linux-stable/fs/btrfs/sysfs.c

2494 lines
66 KiB
C
Raw Permalink Normal View History

// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (C) 2007 Oracle. All rights reserved.
*/
#include <linux/sched.h>
#include <linux/sched/mm.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/completion.h>
#include <linux/bug.h>
#include <linux/list.h>
#include <crypto/hash.h>
#include "messages.h"
#include "ctree.h"
#include "discard.h"
#include "disk-io.h"
#include "send.h"
#include "transaction.h"
#include "sysfs.h"
#include "volumes.h"
#include "space-info.h"
#include "block-group.h"
#include "qgroup.h"
#include "misc.h"
#include "fs.h"
#include "accessors.h"
/*
* Structure name Path
* --------------------------------------------------------------------------
* btrfs_supported_static_feature_attrs /sys/fs/btrfs/features
* btrfs_supported_feature_attrs /sys/fs/btrfs/features and
* /sys/fs/btrfs/<uuid>/features
* btrfs_attrs /sys/fs/btrfs/<uuid>
* devid_attrs /sys/fs/btrfs/<uuid>/devinfo/<devid>
* allocation_attrs /sys/fs/btrfs/<uuid>/allocation
* qgroup_attrs /sys/fs/btrfs/<uuid>/qgroups/<level>_<qgroupid>
* space_info_attrs /sys/fs/btrfs/<uuid>/allocation/<bg-type>
* raid_attrs /sys/fs/btrfs/<uuid>/allocation/<bg-type>/<bg-profile>
* discard_attrs /sys/fs/btrfs/<uuid>/discard
*
* When built with BTRFS_CONFIG_DEBUG:
*
* btrfs_debug_feature_attrs /sys/fs/btrfs/debug
* btrfs_debug_mount_attrs /sys/fs/btrfs/<uuid>/debug
*/
struct btrfs_feature_attr {
struct kobj_attribute kobj_attr;
enum btrfs_feature_set feature_set;
u64 feature_bit;
};
/* For raid type sysfs entries */
struct raid_kobject {
u64 flags;
struct kobject kobj;
};
#define __INIT_KOBJ_ATTR(_name, _mode, _show, _store) \
{ \
.attr = { .name = __stringify(_name), .mode = _mode }, \
.show = _show, \
.store = _store, \
}
#define BTRFS_ATTR_W(_prefix, _name, _store) \
static struct kobj_attribute btrfs_attr_##_prefix##_##_name = \
__INIT_KOBJ_ATTR(_name, 0200, NULL, _store)
#define BTRFS_ATTR_RW(_prefix, _name, _show, _store) \
static struct kobj_attribute btrfs_attr_##_prefix##_##_name = \
__INIT_KOBJ_ATTR(_name, 0644, _show, _store)
#define BTRFS_ATTR(_prefix, _name, _show) \
static struct kobj_attribute btrfs_attr_##_prefix##_##_name = \
__INIT_KOBJ_ATTR(_name, 0444, _show, NULL)
#define BTRFS_ATTR_PTR(_prefix, _name) \
(&btrfs_attr_##_prefix##_##_name.attr)
#define BTRFS_FEAT_ATTR(_name, _feature_set, _feature_prefix, _feature_bit) \
static struct btrfs_feature_attr btrfs_attr_features_##_name = { \
.kobj_attr = __INIT_KOBJ_ATTR(_name, S_IRUGO, \
btrfs_feature_attr_show, \
btrfs_feature_attr_store), \
.feature_set = _feature_set, \
.feature_bit = _feature_prefix ##_## _feature_bit, \
}
#define BTRFS_FEAT_ATTR_PTR(_name) \
(&btrfs_attr_features_##_name.kobj_attr.attr)
#define BTRFS_FEAT_ATTR_COMPAT(name, feature) \
BTRFS_FEAT_ATTR(name, FEAT_COMPAT, BTRFS_FEATURE_COMPAT, feature)
#define BTRFS_FEAT_ATTR_COMPAT_RO(name, feature) \
BTRFS_FEAT_ATTR(name, FEAT_COMPAT_RO, BTRFS_FEATURE_COMPAT_RO, feature)
#define BTRFS_FEAT_ATTR_INCOMPAT(name, feature) \
BTRFS_FEAT_ATTR(name, FEAT_INCOMPAT, BTRFS_FEATURE_INCOMPAT, feature)
static inline struct btrfs_fs_info *to_fs_info(struct kobject *kobj);
static inline struct btrfs_fs_devices *to_fs_devs(struct kobject *kobj);
static struct kobject *get_btrfs_kobj(struct kobject *kobj);
static struct btrfs_feature_attr *to_btrfs_feature_attr(struct kobj_attribute *a)
{
return container_of(a, struct btrfs_feature_attr, kobj_attr);
}
static struct kobj_attribute *attr_to_btrfs_attr(struct attribute *attr)
{
return container_of(attr, struct kobj_attribute, attr);
}
static struct btrfs_feature_attr *attr_to_btrfs_feature_attr(
struct attribute *attr)
{
return to_btrfs_feature_attr(attr_to_btrfs_attr(attr));
}
static u64 get_features(struct btrfs_fs_info *fs_info,
enum btrfs_feature_set set)
{
struct btrfs_super_block *disk_super = fs_info->super_copy;
if (set == FEAT_COMPAT)
return btrfs_super_compat_flags(disk_super);
else if (set == FEAT_COMPAT_RO)
return btrfs_super_compat_ro_flags(disk_super);
else
return btrfs_super_incompat_flags(disk_super);
}
static void set_features(struct btrfs_fs_info *fs_info,
enum btrfs_feature_set set, u64 features)
{
struct btrfs_super_block *disk_super = fs_info->super_copy;
if (set == FEAT_COMPAT)
btrfs_set_super_compat_flags(disk_super, features);
else if (set == FEAT_COMPAT_RO)
btrfs_set_super_compat_ro_flags(disk_super, features);
else
btrfs_set_super_incompat_flags(disk_super, features);
}
static int can_modify_feature(struct btrfs_feature_attr *fa)
{
int val = 0;
u64 set, clear;
switch (fa->feature_set) {
case FEAT_COMPAT:
set = BTRFS_FEATURE_COMPAT_SAFE_SET;
clear = BTRFS_FEATURE_COMPAT_SAFE_CLEAR;
break;
case FEAT_COMPAT_RO:
set = BTRFS_FEATURE_COMPAT_RO_SAFE_SET;
clear = BTRFS_FEATURE_COMPAT_RO_SAFE_CLEAR;
break;
case FEAT_INCOMPAT:
set = BTRFS_FEATURE_INCOMPAT_SAFE_SET;
clear = BTRFS_FEATURE_INCOMPAT_SAFE_CLEAR;
break;
default:
pr_warn("btrfs: sysfs: unknown feature set %d\n",
fa->feature_set);
return 0;
}
if (set & fa->feature_bit)
val |= 1;
if (clear & fa->feature_bit)
val |= 2;
return val;
}
static ssize_t btrfs_feature_attr_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
int val = 0;
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
struct btrfs_feature_attr *fa = to_btrfs_feature_attr(a);
if (fs_info) {
u64 features = get_features(fs_info, fa->feature_set);
if (features & fa->feature_bit)
val = 1;
} else
val = can_modify_feature(fa);
return sysfs_emit(buf, "%d\n", val);
}
static ssize_t btrfs_feature_attr_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t count)
{
struct btrfs_fs_info *fs_info;
struct btrfs_feature_attr *fa = to_btrfs_feature_attr(a);
u64 features, set, clear;
unsigned long val;
int ret;
fs_info = to_fs_info(kobj);
if (!fs_info)
return -EPERM;
if (sb_rdonly(fs_info->sb))
return -EROFS;
ret = kstrtoul(skip_spaces(buf), 0, &val);
if (ret)
return ret;
if (fa->feature_set == FEAT_COMPAT) {
set = BTRFS_FEATURE_COMPAT_SAFE_SET;
clear = BTRFS_FEATURE_COMPAT_SAFE_CLEAR;
} else if (fa->feature_set == FEAT_COMPAT_RO) {
set = BTRFS_FEATURE_COMPAT_RO_SAFE_SET;
clear = BTRFS_FEATURE_COMPAT_RO_SAFE_CLEAR;
} else {
set = BTRFS_FEATURE_INCOMPAT_SAFE_SET;
clear = BTRFS_FEATURE_INCOMPAT_SAFE_CLEAR;
}
features = get_features(fs_info, fa->feature_set);
/* Nothing to do */
if ((val && (features & fa->feature_bit)) ||
(!val && !(features & fa->feature_bit)))
return count;
if ((val && !(set & fa->feature_bit)) ||
(!val && !(clear & fa->feature_bit))) {
btrfs_info(fs_info,
"%sabling feature %s on mounted fs is not supported.",
val ? "En" : "Dis", fa->kobj_attr.attr.name);
return -EPERM;
}
btrfs_info(fs_info, "%s %s feature flag",
val ? "Setting" : "Clearing", fa->kobj_attr.attr.name);
spin_lock(&fs_info->super_lock);
features = get_features(fs_info, fa->feature_set);
if (val)
features |= fa->feature_bit;
else
features &= ~fa->feature_bit;
set_features(fs_info, fa->feature_set, features);
spin_unlock(&fs_info->super_lock);
/*
* We don't want to do full transaction commit from inside sysfs
*/
set_bit(BTRFS_FS_NEED_TRANS_COMMIT, &fs_info->flags);
wake_up_process(fs_info->transaction_kthread);
return count;
}
static umode_t btrfs_feature_visible(struct kobject *kobj,
struct attribute *attr, int unused)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
umode_t mode = attr->mode;
if (fs_info) {
struct btrfs_feature_attr *fa;
u64 features;
fa = attr_to_btrfs_feature_attr(attr);
features = get_features(fs_info, fa->feature_set);
if (can_modify_feature(fa))
mode |= S_IWUSR;
else if (!(features & fa->feature_bit))
mode = 0;
}
return mode;
}
BTRFS_FEAT_ATTR_INCOMPAT(default_subvol, DEFAULT_SUBVOL);
BTRFS_FEAT_ATTR_INCOMPAT(mixed_groups, MIXED_GROUPS);
BTRFS_FEAT_ATTR_INCOMPAT(compress_lzo, COMPRESS_LZO);
btrfs: Add zstd support Add zstd compression and decompression support to BtrFS. zstd at its fastest level compresses almost as well as zlib, while offering much faster compression and decompression, approaching lzo speeds. I benchmarked btrfs with zstd compression against no compression, lzo compression, and zlib compression. I benchmarked two scenarios. Copying a set of files to btrfs, and then reading the files. Copying a tarball to btrfs, extracting it to btrfs, and then reading the extracted files. After every operation, I call `sync` and include the sync time. Between every pair of operations I unmount and remount the filesystem to avoid caching. The benchmark files can be found in the upstream zstd source repository under `contrib/linux-kernel/{btrfs-benchmark.sh,btrfs-extract-benchmark.sh}` [1] [2]. I ran the benchmarks on a Ubuntu 14.04 VM with 2 cores and 4 GiB of RAM. The VM is running on a MacBook Pro with a 3.1 GHz Intel Core i7 processor, 16 GB of RAM, and a SSD. The first compression benchmark is copying 10 copies of the unzipped Silesia corpus [3] into a BtrFS filesystem mounted with `-o compress-force=Method`. The decompression benchmark times how long it takes to `tar` all 10 copies into `/dev/null`. The compression ratio is measured by comparing the output of `df` and `du`. See the benchmark file [1] for details. I benchmarked multiple zstd compression levels, although the patch uses zstd level 1. | Method | Ratio | Compression MB/s | Decompression speed | |---------|-------|------------------|---------------------| | None | 0.99 | 504 | 686 | | lzo | 1.66 | 398 | 442 | | zlib | 2.58 | 65 | 241 | | zstd 1 | 2.57 | 260 | 383 | | zstd 3 | 2.71 | 174 | 408 | | zstd 6 | 2.87 | 70 | 398 | | zstd 9 | 2.92 | 43 | 406 | | zstd 12 | 2.93 | 21 | 408 | | zstd 15 | 3.01 | 11 | 354 | The next benchmark first copies `linux-4.11.6.tar` [4] to btrfs. Then it measures the compression ratio, extracts the tar, and deletes the tar. Then it measures the compression ratio again, and `tar`s the extracted files into `/dev/null`. See the benchmark file [2] for details. | Method | Tar Ratio | Extract Ratio | Copy (s) | Extract (s)| Read (s) | |--------|-----------|---------------|----------|------------|----------| | None | 0.97 | 0.78 | 0.981 | 5.501 | 8.807 | | lzo | 2.06 | 1.38 | 1.631 | 8.458 | 8.585 | | zlib | 3.40 | 1.86 | 7.750 | 21.544 | 11.744 | | zstd 1 | 3.57 | 1.85 | 2.579 | 11.479 | 9.389 | [1] https://github.com/facebook/zstd/blob/dev/contrib/linux-kernel/btrfs-benchmark.sh [2] https://github.com/facebook/zstd/blob/dev/contrib/linux-kernel/btrfs-extract-benchmark.sh [3] http://sun.aei.polsl.pl/~sdeor/index.php?page=silesia [4] https://cdn.kernel.org/pub/linux/kernel/v4.x/linux-4.11.6.tar.xz zstd source repository: https://github.com/facebook/zstd Signed-off-by: Nick Terrell <terrelln@fb.com> Signed-off-by: Chris Mason <clm@fb.com>
2017-08-10 02:39:02 +00:00
BTRFS_FEAT_ATTR_INCOMPAT(compress_zstd, COMPRESS_ZSTD);
BTRFS_FEAT_ATTR_INCOMPAT(extended_iref, EXTENDED_IREF);
BTRFS_FEAT_ATTR_INCOMPAT(raid56, RAID56);
BTRFS_FEAT_ATTR_INCOMPAT(skinny_metadata, SKINNY_METADATA);
BTRFS_FEAT_ATTR_INCOMPAT(no_holes, NO_HOLES);
BTRFS_FEAT_ATTR_INCOMPAT(metadata_uuid, METADATA_UUID);
BTRFS_FEAT_ATTR_COMPAT_RO(free_space_tree, FREE_SPACE_TREE);
BTRFS_FEAT_ATTR_COMPAT_RO(block_group_tree, BLOCK_GROUP_TREE);
BTRFS_FEAT_ATTR_INCOMPAT(raid1c34, RAID1C34);
BTRFS_FEAT_ATTR_INCOMPAT(simple_quota, SIMPLE_QUOTA);
#ifdef CONFIG_BLK_DEV_ZONED
BTRFS_FEAT_ATTR_INCOMPAT(zoned, ZONED);
#endif
#ifdef CONFIG_BTRFS_DEBUG
/* Remove once support for extent tree v2 is feature complete */
BTRFS_FEAT_ATTR_INCOMPAT(extent_tree_v2, EXTENT_TREE_V2);
/* Remove once support for raid stripe tree is feature complete. */
BTRFS_FEAT_ATTR_INCOMPAT(raid_stripe_tree, RAID_STRIPE_TREE);
#endif
btrfs: initial fsverity support Add support for fsverity in btrfs. To support the generic interface in fs/verity, we add two new item types in the fs tree for inodes with verity enabled. One stores the per-file verity descriptor and btrfs verity item and the other stores the Merkle tree data itself. Verity checking is done in end_page_read just before a page is marked uptodate. This naturally handles a variety of edge cases like holes, preallocated extents, and inline extents. Some care needs to be taken to not try to verity pages past the end of the file, which are accessed by the generic buffered file reading code under some circumstances like reading to the end of the last page and trying to read again. Direct IO on a verity file falls back to buffered reads. Verity relies on PageChecked for the Merkle tree data itself to avoid re-walking up shared paths in the tree. For this reason, we need to cache the Merkle tree data. Since the file is immutable after verity is turned on, we can cache it at an index past EOF. Use the new inode ro_flags to store verity on the inode item, so that we can enable verity on a file, then rollback to an older kernel and still mount the file system and read the file. Since we can't safely write the file anymore without ruining the invariants of the Merkle tree, we mark a ro_compat flag on the file system when a file has verity enabled. Acked-by: Eric Biggers <ebiggers@google.com> Co-developed-by: Chris Mason <clm@fb.com> Signed-off-by: Chris Mason <clm@fb.com> Signed-off-by: Boris Burkov <boris@bur.io> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2021-06-30 20:01:49 +00:00
#ifdef CONFIG_FS_VERITY
BTRFS_FEAT_ATTR_COMPAT_RO(verity, VERITY);
#endif
/*
* Features which depend on feature bits and may differ between each fs.
*
* /sys/fs/btrfs/features - all available features implemented by this version
* /sys/fs/btrfs/UUID/features - features of the fs which are enabled or
* can be changed on a mounted filesystem.
*/
static struct attribute *btrfs_supported_feature_attrs[] = {
BTRFS_FEAT_ATTR_PTR(default_subvol),
BTRFS_FEAT_ATTR_PTR(mixed_groups),
BTRFS_FEAT_ATTR_PTR(compress_lzo),
btrfs: Add zstd support Add zstd compression and decompression support to BtrFS. zstd at its fastest level compresses almost as well as zlib, while offering much faster compression and decompression, approaching lzo speeds. I benchmarked btrfs with zstd compression against no compression, lzo compression, and zlib compression. I benchmarked two scenarios. Copying a set of files to btrfs, and then reading the files. Copying a tarball to btrfs, extracting it to btrfs, and then reading the extracted files. After every operation, I call `sync` and include the sync time. Between every pair of operations I unmount and remount the filesystem to avoid caching. The benchmark files can be found in the upstream zstd source repository under `contrib/linux-kernel/{btrfs-benchmark.sh,btrfs-extract-benchmark.sh}` [1] [2]. I ran the benchmarks on a Ubuntu 14.04 VM with 2 cores and 4 GiB of RAM. The VM is running on a MacBook Pro with a 3.1 GHz Intel Core i7 processor, 16 GB of RAM, and a SSD. The first compression benchmark is copying 10 copies of the unzipped Silesia corpus [3] into a BtrFS filesystem mounted with `-o compress-force=Method`. The decompression benchmark times how long it takes to `tar` all 10 copies into `/dev/null`. The compression ratio is measured by comparing the output of `df` and `du`. See the benchmark file [1] for details. I benchmarked multiple zstd compression levels, although the patch uses zstd level 1. | Method | Ratio | Compression MB/s | Decompression speed | |---------|-------|------------------|---------------------| | None | 0.99 | 504 | 686 | | lzo | 1.66 | 398 | 442 | | zlib | 2.58 | 65 | 241 | | zstd 1 | 2.57 | 260 | 383 | | zstd 3 | 2.71 | 174 | 408 | | zstd 6 | 2.87 | 70 | 398 | | zstd 9 | 2.92 | 43 | 406 | | zstd 12 | 2.93 | 21 | 408 | | zstd 15 | 3.01 | 11 | 354 | The next benchmark first copies `linux-4.11.6.tar` [4] to btrfs. Then it measures the compression ratio, extracts the tar, and deletes the tar. Then it measures the compression ratio again, and `tar`s the extracted files into `/dev/null`. See the benchmark file [2] for details. | Method | Tar Ratio | Extract Ratio | Copy (s) | Extract (s)| Read (s) | |--------|-----------|---------------|----------|------------|----------| | None | 0.97 | 0.78 | 0.981 | 5.501 | 8.807 | | lzo | 2.06 | 1.38 | 1.631 | 8.458 | 8.585 | | zlib | 3.40 | 1.86 | 7.750 | 21.544 | 11.744 | | zstd 1 | 3.57 | 1.85 | 2.579 | 11.479 | 9.389 | [1] https://github.com/facebook/zstd/blob/dev/contrib/linux-kernel/btrfs-benchmark.sh [2] https://github.com/facebook/zstd/blob/dev/contrib/linux-kernel/btrfs-extract-benchmark.sh [3] http://sun.aei.polsl.pl/~sdeor/index.php?page=silesia [4] https://cdn.kernel.org/pub/linux/kernel/v4.x/linux-4.11.6.tar.xz zstd source repository: https://github.com/facebook/zstd Signed-off-by: Nick Terrell <terrelln@fb.com> Signed-off-by: Chris Mason <clm@fb.com>
2017-08-10 02:39:02 +00:00
BTRFS_FEAT_ATTR_PTR(compress_zstd),
BTRFS_FEAT_ATTR_PTR(extended_iref),
BTRFS_FEAT_ATTR_PTR(raid56),
BTRFS_FEAT_ATTR_PTR(skinny_metadata),
BTRFS_FEAT_ATTR_PTR(no_holes),
BTRFS_FEAT_ATTR_PTR(metadata_uuid),
BTRFS_FEAT_ATTR_PTR(free_space_tree),
BTRFS_FEAT_ATTR_PTR(raid1c34),
BTRFS_FEAT_ATTR_PTR(block_group_tree),
BTRFS_FEAT_ATTR_PTR(simple_quota),
#ifdef CONFIG_BLK_DEV_ZONED
BTRFS_FEAT_ATTR_PTR(zoned),
#endif
#ifdef CONFIG_BTRFS_DEBUG
BTRFS_FEAT_ATTR_PTR(extent_tree_v2),
BTRFS_FEAT_ATTR_PTR(raid_stripe_tree),
btrfs: initial fsverity support Add support for fsverity in btrfs. To support the generic interface in fs/verity, we add two new item types in the fs tree for inodes with verity enabled. One stores the per-file verity descriptor and btrfs verity item and the other stores the Merkle tree data itself. Verity checking is done in end_page_read just before a page is marked uptodate. This naturally handles a variety of edge cases like holes, preallocated extents, and inline extents. Some care needs to be taken to not try to verity pages past the end of the file, which are accessed by the generic buffered file reading code under some circumstances like reading to the end of the last page and trying to read again. Direct IO on a verity file falls back to buffered reads. Verity relies on PageChecked for the Merkle tree data itself to avoid re-walking up shared paths in the tree. For this reason, we need to cache the Merkle tree data. Since the file is immutable after verity is turned on, we can cache it at an index past EOF. Use the new inode ro_flags to store verity on the inode item, so that we can enable verity on a file, then rollback to an older kernel and still mount the file system and read the file. Since we can't safely write the file anymore without ruining the invariants of the Merkle tree, we mark a ro_compat flag on the file system when a file has verity enabled. Acked-by: Eric Biggers <ebiggers@google.com> Co-developed-by: Chris Mason <clm@fb.com> Signed-off-by: Chris Mason <clm@fb.com> Signed-off-by: Boris Burkov <boris@bur.io> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2021-06-30 20:01:49 +00:00
#endif
#ifdef CONFIG_FS_VERITY
BTRFS_FEAT_ATTR_PTR(verity),
#endif
NULL
};
static const struct attribute_group btrfs_feature_attr_group = {
.name = "features",
.is_visible = btrfs_feature_visible,
.attrs = btrfs_supported_feature_attrs,
};
static ssize_t rmdir_subvol_show(struct kobject *kobj,
struct kobj_attribute *ka, char *buf)
{
return sysfs_emit(buf, "0\n");
}
BTRFS_ATTR(static_feature, rmdir_subvol, rmdir_subvol_show);
static ssize_t supported_checksums_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
ssize_t ret = 0;
int i;
for (i = 0; i < btrfs_get_num_csums(); i++) {
/*
* This "trick" only works as long as 'enum btrfs_csum_type' has
* no holes in it
*/
ret += sysfs_emit_at(buf, ret, "%s%s", (i == 0 ? "" : " "),
btrfs_super_csum_name(i));
}
ret += sysfs_emit_at(buf, ret, "\n");
return ret;
}
BTRFS_ATTR(static_feature, supported_checksums, supported_checksums_show);
static ssize_t send_stream_version_show(struct kobject *kobj,
struct kobj_attribute *ka, char *buf)
{
return sysfs_emit(buf, "%d\n", BTRFS_SEND_STREAM_VERSION);
}
BTRFS_ATTR(static_feature, send_stream_version, send_stream_version_show);
static const char *rescue_opts[] = {
"usebackuproot",
"nologreplay",
"ignorebadroots",
"ignoredatacsums",
"all",
};
static ssize_t supported_rescue_options_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
ssize_t ret = 0;
int i;
for (i = 0; i < ARRAY_SIZE(rescue_opts); i++)
ret += sysfs_emit_at(buf, ret, "%s%s", (i ? " " : ""), rescue_opts[i]);
ret += sysfs_emit_at(buf, ret, "\n");
return ret;
}
BTRFS_ATTR(static_feature, supported_rescue_options,
supported_rescue_options_show);
static ssize_t supported_sectorsizes_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
ssize_t ret = 0;
/* An artificial limit to only support 4K and PAGE_SIZE */
if (PAGE_SIZE > SZ_4K)
ret += sysfs_emit_at(buf, ret, "%u ", SZ_4K);
ret += sysfs_emit_at(buf, ret, "%lu\n", PAGE_SIZE);
return ret;
}
BTRFS_ATTR(static_feature, supported_sectorsizes,
supported_sectorsizes_show);
static ssize_t acl_show(struct kobject *kobj, struct kobj_attribute *a, char *buf)
{
return sysfs_emit(buf, "%d\n", IS_ENABLED(CONFIG_BTRFS_FS_POSIX_ACL));
}
BTRFS_ATTR(static_feature, acl, acl_show);
static ssize_t temp_fsid_supported_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
return sysfs_emit(buf, "0\n");
}
BTRFS_ATTR(static_feature, temp_fsid, temp_fsid_supported_show);
/*
* Features which only depend on kernel version.
*
* These are listed in /sys/fs/btrfs/features along with
* btrfs_supported_feature_attrs.
*/
static struct attribute *btrfs_supported_static_feature_attrs[] = {
BTRFS_ATTR_PTR(static_feature, acl),
BTRFS_ATTR_PTR(static_feature, rmdir_subvol),
BTRFS_ATTR_PTR(static_feature, supported_checksums),
BTRFS_ATTR_PTR(static_feature, send_stream_version),
BTRFS_ATTR_PTR(static_feature, supported_rescue_options),
BTRFS_ATTR_PTR(static_feature, supported_sectorsizes),
BTRFS_ATTR_PTR(static_feature, temp_fsid),
NULL
};
static const struct attribute_group btrfs_static_feature_attr_group = {
.name = "features",
.attrs = btrfs_supported_static_feature_attrs,
};
/*
* Discard statistics and tunables
*/
#define discard_to_fs_info(_kobj) to_fs_info(get_btrfs_kobj(_kobj))
static ssize_t btrfs_discardable_bytes_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
return sysfs_emit(buf, "%lld\n",
atomic64_read(&fs_info->discard_ctl.discardable_bytes));
}
BTRFS_ATTR(discard, discardable_bytes, btrfs_discardable_bytes_show);
static ssize_t btrfs_discardable_extents_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
return sysfs_emit(buf, "%d\n",
atomic_read(&fs_info->discard_ctl.discardable_extents));
}
BTRFS_ATTR(discard, discardable_extents, btrfs_discardable_extents_show);
static ssize_t btrfs_discard_bitmap_bytes_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
return sysfs_emit(buf, "%llu\n",
fs_info->discard_ctl.discard_bitmap_bytes);
}
BTRFS_ATTR(discard, discard_bitmap_bytes, btrfs_discard_bitmap_bytes_show);
static ssize_t btrfs_discard_bytes_saved_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
return sysfs_emit(buf, "%lld\n",
atomic64_read(&fs_info->discard_ctl.discard_bytes_saved));
}
BTRFS_ATTR(discard, discard_bytes_saved, btrfs_discard_bytes_saved_show);
static ssize_t btrfs_discard_extent_bytes_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
return sysfs_emit(buf, "%llu\n",
fs_info->discard_ctl.discard_extent_bytes);
}
BTRFS_ATTR(discard, discard_extent_bytes, btrfs_discard_extent_bytes_show);
static ssize_t btrfs_discard_iops_limit_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
return sysfs_emit(buf, "%u\n",
READ_ONCE(fs_info->discard_ctl.iops_limit));
}
static ssize_t btrfs_discard_iops_limit_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
struct btrfs_discard_ctl *discard_ctl = &fs_info->discard_ctl;
u32 iops_limit;
int ret;
ret = kstrtou32(buf, 10, &iops_limit);
if (ret)
return -EINVAL;
WRITE_ONCE(discard_ctl->iops_limit, iops_limit);
btrfs_discard_calc_delay(discard_ctl);
btrfs_discard_schedule_work(discard_ctl, true);
return len;
}
BTRFS_ATTR_RW(discard, iops_limit, btrfs_discard_iops_limit_show,
btrfs_discard_iops_limit_store);
static ssize_t btrfs_discard_kbps_limit_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
return sysfs_emit(buf, "%u\n",
READ_ONCE(fs_info->discard_ctl.kbps_limit));
}
static ssize_t btrfs_discard_kbps_limit_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
struct btrfs_discard_ctl *discard_ctl = &fs_info->discard_ctl;
u32 kbps_limit;
int ret;
ret = kstrtou32(buf, 10, &kbps_limit);
if (ret)
return -EINVAL;
WRITE_ONCE(discard_ctl->kbps_limit, kbps_limit);
btrfs_discard_schedule_work(discard_ctl, true);
return len;
}
BTRFS_ATTR_RW(discard, kbps_limit, btrfs_discard_kbps_limit_show,
btrfs_discard_kbps_limit_store);
static ssize_t btrfs_discard_max_discard_size_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
return sysfs_emit(buf, "%llu\n",
READ_ONCE(fs_info->discard_ctl.max_discard_size));
}
static ssize_t btrfs_discard_max_discard_size_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_fs_info *fs_info = discard_to_fs_info(kobj);
struct btrfs_discard_ctl *discard_ctl = &fs_info->discard_ctl;
u64 max_discard_size;
int ret;
ret = kstrtou64(buf, 10, &max_discard_size);
if (ret)
return -EINVAL;
WRITE_ONCE(discard_ctl->max_discard_size, max_discard_size);
return len;
}
BTRFS_ATTR_RW(discard, max_discard_size, btrfs_discard_max_discard_size_show,
btrfs_discard_max_discard_size_store);
/*
* Per-filesystem stats for discard (when mounted with discard=async).
*
* Path: /sys/fs/btrfs/<uuid>/discard/
*/
static const struct attribute *discard_attrs[] = {
BTRFS_ATTR_PTR(discard, discardable_bytes),
BTRFS_ATTR_PTR(discard, discardable_extents),
BTRFS_ATTR_PTR(discard, discard_bitmap_bytes),
BTRFS_ATTR_PTR(discard, discard_bytes_saved),
BTRFS_ATTR_PTR(discard, discard_extent_bytes),
BTRFS_ATTR_PTR(discard, iops_limit),
BTRFS_ATTR_PTR(discard, kbps_limit),
BTRFS_ATTR_PTR(discard, max_discard_size),
NULL,
};
#ifdef CONFIG_BTRFS_DEBUG
/*
* Per-filesystem runtime debugging exported via sysfs.
*
* Path: /sys/fs/btrfs/UUID/debug/
*/
static const struct attribute *btrfs_debug_mount_attrs[] = {
NULL,
};
/*
* Runtime debugging exported via sysfs, applies to all mounted filesystems.
*
* Path: /sys/fs/btrfs/debug
*/
static struct attribute *btrfs_debug_feature_attrs[] = {
NULL
};
static const struct attribute_group btrfs_debug_feature_attr_group = {
.name = "debug",
.attrs = btrfs_debug_feature_attrs,
};
#endif
static ssize_t btrfs_show_u64(u64 *value_ptr, spinlock_t *lock, char *buf)
{
u64 val;
if (lock)
spin_lock(lock);
val = *value_ptr;
if (lock)
spin_unlock(lock);
return sysfs_emit(buf, "%llu\n", val);
}
static ssize_t global_rsv_size_show(struct kobject *kobj,
struct kobj_attribute *ka, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj->parent);
struct btrfs_block_rsv *block_rsv = &fs_info->global_block_rsv;
return btrfs_show_u64(&block_rsv->size, &block_rsv->lock, buf);
}
BTRFS_ATTR(allocation, global_rsv_size, global_rsv_size_show);
static ssize_t global_rsv_reserved_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj->parent);
struct btrfs_block_rsv *block_rsv = &fs_info->global_block_rsv;
return btrfs_show_u64(&block_rsv->reserved, &block_rsv->lock, buf);
}
BTRFS_ATTR(allocation, global_rsv_reserved, global_rsv_reserved_show);
#define to_space_info(_kobj) container_of(_kobj, struct btrfs_space_info, kobj)
#define to_raid_kobj(_kobj) container_of(_kobj, struct raid_kobject, kobj)
static ssize_t raid_bytes_show(struct kobject *kobj,
struct kobj_attribute *attr, char *buf);
BTRFS_ATTR(raid, total_bytes, raid_bytes_show);
BTRFS_ATTR(raid, used_bytes, raid_bytes_show);
static ssize_t raid_bytes_show(struct kobject *kobj,
struct kobj_attribute *attr, char *buf)
{
struct btrfs_space_info *sinfo = to_space_info(kobj->parent);
struct btrfs_block_group *block_group;
int index = btrfs_bg_flags_to_raid_index(to_raid_kobj(kobj)->flags);
u64 val = 0;
down_read(&sinfo->groups_sem);
list_for_each_entry(block_group, &sinfo->block_groups[index], list) {
if (&attr->attr == BTRFS_ATTR_PTR(raid, total_bytes))
val += block_group->length;
else
val += block_group->used;
}
up_read(&sinfo->groups_sem);
return sysfs_emit(buf, "%llu\n", val);
}
/*
* Allocation information about block group profiles.
*
* Path: /sys/fs/btrfs/<uuid>/allocation/<bg-type>/<bg-profile>/
*/
static struct attribute *raid_attrs[] = {
BTRFS_ATTR_PTR(raid, total_bytes),
BTRFS_ATTR_PTR(raid, used_bytes),
NULL
};
ATTRIBUTE_GROUPS(raid);
static void release_raid_kobj(struct kobject *kobj)
{
kfree(to_raid_kobj(kobj));
}
static const struct kobj_type btrfs_raid_ktype = {
.sysfs_ops = &kobj_sysfs_ops,
.release = release_raid_kobj,
.default_groups = raid_groups,
};
#define SPACE_INFO_ATTR(field) \
static ssize_t btrfs_space_info_show_##field(struct kobject *kobj, \
struct kobj_attribute *a, \
char *buf) \
{ \
struct btrfs_space_info *sinfo = to_space_info(kobj); \
return btrfs_show_u64(&sinfo->field, &sinfo->lock, buf); \
} \
BTRFS_ATTR(space_info, field, btrfs_space_info_show_##field)
static ssize_t btrfs_chunk_size_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_space_info *sinfo = to_space_info(kobj);
return sysfs_emit(buf, "%llu\n", READ_ONCE(sinfo->chunk_size));
}
/*
* Store new chunk size in space info. Can be called on a read-only filesystem.
*
* If the new chunk size value is larger than 10% of free space it is reduced
* to match that limit. Alignment must be to 256M and the system chunk size
* cannot be set.
*/
static ssize_t btrfs_chunk_size_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_space_info *space_info = to_space_info(kobj);
struct btrfs_fs_info *fs_info = to_fs_info(get_btrfs_kobj(kobj));
char *retptr;
u64 val;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
if (!fs_info->fs_devices)
return -EINVAL;
if (btrfs_is_zoned(fs_info))
return -EINVAL;
/* System block type must not be changed. */
if (space_info->flags & BTRFS_BLOCK_GROUP_SYSTEM)
return -EPERM;
val = memparse(buf, &retptr);
/* There could be trailing '\n', also catch any typos after the value */
retptr = skip_spaces(retptr);
if (*retptr != 0 || val == 0)
return -EINVAL;
val = min(val, BTRFS_MAX_DATA_CHUNK_SIZE);
/* Limit stripe size to 10% of available space. */
val = min(mult_perc(fs_info->fs_devices->total_rw_bytes, 10), val);
/* Must be multiple of 256M. */
val &= ~((u64)SZ_256M - 1);
/* Must be at least 256M. */
if (val < SZ_256M)
return -EINVAL;
btrfs_update_space_info_chunk_size(space_info, val);
return len;
}
static ssize_t btrfs_size_classes_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_space_info *sinfo = to_space_info(kobj);
struct btrfs_block_group *bg;
u32 none = 0;
u32 small = 0;
u32 medium = 0;
u32 large = 0;
for (int i = 0; i < BTRFS_NR_RAID_TYPES; ++i) {
down_read(&sinfo->groups_sem);
list_for_each_entry(bg, &sinfo->block_groups[i], list) {
if (!btrfs_block_group_should_use_size_class(bg))
continue;
switch (bg->size_class) {
case BTRFS_BG_SZ_NONE:
none++;
break;
case BTRFS_BG_SZ_SMALL:
small++;
break;
case BTRFS_BG_SZ_MEDIUM:
medium++;
break;
case BTRFS_BG_SZ_LARGE:
large++;
break;
}
}
up_read(&sinfo->groups_sem);
}
return sysfs_emit(buf, "none %u\n"
"small %u\n"
"medium %u\n"
"large %u\n",
none, small, medium, large);
}
#ifdef CONFIG_BTRFS_DEBUG
/*
* Request chunk allocation with current chunk size.
*/
static ssize_t btrfs_force_chunk_alloc_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_space_info *space_info = to_space_info(kobj);
struct btrfs_fs_info *fs_info = to_fs_info(get_btrfs_kobj(kobj));
struct btrfs_trans_handle *trans;
bool val;
int ret;
if (!capable(CAP_SYS_ADMIN))
return -EPERM;
if (sb_rdonly(fs_info->sb))
return -EROFS;
ret = kstrtobool(buf, &val);
if (ret)
return ret;
if (!val)
return -EINVAL;
/*
* This is unsafe to be called from sysfs context and may cause
* unexpected problems.
*/
trans = btrfs_start_transaction(fs_info->tree_root, 0);
if (IS_ERR(trans))
return PTR_ERR(trans);
ret = btrfs_force_chunk_alloc(trans, space_info->flags);
btrfs_end_transaction(trans);
if (ret == 1)
return len;
return -ENOSPC;
}
BTRFS_ATTR_W(space_info, force_chunk_alloc, btrfs_force_chunk_alloc_store);
#endif
SPACE_INFO_ATTR(flags);
SPACE_INFO_ATTR(total_bytes);
SPACE_INFO_ATTR(bytes_used);
SPACE_INFO_ATTR(bytes_pinned);
SPACE_INFO_ATTR(bytes_reserved);
SPACE_INFO_ATTR(bytes_may_use);
SPACE_INFO_ATTR(bytes_readonly);
SPACE_INFO_ATTR(bytes_zone_unusable);
SPACE_INFO_ATTR(disk_used);
SPACE_INFO_ATTR(disk_total);
BTRFS_ATTR_RW(space_info, chunk_size, btrfs_chunk_size_show, btrfs_chunk_size_store);
BTRFS_ATTR(space_info, size_classes, btrfs_size_classes_show);
static ssize_t btrfs_sinfo_bg_reclaim_threshold_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_space_info *space_info = to_space_info(kobj);
return sysfs_emit(buf, "%d\n", READ_ONCE(space_info->bg_reclaim_threshold));
}
static ssize_t btrfs_sinfo_bg_reclaim_threshold_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_space_info *space_info = to_space_info(kobj);
int thresh;
int ret;
ret = kstrtoint(buf, 10, &thresh);
if (ret)
return ret;
if (thresh < 0 || thresh > 100)
return -EINVAL;
WRITE_ONCE(space_info->bg_reclaim_threshold, thresh);
return len;
}
BTRFS_ATTR_RW(space_info, bg_reclaim_threshold,
btrfs_sinfo_bg_reclaim_threshold_show,
btrfs_sinfo_bg_reclaim_threshold_store);
/*
* Allocation information about block group types.
*
* Path: /sys/fs/btrfs/<uuid>/allocation/<bg-type>/
*/
static struct attribute *space_info_attrs[] = {
BTRFS_ATTR_PTR(space_info, flags),
BTRFS_ATTR_PTR(space_info, total_bytes),
BTRFS_ATTR_PTR(space_info, bytes_used),
BTRFS_ATTR_PTR(space_info, bytes_pinned),
BTRFS_ATTR_PTR(space_info, bytes_reserved),
BTRFS_ATTR_PTR(space_info, bytes_may_use),
BTRFS_ATTR_PTR(space_info, bytes_readonly),
BTRFS_ATTR_PTR(space_info, bytes_zone_unusable),
BTRFS_ATTR_PTR(space_info, disk_used),
BTRFS_ATTR_PTR(space_info, disk_total),
BTRFS_ATTR_PTR(space_info, bg_reclaim_threshold),
BTRFS_ATTR_PTR(space_info, chunk_size),
BTRFS_ATTR_PTR(space_info, size_classes),
#ifdef CONFIG_BTRFS_DEBUG
BTRFS_ATTR_PTR(space_info, force_chunk_alloc),
#endif
NULL,
};
ATTRIBUTE_GROUPS(space_info);
static void space_info_release(struct kobject *kobj)
{
struct btrfs_space_info *sinfo = to_space_info(kobj);
kfree(sinfo);
}
static const struct kobj_type space_info_ktype = {
.sysfs_ops = &kobj_sysfs_ops,
.release = space_info_release,
.default_groups = space_info_groups,
};
/*
* Allocation information about block groups.
*
* Path: /sys/fs/btrfs/<uuid>/allocation/
*/
static const struct attribute *allocation_attrs[] = {
BTRFS_ATTR_PTR(allocation, global_rsv_reserved),
BTRFS_ATTR_PTR(allocation, global_rsv_size),
NULL,
};
static ssize_t btrfs_label_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
char *label = fs_info->super_copy->label;
ssize_t ret;
spin_lock(&fs_info->super_lock);
ret = sysfs_emit(buf, label[0] ? "%s\n" : "%s", label);
spin_unlock(&fs_info->super_lock);
return ret;
}
static ssize_t btrfs_label_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
size_t p_len;
if (!fs_info)
return -EPERM;
if (sb_rdonly(fs_info->sb))
return -EROFS;
/*
* p_len is the len until the first occurrence of either
* '\n' or '\0'
*/
p_len = strcspn(buf, "\n");
if (p_len >= BTRFS_LABEL_SIZE)
return -EINVAL;
spin_lock(&fs_info->super_lock);
memset(fs_info->super_copy->label, 0, BTRFS_LABEL_SIZE);
memcpy(fs_info->super_copy->label, buf, p_len);
spin_unlock(&fs_info->super_lock);
/*
* We don't want to do full transaction commit from inside sysfs
*/
set_bit(BTRFS_FS_NEED_TRANS_COMMIT, &fs_info->flags);
wake_up_process(fs_info->transaction_kthread);
return len;
}
BTRFS_ATTR_RW(, label, btrfs_label_show, btrfs_label_store);
static ssize_t btrfs_nodesize_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
return sysfs_emit(buf, "%u\n", fs_info->super_copy->nodesize);
}
BTRFS_ATTR(, nodesize, btrfs_nodesize_show);
static ssize_t btrfs_sectorsize_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
return sysfs_emit(buf, "%u\n", fs_info->super_copy->sectorsize);
}
BTRFS_ATTR(, sectorsize, btrfs_sectorsize_show);
static ssize_t btrfs_commit_stats_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
return sysfs_emit(buf,
"commits %llu\n"
"last_commit_ms %llu\n"
"max_commit_ms %llu\n"
"total_commit_ms %llu\n",
fs_info->commit_stats.commit_count,
div_u64(fs_info->commit_stats.last_commit_dur, NSEC_PER_MSEC),
div_u64(fs_info->commit_stats.max_commit_dur, NSEC_PER_MSEC),
div_u64(fs_info->commit_stats.total_commit_dur, NSEC_PER_MSEC));
}
static ssize_t btrfs_commit_stats_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
unsigned long val;
int ret;
if (!fs_info)
return -EPERM;
if (!capable(CAP_SYS_RESOURCE))
return -EPERM;
ret = kstrtoul(buf, 10, &val);
if (ret)
return ret;
if (val)
return -EINVAL;
WRITE_ONCE(fs_info->commit_stats.max_commit_dur, 0);
return len;
}
BTRFS_ATTR_RW(, commit_stats, btrfs_commit_stats_show, btrfs_commit_stats_store);
static ssize_t btrfs_clone_alignment_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
return sysfs_emit(buf, "%u\n", fs_info->super_copy->sectorsize);
}
BTRFS_ATTR(, clone_alignment, btrfs_clone_alignment_show);
static ssize_t quota_override_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
int quota_override;
quota_override = test_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags);
return sysfs_emit(buf, "%d\n", quota_override);
}
static ssize_t quota_override_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
unsigned long knob;
int err;
if (!fs_info)
return -EPERM;
if (!capable(CAP_SYS_RESOURCE))
return -EPERM;
err = kstrtoul(buf, 10, &knob);
if (err)
return err;
if (knob > 1)
return -EINVAL;
if (knob)
set_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags);
else
clear_bit(BTRFS_FS_QUOTA_OVERRIDE, &fs_info->flags);
return len;
}
BTRFS_ATTR_RW(, quota_override, quota_override_show, quota_override_store);
static ssize_t btrfs_metadata_uuid_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
return sysfs_emit(buf, "%pU\n", fs_info->fs_devices->metadata_uuid);
}
BTRFS_ATTR(, metadata_uuid, btrfs_metadata_uuid_show);
static ssize_t btrfs_checksum_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
u16 csum_type = btrfs_super_csum_type(fs_info->super_copy);
return sysfs_emit(buf, "%s (%s)\n",
btrfs_super_csum_name(csum_type),
crypto_shash_driver_name(fs_info->csum_shash));
}
BTRFS_ATTR(, checksum, btrfs_checksum_show);
static ssize_t btrfs_exclusive_operation_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
const char *str;
switch (READ_ONCE(fs_info->exclusive_operation)) {
case BTRFS_EXCLOP_NONE:
str = "none\n";
break;
case BTRFS_EXCLOP_BALANCE:
str = "balance\n";
break;
case BTRFS_EXCLOP_BALANCE_PAUSED:
str = "balance paused\n";
break;
case BTRFS_EXCLOP_DEV_ADD:
str = "device add\n";
break;
case BTRFS_EXCLOP_DEV_REMOVE:
str = "device remove\n";
break;
case BTRFS_EXCLOP_DEV_REPLACE:
str = "device replace\n";
break;
case BTRFS_EXCLOP_RESIZE:
str = "resize\n";
break;
case BTRFS_EXCLOP_SWAP_ACTIVATE:
str = "swap activate\n";
break;
default:
str = "UNKNOWN\n";
break;
}
return sysfs_emit(buf, "%s", str);
}
BTRFS_ATTR(, exclusive_operation, btrfs_exclusive_operation_show);
static ssize_t btrfs_generation_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
btrfs: add and use helpers for reading and writing fs_info->generation Currently the generation field of struct btrfs_fs_info is always modified while holding fs_info->trans_lock locked. Most readers will access this field without taking that lock but while holding a transaction handle, which is safe to do due to the transaction life cycle. However there are other readers that are neither holding the lock nor holding a transaction handle open: 1) When reading an inode from disk, at btrfs_read_locked_inode(); 2) When reading the generation to expose it to sysfs, at btrfs_generation_show(); 3) Early in the fsync path, at skip_inode_logging(); 4) When creating a hole at btrfs_cont_expand(), during write paths, truncate and reflinking; 5) In the fs_info ioctl (btrfs_ioctl_fs_info()); 6) While mounting the filesystem, in the open_ctree() path. In these cases it's safe to directly read fs_info->generation as no one can concurrently start a transaction and update fs_info->generation. In case of the fsync path, races here should be harmless, and in the worst case they may cause a fsync to log an inode when it's not really needed, so nothing bad from a functional perspective. In the other cases it's not so clear if functional problems may arise, though in case 1 rare things like a load/store tearing [1] may cause the BTRFS_INODE_NEEDS_FULL_SYNC flag not being set on an inode and therefore result in incorrect logging later on in case a fsync call is made. To avoid data race warnings from tools like KCSAN and other issues such as load and store tearing (amongst others, see [1]), create helpers to access the generation field of struct btrfs_fs_info using READ_ONCE() and WRITE_ONCE(), and use these helpers where needed. [1] https://lwn.net/Articles/793253/ Signed-off-by: Filipe Manana <fdmanana@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-10-04 10:38:50 +00:00
return sysfs_emit(buf, "%llu\n", btrfs_get_fs_generation(fs_info));
}
BTRFS_ATTR(, generation, btrfs_generation_show);
static ssize_t btrfs_temp_fsid_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
return sysfs_emit(buf, "%d\n", fs_info->fs_devices->temp_fsid);
}
BTRFS_ATTR(, temp_fsid, btrfs_temp_fsid_show);
static const char * const btrfs_read_policy_name[] = { "pid" };
static ssize_t btrfs_read_policy_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_devices *fs_devices = to_fs_devs(kobj);
const enum btrfs_read_policy policy = READ_ONCE(fs_devices->read_policy);
ssize_t ret = 0;
int i;
for (i = 0; i < BTRFS_NR_READ_POLICY; i++) {
if (policy == i)
ret += sysfs_emit_at(buf, ret, "%s[%s]",
(ret == 0 ? "" : " "),
btrfs_read_policy_name[i]);
else
ret += sysfs_emit_at(buf, ret, "%s%s",
(ret == 0 ? "" : " "),
btrfs_read_policy_name[i]);
}
ret += sysfs_emit_at(buf, ret, "\n");
return ret;
}
static ssize_t btrfs_read_policy_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_fs_devices *fs_devices = to_fs_devs(kobj);
int i;
for (i = 0; i < BTRFS_NR_READ_POLICY; i++) {
if (sysfs_streq(buf, btrfs_read_policy_name[i])) {
if (i != READ_ONCE(fs_devices->read_policy)) {
WRITE_ONCE(fs_devices->read_policy, i);
btrfs_info(fs_devices->fs_info,
"read policy set to '%s'",
btrfs_read_policy_name[i]);
}
return len;
}
}
return -EINVAL;
}
BTRFS_ATTR_RW(, read_policy, btrfs_read_policy_show, btrfs_read_policy_store);
static ssize_t btrfs_bg_reclaim_threshold_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
return sysfs_emit(buf, "%d\n", READ_ONCE(fs_info->bg_reclaim_threshold));
}
static ssize_t btrfs_bg_reclaim_threshold_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_fs_info *fs_info = to_fs_info(kobj);
int thresh;
int ret;
ret = kstrtoint(buf, 10, &thresh);
if (ret)
return ret;
#ifdef CONFIG_BTRFS_DEBUG
if (thresh != 0 && (thresh > 100))
return -EINVAL;
#else
if (thresh != 0 && (thresh <= 50 || thresh > 100))
return -EINVAL;
#endif
WRITE_ONCE(fs_info->bg_reclaim_threshold, thresh);
return len;
}
BTRFS_ATTR_RW(, bg_reclaim_threshold, btrfs_bg_reclaim_threshold_show,
btrfs_bg_reclaim_threshold_store);
#ifdef CONFIG_BTRFS_DEBUG
static ssize_t btrfs_offload_csum_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_fs_devices *fs_devices = to_fs_devs(kobj);
switch (READ_ONCE(fs_devices->offload_csum_mode)) {
case BTRFS_OFFLOAD_CSUM_AUTO:
return sysfs_emit(buf, "auto\n");
case BTRFS_OFFLOAD_CSUM_FORCE_ON:
return sysfs_emit(buf, "1\n");
case BTRFS_OFFLOAD_CSUM_FORCE_OFF:
return sysfs_emit(buf, "0\n");
default:
WARN_ON(1);
return -EINVAL;
}
}
static ssize_t btrfs_offload_csum_store(struct kobject *kobj,
struct kobj_attribute *a, const char *buf,
size_t len)
{
struct btrfs_fs_devices *fs_devices = to_fs_devs(kobj);
int ret;
bool val;
ret = kstrtobool(buf, &val);
if (ret == 0)
WRITE_ONCE(fs_devices->offload_csum_mode,
val ? BTRFS_OFFLOAD_CSUM_FORCE_ON : BTRFS_OFFLOAD_CSUM_FORCE_OFF);
else if (ret == -EINVAL && sysfs_streq(buf, "auto"))
WRITE_ONCE(fs_devices->offload_csum_mode, BTRFS_OFFLOAD_CSUM_AUTO);
else
return -EINVAL;
return len;
}
BTRFS_ATTR_RW(, offload_csum, btrfs_offload_csum_show, btrfs_offload_csum_store);
#endif
/*
* Per-filesystem information and stats.
*
* Path: /sys/fs/btrfs/<uuid>/
*/
static const struct attribute *btrfs_attrs[] = {
BTRFS_ATTR_PTR(, label),
BTRFS_ATTR_PTR(, nodesize),
BTRFS_ATTR_PTR(, sectorsize),
BTRFS_ATTR_PTR(, clone_alignment),
BTRFS_ATTR_PTR(, quota_override),
BTRFS_ATTR_PTR(, metadata_uuid),
BTRFS_ATTR_PTR(, checksum),
BTRFS_ATTR_PTR(, exclusive_operation),
BTRFS_ATTR_PTR(, generation),
BTRFS_ATTR_PTR(, read_policy),
BTRFS_ATTR_PTR(, bg_reclaim_threshold),
BTRFS_ATTR_PTR(, commit_stats),
BTRFS_ATTR_PTR(, temp_fsid),
#ifdef CONFIG_BTRFS_DEBUG
BTRFS_ATTR_PTR(, offload_csum),
#endif
NULL,
};
static void btrfs_release_fsid_kobj(struct kobject *kobj)
{
struct btrfs_fs_devices *fs_devs = to_fs_devs(kobj);
memset(&fs_devs->fsid_kobj, 0, sizeof(struct kobject));
complete(&fs_devs->kobj_unregister);
}
static const struct kobj_type btrfs_ktype = {
.sysfs_ops = &kobj_sysfs_ops,
.release = btrfs_release_fsid_kobj,
};
static inline struct btrfs_fs_devices *to_fs_devs(struct kobject *kobj)
{
if (kobj->ktype != &btrfs_ktype)
return NULL;
return container_of(kobj, struct btrfs_fs_devices, fsid_kobj);
}
static inline struct btrfs_fs_info *to_fs_info(struct kobject *kobj)
{
if (kobj->ktype != &btrfs_ktype)
return NULL;
return to_fs_devs(kobj)->fs_info;
}
static struct kobject *get_btrfs_kobj(struct kobject *kobj)
{
while (kobj) {
if (kobj->ktype == &btrfs_ktype)
return kobj;
kobj = kobj->parent;
}
return NULL;
}
#define NUM_FEATURE_BITS 64
#define BTRFS_FEATURE_NAME_MAX 13
static char btrfs_unknown_feature_names[FEAT_MAX][NUM_FEATURE_BITS][BTRFS_FEATURE_NAME_MAX];
static struct btrfs_feature_attr btrfs_feature_attrs[FEAT_MAX][NUM_FEATURE_BITS];
static_assert(ARRAY_SIZE(btrfs_unknown_feature_names) ==
ARRAY_SIZE(btrfs_feature_attrs));
static_assert(ARRAY_SIZE(btrfs_unknown_feature_names[0]) ==
ARRAY_SIZE(btrfs_feature_attrs[0]));
static const u64 supported_feature_masks[FEAT_MAX] = {
[FEAT_COMPAT] = BTRFS_FEATURE_COMPAT_SUPP,
[FEAT_COMPAT_RO] = BTRFS_FEATURE_COMPAT_RO_SUPP,
[FEAT_INCOMPAT] = BTRFS_FEATURE_INCOMPAT_SUPP,
};
static int addrm_unknown_feature_attrs(struct btrfs_fs_info *fs_info, bool add)
{
int set;
for (set = 0; set < FEAT_MAX; set++) {
int i;
struct attribute *attrs[2];
struct attribute_group agroup = {
.name = "features",
.attrs = attrs,
};
u64 features = get_features(fs_info, set);
features &= ~supported_feature_masks[set];
if (!features)
continue;
attrs[1] = NULL;
for (i = 0; i < NUM_FEATURE_BITS; i++) {
struct btrfs_feature_attr *fa;
if (!(features & (1ULL << i)))
continue;
fa = &btrfs_feature_attrs[set][i];
attrs[0] = &fa->kobj_attr.attr;
if (add) {
int ret;
ret = sysfs_merge_group(&fs_info->fs_devices->fsid_kobj,
&agroup);
if (ret)
return ret;
} else
sysfs_unmerge_group(&fs_info->fs_devices->fsid_kobj,
&agroup);
}
}
return 0;
}
static void __btrfs_sysfs_remove_fsid(struct btrfs_fs_devices *fs_devs)
{
if (fs_devs->devinfo_kobj) {
kobject_del(fs_devs->devinfo_kobj);
kobject_put(fs_devs->devinfo_kobj);
fs_devs->devinfo_kobj = NULL;
}
if (fs_devs->devices_kobj) {
kobject_del(fs_devs->devices_kobj);
kobject_put(fs_devs->devices_kobj);
fs_devs->devices_kobj = NULL;
}
if (fs_devs->fsid_kobj.state_initialized) {
kobject_del(&fs_devs->fsid_kobj);
kobject_put(&fs_devs->fsid_kobj);
wait_for_completion(&fs_devs->kobj_unregister);
}
}
/* when fs_devs is NULL it will remove all fsid kobject */
void btrfs_sysfs_remove_fsid(struct btrfs_fs_devices *fs_devs)
{
struct list_head *fs_uuids = btrfs_get_fs_uuids();
if (fs_devs) {
__btrfs_sysfs_remove_fsid(fs_devs);
return;
}
list_for_each_entry(fs_devs, fs_uuids, fs_list) {
__btrfs_sysfs_remove_fsid(fs_devs);
}
}
static void btrfs_sysfs_remove_fs_devices(struct btrfs_fs_devices *fs_devices)
{
struct btrfs_device *device;
struct btrfs_fs_devices *seed;
list_for_each_entry(device, &fs_devices->devices, dev_list)
btrfs_sysfs_remove_device(device);
list_for_each_entry(seed, &fs_devices->seed_list, seed_list) {
list_for_each_entry(device, &seed->devices, dev_list)
btrfs_sysfs_remove_device(device);
}
}
void btrfs_sysfs_remove_mounted(struct btrfs_fs_info *fs_info)
{
struct kobject *fsid_kobj = &fs_info->fs_devices->fsid_kobj;
sysfs_remove_link(fsid_kobj, "bdi");
if (fs_info->space_info_kobj) {
sysfs_remove_files(fs_info->space_info_kobj, allocation_attrs);
kobject_del(fs_info->space_info_kobj);
kobject_put(fs_info->space_info_kobj);
}
if (fs_info->discard_kobj) {
sysfs_remove_files(fs_info->discard_kobj, discard_attrs);
kobject_del(fs_info->discard_kobj);
kobject_put(fs_info->discard_kobj);
}
#ifdef CONFIG_BTRFS_DEBUG
if (fs_info->debug_kobj) {
sysfs_remove_files(fs_info->debug_kobj, btrfs_debug_mount_attrs);
kobject_del(fs_info->debug_kobj);
kobject_put(fs_info->debug_kobj);
}
#endif
addrm_unknown_feature_attrs(fs_info, false);
sysfs_remove_group(fsid_kobj, &btrfs_feature_attr_group);
sysfs_remove_files(fsid_kobj, btrfs_attrs);
btrfs_sysfs_remove_fs_devices(fs_info->fs_devices);
}
static const char * const btrfs_feature_set_names[FEAT_MAX] = {
[FEAT_COMPAT] = "compat",
[FEAT_COMPAT_RO] = "compat_ro",
[FEAT_INCOMPAT] = "incompat",
};
const char *btrfs_feature_set_name(enum btrfs_feature_set set)
{
return btrfs_feature_set_names[set];
}
char *btrfs_printable_features(enum btrfs_feature_set set, u64 flags)
{
size_t bufsize = 4096; /* safe max, 64 names * 64 bytes */
int len = 0;
int i;
char *str;
str = kmalloc(bufsize, GFP_KERNEL);
if (!str)
return str;
for (i = 0; i < ARRAY_SIZE(btrfs_feature_attrs[set]); i++) {
const char *name;
if (!(flags & (1ULL << i)))
continue;
name = btrfs_feature_attrs[set][i].kobj_attr.attr.name;
len += scnprintf(str + len, bufsize - len, "%s%s",
len ? "," : "", name);
}
return str;
}
static void init_feature_attrs(void)
{
struct btrfs_feature_attr *fa;
int set, i;
memset(btrfs_feature_attrs, 0, sizeof(btrfs_feature_attrs));
memset(btrfs_unknown_feature_names, 0,
sizeof(btrfs_unknown_feature_names));
for (i = 0; btrfs_supported_feature_attrs[i]; i++) {
struct btrfs_feature_attr *sfa;
struct attribute *a = btrfs_supported_feature_attrs[i];
int bit;
sfa = attr_to_btrfs_feature_attr(a);
bit = ilog2(sfa->feature_bit);
fa = &btrfs_feature_attrs[sfa->feature_set][bit];
fa->kobj_attr.attr.name = sfa->kobj_attr.attr.name;
}
for (set = 0; set < FEAT_MAX; set++) {
for (i = 0; i < ARRAY_SIZE(btrfs_feature_attrs[set]); i++) {
char *name = btrfs_unknown_feature_names[set][i];
fa = &btrfs_feature_attrs[set][i];
if (fa->kobj_attr.attr.name)
continue;
snprintf(name, BTRFS_FEATURE_NAME_MAX, "%s:%u",
btrfs_feature_set_names[set], i);
fa->kobj_attr.attr.name = name;
fa->kobj_attr.attr.mode = S_IRUGO;
fa->feature_set = set;
fa->feature_bit = 1ULL << i;
}
}
}
/*
* Create a sysfs entry for a given block group type at path
* /sys/fs/btrfs/UUID/allocation/data/TYPE
*/
void btrfs_sysfs_add_block_group_type(struct btrfs_block_group *cache)
{
struct btrfs_fs_info *fs_info = cache->fs_info;
struct btrfs_space_info *space_info = cache->space_info;
struct raid_kobject *rkobj;
const int index = btrfs_bg_flags_to_raid_index(cache->flags);
unsigned int nofs_flag;
int ret;
/*
* Setup a NOFS context because kobject_add(), deep in its call chain,
* does GFP_KERNEL allocations, and we are often called in a context
* where if reclaim is triggered we can deadlock (we are either holding
* a transaction handle or some lock required for a transaction
* commit).
*/
nofs_flag = memalloc_nofs_save();
rkobj = kzalloc(sizeof(*rkobj), GFP_NOFS);
if (!rkobj) {
memalloc_nofs_restore(nofs_flag);
btrfs_warn(cache->fs_info,
"couldn't alloc memory for raid level kobject");
return;
}
rkobj->flags = cache->flags;
kobject_init(&rkobj->kobj, &btrfs_raid_ktype);
btrfs: do not create raid sysfs entries under any locks While running xfstests btrfs/177 I got the following lockdep splat ====================================================== WARNING: possible circular locking dependency detected 5.9.0-rc3+ #5 Not tainted ------------------------------------------------------ kswapd0/100 is trying to acquire lock: ffff97066aa56760 (&delayed_node->mutex){+.+.}-{3:3}, at: __btrfs_release_delayed_node.part.0+0x3f/0x330 but task is already holding lock: ffffffff9fd74700 (fs_reclaim){+.+.}-{0:0}, at: __fs_reclaim_acquire+0x5/0x30 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #3 (fs_reclaim){+.+.}-{0:0}: fs_reclaim_acquire+0x65/0x80 slab_pre_alloc_hook.constprop.0+0x20/0x200 kmem_cache_alloc+0x37/0x270 alloc_inode+0x82/0xb0 iget_locked+0x10d/0x2c0 kernfs_get_inode+0x1b/0x130 kernfs_get_tree+0x136/0x240 sysfs_get_tree+0x16/0x40 vfs_get_tree+0x28/0xc0 path_mount+0x434/0xc00 __x64_sys_mount+0xe3/0x120 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xa9 -> #2 (kernfs_mutex){+.+.}-{3:3}: __mutex_lock+0x7e/0x7e0 kernfs_add_one+0x23/0x150 kernfs_create_dir_ns+0x7a/0xb0 sysfs_create_dir_ns+0x60/0xb0 kobject_add_internal+0xc0/0x2c0 kobject_add+0x6e/0x90 btrfs_sysfs_add_block_group_type+0x102/0x160 btrfs_make_block_group+0x167/0x230 btrfs_alloc_chunk+0x54f/0xb80 btrfs_chunk_alloc+0x18e/0x3a0 find_free_extent+0xdf6/0x1210 btrfs_reserve_extent+0xb3/0x1b0 btrfs_alloc_tree_block+0xb0/0x310 alloc_tree_block_no_bg_flush+0x4a/0x60 __btrfs_cow_block+0x11a/0x530 btrfs_cow_block+0x104/0x220 btrfs_search_slot+0x52e/0x9d0 btrfs_insert_empty_items+0x64/0xb0 btrfs_new_inode+0x225/0x730 btrfs_create+0xab/0x1f0 lookup_open.isra.0+0x52d/0x690 path_openat+0x2a7/0x9e0 do_filp_open+0x75/0x100 do_sys_openat2+0x7b/0x130 __x64_sys_openat+0x46/0x70 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xa9 -> #1 (&fs_info->chunk_mutex){+.+.}-{3:3}: __mutex_lock+0x7e/0x7e0 btrfs_chunk_alloc+0x125/0x3a0 find_free_extent+0xdf6/0x1210 btrfs_reserve_extent+0xb3/0x1b0 btrfs_alloc_tree_block+0xb0/0x310 alloc_tree_block_no_bg_flush+0x4a/0x60 __btrfs_cow_block+0x11a/0x530 btrfs_cow_block+0x104/0x220 btrfs_search_slot+0x52e/0x9d0 btrfs_lookup_inode+0x2a/0x8f __btrfs_update_delayed_inode+0x80/0x240 btrfs_commit_inode_delayed_inode+0x119/0x120 btrfs_evict_inode+0x357/0x500 evict+0xcf/0x1f0 do_unlinkat+0x1a9/0x2b0 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xa9 -> #0 (&delayed_node->mutex){+.+.}-{3:3}: __lock_acquire+0x119c/0x1fc0 lock_acquire+0xa7/0x3d0 __mutex_lock+0x7e/0x7e0 __btrfs_release_delayed_node.part.0+0x3f/0x330 btrfs_evict_inode+0x24c/0x500 evict+0xcf/0x1f0 dispose_list+0x48/0x70 prune_icache_sb+0x44/0x50 super_cache_scan+0x161/0x1e0 do_shrink_slab+0x178/0x3c0 shrink_slab+0x17c/0x290 shrink_node+0x2b2/0x6d0 balance_pgdat+0x30a/0x670 kswapd+0x213/0x4c0 kthread+0x138/0x160 ret_from_fork+0x1f/0x30 other info that might help us debug this: Chain exists of: &delayed_node->mutex --> kernfs_mutex --> fs_reclaim Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(fs_reclaim); lock(kernfs_mutex); lock(fs_reclaim); lock(&delayed_node->mutex); *** DEADLOCK *** 3 locks held by kswapd0/100: #0: ffffffff9fd74700 (fs_reclaim){+.+.}-{0:0}, at: __fs_reclaim_acquire+0x5/0x30 #1: ffffffff9fd65c50 (shrinker_rwsem){++++}-{3:3}, at: shrink_slab+0x115/0x290 #2: ffff9706629780e0 (&type->s_umount_key#36){++++}-{3:3}, at: super_cache_scan+0x38/0x1e0 stack backtrace: CPU: 1 PID: 100 Comm: kswapd0 Not tainted 5.9.0-rc3+ #5 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.13.0-2.fc32 04/01/2014 Call Trace: dump_stack+0x8b/0xb8 check_noncircular+0x12d/0x150 __lock_acquire+0x119c/0x1fc0 lock_acquire+0xa7/0x3d0 ? __btrfs_release_delayed_node.part.0+0x3f/0x330 __mutex_lock+0x7e/0x7e0 ? __btrfs_release_delayed_node.part.0+0x3f/0x330 ? __btrfs_release_delayed_node.part.0+0x3f/0x330 ? lock_acquire+0xa7/0x3d0 ? find_held_lock+0x2b/0x80 __btrfs_release_delayed_node.part.0+0x3f/0x330 btrfs_evict_inode+0x24c/0x500 evict+0xcf/0x1f0 dispose_list+0x48/0x70 prune_icache_sb+0x44/0x50 super_cache_scan+0x161/0x1e0 do_shrink_slab+0x178/0x3c0 shrink_slab+0x17c/0x290 shrink_node+0x2b2/0x6d0 balance_pgdat+0x30a/0x670 kswapd+0x213/0x4c0 ? _raw_spin_unlock_irqrestore+0x41/0x50 ? add_wait_queue_exclusive+0x70/0x70 ? balance_pgdat+0x670/0x670 kthread+0x138/0x160 ? kthread_create_worker_on_cpu+0x40/0x40 ret_from_fork+0x1f/0x30 This happens because when we link in a block group with a new raid index type we'll create the corresponding sysfs entries for it. This is problematic because while restriping we're holding the chunk_mutex, and while mounting we're holding the tree locks. Fixing this isn't pretty, we move the call to the sysfs stuff into the btrfs_create_pending_block_groups() work, where we're not holding any locks. This creates a slight race where other threads could see that there's no sysfs kobj for that raid type, and race to create the sysfs dir. Fix this by wrapping the creation in space_info->lock, so we only get one thread calling kobject_add() for the new directory. We don't worry about the lock on cleanup as it only gets deleted on unmount. On mount it's more straightforward, we loop through the space_infos already, just check every raid index in each space_info and added the sysfs entries for the corresponding block groups. Signed-off-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: David Sterba <dsterba@suse.com>
2020-09-01 21:40:38 +00:00
/*
* We call this either on mount, or if we've created a block group for a
* new index type while running (i.e. when restriping). The running
* case is tricky because we could race with other threads, so we need
* to have this check to make sure we didn't already init the kobject.
*
* We don't have to protect on the free side because it only happens on
* unmount.
*/
spin_lock(&space_info->lock);
if (space_info->block_group_kobjs[index]) {
spin_unlock(&space_info->lock);
kobject_put(&rkobj->kobj);
return;
} else {
space_info->block_group_kobjs[index] = &rkobj->kobj;
}
spin_unlock(&space_info->lock);
ret = kobject_add(&rkobj->kobj, &space_info->kobj, "%s",
btrfs_bg_type_to_raid_name(rkobj->flags));
memalloc_nofs_restore(nofs_flag);
if (ret) {
btrfs: do not create raid sysfs entries under any locks While running xfstests btrfs/177 I got the following lockdep splat ====================================================== WARNING: possible circular locking dependency detected 5.9.0-rc3+ #5 Not tainted ------------------------------------------------------ kswapd0/100 is trying to acquire lock: ffff97066aa56760 (&delayed_node->mutex){+.+.}-{3:3}, at: __btrfs_release_delayed_node.part.0+0x3f/0x330 but task is already holding lock: ffffffff9fd74700 (fs_reclaim){+.+.}-{0:0}, at: __fs_reclaim_acquire+0x5/0x30 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #3 (fs_reclaim){+.+.}-{0:0}: fs_reclaim_acquire+0x65/0x80 slab_pre_alloc_hook.constprop.0+0x20/0x200 kmem_cache_alloc+0x37/0x270 alloc_inode+0x82/0xb0 iget_locked+0x10d/0x2c0 kernfs_get_inode+0x1b/0x130 kernfs_get_tree+0x136/0x240 sysfs_get_tree+0x16/0x40 vfs_get_tree+0x28/0xc0 path_mount+0x434/0xc00 __x64_sys_mount+0xe3/0x120 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xa9 -> #2 (kernfs_mutex){+.+.}-{3:3}: __mutex_lock+0x7e/0x7e0 kernfs_add_one+0x23/0x150 kernfs_create_dir_ns+0x7a/0xb0 sysfs_create_dir_ns+0x60/0xb0 kobject_add_internal+0xc0/0x2c0 kobject_add+0x6e/0x90 btrfs_sysfs_add_block_group_type+0x102/0x160 btrfs_make_block_group+0x167/0x230 btrfs_alloc_chunk+0x54f/0xb80 btrfs_chunk_alloc+0x18e/0x3a0 find_free_extent+0xdf6/0x1210 btrfs_reserve_extent+0xb3/0x1b0 btrfs_alloc_tree_block+0xb0/0x310 alloc_tree_block_no_bg_flush+0x4a/0x60 __btrfs_cow_block+0x11a/0x530 btrfs_cow_block+0x104/0x220 btrfs_search_slot+0x52e/0x9d0 btrfs_insert_empty_items+0x64/0xb0 btrfs_new_inode+0x225/0x730 btrfs_create+0xab/0x1f0 lookup_open.isra.0+0x52d/0x690 path_openat+0x2a7/0x9e0 do_filp_open+0x75/0x100 do_sys_openat2+0x7b/0x130 __x64_sys_openat+0x46/0x70 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xa9 -> #1 (&fs_info->chunk_mutex){+.+.}-{3:3}: __mutex_lock+0x7e/0x7e0 btrfs_chunk_alloc+0x125/0x3a0 find_free_extent+0xdf6/0x1210 btrfs_reserve_extent+0xb3/0x1b0 btrfs_alloc_tree_block+0xb0/0x310 alloc_tree_block_no_bg_flush+0x4a/0x60 __btrfs_cow_block+0x11a/0x530 btrfs_cow_block+0x104/0x220 btrfs_search_slot+0x52e/0x9d0 btrfs_lookup_inode+0x2a/0x8f __btrfs_update_delayed_inode+0x80/0x240 btrfs_commit_inode_delayed_inode+0x119/0x120 btrfs_evict_inode+0x357/0x500 evict+0xcf/0x1f0 do_unlinkat+0x1a9/0x2b0 do_syscall_64+0x33/0x40 entry_SYSCALL_64_after_hwframe+0x44/0xa9 -> #0 (&delayed_node->mutex){+.+.}-{3:3}: __lock_acquire+0x119c/0x1fc0 lock_acquire+0xa7/0x3d0 __mutex_lock+0x7e/0x7e0 __btrfs_release_delayed_node.part.0+0x3f/0x330 btrfs_evict_inode+0x24c/0x500 evict+0xcf/0x1f0 dispose_list+0x48/0x70 prune_icache_sb+0x44/0x50 super_cache_scan+0x161/0x1e0 do_shrink_slab+0x178/0x3c0 shrink_slab+0x17c/0x290 shrink_node+0x2b2/0x6d0 balance_pgdat+0x30a/0x670 kswapd+0x213/0x4c0 kthread+0x138/0x160 ret_from_fork+0x1f/0x30 other info that might help us debug this: Chain exists of: &delayed_node->mutex --> kernfs_mutex --> fs_reclaim Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(fs_reclaim); lock(kernfs_mutex); lock(fs_reclaim); lock(&delayed_node->mutex); *** DEADLOCK *** 3 locks held by kswapd0/100: #0: ffffffff9fd74700 (fs_reclaim){+.+.}-{0:0}, at: __fs_reclaim_acquire+0x5/0x30 #1: ffffffff9fd65c50 (shrinker_rwsem){++++}-{3:3}, at: shrink_slab+0x115/0x290 #2: ffff9706629780e0 (&type->s_umount_key#36){++++}-{3:3}, at: super_cache_scan+0x38/0x1e0 stack backtrace: CPU: 1 PID: 100 Comm: kswapd0 Not tainted 5.9.0-rc3+ #5 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.13.0-2.fc32 04/01/2014 Call Trace: dump_stack+0x8b/0xb8 check_noncircular+0x12d/0x150 __lock_acquire+0x119c/0x1fc0 lock_acquire+0xa7/0x3d0 ? __btrfs_release_delayed_node.part.0+0x3f/0x330 __mutex_lock+0x7e/0x7e0 ? __btrfs_release_delayed_node.part.0+0x3f/0x330 ? __btrfs_release_delayed_node.part.0+0x3f/0x330 ? lock_acquire+0xa7/0x3d0 ? find_held_lock+0x2b/0x80 __btrfs_release_delayed_node.part.0+0x3f/0x330 btrfs_evict_inode+0x24c/0x500 evict+0xcf/0x1f0 dispose_list+0x48/0x70 prune_icache_sb+0x44/0x50 super_cache_scan+0x161/0x1e0 do_shrink_slab+0x178/0x3c0 shrink_slab+0x17c/0x290 shrink_node+0x2b2/0x6d0 balance_pgdat+0x30a/0x670 kswapd+0x213/0x4c0 ? _raw_spin_unlock_irqrestore+0x41/0x50 ? add_wait_queue_exclusive+0x70/0x70 ? balance_pgdat+0x670/0x670 kthread+0x138/0x160 ? kthread_create_worker_on_cpu+0x40/0x40 ret_from_fork+0x1f/0x30 This happens because when we link in a block group with a new raid index type we'll create the corresponding sysfs entries for it. This is problematic because while restriping we're holding the chunk_mutex, and while mounting we're holding the tree locks. Fixing this isn't pretty, we move the call to the sysfs stuff into the btrfs_create_pending_block_groups() work, where we're not holding any locks. This creates a slight race where other threads could see that there's no sysfs kobj for that raid type, and race to create the sysfs dir. Fix this by wrapping the creation in space_info->lock, so we only get one thread calling kobject_add() for the new directory. We don't worry about the lock on cleanup as it only gets deleted on unmount. On mount it's more straightforward, we loop through the space_infos already, just check every raid index in each space_info and added the sysfs entries for the corresponding block groups. Signed-off-by: Josef Bacik <josef@toxicpanda.com> Signed-off-by: David Sterba <dsterba@suse.com>
2020-09-01 21:40:38 +00:00
spin_lock(&space_info->lock);
space_info->block_group_kobjs[index] = NULL;
spin_unlock(&space_info->lock);
kobject_put(&rkobj->kobj);
btrfs_warn(fs_info,
"failed to add kobject for block cache, ignoring");
return;
}
}
/*
* Remove sysfs directories for all block group types of a given space info and
* the space info as well
*/
void btrfs_sysfs_remove_space_info(struct btrfs_space_info *space_info)
{
int i;
for (i = 0; i < BTRFS_NR_RAID_TYPES; i++) {
struct kobject *kobj;
kobj = space_info->block_group_kobjs[i];
space_info->block_group_kobjs[i] = NULL;
if (kobj) {
kobject_del(kobj);
kobject_put(kobj);
}
}
kobject_del(&space_info->kobj);
kobject_put(&space_info->kobj);
}
static const char *alloc_name(u64 flags)
{
switch (flags) {
case BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA:
return "mixed";
case BTRFS_BLOCK_GROUP_METADATA:
return "metadata";
case BTRFS_BLOCK_GROUP_DATA:
return "data";
case BTRFS_BLOCK_GROUP_SYSTEM:
return "system";
default:
WARN_ON(1);
return "invalid-combination";
}
}
/*
* Create a sysfs entry for a space info type at path
* /sys/fs/btrfs/UUID/allocation/TYPE
*/
int btrfs_sysfs_add_space_info_type(struct btrfs_fs_info *fs_info,
struct btrfs_space_info *space_info)
{
int ret;
ret = kobject_init_and_add(&space_info->kobj, &space_info_ktype,
fs_info->space_info_kobj, "%s",
alloc_name(space_info->flags));
if (ret) {
kobject_put(&space_info->kobj);
return ret;
}
return 0;
}
void btrfs_sysfs_remove_device(struct btrfs_device *device)
{
struct kobject *devices_kobj;
/*
* Seed fs_devices devices_kobj aren't used, fetch kobject from the
* fs_info::fs_devices.
*/
devices_kobj = device->fs_info->fs_devices->devices_kobj;
ASSERT(devices_kobj);
if (device->bdev)
sysfs_remove_link(devices_kobj, bdev_kobj(device->bdev)->name);
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
if (device->devid_kobj.state_initialized) {
kobject_del(&device->devid_kobj);
kobject_put(&device->devid_kobj);
wait_for_completion(&device->kobj_unregister);
}
}
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
static ssize_t btrfs_devinfo_in_fs_metadata_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
int val;
struct btrfs_device *device = container_of(kobj, struct btrfs_device,
devid_kobj);
val = !!test_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state);
return sysfs_emit(buf, "%d\n", val);
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
}
BTRFS_ATTR(devid, in_fs_metadata, btrfs_devinfo_in_fs_metadata_show);
static ssize_t btrfs_devinfo_missing_show(struct kobject *kobj,
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
struct kobj_attribute *a, char *buf)
{
int val;
struct btrfs_device *device = container_of(kobj, struct btrfs_device,
devid_kobj);
val = !!test_bit(BTRFS_DEV_STATE_MISSING, &device->dev_state);
return sysfs_emit(buf, "%d\n", val);
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
}
BTRFS_ATTR(devid, missing, btrfs_devinfo_missing_show);
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
static ssize_t btrfs_devinfo_replace_target_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
int val;
struct btrfs_device *device = container_of(kobj, struct btrfs_device,
devid_kobj);
val = !!test_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state);
return sysfs_emit(buf, "%d\n", val);
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
}
BTRFS_ATTR(devid, replace_target, btrfs_devinfo_replace_target_show);
btrfs: scrub: per-device bandwidth control Add sysfs interface to limit io during scrub. We relied on the ionice interface to do that, eg. the idle class let the system usable while scrub was running. This has changed when mq-deadline got widespread and did not implement the scheduling classes. That was a CFQ thing that got deleted. We've got numerous complaints from users about degraded performance. Currently only BFQ supports that but it's not a common scheduler and we can't ask everybody to switch to it. Alternatively the cgroup io limiting can be used but that also a non-trivial setup (v2 required, the controller must be enabled on the system). This can still be used if desired. Other ideas that have been explored: piggy-back on ionice (that is set per-process and is accessible) and interpret the class and classdata as bandwidth limits, but this does not have enough flexibility as there are only 8 allowed and we'd have to map fixed limits to each value. Also adjusting the value would need to lookup the process that currently runs scrub on the given device, and the value is not sticky so would have to be adjusted each time scrub runs. Running out of options, sysfs does not look that bad: - it's accessible from scripts, or udev rules - the name is similar to what MD-RAID has (/proc/sys/dev/raid/speed_limit_max or /sys/block/mdX/md/sync_speed_max) - the value is sticky at least for filesystem mount time - adjusting the value has immediate effect - sysfs is available in constrained environments (eg. system rescue) - the limit also applies to device replace Sysfs: - raw value is in bytes - values written to the file accept suffixes like K, M - file is in the per-device directory /sys/fs/btrfs/FSID/devinfo/DEVID/scrub_speed_max - 0 means use default priority of IO The scheduler is a simple deadline one and the accuracy is up to nearest 128K. Signed-off-by: David Sterba <dsterba@suse.com>
2019-10-09 11:58:13 +00:00
static ssize_t btrfs_devinfo_scrub_speed_max_show(struct kobject *kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_device *device = container_of(kobj, struct btrfs_device,
devid_kobj);
return sysfs_emit(buf, "%llu\n", READ_ONCE(device->scrub_speed_max));
btrfs: scrub: per-device bandwidth control Add sysfs interface to limit io during scrub. We relied on the ionice interface to do that, eg. the idle class let the system usable while scrub was running. This has changed when mq-deadline got widespread and did not implement the scheduling classes. That was a CFQ thing that got deleted. We've got numerous complaints from users about degraded performance. Currently only BFQ supports that but it's not a common scheduler and we can't ask everybody to switch to it. Alternatively the cgroup io limiting can be used but that also a non-trivial setup (v2 required, the controller must be enabled on the system). This can still be used if desired. Other ideas that have been explored: piggy-back on ionice (that is set per-process and is accessible) and interpret the class and classdata as bandwidth limits, but this does not have enough flexibility as there are only 8 allowed and we'd have to map fixed limits to each value. Also adjusting the value would need to lookup the process that currently runs scrub on the given device, and the value is not sticky so would have to be adjusted each time scrub runs. Running out of options, sysfs does not look that bad: - it's accessible from scripts, or udev rules - the name is similar to what MD-RAID has (/proc/sys/dev/raid/speed_limit_max or /sys/block/mdX/md/sync_speed_max) - the value is sticky at least for filesystem mount time - adjusting the value has immediate effect - sysfs is available in constrained environments (eg. system rescue) - the limit also applies to device replace Sysfs: - raw value is in bytes - values written to the file accept suffixes like K, M - file is in the per-device directory /sys/fs/btrfs/FSID/devinfo/DEVID/scrub_speed_max - 0 means use default priority of IO The scheduler is a simple deadline one and the accuracy is up to nearest 128K. Signed-off-by: David Sterba <dsterba@suse.com>
2019-10-09 11:58:13 +00:00
}
static ssize_t btrfs_devinfo_scrub_speed_max_store(struct kobject *kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_device *device = container_of(kobj, struct btrfs_device,
devid_kobj);
char *endptr;
unsigned long long limit;
limit = memparse(buf, &endptr);
/* There could be trailing '\n', also catch any typos after the value. */
endptr = skip_spaces(endptr);
if (*endptr != 0)
return -EINVAL;
btrfs: scrub: per-device bandwidth control Add sysfs interface to limit io during scrub. We relied on the ionice interface to do that, eg. the idle class let the system usable while scrub was running. This has changed when mq-deadline got widespread and did not implement the scheduling classes. That was a CFQ thing that got deleted. We've got numerous complaints from users about degraded performance. Currently only BFQ supports that but it's not a common scheduler and we can't ask everybody to switch to it. Alternatively the cgroup io limiting can be used but that also a non-trivial setup (v2 required, the controller must be enabled on the system). This can still be used if desired. Other ideas that have been explored: piggy-back on ionice (that is set per-process and is accessible) and interpret the class and classdata as bandwidth limits, but this does not have enough flexibility as there are only 8 allowed and we'd have to map fixed limits to each value. Also adjusting the value would need to lookup the process that currently runs scrub on the given device, and the value is not sticky so would have to be adjusted each time scrub runs. Running out of options, sysfs does not look that bad: - it's accessible from scripts, or udev rules - the name is similar to what MD-RAID has (/proc/sys/dev/raid/speed_limit_max or /sys/block/mdX/md/sync_speed_max) - the value is sticky at least for filesystem mount time - adjusting the value has immediate effect - sysfs is available in constrained environments (eg. system rescue) - the limit also applies to device replace Sysfs: - raw value is in bytes - values written to the file accept suffixes like K, M - file is in the per-device directory /sys/fs/btrfs/FSID/devinfo/DEVID/scrub_speed_max - 0 means use default priority of IO The scheduler is a simple deadline one and the accuracy is up to nearest 128K. Signed-off-by: David Sterba <dsterba@suse.com>
2019-10-09 11:58:13 +00:00
WRITE_ONCE(device->scrub_speed_max, limit);
return len;
}
BTRFS_ATTR_RW(devid, scrub_speed_max, btrfs_devinfo_scrub_speed_max_show,
btrfs_devinfo_scrub_speed_max_store);
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
static ssize_t btrfs_devinfo_writeable_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
int val;
struct btrfs_device *device = container_of(kobj, struct btrfs_device,
devid_kobj);
val = !!test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state);
return sysfs_emit(buf, "%d\n", val);
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
}
BTRFS_ATTR(devid, writeable, btrfs_devinfo_writeable_show);
static ssize_t btrfs_devinfo_fsid_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_device *device = container_of(kobj, struct btrfs_device,
devid_kobj);
return sysfs_emit(buf, "%pU\n", device->fs_devices->fsid);
}
BTRFS_ATTR(devid, fsid, btrfs_devinfo_fsid_show);
static ssize_t btrfs_devinfo_error_stats_show(struct kobject *kobj,
struct kobj_attribute *a, char *buf)
{
struct btrfs_device *device = container_of(kobj, struct btrfs_device,
devid_kobj);
if (!device->dev_stats_valid)
return sysfs_emit(buf, "invalid\n");
/*
* Print all at once so we get a snapshot of all values from the same
* time. Keep them in sync and in order of definition of
* btrfs_dev_stat_values.
*/
return sysfs_emit(buf,
"write_errs %d\n"
"read_errs %d\n"
"flush_errs %d\n"
"corruption_errs %d\n"
"generation_errs %d\n",
btrfs_dev_stat_read(device, BTRFS_DEV_STAT_WRITE_ERRS),
btrfs_dev_stat_read(device, BTRFS_DEV_STAT_READ_ERRS),
btrfs_dev_stat_read(device, BTRFS_DEV_STAT_FLUSH_ERRS),
btrfs_dev_stat_read(device, BTRFS_DEV_STAT_CORRUPTION_ERRS),
btrfs_dev_stat_read(device, BTRFS_DEV_STAT_GENERATION_ERRS));
}
BTRFS_ATTR(devid, error_stats, btrfs_devinfo_error_stats_show);
/*
* Information about one device.
*
* Path: /sys/fs/btrfs/<uuid>/devinfo/<devid>/
*/
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
static struct attribute *devid_attrs[] = {
BTRFS_ATTR_PTR(devid, error_stats),
BTRFS_ATTR_PTR(devid, fsid),
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
BTRFS_ATTR_PTR(devid, in_fs_metadata),
BTRFS_ATTR_PTR(devid, missing),
BTRFS_ATTR_PTR(devid, replace_target),
btrfs: scrub: per-device bandwidth control Add sysfs interface to limit io during scrub. We relied on the ionice interface to do that, eg. the idle class let the system usable while scrub was running. This has changed when mq-deadline got widespread and did not implement the scheduling classes. That was a CFQ thing that got deleted. We've got numerous complaints from users about degraded performance. Currently only BFQ supports that but it's not a common scheduler and we can't ask everybody to switch to it. Alternatively the cgroup io limiting can be used but that also a non-trivial setup (v2 required, the controller must be enabled on the system). This can still be used if desired. Other ideas that have been explored: piggy-back on ionice (that is set per-process and is accessible) and interpret the class and classdata as bandwidth limits, but this does not have enough flexibility as there are only 8 allowed and we'd have to map fixed limits to each value. Also adjusting the value would need to lookup the process that currently runs scrub on the given device, and the value is not sticky so would have to be adjusted each time scrub runs. Running out of options, sysfs does not look that bad: - it's accessible from scripts, or udev rules - the name is similar to what MD-RAID has (/proc/sys/dev/raid/speed_limit_max or /sys/block/mdX/md/sync_speed_max) - the value is sticky at least for filesystem mount time - adjusting the value has immediate effect - sysfs is available in constrained environments (eg. system rescue) - the limit also applies to device replace Sysfs: - raw value is in bytes - values written to the file accept suffixes like K, M - file is in the per-device directory /sys/fs/btrfs/FSID/devinfo/DEVID/scrub_speed_max - 0 means use default priority of IO The scheduler is a simple deadline one and the accuracy is up to nearest 128K. Signed-off-by: David Sterba <dsterba@suse.com>
2019-10-09 11:58:13 +00:00
BTRFS_ATTR_PTR(devid, scrub_speed_max),
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
BTRFS_ATTR_PTR(devid, writeable),
NULL
};
ATTRIBUTE_GROUPS(devid);
static void btrfs_release_devid_kobj(struct kobject *kobj)
{
struct btrfs_device *device = container_of(kobj, struct btrfs_device,
devid_kobj);
memset(&device->devid_kobj, 0, sizeof(struct kobject));
complete(&device->kobj_unregister);
}
static const struct kobj_type devid_ktype = {
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
.sysfs_ops = &kobj_sysfs_ops,
.default_groups = devid_groups,
.release = btrfs_release_devid_kobj,
};
int btrfs_sysfs_add_device(struct btrfs_device *device)
{
int ret;
btrfs: sysfs: use NOFS for device creation Dave hit this splat during testing btrfs/078: ====================================================== WARNING: possible circular locking dependency detected 5.8.0-rc6-default+ #1191 Not tainted ------------------------------------------------------ kswapd0/75 is trying to acquire lock: ffffa040e9d04ff8 (&delayed_node->mutex){+.+.}-{3:3}, at: __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] but task is already holding lock: ffffffff8b0c8040 (fs_reclaim){+.+.}-{0:0}, at: __fs_reclaim_acquire+0x5/0x30 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #2 (fs_reclaim){+.+.}-{0:0}: __lock_acquire+0x56f/0xaa0 lock_acquire+0xa3/0x440 fs_reclaim_acquire.part.0+0x25/0x30 __kmalloc_track_caller+0x49/0x330 kstrdup+0x2e/0x60 __kernfs_new_node.constprop.0+0x44/0x250 kernfs_new_node+0x25/0x50 kernfs_create_link+0x34/0xa0 sysfs_do_create_link_sd+0x5e/0xd0 btrfs_sysfs_add_devices_dir+0x65/0x100 [btrfs] btrfs_init_new_device+0x44c/0x12b0 [btrfs] btrfs_ioctl+0xc3c/0x25c0 [btrfs] ksys_ioctl+0x68/0xa0 __x64_sys_ioctl+0x16/0x20 do_syscall_64+0x50/0xe0 entry_SYSCALL_64_after_hwframe+0x44/0xa9 -> #1 (&fs_info->chunk_mutex){+.+.}-{3:3}: __lock_acquire+0x56f/0xaa0 lock_acquire+0xa3/0x440 __mutex_lock+0xa0/0xaf0 btrfs_chunk_alloc+0x137/0x3e0 [btrfs] find_free_extent+0xb44/0xfb0 [btrfs] btrfs_reserve_extent+0x9b/0x180 [btrfs] btrfs_alloc_tree_block+0xc1/0x350 [btrfs] alloc_tree_block_no_bg_flush+0x4a/0x60 [btrfs] __btrfs_cow_block+0x143/0x7a0 [btrfs] btrfs_cow_block+0x15f/0x310 [btrfs] push_leaf_right+0x150/0x240 [btrfs] split_leaf+0x3cd/0x6d0 [btrfs] btrfs_search_slot+0xd14/0xf70 [btrfs] btrfs_insert_empty_items+0x64/0xc0 [btrfs] __btrfs_commit_inode_delayed_items+0xb2/0x840 [btrfs] btrfs_async_run_delayed_root+0x10e/0x1d0 [btrfs] btrfs_work_helper+0x2f9/0x650 [btrfs] process_one_work+0x22c/0x600 worker_thread+0x50/0x3b0 kthread+0x137/0x150 ret_from_fork+0x1f/0x30 -> #0 (&delayed_node->mutex){+.+.}-{3:3}: check_prev_add+0x98/0xa20 validate_chain+0xa8c/0x2a00 __lock_acquire+0x56f/0xaa0 lock_acquire+0xa3/0x440 __mutex_lock+0xa0/0xaf0 __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] btrfs_evict_inode+0x3bf/0x560 [btrfs] evict+0xd6/0x1c0 dispose_list+0x48/0x70 prune_icache_sb+0x54/0x80 super_cache_scan+0x121/0x1a0 do_shrink_slab+0x175/0x420 shrink_slab+0xb1/0x2e0 shrink_node+0x192/0x600 balance_pgdat+0x31f/0x750 kswapd+0x206/0x510 kthread+0x137/0x150 ret_from_fork+0x1f/0x30 other info that might help us debug this: Chain exists of: &delayed_node->mutex --> &fs_info->chunk_mutex --> fs_reclaim Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(fs_reclaim); lock(&fs_info->chunk_mutex); lock(fs_reclaim); lock(&delayed_node->mutex); *** DEADLOCK *** 3 locks held by kswapd0/75: #0: ffffffff8b0c8040 (fs_reclaim){+.+.}-{0:0}, at: __fs_reclaim_acquire+0x5/0x30 #1: ffffffff8b0b50b8 (shrinker_rwsem){++++}-{3:3}, at: shrink_slab+0x54/0x2e0 #2: ffffa040e057c0e8 (&type->s_umount_key#26){++++}-{3:3}, at: trylock_super+0x16/0x50 stack backtrace: CPU: 2 PID: 75 Comm: kswapd0 Not tainted 5.8.0-rc6-default+ #1191 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.0-59-gc9ba527-rebuilt.opensuse.org 04/01/2014 Call Trace: dump_stack+0x78/0xa0 check_noncircular+0x16f/0x190 check_prev_add+0x98/0xa20 validate_chain+0xa8c/0x2a00 __lock_acquire+0x56f/0xaa0 lock_acquire+0xa3/0x440 ? __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] __mutex_lock+0xa0/0xaf0 ? __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] ? __lock_acquire+0x56f/0xaa0 ? __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] ? lock_acquire+0xa3/0x440 ? btrfs_evict_inode+0x138/0x560 [btrfs] ? btrfs_evict_inode+0x2fe/0x560 [btrfs] ? __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] btrfs_evict_inode+0x3bf/0x560 [btrfs] evict+0xd6/0x1c0 dispose_list+0x48/0x70 prune_icache_sb+0x54/0x80 super_cache_scan+0x121/0x1a0 do_shrink_slab+0x175/0x420 shrink_slab+0xb1/0x2e0 shrink_node+0x192/0x600 balance_pgdat+0x31f/0x750 kswapd+0x206/0x510 ? _raw_spin_unlock_irqrestore+0x3e/0x50 ? finish_wait+0x90/0x90 ? balance_pgdat+0x750/0x750 kthread+0x137/0x150 ? kthread_stop+0x2a0/0x2a0 ret_from_fork+0x1f/0x30 This is because we're holding the chunk_mutex while adding this device and adding its sysfs entries. We actually hold different locks in different places when calling this function, the dev_replace semaphore for instance in dev replace, so instead of moving this call around simply wrap it's operations in NOFS. CC: stable@vger.kernel.org # 4.14+ Reported-by: David Sterba <dsterba@suse.com> Signed-off-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2020-07-21 14:17:50 +00:00
unsigned int nofs_flag;
struct kobject *devices_kobj;
struct kobject *devinfo_kobj;
/*
* Make sure we use the fs_info::fs_devices to fetch the kobjects even
* for the seed fs_devices
*/
devices_kobj = device->fs_info->fs_devices->devices_kobj;
devinfo_kobj = device->fs_info->fs_devices->devinfo_kobj;
ASSERT(devices_kobj);
ASSERT(devinfo_kobj);
nofs_flag = memalloc_nofs_save();
if (device->bdev) {
struct kobject *disk_kobj = bdev_kobj(device->bdev);
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
ret = sysfs_create_link(devices_kobj, disk_kobj, disk_kobj->name);
if (ret) {
btrfs_warn(device->fs_info,
"creating sysfs device link for devid %llu failed: %d",
device->devid, ret);
goto out;
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
}
}
init_completion(&device->kobj_unregister);
ret = kobject_init_and_add(&device->devid_kobj, &devid_ktype,
devinfo_kobj, "%llu", device->devid);
if (ret) {
kobject_put(&device->devid_kobj);
btrfs_warn(device->fs_info,
"devinfo init for devid %llu failed: %d",
device->devid, ret);
}
out:
btrfs: sysfs: use NOFS for device creation Dave hit this splat during testing btrfs/078: ====================================================== WARNING: possible circular locking dependency detected 5.8.0-rc6-default+ #1191 Not tainted ------------------------------------------------------ kswapd0/75 is trying to acquire lock: ffffa040e9d04ff8 (&delayed_node->mutex){+.+.}-{3:3}, at: __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] but task is already holding lock: ffffffff8b0c8040 (fs_reclaim){+.+.}-{0:0}, at: __fs_reclaim_acquire+0x5/0x30 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #2 (fs_reclaim){+.+.}-{0:0}: __lock_acquire+0x56f/0xaa0 lock_acquire+0xa3/0x440 fs_reclaim_acquire.part.0+0x25/0x30 __kmalloc_track_caller+0x49/0x330 kstrdup+0x2e/0x60 __kernfs_new_node.constprop.0+0x44/0x250 kernfs_new_node+0x25/0x50 kernfs_create_link+0x34/0xa0 sysfs_do_create_link_sd+0x5e/0xd0 btrfs_sysfs_add_devices_dir+0x65/0x100 [btrfs] btrfs_init_new_device+0x44c/0x12b0 [btrfs] btrfs_ioctl+0xc3c/0x25c0 [btrfs] ksys_ioctl+0x68/0xa0 __x64_sys_ioctl+0x16/0x20 do_syscall_64+0x50/0xe0 entry_SYSCALL_64_after_hwframe+0x44/0xa9 -> #1 (&fs_info->chunk_mutex){+.+.}-{3:3}: __lock_acquire+0x56f/0xaa0 lock_acquire+0xa3/0x440 __mutex_lock+0xa0/0xaf0 btrfs_chunk_alloc+0x137/0x3e0 [btrfs] find_free_extent+0xb44/0xfb0 [btrfs] btrfs_reserve_extent+0x9b/0x180 [btrfs] btrfs_alloc_tree_block+0xc1/0x350 [btrfs] alloc_tree_block_no_bg_flush+0x4a/0x60 [btrfs] __btrfs_cow_block+0x143/0x7a0 [btrfs] btrfs_cow_block+0x15f/0x310 [btrfs] push_leaf_right+0x150/0x240 [btrfs] split_leaf+0x3cd/0x6d0 [btrfs] btrfs_search_slot+0xd14/0xf70 [btrfs] btrfs_insert_empty_items+0x64/0xc0 [btrfs] __btrfs_commit_inode_delayed_items+0xb2/0x840 [btrfs] btrfs_async_run_delayed_root+0x10e/0x1d0 [btrfs] btrfs_work_helper+0x2f9/0x650 [btrfs] process_one_work+0x22c/0x600 worker_thread+0x50/0x3b0 kthread+0x137/0x150 ret_from_fork+0x1f/0x30 -> #0 (&delayed_node->mutex){+.+.}-{3:3}: check_prev_add+0x98/0xa20 validate_chain+0xa8c/0x2a00 __lock_acquire+0x56f/0xaa0 lock_acquire+0xa3/0x440 __mutex_lock+0xa0/0xaf0 __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] btrfs_evict_inode+0x3bf/0x560 [btrfs] evict+0xd6/0x1c0 dispose_list+0x48/0x70 prune_icache_sb+0x54/0x80 super_cache_scan+0x121/0x1a0 do_shrink_slab+0x175/0x420 shrink_slab+0xb1/0x2e0 shrink_node+0x192/0x600 balance_pgdat+0x31f/0x750 kswapd+0x206/0x510 kthread+0x137/0x150 ret_from_fork+0x1f/0x30 other info that might help us debug this: Chain exists of: &delayed_node->mutex --> &fs_info->chunk_mutex --> fs_reclaim Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(fs_reclaim); lock(&fs_info->chunk_mutex); lock(fs_reclaim); lock(&delayed_node->mutex); *** DEADLOCK *** 3 locks held by kswapd0/75: #0: ffffffff8b0c8040 (fs_reclaim){+.+.}-{0:0}, at: __fs_reclaim_acquire+0x5/0x30 #1: ffffffff8b0b50b8 (shrinker_rwsem){++++}-{3:3}, at: shrink_slab+0x54/0x2e0 #2: ffffa040e057c0e8 (&type->s_umount_key#26){++++}-{3:3}, at: trylock_super+0x16/0x50 stack backtrace: CPU: 2 PID: 75 Comm: kswapd0 Not tainted 5.8.0-rc6-default+ #1191 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.0-59-gc9ba527-rebuilt.opensuse.org 04/01/2014 Call Trace: dump_stack+0x78/0xa0 check_noncircular+0x16f/0x190 check_prev_add+0x98/0xa20 validate_chain+0xa8c/0x2a00 __lock_acquire+0x56f/0xaa0 lock_acquire+0xa3/0x440 ? __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] __mutex_lock+0xa0/0xaf0 ? __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] ? __lock_acquire+0x56f/0xaa0 ? __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] ? lock_acquire+0xa3/0x440 ? btrfs_evict_inode+0x138/0x560 [btrfs] ? btrfs_evict_inode+0x2fe/0x560 [btrfs] ? __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] __btrfs_release_delayed_node.part.0+0x3f/0x310 [btrfs] btrfs_evict_inode+0x3bf/0x560 [btrfs] evict+0xd6/0x1c0 dispose_list+0x48/0x70 prune_icache_sb+0x54/0x80 super_cache_scan+0x121/0x1a0 do_shrink_slab+0x175/0x420 shrink_slab+0xb1/0x2e0 shrink_node+0x192/0x600 balance_pgdat+0x31f/0x750 kswapd+0x206/0x510 ? _raw_spin_unlock_irqrestore+0x3e/0x50 ? finish_wait+0x90/0x90 ? balance_pgdat+0x750/0x750 kthread+0x137/0x150 ? kthread_stop+0x2a0/0x2a0 ret_from_fork+0x1f/0x30 This is because we're holding the chunk_mutex while adding this device and adding its sysfs entries. We actually hold different locks in different places when calling this function, the dev_replace semaphore for instance in dev replace, so instead of moving this call around simply wrap it's operations in NOFS. CC: stable@vger.kernel.org # 4.14+ Reported-by: David Sterba <dsterba@suse.com> Signed-off-by: Josef Bacik <josef@toxicpanda.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2020-07-21 14:17:50 +00:00
memalloc_nofs_restore(nofs_flag);
return ret;
}
static int btrfs_sysfs_add_fs_devices(struct btrfs_fs_devices *fs_devices)
{
int ret;
struct btrfs_device *device;
struct btrfs_fs_devices *seed;
list_for_each_entry(device, &fs_devices->devices, dev_list) {
ret = btrfs_sysfs_add_device(device);
if (ret)
goto fail;
}
list_for_each_entry(seed, &fs_devices->seed_list, seed_list) {
list_for_each_entry(device, &seed->devices, dev_list) {
ret = btrfs_sysfs_add_device(device);
if (ret)
goto fail;
}
}
return 0;
fail:
btrfs_sysfs_remove_fs_devices(fs_devices);
return ret;
}
void btrfs_kobject_uevent(struct block_device *bdev, enum kobject_action action)
{
int ret;
ret = kobject_uevent(&disk_to_dev(bdev->bd_disk)->kobj, action);
if (ret)
pr_warn("BTRFS: Sending event '%d' to kobject: '%s' (%p): failed\n",
action, kobject_name(&disk_to_dev(bdev->bd_disk)->kobj),
&disk_to_dev(bdev->bd_disk)->kobj);
}
void btrfs_sysfs_update_sprout_fsid(struct btrfs_fs_devices *fs_devices)
{
char fsid_buf[BTRFS_UUID_UNPARSED_SIZE];
/*
* Sprouting changes fsid of the mounted filesystem, rename the fsid
* directory
*/
snprintf(fsid_buf, BTRFS_UUID_UNPARSED_SIZE, "%pU", fs_devices->fsid);
if (kobject_rename(&fs_devices->fsid_kobj, fsid_buf))
btrfs_warn(fs_devices->fs_info,
"sysfs: failed to create fsid for sprout");
}
btrfs: sysfs, add devid/dev_state kobject and device attributes New sysfs attributes that track the filesystem status of devices, stored in the per-filesystem directory in /sys/fs/btrfs/FSID/devinfo . There's a directory for each device, with name corresponding to the numerical device id. in_fs_metadata - device is in the list of fs metadata missing - device is missing (no device node or block device) replace_target - device is target of replace writeable - writes from fs are allowed These attributes reflect the state of the device::dev_state and created at mount time. Sample output: $ pwd /sys/fs/btrfs/6e1961f1-5918-4ecc-a22f-948897b409f7/devinfo/1/ $ ls in_fs_metadata missing replace_target writeable $ cat missing 0 The output from these attributes are 0 or 1. 0 indicates unset and 1 indicates set. These attributes are readonly. It is observed that the device delete thread and sysfs read thread will not race because the delete thread calls sysfs kobject_put() which in turn waits for existing sysfs read to complete. Note for device replace devid swap: During the replace the target device temporarily assumes devid 0 before assigning the devid of the soruce device. In btrfs_dev_replace_finishing() we remove source sysfs devid using the function btrfs_sysfs_remove_devices_attr(), so after that call kobject_rename() to update the devid in the sysfs. This adds and calls btrfs_sysfs_update_devid() helper function to update the device id. Signed-off-by: Anand Jain <anand.jain@oracle.com> Reviewed-by: David Sterba <dsterba@suse.com> [ update changelog ] Signed-off-by: David Sterba <dsterba@suse.com>
2020-01-06 11:38:31 +00:00
void btrfs_sysfs_update_devid(struct btrfs_device *device)
{
char tmp[24];
snprintf(tmp, sizeof(tmp), "%llu", device->devid);
if (kobject_rename(&device->devid_kobj, tmp))
btrfs_warn(device->fs_devices->fs_info,
"sysfs: failed to update devid for %llu",
device->devid);
}
/* /sys/fs/btrfs/ entry */
static struct kset *btrfs_kset;
/*
* Creates:
* /sys/fs/btrfs/UUID
*
* Can be called by the device discovery thread.
*/
int btrfs_sysfs_add_fsid(struct btrfs_fs_devices *fs_devs)
{
int error;
init_completion(&fs_devs->kobj_unregister);
fs_devs->fsid_kobj.kset = btrfs_kset;
error = kobject_init_and_add(&fs_devs->fsid_kobj, &btrfs_ktype, NULL,
"%pU", fs_devs->fsid);
if (error) {
kobject_put(&fs_devs->fsid_kobj);
return error;
}
fs_devs->devices_kobj = kobject_create_and_add("devices",
&fs_devs->fsid_kobj);
if (!fs_devs->devices_kobj) {
btrfs_err(fs_devs->fs_info,
"failed to init sysfs device interface");
btrfs_sysfs_remove_fsid(fs_devs);
return -ENOMEM;
}
fs_devs->devinfo_kobj = kobject_create_and_add("devinfo",
&fs_devs->fsid_kobj);
if (!fs_devs->devinfo_kobj) {
btrfs_err(fs_devs->fs_info,
"failed to init sysfs devinfo kobject");
btrfs_sysfs_remove_fsid(fs_devs);
return -ENOMEM;
}
return 0;
}
int btrfs_sysfs_add_mounted(struct btrfs_fs_info *fs_info)
{
int error;
struct btrfs_fs_devices *fs_devs = fs_info->fs_devices;
struct kobject *fsid_kobj = &fs_devs->fsid_kobj;
error = btrfs_sysfs_add_fs_devices(fs_devs);
if (error)
return error;
error = sysfs_create_files(fsid_kobj, btrfs_attrs);
if (error) {
btrfs_sysfs_remove_fs_devices(fs_devs);
return error;
}
error = sysfs_create_group(fsid_kobj,
&btrfs_feature_attr_group);
if (error)
goto failure;
#ifdef CONFIG_BTRFS_DEBUG
fs_info->debug_kobj = kobject_create_and_add("debug", fsid_kobj);
if (!fs_info->debug_kobj) {
error = -ENOMEM;
goto failure;
}
error = sysfs_create_files(fs_info->debug_kobj, btrfs_debug_mount_attrs);
if (error)
goto failure;
#endif
/* Discard directory */
fs_info->discard_kobj = kobject_create_and_add("discard", fsid_kobj);
if (!fs_info->discard_kobj) {
error = -ENOMEM;
goto failure;
}
error = sysfs_create_files(fs_info->discard_kobj, discard_attrs);
if (error)
goto failure;
error = addrm_unknown_feature_attrs(fs_info, true);
if (error)
goto failure;
error = sysfs_create_link(fsid_kobj, &fs_info->sb->s_bdi->dev->kobj, "bdi");
if (error)
goto failure;
fs_info->space_info_kobj = kobject_create_and_add("allocation",
fsid_kobj);
if (!fs_info->space_info_kobj) {
error = -ENOMEM;
goto failure;
}
error = sysfs_create_files(fs_info->space_info_kobj, allocation_attrs);
if (error)
goto failure;
return 0;
failure:
btrfs_sysfs_remove_mounted(fs_info);
return error;
}
static ssize_t qgroup_enabled_show(struct kobject *qgroups_kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(qgroups_kobj->parent);
bool enabled;
spin_lock(&fs_info->qgroup_lock);
enabled = fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_ON;
spin_unlock(&fs_info->qgroup_lock);
return sysfs_emit(buf, "%d\n", enabled);
}
BTRFS_ATTR(qgroups, enabled, qgroup_enabled_show);
static ssize_t qgroup_mode_show(struct kobject *qgroups_kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(qgroups_kobj->parent);
ssize_t ret = 0;
spin_lock(&fs_info->qgroup_lock);
ASSERT(btrfs_qgroup_enabled(fs_info));
switch (btrfs_qgroup_mode(fs_info)) {
case BTRFS_QGROUP_MODE_FULL:
ret = sysfs_emit(buf, "qgroup\n");
break;
case BTRFS_QGROUP_MODE_SIMPLE:
ret = sysfs_emit(buf, "squota\n");
break;
default:
btrfs_warn(fs_info, "unexpected qgroup mode %d\n",
btrfs_qgroup_mode(fs_info));
break;
}
spin_unlock(&fs_info->qgroup_lock);
return ret;
}
BTRFS_ATTR(qgroups, mode, qgroup_mode_show);
static ssize_t qgroup_inconsistent_show(struct kobject *qgroups_kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(qgroups_kobj->parent);
bool inconsistent;
spin_lock(&fs_info->qgroup_lock);
inconsistent = (fs_info->qgroup_flags & BTRFS_QGROUP_STATUS_FLAG_INCONSISTENT);
spin_unlock(&fs_info->qgroup_lock);
return sysfs_emit(buf, "%d\n", inconsistent);
}
BTRFS_ATTR(qgroups, inconsistent, qgroup_inconsistent_show);
static ssize_t qgroup_drop_subtree_thres_show(struct kobject *qgroups_kobj,
struct kobj_attribute *a,
char *buf)
{
struct btrfs_fs_info *fs_info = to_fs_info(qgroups_kobj->parent);
u8 result;
spin_lock(&fs_info->qgroup_lock);
result = fs_info->qgroup_drop_subtree_thres;
spin_unlock(&fs_info->qgroup_lock);
return sysfs_emit(buf, "%d\n", result);
}
static ssize_t qgroup_drop_subtree_thres_store(struct kobject *qgroups_kobj,
struct kobj_attribute *a,
const char *buf, size_t len)
{
struct btrfs_fs_info *fs_info = to_fs_info(qgroups_kobj->parent);
u8 new_thres;
int ret;
ret = kstrtou8(buf, 10, &new_thres);
if (ret)
return -EINVAL;
if (new_thres > BTRFS_MAX_LEVEL)
return -EINVAL;
spin_lock(&fs_info->qgroup_lock);
fs_info->qgroup_drop_subtree_thres = new_thres;
spin_unlock(&fs_info->qgroup_lock);
return len;
}
BTRFS_ATTR_RW(qgroups, drop_subtree_threshold, qgroup_drop_subtree_thres_show,
qgroup_drop_subtree_thres_store);
/*
* Qgroups global info
*
* Path: /sys/fs/btrfs/<uuid>/qgroups/
*/
static struct attribute *qgroups_attrs[] = {
BTRFS_ATTR_PTR(qgroups, enabled),
BTRFS_ATTR_PTR(qgroups, inconsistent),
BTRFS_ATTR_PTR(qgroups, drop_subtree_threshold),
BTRFS_ATTR_PTR(qgroups, mode),
NULL
};
ATTRIBUTE_GROUPS(qgroups);
static void qgroups_release(struct kobject *kobj)
{
kfree(kobj);
}
static const struct kobj_type qgroups_ktype = {
.sysfs_ops = &kobj_sysfs_ops,
.default_groups = qgroups_groups,
.release = qgroups_release,
};
static inline struct btrfs_fs_info *qgroup_kobj_to_fs_info(struct kobject *kobj)
{
return to_fs_info(kobj->parent->parent);
}
#define QGROUP_ATTR(_member, _show_name) \
static ssize_t btrfs_qgroup_show_##_member(struct kobject *qgroup_kobj, \
struct kobj_attribute *a, \
char *buf) \
{ \
struct btrfs_fs_info *fs_info = qgroup_kobj_to_fs_info(qgroup_kobj); \
struct btrfs_qgroup *qgroup = container_of(qgroup_kobj, \
struct btrfs_qgroup, kobj); \
return btrfs_show_u64(&qgroup->_member, &fs_info->qgroup_lock, buf); \
} \
BTRFS_ATTR(qgroup, _show_name, btrfs_qgroup_show_##_member)
#define QGROUP_RSV_ATTR(_name, _type) \
static ssize_t btrfs_qgroup_rsv_show_##_name(struct kobject *qgroup_kobj, \
struct kobj_attribute *a, \
char *buf) \
{ \
struct btrfs_fs_info *fs_info = qgroup_kobj_to_fs_info(qgroup_kobj); \
struct btrfs_qgroup *qgroup = container_of(qgroup_kobj, \
struct btrfs_qgroup, kobj); \
return btrfs_show_u64(&qgroup->rsv.values[_type], \
&fs_info->qgroup_lock, buf); \
} \
BTRFS_ATTR(qgroup, rsv_##_name, btrfs_qgroup_rsv_show_##_name)
QGROUP_ATTR(rfer, referenced);
QGROUP_ATTR(excl, exclusive);
QGROUP_ATTR(max_rfer, max_referenced);
QGROUP_ATTR(max_excl, max_exclusive);
QGROUP_ATTR(lim_flags, limit_flags);
QGROUP_RSV_ATTR(data, BTRFS_QGROUP_RSV_DATA);
QGROUP_RSV_ATTR(meta_pertrans, BTRFS_QGROUP_RSV_META_PERTRANS);
QGROUP_RSV_ATTR(meta_prealloc, BTRFS_QGROUP_RSV_META_PREALLOC);
/*
* Qgroup information.
*
* Path: /sys/fs/btrfs/<uuid>/qgroups/<level>_<qgroupid>/
*/
static struct attribute *qgroup_attrs[] = {
BTRFS_ATTR_PTR(qgroup, referenced),
BTRFS_ATTR_PTR(qgroup, exclusive),
BTRFS_ATTR_PTR(qgroup, max_referenced),
BTRFS_ATTR_PTR(qgroup, max_exclusive),
BTRFS_ATTR_PTR(qgroup, limit_flags),
BTRFS_ATTR_PTR(qgroup, rsv_data),
BTRFS_ATTR_PTR(qgroup, rsv_meta_pertrans),
BTRFS_ATTR_PTR(qgroup, rsv_meta_prealloc),
NULL
};
ATTRIBUTE_GROUPS(qgroup);
static void qgroup_release(struct kobject *kobj)
{
struct btrfs_qgroup *qgroup = container_of(kobj, struct btrfs_qgroup, kobj);
memset(&qgroup->kobj, 0, sizeof(*kobj));
}
static const struct kobj_type qgroup_ktype = {
.sysfs_ops = &kobj_sysfs_ops,
.release = qgroup_release,
.default_groups = qgroup_groups,
};
int btrfs_sysfs_add_one_qgroup(struct btrfs_fs_info *fs_info,
struct btrfs_qgroup *qgroup)
{
struct kobject *qgroups_kobj = fs_info->qgroups_kobj;
int ret;
if (btrfs_is_testing(fs_info))
return 0;
if (qgroup->kobj.state_initialized)
return 0;
if (!qgroups_kobj)
return -EINVAL;
ret = kobject_init_and_add(&qgroup->kobj, &qgroup_ktype, qgroups_kobj,
"%hu_%llu", btrfs_qgroup_level(qgroup->qgroupid),
btrfs_qgroup_subvolid(qgroup->qgroupid));
if (ret < 0)
kobject_put(&qgroup->kobj);
return ret;
}
void btrfs_sysfs_del_qgroups(struct btrfs_fs_info *fs_info)
{
struct btrfs_qgroup *qgroup;
struct btrfs_qgroup *next;
if (btrfs_is_testing(fs_info))
return;
rbtree_postorder_for_each_entry_safe(qgroup, next,
&fs_info->qgroup_tree, node)
btrfs_sysfs_del_one_qgroup(fs_info, qgroup);
btrfs: sysfs: fix NULL pointer dereference at btrfs_sysfs_del_qgroups() [BUG] Unmounting a btrfs filesystem with quota disabled will cause the following NULL pointer dereference: BTRFS info (device dm-5): has skinny extents BUG: kernel NULL pointer dereference, address: 0000000000000018 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page CPU: 7 PID: 637 Comm: umount Not tainted 5.8.0-rc7-next-20200731-custom #76 RIP: 0010:kobject_del+0x6/0x20 Call Trace: btrfs_sysfs_del_qgroups+0xac/0xf0 [btrfs] btrfs_free_qgroup_config+0x63/0x70 [btrfs] close_ctree+0x1f5/0x323 [btrfs] btrfs_put_super+0x15/0x17 [btrfs] generic_shutdown_super+0x72/0x110 kill_anon_super+0x18/0x30 btrfs_kill_super+0x17/0x30 [btrfs] deactivate_locked_super+0x3b/0xa0 deactivate_super+0x40/0x50 cleanup_mnt+0x135/0x190 __cleanup_mnt+0x12/0x20 task_work_run+0x64/0xb0 exit_to_user_mode_prepare+0x18a/0x190 syscall_exit_to_user_mode+0x4f/0x270 do_syscall_64+0x45/0x50 entry_SYSCALL_64_after_hwframe+0x44/0xa9 ---[ end trace 37b7adca5c1d5c5d ]--- [CAUSE] Commit 079ad2fb4bf9 ("kobject: Avoid premature parent object freeing in kobject_cleanup()") changed kobject_del() that it no longer accepts NULL pointer. Before that commit, kobject_del() and kobject_put() all accept NULL pointers and just ignore such NULL pointers. But that mentioned commit needs to access the parent node, killing the old NULL pointer behavior. Unfortunately btrfs is relying on that hidden feature thus we will trigger such NULL pointer dereference. [FIX] Instead of just saving several lines, do proper fs_info->qgroups_kobj check before calling kobject_del() and kobject_put(). Reviewed-by: Nikolay Borisov <nborisov@suse.com> Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2020-08-03 06:20:11 +00:00
if (fs_info->qgroups_kobj) {
kobject_del(fs_info->qgroups_kobj);
kobject_put(fs_info->qgroups_kobj);
fs_info->qgroups_kobj = NULL;
}
}
/* Called when qgroups get initialized, thus there is no need for locking */
int btrfs_sysfs_add_qgroups(struct btrfs_fs_info *fs_info)
{
struct kobject *fsid_kobj = &fs_info->fs_devices->fsid_kobj;
struct btrfs_qgroup *qgroup;
struct btrfs_qgroup *next;
int ret = 0;
if (btrfs_is_testing(fs_info))
return 0;
ASSERT(fsid_kobj);
if (fs_info->qgroups_kobj)
return 0;
fs_info->qgroups_kobj = kzalloc(sizeof(struct kobject), GFP_KERNEL);
if (!fs_info->qgroups_kobj)
return -ENOMEM;
ret = kobject_init_and_add(fs_info->qgroups_kobj, &qgroups_ktype,
fsid_kobj, "qgroups");
if (ret < 0)
goto out;
rbtree_postorder_for_each_entry_safe(qgroup, next,
&fs_info->qgroup_tree, node) {
ret = btrfs_sysfs_add_one_qgroup(fs_info, qgroup);
if (ret < 0)
goto out;
}
out:
if (ret < 0)
btrfs_sysfs_del_qgroups(fs_info);
return ret;
}
void btrfs_sysfs_del_one_qgroup(struct btrfs_fs_info *fs_info,
struct btrfs_qgroup *qgroup)
{
if (btrfs_is_testing(fs_info))
return;
if (qgroup->kobj.state_initialized) {
kobject_del(&qgroup->kobj);
kobject_put(&qgroup->kobj);
}
}
/*
* Change per-fs features in /sys/fs/btrfs/UUID/features to match current
* values in superblock. Call after any changes to incompat/compat_ro flags
*/
btrfs: sysfs: update fs features directory asynchronously [BUG] Since the introduction of per-fs feature sysfs interface (/sys/fs/btrfs/<UUID>/features/), the content of that directory is never updated. Thus for the following case, that directory will not show the new features like RAID56: # mkfs.btrfs -f $dev1 $dev2 $dev3 # mount $dev1 $mnt # btrfs balance start -f -mconvert=raid5 $mnt # ls /sys/fs/btrfs/$uuid/features/ extended_iref free_space_tree no_holes skinny_metadata While after unmount and mount, we got the correct features: # umount $mnt # mount $dev1 $mnt # ls /sys/fs/btrfs/$uuid/features/ extended_iref free_space_tree no_holes raid56 skinny_metadata [CAUSE] Because we never really try to update the content of per-fs features/ directory. We had an attempt to update the features directory dynamically in commit 14e46e04958d ("btrfs: synchronize incompat feature bits with sysfs files"), but unfortunately it get reverted in commit e410e34fad91 ("Revert "btrfs: synchronize incompat feature bits with sysfs files""). The problem in the original patch is, in the context of btrfs_create_chunk(), we can not afford to update the sysfs group. The exported but never utilized function, btrfs_sysfs_feature_update() is the leftover of such attempt. As even if we go sysfs_update_group(), new files will need extra memory allocation, and we have no way to specify the sysfs update to go GFP_NOFS. [FIX] This patch will address the old problem by doing asynchronous sysfs update in the cleaner thread. This involves the following changes: - Make __btrfs_(set|clear)_fs_(incompat|compat_ro) helpers to set BTRFS_FS_FEATURE_CHANGED flag when needed - Update btrfs_sysfs_feature_update() to use sysfs_update_group() And drop unnecessary arguments. - Call btrfs_sysfs_feature_update() in cleaner_kthread If we have the BTRFS_FS_FEATURE_CHANGED flag set. - Wake up cleaner_kthread in btrfs_commit_transaction if we have BTRFS_FS_FEATURE_CHANGED flag By this, all the previously dangerous call sites like btrfs_create_chunk() need no new changes, as above helpers would have already set the BTRFS_FS_FEATURE_CHANGED flag. The real work happens at cleaner_kthread, thus we pay the cost of delaying the update to sysfs directory, but the delayed time should be small enough that end user can not distinguish though it might get delayed if the cleaner thread is busy with removing subvolumes or defrag. CC: stable@vger.kernel.org # 4.14+ Reviewed-by: Anand Jain <anand.jain@oracle.com> Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-13 11:11:39 +00:00
void btrfs_sysfs_feature_update(struct btrfs_fs_info *fs_info)
{
struct kobject *fsid_kobj;
btrfs: sysfs: update fs features directory asynchronously [BUG] Since the introduction of per-fs feature sysfs interface (/sys/fs/btrfs/<UUID>/features/), the content of that directory is never updated. Thus for the following case, that directory will not show the new features like RAID56: # mkfs.btrfs -f $dev1 $dev2 $dev3 # mount $dev1 $mnt # btrfs balance start -f -mconvert=raid5 $mnt # ls /sys/fs/btrfs/$uuid/features/ extended_iref free_space_tree no_holes skinny_metadata While after unmount and mount, we got the correct features: # umount $mnt # mount $dev1 $mnt # ls /sys/fs/btrfs/$uuid/features/ extended_iref free_space_tree no_holes raid56 skinny_metadata [CAUSE] Because we never really try to update the content of per-fs features/ directory. We had an attempt to update the features directory dynamically in commit 14e46e04958d ("btrfs: synchronize incompat feature bits with sysfs files"), but unfortunately it get reverted in commit e410e34fad91 ("Revert "btrfs: synchronize incompat feature bits with sysfs files""). The problem in the original patch is, in the context of btrfs_create_chunk(), we can not afford to update the sysfs group. The exported but never utilized function, btrfs_sysfs_feature_update() is the leftover of such attempt. As even if we go sysfs_update_group(), new files will need extra memory allocation, and we have no way to specify the sysfs update to go GFP_NOFS. [FIX] This patch will address the old problem by doing asynchronous sysfs update in the cleaner thread. This involves the following changes: - Make __btrfs_(set|clear)_fs_(incompat|compat_ro) helpers to set BTRFS_FS_FEATURE_CHANGED flag when needed - Update btrfs_sysfs_feature_update() to use sysfs_update_group() And drop unnecessary arguments. - Call btrfs_sysfs_feature_update() in cleaner_kthread If we have the BTRFS_FS_FEATURE_CHANGED flag set. - Wake up cleaner_kthread in btrfs_commit_transaction if we have BTRFS_FS_FEATURE_CHANGED flag By this, all the previously dangerous call sites like btrfs_create_chunk() need no new changes, as above helpers would have already set the BTRFS_FS_FEATURE_CHANGED flag. The real work happens at cleaner_kthread, thus we pay the cost of delaying the update to sysfs directory, but the delayed time should be small enough that end user can not distinguish though it might get delayed if the cleaner thread is busy with removing subvolumes or defrag. CC: stable@vger.kernel.org # 4.14+ Reviewed-by: Anand Jain <anand.jain@oracle.com> Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-13 11:11:39 +00:00
int ret;
if (!fs_info)
return;
btrfs: sysfs: update fs features directory asynchronously [BUG] Since the introduction of per-fs feature sysfs interface (/sys/fs/btrfs/<UUID>/features/), the content of that directory is never updated. Thus for the following case, that directory will not show the new features like RAID56: # mkfs.btrfs -f $dev1 $dev2 $dev3 # mount $dev1 $mnt # btrfs balance start -f -mconvert=raid5 $mnt # ls /sys/fs/btrfs/$uuid/features/ extended_iref free_space_tree no_holes skinny_metadata While after unmount and mount, we got the correct features: # umount $mnt # mount $dev1 $mnt # ls /sys/fs/btrfs/$uuid/features/ extended_iref free_space_tree no_holes raid56 skinny_metadata [CAUSE] Because we never really try to update the content of per-fs features/ directory. We had an attempt to update the features directory dynamically in commit 14e46e04958d ("btrfs: synchronize incompat feature bits with sysfs files"), but unfortunately it get reverted in commit e410e34fad91 ("Revert "btrfs: synchronize incompat feature bits with sysfs files""). The problem in the original patch is, in the context of btrfs_create_chunk(), we can not afford to update the sysfs group. The exported but never utilized function, btrfs_sysfs_feature_update() is the leftover of such attempt. As even if we go sysfs_update_group(), new files will need extra memory allocation, and we have no way to specify the sysfs update to go GFP_NOFS. [FIX] This patch will address the old problem by doing asynchronous sysfs update in the cleaner thread. This involves the following changes: - Make __btrfs_(set|clear)_fs_(incompat|compat_ro) helpers to set BTRFS_FS_FEATURE_CHANGED flag when needed - Update btrfs_sysfs_feature_update() to use sysfs_update_group() And drop unnecessary arguments. - Call btrfs_sysfs_feature_update() in cleaner_kthread If we have the BTRFS_FS_FEATURE_CHANGED flag set. - Wake up cleaner_kthread in btrfs_commit_transaction if we have BTRFS_FS_FEATURE_CHANGED flag By this, all the previously dangerous call sites like btrfs_create_chunk() need no new changes, as above helpers would have already set the BTRFS_FS_FEATURE_CHANGED flag. The real work happens at cleaner_kthread, thus we pay the cost of delaying the update to sysfs directory, but the delayed time should be small enough that end user can not distinguish though it might get delayed if the cleaner thread is busy with removing subvolumes or defrag. CC: stable@vger.kernel.org # 4.14+ Reviewed-by: Anand Jain <anand.jain@oracle.com> Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-13 11:11:39 +00:00
fsid_kobj = &fs_info->fs_devices->fsid_kobj;
if (!fsid_kobj->state_initialized)
return;
btrfs: sysfs: update fs features directory asynchronously [BUG] Since the introduction of per-fs feature sysfs interface (/sys/fs/btrfs/<UUID>/features/), the content of that directory is never updated. Thus for the following case, that directory will not show the new features like RAID56: # mkfs.btrfs -f $dev1 $dev2 $dev3 # mount $dev1 $mnt # btrfs balance start -f -mconvert=raid5 $mnt # ls /sys/fs/btrfs/$uuid/features/ extended_iref free_space_tree no_holes skinny_metadata While after unmount and mount, we got the correct features: # umount $mnt # mount $dev1 $mnt # ls /sys/fs/btrfs/$uuid/features/ extended_iref free_space_tree no_holes raid56 skinny_metadata [CAUSE] Because we never really try to update the content of per-fs features/ directory. We had an attempt to update the features directory dynamically in commit 14e46e04958d ("btrfs: synchronize incompat feature bits with sysfs files"), but unfortunately it get reverted in commit e410e34fad91 ("Revert "btrfs: synchronize incompat feature bits with sysfs files""). The problem in the original patch is, in the context of btrfs_create_chunk(), we can not afford to update the sysfs group. The exported but never utilized function, btrfs_sysfs_feature_update() is the leftover of such attempt. As even if we go sysfs_update_group(), new files will need extra memory allocation, and we have no way to specify the sysfs update to go GFP_NOFS. [FIX] This patch will address the old problem by doing asynchronous sysfs update in the cleaner thread. This involves the following changes: - Make __btrfs_(set|clear)_fs_(incompat|compat_ro) helpers to set BTRFS_FS_FEATURE_CHANGED flag when needed - Update btrfs_sysfs_feature_update() to use sysfs_update_group() And drop unnecessary arguments. - Call btrfs_sysfs_feature_update() in cleaner_kthread If we have the BTRFS_FS_FEATURE_CHANGED flag set. - Wake up cleaner_kthread in btrfs_commit_transaction if we have BTRFS_FS_FEATURE_CHANGED flag By this, all the previously dangerous call sites like btrfs_create_chunk() need no new changes, as above helpers would have already set the BTRFS_FS_FEATURE_CHANGED flag. The real work happens at cleaner_kthread, thus we pay the cost of delaying the update to sysfs directory, but the delayed time should be small enough that end user can not distinguish though it might get delayed if the cleaner thread is busy with removing subvolumes or defrag. CC: stable@vger.kernel.org # 4.14+ Reviewed-by: Anand Jain <anand.jain@oracle.com> Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com> Signed-off-by: David Sterba <dsterba@suse.com>
2023-01-13 11:11:39 +00:00
ret = sysfs_update_group(fsid_kobj, &btrfs_feature_attr_group);
if (ret < 0)
btrfs_warn(fs_info,
"failed to update /sys/fs/btrfs/%pU/features: %d",
fs_info->fs_devices->fsid, ret);
}
int __init btrfs_init_sysfs(void)
{
int ret;
btrfs_kset = kset_create_and_add("btrfs", NULL, fs_kobj);
if (!btrfs_kset)
return -ENOMEM;
init_feature_attrs();
ret = sysfs_create_group(&btrfs_kset->kobj, &btrfs_feature_attr_group);
Btrfs: add missing cleanup on sysfs init failure If we failed during initialization of sysfs, we weren't unregistering the top level btrfs sysfs entry nor the debugfs stuff. Not unregistering the top level sysfs entry makes future attempts to reload the btrfs module impossible and the following is reported in dmesg: [ 2246.451296] WARNING: CPU: 3 PID: 10999 at fs/sysfs/dir.c:486 sysfs_warn_dup+0x91/0xb0() [ 2246.451298] sysfs: cannot create duplicate filename '/fs/btrfs' [ 2246.451298] Modules linked in: btrfs(+) raid6_pq xor bnep rfcomm bluetooth binfmt_misc nfsd auth_rpcgss oid_registry nfs_acl nfs lockd fscache sunrpc parport_pc parport psmouse serio_raw pcspkr evbug i2c_piix4 e1000 floppy [last unloaded: btrfs] [ 2246.451310] CPU: 3 PID: 10999 Comm: modprobe Tainted: G W 3.13.0-fdm-btrfs-next-24+ #7 [ 2246.451311] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 2246.451312] 0000000000000009 ffff8800d353fa08 ffffffff816f1da6 0000000000000410 [ 2246.451314] ffff8800d353fa58 ffff8800d353fa48 ffffffff8104a32c ffff88020821a290 [ 2246.451316] ffff88020821a290 ffff88020821a290 ffff8802148f0000 ffff8800d353fb80 [ 2246.451318] Call Trace: [ 2246.451322] [<ffffffff816f1da6>] dump_stack+0x4e/0x68 [ 2246.451324] [<ffffffff8104a32c>] warn_slowpath_common+0x8c/0xc0 [ 2246.451325] [<ffffffff8104a416>] warn_slowpath_fmt+0x46/0x50 [ 2246.451328] [<ffffffff81367dc5>] ? strlcat+0x65/0x90 (....) This fixes the following change: btrfs: add simple debugfs interface commit 1bae30982bc86ab66d61ccb6e22792593b45d44d Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: Chris Mason <clm@fb.com>
2015-01-23 18:27:00 +00:00
if (ret)
goto out2;
ret = sysfs_merge_group(&btrfs_kset->kobj,
&btrfs_static_feature_attr_group);
if (ret)
goto out_remove_group;
Btrfs: add missing cleanup on sysfs init failure If we failed during initialization of sysfs, we weren't unregistering the top level btrfs sysfs entry nor the debugfs stuff. Not unregistering the top level sysfs entry makes future attempts to reload the btrfs module impossible and the following is reported in dmesg: [ 2246.451296] WARNING: CPU: 3 PID: 10999 at fs/sysfs/dir.c:486 sysfs_warn_dup+0x91/0xb0() [ 2246.451298] sysfs: cannot create duplicate filename '/fs/btrfs' [ 2246.451298] Modules linked in: btrfs(+) raid6_pq xor bnep rfcomm bluetooth binfmt_misc nfsd auth_rpcgss oid_registry nfs_acl nfs lockd fscache sunrpc parport_pc parport psmouse serio_raw pcspkr evbug i2c_piix4 e1000 floppy [last unloaded: btrfs] [ 2246.451310] CPU: 3 PID: 10999 Comm: modprobe Tainted: G W 3.13.0-fdm-btrfs-next-24+ #7 [ 2246.451311] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 2246.451312] 0000000000000009 ffff8800d353fa08 ffffffff816f1da6 0000000000000410 [ 2246.451314] ffff8800d353fa58 ffff8800d353fa48 ffffffff8104a32c ffff88020821a290 [ 2246.451316] ffff88020821a290 ffff88020821a290 ffff8802148f0000 ffff8800d353fb80 [ 2246.451318] Call Trace: [ 2246.451322] [<ffffffff816f1da6>] dump_stack+0x4e/0x68 [ 2246.451324] [<ffffffff8104a32c>] warn_slowpath_common+0x8c/0xc0 [ 2246.451325] [<ffffffff8104a416>] warn_slowpath_fmt+0x46/0x50 [ 2246.451328] [<ffffffff81367dc5>] ? strlcat+0x65/0x90 (....) This fixes the following change: btrfs: add simple debugfs interface commit 1bae30982bc86ab66d61ccb6e22792593b45d44d Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: Chris Mason <clm@fb.com>
2015-01-23 18:27:00 +00:00
#ifdef CONFIG_BTRFS_DEBUG
ret = sysfs_create_group(&btrfs_kset->kobj, &btrfs_debug_feature_attr_group);
if (ret) {
sysfs_unmerge_group(&btrfs_kset->kobj,
&btrfs_static_feature_attr_group);
goto out_remove_group;
}
#endif
Btrfs: add missing cleanup on sysfs init failure If we failed during initialization of sysfs, we weren't unregistering the top level btrfs sysfs entry nor the debugfs stuff. Not unregistering the top level sysfs entry makes future attempts to reload the btrfs module impossible and the following is reported in dmesg: [ 2246.451296] WARNING: CPU: 3 PID: 10999 at fs/sysfs/dir.c:486 sysfs_warn_dup+0x91/0xb0() [ 2246.451298] sysfs: cannot create duplicate filename '/fs/btrfs' [ 2246.451298] Modules linked in: btrfs(+) raid6_pq xor bnep rfcomm bluetooth binfmt_misc nfsd auth_rpcgss oid_registry nfs_acl nfs lockd fscache sunrpc parport_pc parport psmouse serio_raw pcspkr evbug i2c_piix4 e1000 floppy [last unloaded: btrfs] [ 2246.451310] CPU: 3 PID: 10999 Comm: modprobe Tainted: G W 3.13.0-fdm-btrfs-next-24+ #7 [ 2246.451311] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 2246.451312] 0000000000000009 ffff8800d353fa08 ffffffff816f1da6 0000000000000410 [ 2246.451314] ffff8800d353fa58 ffff8800d353fa48 ffffffff8104a32c ffff88020821a290 [ 2246.451316] ffff88020821a290 ffff88020821a290 ffff8802148f0000 ffff8800d353fb80 [ 2246.451318] Call Trace: [ 2246.451322] [<ffffffff816f1da6>] dump_stack+0x4e/0x68 [ 2246.451324] [<ffffffff8104a32c>] warn_slowpath_common+0x8c/0xc0 [ 2246.451325] [<ffffffff8104a416>] warn_slowpath_fmt+0x46/0x50 [ 2246.451328] [<ffffffff81367dc5>] ? strlcat+0x65/0x90 (....) This fixes the following change: btrfs: add simple debugfs interface commit 1bae30982bc86ab66d61ccb6e22792593b45d44d Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: Chris Mason <clm@fb.com>
2015-01-23 18:27:00 +00:00
return 0;
out_remove_group:
sysfs_remove_group(&btrfs_kset->kobj, &btrfs_feature_attr_group);
Btrfs: add missing cleanup on sysfs init failure If we failed during initialization of sysfs, we weren't unregistering the top level btrfs sysfs entry nor the debugfs stuff. Not unregistering the top level sysfs entry makes future attempts to reload the btrfs module impossible and the following is reported in dmesg: [ 2246.451296] WARNING: CPU: 3 PID: 10999 at fs/sysfs/dir.c:486 sysfs_warn_dup+0x91/0xb0() [ 2246.451298] sysfs: cannot create duplicate filename '/fs/btrfs' [ 2246.451298] Modules linked in: btrfs(+) raid6_pq xor bnep rfcomm bluetooth binfmt_misc nfsd auth_rpcgss oid_registry nfs_acl nfs lockd fscache sunrpc parport_pc parport psmouse serio_raw pcspkr evbug i2c_piix4 e1000 floppy [last unloaded: btrfs] [ 2246.451310] CPU: 3 PID: 10999 Comm: modprobe Tainted: G W 3.13.0-fdm-btrfs-next-24+ #7 [ 2246.451311] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 2246.451312] 0000000000000009 ffff8800d353fa08 ffffffff816f1da6 0000000000000410 [ 2246.451314] ffff8800d353fa58 ffff8800d353fa48 ffffffff8104a32c ffff88020821a290 [ 2246.451316] ffff88020821a290 ffff88020821a290 ffff8802148f0000 ffff8800d353fb80 [ 2246.451318] Call Trace: [ 2246.451322] [<ffffffff816f1da6>] dump_stack+0x4e/0x68 [ 2246.451324] [<ffffffff8104a32c>] warn_slowpath_common+0x8c/0xc0 [ 2246.451325] [<ffffffff8104a416>] warn_slowpath_fmt+0x46/0x50 [ 2246.451328] [<ffffffff81367dc5>] ? strlcat+0x65/0x90 (....) This fixes the following change: btrfs: add simple debugfs interface commit 1bae30982bc86ab66d61ccb6e22792593b45d44d Signed-off-by: Filipe Manana <fdmanana@suse.com> Signed-off-by: Chris Mason <clm@fb.com>
2015-01-23 18:27:00 +00:00
out2:
kset_unregister(btrfs_kset);
return ret;
}
void __cold btrfs_exit_sysfs(void)
{
sysfs_unmerge_group(&btrfs_kset->kobj,
&btrfs_static_feature_attr_group);
sysfs_remove_group(&btrfs_kset->kobj, &btrfs_feature_attr_group);
#ifdef CONFIG_BTRFS_DEBUG
sysfs_remove_group(&btrfs_kset->kobj, &btrfs_debug_feature_attr_group);
#endif
kset_unregister(btrfs_kset);
}