store, \ +} + static ssize_t f2fs_sb_feature_show(struct f2fs_attr *a, struct f2fs_sb_info *sbi, char *buf) { @@ -1368,6 +1401,14 @@ static struct attribute *f2fs_sb_feat_attrs[] =3D { }; ATTRIBUTE_GROUPS(f2fs_sb_feat); =20 +F2FS_TUNE_RW_ATTR(reclaim_caches_kb); + +static struct attribute *f2fs_tune_attrs[] =3D { + BASE_ATTR_LIST(reclaim_caches_kb), + NULL, +}; +ATTRIBUTE_GROUPS(f2fs_tune); + static const struct sysfs_ops f2fs_attr_ops =3D { .show =3D f2fs_attr_show, .store =3D f2fs_attr_store, @@ -1401,6 +1442,20 @@ static struct kobject f2fs_feat =3D { .kset =3D &f2fs_kset, }; =20 +static const struct sysfs_ops f2fs_tune_attr_ops =3D { + .show =3D f2fs_base_attr_show, + .store =3D f2fs_base_attr_store, +}; + +static const struct kobj_type f2fs_tune_ktype =3D { + .default_groups =3D f2fs_tune_groups, + .sysfs_ops =3D &f2fs_tune_attr_ops, +}; + +static struct kobject f2fs_tune =3D { + .kset =3D &f2fs_kset, +}; + static ssize_t f2fs_stat_attr_show(struct kobject *kobj, struct attribute *attr, char *buf) { @@ -1637,6 +1692,11 @@ int __init f2fs_init_sysfs(void) if (ret) goto unregister_out; =20 + ret =3D kobject_init_and_add(&f2fs_tune, &f2fs_tune_ktype, + NULL, "tuning"); + if (ret) + goto put_feat; + f2fs_proc_root =3D proc_mkdir("fs/f2fs", NULL); if (!f2fs_proc_root) { ret =3D -ENOMEM; @@ -1645,6 +1705,8 @@ int __init f2fs_init_sysfs(void) =20 return 0; put_kobject: + kobject_put(&f2fs_tune); +put_feat: kobject_put(&f2fs_feat); unregister_out: kset_unregister(&f2fs_kset); @@ -1653,6 +1715,7 @@ int __init f2fs_init_sysfs(void) =20 void f2fs_exit_sysfs(void) { + kobject_put(&f2fs_tune); kobject_put(&f2fs_feat); kset_unregister(&f2fs_kset); remove_proc_entry("fs/f2fs", NULL); --=20 2.48.1.362.g079036d154-goog