diff options
Diffstat (limited to 'fs')
34 files changed, 867 insertions, 220 deletions
@@ -25,7 +25,6 @@ #include <linux/module.h> #include <linux/mempool.h> #include <linux/workqueue.h> -#include <linux/blktrace_api.h> #include <scsi/sg.h> /* for struct sg_iovec */ #include <trace/events/block.h> diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 0d50d49d990a..d28d29c95f7c 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -42,6 +42,8 @@ static struct extent_io_ops btree_extent_io_ops; static void end_workqueue_fn(struct btrfs_work *work); +static atomic_t btrfs_bdi_num = ATOMIC_INIT(0); + /* * end_io_wq structs are used to do processing in task context when an IO is * complete. This is used during reads to verify checksums, and it is used @@ -1342,12 +1344,25 @@ static void btrfs_unplug_io_fn(struct backing_dev_info *bdi, struct page *page) free_extent_map(em); } +/* + * If this fails, caller must call bdi_destroy() to get rid of the + * bdi again. + */ static int setup_bdi(struct btrfs_fs_info *info, struct backing_dev_info *bdi) { - bdi_init(bdi); + int err; + + bdi->capabilities = BDI_CAP_MAP_COPY; + err = bdi_init(bdi); + if (err) + return err; + + err = bdi_register(bdi, NULL, "btrfs-%d", + atomic_inc_return(&btrfs_bdi_num)); + if (err) + return err; + bdi->ra_pages = default_backing_dev_info.ra_pages; - bdi->state = 0; - bdi->capabilities = default_backing_dev_info.capabilities; bdi->unplug_io_fn = btrfs_unplug_io_fn; bdi->unplug_io_data = info; bdi->congested_fn = btrfs_congested_fn; @@ -1569,7 +1584,8 @@ struct btrfs_root *open_ctree(struct super_block *sb, fs_info->sb = sb; fs_info->max_extent = (u64)-1; fs_info->max_inline = 8192 * 1024; - setup_bdi(fs_info, &fs_info->bdi); + if (setup_bdi(fs_info, &fs_info->bdi)) + goto fail_bdi; fs_info->btree_inode = new_inode(sb); fs_info->btree_inode->i_ino = 1; fs_info->btree_inode->i_nlink = 1; @@ -1946,8 +1962,8 @@ fail_iput: btrfs_close_devices(fs_info->fs_devices); btrfs_mapping_tree_free(&fs_info->mapping_tree); +fail_bdi: bdi_destroy(&fs_info->bdi); - fail: kfree(extent_root); kfree(tree_root); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 2e177d7f4bb9..4e83457ea253 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -543,13 +543,13 @@ static noinline int commit_fs_roots(struct btrfs_trans_handle *trans, btrfs_free_log(trans, root); btrfs_update_reloc_root(trans, root); - if (root->commit_root == root->node) - continue; - - free_extent_buffer(root->commit_root); - root->commit_root = btrfs_root_node(root); + if (root->commit_root != root->node) { + free_extent_buffer(root->commit_root); + root->commit_root = btrfs_root_node(root); + btrfs_set_root_node(&root->root_item, + root->node); + } - btrfs_set_root_node(&root->root_item, root->node); err = btrfs_update_root(trans, fs_info->tree_root, &root->root_key, &root->root_item); diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c index 33a90120f6ad..4d74fc72c195 100644 --- a/fs/debugfs/file.c +++ b/fs/debugfs/file.c @@ -67,6 +67,8 @@ static int debugfs_u8_get(void *data, u64 *val) return 0; } DEFINE_SIMPLE_ATTRIBUTE(fops_u8, debugfs_u8_get, debugfs_u8_set, "%llu\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_u8_ro, debugfs_u8_get, NULL, "%llu\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_u8_wo, NULL, debugfs_u8_set, "%llu\n"); /** * debugfs_create_u8 - create a debugfs file that is used to read and write an unsigned 8-bit value @@ -95,6 +97,13 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u8, debugfs_u8_get, debugfs_u8_set, "%llu\n"); struct dentry *debugfs_create_u8(const char *name, mode_t mode, struct dentry *parent, u8 *value) { + /* if there are no write bits set, make read only */ + if (!(mode & S_IWUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_u8_ro); + /* if there are no read bits set, make write only */ + if (!(mode & S_IRUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_u8_wo); + return debugfs_create_file(name, mode, parent, value, &fops_u8); } EXPORT_SYMBOL_GPL(debugfs_create_u8); @@ -110,6 +119,8 @@ static int debugfs_u16_get(void *data, u64 *val) return 0; } DEFINE_SIMPLE_ATTRIBUTE(fops_u16, debugfs_u16_get, debugfs_u16_set, "%llu\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_u16_ro, debugfs_u16_get, NULL, "%llu\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_u16_wo, NULL, debugfs_u16_set, "%llu\n"); /** * debugfs_create_u16 - create a debugfs file that is used to read and write an unsigned 16-bit value @@ -138,6 +149,13 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u16, debugfs_u16_get, debugfs_u16_set, "%llu\n"); struct dentry *debugfs_create_u16(const char *name, mode_t mode, struct dentry *parent, u16 *value) { + /* if there are no write bits set, make read only */ + if (!(mode & S_IWUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_u16_ro); + /* if there are no read bits set, make write only */ + if (!(mode & S_IRUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_u16_wo); + return debugfs_create_file(name, mode, parent, value, &fops_u16); } EXPORT_SYMBOL_GPL(debugfs_create_u16); @@ -153,6 +171,8 @@ static int debugfs_u32_get(void *data, u64 *val) return 0; } DEFINE_SIMPLE_ATTRIBUTE(fops_u32, debugfs_u32_get, debugfs_u32_set, "%llu\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_u32_ro, debugfs_u32_get, NULL, "%llu\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_u32_wo, NULL, debugfs_u32_set, "%llu\n"); /** * debugfs_create_u32 - create a debugfs file that is used to read and write an unsigned 32-bit value @@ -181,6 +201,13 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u32, debugfs_u32_get, debugfs_u32_set, "%llu\n"); struct dentry *debugfs_create_u32(const char *name, mode_t mode, struct dentry *parent, u32 *value) { + /* if there are no write bits set, make read only */ + if (!(mode & S_IWUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_u32_ro); + /* if there are no read bits set, make write only */ + if (!(mode & S_IRUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_u32_wo); + return debugfs_create_file(name, mode, parent, value, &fops_u32); } EXPORT_SYMBOL_GPL(debugfs_create_u32); @@ -197,6 +224,8 @@ static int debugfs_u64_get(void *data, u64 *val) return 0; } DEFINE_SIMPLE_ATTRIBUTE(fops_u64, debugfs_u64_get, debugfs_u64_set, "%llu\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_u64_ro, debugfs_u64_get, NULL, "%llu\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_u64_wo, NULL, debugfs_u64_set, "%llu\n"); /** * debugfs_create_u64 - create a debugfs file that is used to read and write an unsigned 64-bit value @@ -225,15 +254,28 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_u64, debugfs_u64_get, debugfs_u64_set, "%llu\n"); struct dentry *debugfs_create_u64(const char *name, mode_t mode, struct dentry *parent, u64 *value) { + /* if there are no write bits set, make read only */ + if (!(mode & S_IWUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_u64_ro); + /* if there are no read bits set, make write only */ + if (!(mode & S_IRUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_u64_wo); + return debugfs_create_file(name, mode, parent, value, &fops_u64); } EXPORT_SYMBOL_GPL(debugfs_create_u64); DEFINE_SIMPLE_ATTRIBUTE(fops_x8, debugfs_u8_get, debugfs_u8_set, "0x%02llx\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_x8_ro, debugfs_u8_get, NULL, "0x%02llx\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_x8_wo, NULL, debugfs_u8_set, "0x%02llx\n"); DEFINE_SIMPLE_ATTRIBUTE(fops_x16, debugfs_u16_get, debugfs_u16_set, "0x%04llx\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_x16_ro, debugfs_u16_get, NULL, "0x%04llx\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_x16_wo, NULL, debugfs_u16_set, "0x%04llx\n"); DEFINE_SIMPLE_ATTRIBUTE(fops_x32, debugfs_u32_get, debugfs_u32_set, "0x%08llx\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_x32_ro, debugfs_u32_get, NULL, "0x%08llx\n"); +DEFINE_SIMPLE_ATTRIBUTE(fops_x32_wo, NULL, debugfs_u32_set, "0x%08llx\n"); /* * debugfs_create_x{8,16,32} - create a debugfs file that is used to read and write an unsigned {8,16,32}-bit value @@ -256,6 +298,13 @@ DEFINE_SIMPLE_ATTRIBUTE(fops_x32, debugfs_u32_get, debugfs_u32_set, "0x%08llx\n" struct dentry *debugfs_create_x8(const char *name, mode_t mode, struct dentry *parent, u8 *value) { + /* if there are no write bits set, make read only */ + if (!(mode & S_IWUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_x8_ro); + /* if there are no read bits set, make write only */ + if (!(mode & S_IRUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_x8_wo); + return debugfs_create_file(name, mode, parent, value, &fops_x8); } EXPORT_SYMBOL_GPL(debugfs_create_x8); @@ -273,6 +322,13 @@ EXPORT_SYMBOL_GPL(debugfs_create_x8); struct dentry *debugfs_create_x16(const char *name, mode_t mode, struct dentry *parent, u16 *value) { + /* if there are no write bits set, make read only */ + if (!(mode & S_IWUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_x16_ro); + /* if there are no read bits set, make write only */ + if (!(mode & S_IRUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_x16_wo); + return debugfs_create_file(name, mode, parent, value, &fops_x16); } EXPORT_SYMBOL_GPL(debugfs_create_x16); @@ -290,6 +346,13 @@ EXPORT_SYMBOL_GPL(debugfs_create_x16); struct dentry *debugfs_create_x32(const char *name, mode_t mode, struct dentry *parent, u32 *value) { + /* if there are no write bits set, make read only */ + if (!(mode & S_IWUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_x32_ro); + /* if there are no read bits set, make write only */ + if (!(mode & S_IRUGO)) + return debugfs_create_file(name, mode, parent, value, &fops_x32_wo); + return debugfs_create_file(name, mode, parent, value, &fops_x32); } EXPORT_SYMBOL_GPL(debugfs_create_x32); @@ -419,7 +482,7 @@ static const struct file_operations fops_blob = { }; /** - * debugfs_create_blob - create a debugfs file that is used to read and write a binary blob + * debugfs_create_blob - create a debugfs file that is used to read a binary blob * @name: a pointer to a string containing the name of the file to create. * @mode: the permission that the file should have * @parent: a pointer to the parent dentry for this file. This should be a diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index 0662ba6de85a..d22438ef7674 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -403,6 +403,7 @@ void debugfs_remove_recursive(struct dentry *dentry) } child = list_entry(parent->d_subdirs.next, struct dentry, d_u.d_child); + next_sibling: /* * If "child" isn't empty, walk down the tree and @@ -417,6 +418,16 @@ void debugfs_remove_recursive(struct dentry *dentry) __debugfs_remove(child, parent); if (parent->d_subdirs.next == &child->d_u.d_child) { /* + * Try the next sibling. + */ + if (child->d_u.d_child.next != &parent->d_subdirs) { + child = list_entry(child->d_u.d_child.next, + struct dentry, + d_u.d_child); + goto next_sibling; + } + + /* * Avoid infinite loop if we fail to remove * one dentry. */ diff --git a/fs/fat/cache.c b/fs/fat/cache.c index b42602298087..923990e4f16e 100644 --- a/fs/fat/cache.c +++ b/fs/fat/cache.c @@ -241,7 +241,7 @@ int fat_get_cluster(struct inode *inode, int cluster, int *fclus, int *dclus) while (*fclus < cluster) { /* prevent the infinite loop of cluster chain */ if (*fclus > limit) { - fat_fs_panic(sb, "%s: detected the cluster chain loop" + fat_fs_error(sb, "%s: detected the cluster chain loop" " (i_pos %lld)", __func__, MSDOS_I(inode)->i_pos); nr = -EIO; @@ -252,7 +252,7 @@ int fat_get_cluster(struct inode *inode, int cluster, int *fclus, int *dclus) if (nr < 0) goto out; else if (nr == FAT_ENT_FREE) { - fat_fs_panic(sb, "%s: invalid cluster chain" + fat_fs_error(sb, "%s: invalid cluster chain" " (i_pos %lld)", __func__, MSDOS_I(inode)->i_pos); nr = -EIO; @@ -285,7 +285,7 @@ static int fat_bmap_cluster(struct inode *inode, int cluster) if (ret < 0) return ret; else if (ret == FAT_ENT_EOF) { - fat_fs_panic(sb, "%s: request beyond EOF (i_pos %lld)", + fat_fs_error(sb, "%s: request beyond EOF (i_pos %lld)", __func__, MSDOS_I(inode)->i_pos); return -EIO; } diff --git a/fs/fat/dir.c b/fs/fat/dir.c index 7c14c8cbbaba..38ff75a0fe22 100644 --- a/fs/fat/dir.c +++ b/fs/fat/dir.c @@ -1335,7 +1335,7 @@ found: goto error_remove; } if (dir->i_size & (sbi->cluster_size - 1)) { - fat_fs_panic(sb, "Odd directory size"); + fat_fs_error(sb, "Odd directory size"); dir->i_size = (dir->i_size + sbi->cluster_size - 1) & ~((loff_t)sbi->cluster_size - 1); } diff --git a/fs/fat/fat.h b/fs/fat/fat.h index e4d88527b5dd..adb0e72a176d 100644 --- a/fs/fat/fat.h +++ b/fs/fat/fat.h @@ -17,6 +17,10 @@ #define VFAT_SFN_CREATE_WIN95 0x0100 /* emulate win95 rule for create */ #define VFAT_SFN_CREATE_WINNT 0x0200 /* emulate winnt rule for create */ +#define FAT_ERRORS_CONT 1 /* ignore error and continue */ +#define FAT_ERRORS_PANIC 2 /* panic on error */ +#define FAT_ERRORS_RO 3 /* remount r/o on error */ + struct fat_mount_options { uid_t fs_uid; gid_t fs_gid; @@ -26,6 +30,7 @@ struct fat_mount_options { char *iocharset; /* Charset used for filename input/display */ unsigned short shortname; /* flags for shortname display/create rule */ unsigned char name_check; /* r = relaxed, n = normal, s = strict */ + unsigned char errors; /* On error: continue, panic, remount-ro */ unsigned short allow_utime;/* permission for setting the [am]time */ unsigned quiet:1, /* set = fake successful chmods and chowns */ showexec:1, /* set = only set x bit for com/exe/bat */ @@ -316,7 +321,7 @@ extern int fat_fill_super(struct super_block *sb, void *data, int silent, extern int fat_flush_inodes(struct super_block *sb, struct inode *i1, struct inode *i2); /* fat/misc.c */ -extern void fat_fs_panic(struct super_block *s, const char *fmt, ...) +extern void fat_fs_error(struct super_block *s, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))) __cold; extern void fat_clusters_flush(struct super_block *sb); extern int fat_chain_add(struct inode *inode, int new_dclus, int nr_cluster); diff --git a/fs/fat/fatent.c b/fs/fat/fatent.c index 618f5305c2e4..a81037721a6f 100644 --- a/fs/fat/fatent.c +++ b/fs/fat/fatent.c @@ -348,7 +348,7 @@ int fat_ent_read(struct inode *inode, struct fat_entry *fatent, int entry) if (entry < FAT_START_ENT || sbi->max_cluster <= entry) { fatent_brelse(fatent); - fat_fs_panic(sb, "invalid access to FAT (entry 0x%08x)", entry); + fat_fs_error(sb, "invalid access to FAT (entry 0x%08x)", entry); return -EIO; } @@ -560,7 +560,7 @@ int fat_free_clusters(struct inode *inode, int cluster) err = cluster; goto error; } else if (cluster == FAT_ENT_FREE) { - fat_fs_panic(sb, "%s: deleting FAT entry beyond EOF", + fat_fs_error(sb, "%s: deleting FAT entry beyond EOF", __func__); err = -EIO; goto error; diff --git a/fs/fat/file.c b/fs/fat/file.c index e955a56b4e5e..b28ea646ff60 100644 --- a/fs/fat/file.c +++ b/fs/fat/file.c @@ -18,106 +18,112 @@ #include <linux/security.h> #include "fat.h" -int fat_generic_ioctl(struct inode *inode, struct file *filp, - unsigned int cmd, unsigned long arg) +static int fat_ioctl_get_attributes(struct inode *inode, u32 __user *user_attr) { + u32 attr; + + mutex_lock(&inode->i_mutex); + attr = fat_make_attrs(inode); + mutex_unlock(&inode->i_mutex); + + return put_user(attr, user_attr); +} + +static int fat_ioctl_set_attributes(struct file *file, u32 __user *user_attr) +{ + struct inode *inode = file->f_path.dentry->d_inode; struct msdos_sb_info *sbi = MSDOS_SB(inode->i_sb); - u32 __user *user_attr = (u32 __user *)arg; + int is_dir = S_ISDIR(inode->i_mode); + u32 attr, oldattr; + struct iattr ia; + int err; - switch (cmd) { - case FAT_IOCTL_GET_ATTRIBUTES: - { - u32 attr; + err = get_user(attr, user_attr); + if (err) + goto out; - mutex_lock(&inode->i_mutex); - attr = fat_make_attrs(inode); - mutex_unlock(&inode->i_mutex); + mutex_lock(&inode->i_mutex); + err = mnt_want_write(file->f_path.mnt); + if (err) + goto out_unlock_inode; - return put_user(attr, user_attr); + /* + * ATTR_VOLUME and ATTR_DIR cannot be changed; this also + * prevents the user from turning us into a VFAT + * longname entry. Also, we obviously can't set + * any of the NTFS attributes in the high 24 bits. + */ + attr &= 0xff & ~(ATTR_VOLUME | ATTR_DIR); + /* Merge in ATTR_VOLUME and ATTR_DIR */ + attr |= (MSDOS_I(inode)->i_attrs & ATTR_VOLUME) | + (is_dir ? ATTR_DIR : 0); + oldattr = fat_make_attrs(inode); + + /* Equivalent to a chmod() */ + ia.ia_valid = ATTR_MODE | ATTR_CTIME; + ia.ia_ctime = current_fs_time(inode->i_sb); + if (is_dir) + ia.ia_mode = fat_make_mode(sbi, attr, S_IRWXUGO); + else { + ia.ia_mode = fat_make_mode(sbi, attr, + S_IRUGO | S_IWUGO | (inode->i_mode & S_IXUGO)); } - case FAT_IOCTL_SET_ATTRIBUTES: - { - u32 attr, oldattr; - int err, is_dir = S_ISDIR(inode->i_mode); - struct iattr ia; - err = get_user(attr, user_attr); - if (err) - return err; + /* The root directory has no attributes */ + if (inode->i_ino == MSDOS_ROOT_INO && attr != ATTR_DIR) { + err = -EINVAL; + goto out_drop_write; + } - mutex_lock(&inode->i_mutex); - - err = mnt_want_write(filp->f_path.mnt); - if (err) - goto up_no_drop_write; - - /* - * ATTR_VOLUME and ATTR_DIR cannot be changed; this also - * prevents the user from turning us into a VFAT - * longname entry. Also, we obviously can't set - * any of the NTFS attributes in the high 24 bits. - */ - attr &= 0xff & ~(ATTR_VOLUME | ATTR_DIR); - /* Merge in ATTR_VOLUME and ATTR_DIR */ - attr |= (MSDOS_I(inode)->i_attrs & ATTR_VOLUME) | - (is_dir ? ATTR_DIR : 0); - oldattr = fat_make_attrs(inode); - - /* Equivalent to a chmod() */ - ia.ia_valid = ATTR_MODE | ATTR_CTIME; - ia.ia_ctime = current_fs_time(inode->i_sb); - if (is_dir) - ia.ia_mode = fat_make_mode(sbi, attr, S_IRWXUGO); - else { - ia.ia_mode = fat_make_mode(sbi, attr, - S_IRUGO | S_IWUGO | (inode->i_mode & S_IXUGO)); - } + if (sbi->options.sys_immutable && + ((attr | oldattr) & ATTR_SYS) && + !capable(CAP_LINUX_IMMUTABLE)) { + err = -EPERM; + goto out_drop_write; + } - /* The root directory has no attributes */ - if (inode->i_ino == MSDOS_ROOT_INO && attr != ATTR_DIR) { - err = -EINVAL; - goto up; - } + /* + * The security check is questionable... We single + * out the RO attribute for checking by the security + * module, just because it maps to a file mode. + */ + err = security_inode_setattr(file->f_path.dentry, &ia); + if (err) + goto out_drop_write; - if (sbi->options.sys_immutable) { - if ((attr | oldattr) & ATTR_SYS) { - if (!capable(CAP_LINUX_IMMUTABLE)) { - err = -EPERM; - goto up; - } - } - } + /* This MUST be done before doing anything irreversible... */ + err = fat_setattr(file->f_path.dentry, &ia); + if (err) + goto out_drop_write; + + fsnotify_change(file->f_path.dentry, ia.ia_valid); + if (sbi->options.sys_immutable) { + if (attr & ATTR_SYS) + inode->i_flags |= S_IMMUTABLE; + else + inode->i_flags &= S_IMMUTABLE; + } - /* - * The security check is questionable... We single - * out the RO attribute for checking by the security - * module, just because it maps to a file mode. - */ - err = security_inode_setattr(filp->f_path.dentry, &ia); - if (err) - goto up; - - /* This MUST be done before doing anything irreversible... */ - err = fat_setattr(filp->f_path.dentry, &ia); - if (err) - goto up; - - fsnotify_change(filp->f_path.dentry, ia.ia_valid); - if (sbi->options.sys_immutable) { - if (attr & ATTR_SYS) - inode->i_flags |= S_IMMUTABLE; - else - inode->i_flags &= S_IMMUTABLE; - } + fat_save_attrs(inode, attr); + mark_inode_dirty(inode); +out_drop_write: + mnt_drop_write(file->f_path.mnt); +out_unlock_inode: + mutex_unlock(&inode->i_mutex); +out: + return err; +} - fat_save_attrs(inode, attr); - mark_inode_dirty(inode); -up: - mnt_drop_write(filp->f_path.mnt); -up_no_drop_write: - mutex_unlock(&inode->i_mutex); - return err; - } +int fat_generic_ioctl(struct inode *inode, struct file *filp, + unsigned int cmd, unsigned long arg) +{ + u32 __user *user_attr = (u32 __user *)arg; + + switch (cmd) { + case FAT_IOCTL_GET_ATTRIBUTES: + return fat_ioctl_get_attributes(inode, user_attr); + case FAT_IOCTL_SET_ATTRIBUTES: + return fat_ioctl_set_attributes(filp, user_attr); default: return -ENOTTY; /* Inappropriate ioctl for device */ } @@ -225,7 +231,7 @@ static int fat_free(struct inode *inode, int skip) fatent_brelse(&fatent); return 0; } else if (ret == FAT_ENT_FREE) { - fat_fs_panic(sb, + fat_fs_error(sb, "%s: invalid cluster chain (i_pos %lld)", __func__, MSDOS_I(inode)->i_pos); ret = -EIO; diff --git a/fs/fat/inode.c b/fs/fat/inode.c index 51a5ecf9000a..304b411cb8bc 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -76,7 +76,7 @@ static inline int __fat_get_block(struct inode *inode, sector_t iblock, return 0; if (iblock != MSDOS_I(inode)->mmu_private >> sb->s_blocksize_bits) { - fat_fs_panic(sb, "corrupted file size (i_pos %lld, %lld)", + fat_fs_error(sb, "corrupted file size (i_pos %lld, %lld)", MSDOS_I(inode)->i_pos, MSDOS_I(inode)->mmu_private); return -EIO; } @@ -856,6 +856,12 @@ static int fat_show_options(struct seq_file *m, struct vfsmount *mnt) seq_puts(m, ",flush"); if (opts->tz_utc) seq_puts(m, ",tz=UTC"); + if (opts->errors == FAT_ERRORS_CONT) + seq_puts(m, ",errors=continue"); + else if (opts->errors == FAT_ERRORS_PANIC) + seq_puts(m, ",errors=panic"); + else + seq_puts(m, ",errors=remount-ro"); return 0; } @@ -868,7 +874,8 @@ enum { Opt_charset, Opt_shortname_lower, Opt_shortname_win95, Opt_shortname_winnt, Opt_shortname_mixed, Opt_utf8_no, Opt_utf8_yes, Opt_uni_xl_no, Opt_uni_xl_yes, Opt_nonumtail_no, Opt_nonumtail_yes, - Opt_obsolate, Opt_flush, Opt_tz_utc, Opt_rodir, Opt_err, + Opt_obsolate, Opt_flush, Opt_tz_utc, Opt_rodir, Opt_err_cont, + Opt_err_panic, Opt_err_ro, Opt_err, }; static const match_table_t fat_tokens = { @@ -891,6 +898,11 @@ static const match_table_t fat_tokens = { {Opt_showexec, "showexec"}, {Opt_debug, "debug"}, {Opt_immutable, "sys_immutable"}, + {Opt_flush, "flush"}, + {Opt_tz_utc, "tz=UTC"}, + {Opt_err_cont, "errors=continue"}, + {Opt_err_panic, "errors=panic"}, + {Opt_err_ro, "errors=remount-ro"}, {Opt_obsolate, "conv=binary"}, {Opt_obsolate, "conv=text"}, {Opt_obsolate, "conv=auto"}, @@ -902,8 +914,6 @@ static const match_table_t fat_tokens = { {Opt_obsolate, "cvf_format=%20s"}, {Opt_obsolate, "cvf_options=%100s"}, {Opt_obsolate, "posix"}, - {Opt_flush, "flush"}, - {Opt_tz_utc, "tz=UTC"}, {Opt_err, NULL}, }; static const match_table_t msdos_tokens = { @@ -973,6 +983,7 @@ static int parse_options(char *options, int is_vfat, int silent, int *debug, opts->numtail = 1; opts->usefree = opts->nocase = 0; opts->tz_utc = 0; + opts->errors = FAT_ERRORS_RO; *debug = 0; if (!options) @@ -1065,6 +1076,15 @@ static int parse_options(char *options, int is_vfat, int silent, int *debug, case Opt_tz_utc: opts->tz_utc = 1; break; + case Opt_err_cont: + opts->errors = FAT_ERRORS_CONT; + break; + case Opt_err_panic: + opts->errors = FAT_ERRORS_PANIC; + break; + case Opt_err_ro: + opts->errors = FAT_ERRORS_RO; + break; /* msdos specific */ case Opt_dots: diff --git a/fs/fat/misc.c b/fs/fat/misc.c index ac39ebcc1496..a6c20473dfd7 100644 --- a/fs/fat/misc.c +++ b/fs/fat/misc.c @@ -12,14 +12,19 @@ #include "fat.h" /* - * fat_fs_panic reports a severe file system problem and sets the file system - * read-only. The file system can be made writable again by remounting it. + * fat_fs_error reports a file system problem that might indicate fa data + * corruption/inconsistency. Depending on 'errors' mount option the + * panic() is called, or error message is printed FAT and nothing is done, + * or filesystem is remounted read-only (default behavior). + * In case the file system is remounted read-only, it can be made writable + * again by remounting it. */ -void fat_fs_panic(struct super_block *s, const char *fmt, ...) +void fat_fs_error(struct super_block *s, const char *fmt, ...) { + struct fat_mount_options *opts = &MSDOS_SB(s)->options; va_list args; - printk(KERN_ERR "FAT: Filesystem panic (dev %s)\n", s->s_id); + printk(KERN_ERR "FAT: Filesystem error (dev %s)\n", s->s_id); printk(KERN_ERR " "); va_start(args, fmt); @@ -27,13 +32,14 @@ void fat_fs_panic(struct super_block *s, const char *fmt, ...) va_end(args); printk("\n"); - if (!(s->s_flags & MS_RDONLY)) { + if (opts->errors == FAT_ERRORS_PANIC) + panic(" FAT fs panic from previous error\n"); + else if (opts->errors == FAT_ERRORS_RO && !(s->s_flags & MS_RDONLY)) { s->s_flags |= MS_RDONLY; printk(KERN_ERR " File system has been set read-only\n"); } } - -EXPORT_SYMBOL_GPL(fat_fs_panic); +EXPORT_SYMBOL_GPL(fat_fs_error); /* Flushes the number of free clusters on FAT32 */ /* XXX: Need to write one per FSINFO block. Currently only writes 1 */ @@ -124,7 +130,7 @@ int fat_chain_add(struct inode *inode, int new_dclus, int nr_cluster) mark_inode_dirty(inode); } if (new_fclus != (inode->i_blocks >> (sbi->cluster_bits - 9))) { - fat_fs_panic(sb, "clusters badly computed (%d != %llu)", + fat_fs_error(sb, "clusters badly computed (%d != %llu)", new_fclus, (llu)(inode->i_blocks >> (sbi->cluster_bits - 9))); fat_cache_inval_inode(inode); diff --git a/fs/fat/namei_msdos.c b/fs/fat/namei_msdos.c index 20f522861355..82f88733b681 100644 --- a/fs/fat/namei_msdos.c +++ b/fs/fat/namei_msdos.c @@ -608,7 +608,7 @@ error_inode: sinfo.bh = NULL; } if (corrupt < 0) { - fat_fs_panic(new_dir->i_sb, + fat_fs_error(new_dir->i_sb, "%s: Filesystem corrupted (i_pos %lld)", __func__, sinfo.i_pos); } diff --git a/fs/fat/namei_vfat.c b/fs/fat/namei_vfat.c index f92ad9995356..73471b7ecc8c 100644 --- a/fs/fat/namei_vfat.c +++ b/fs/fat/namei_vfat.c @@ -1030,7 +1030,7 @@ error_inode: sinfo.bh = NULL; } if (corrupt < 0) { - fat_fs_panic(new_dir->i_sb, + fat_fs_error(new_dir->i_sb, "%s: Filesystem corrupted (i_pos %lld)", __func__, sinfo.i_pos); } diff --git a/fs/jfs/jfs_extent.c b/fs/jfs/jfs_extent.c index bbbd5f202e37..41d6045dbeb0 100644 --- a/fs/jfs/jfs_extent.c +++ b/fs/jfs/jfs_extent.c @@ -391,6 +391,7 @@ int extHint(struct inode *ip, s64 offset, xad_t * xp) } XADaddress(xp, xaddr); XADlength(xp, xlen); + XADoffset(xp, prev); /* * only preserve the abnr flag within the xad flags * of the returned hint. diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index 678a067d9251..9edcde4974aa 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -475,6 +475,12 @@ struct ocfs2_path { #define path_leaf_el(_path) ((_path)->p_node[(_path)->p_tree_depth].el) #define path_num_items(_path) ((_path)->p_tree_depth + 1) +static int ocfs2_find_path(struct inode *inode, struct ocfs2_path *path, + u32 cpos); +static void ocfs2_adjust_rightmost_records(struct inode *inode, + handle_t *handle, + struct ocfs2_path *path, + struct ocfs2_extent_rec *insert_rec); /* * Reset the actual path elements so that we can re-use the structure * to build another path. Generally, this involves freeing the buffer @@ -1013,6 +1019,54 @@ static inline u32 ocfs2_sum_rightmost_rec(struct ocfs2_extent_list *el) } /* + * Change range of the branches in the right most path according to the leaf + * extent block's rightmost record. + */ +static int ocfs2_adjust_rightmost_branch(handle_t *handle, + struct inode *inode, + struct ocfs2_extent_tree *et) +{ + int status; + struct ocfs2_path *path = NULL; + struct ocfs2_extent_list *el; + struct ocfs2_extent_rec *rec; + + path = ocfs2_new_path_from_et(et); + if (!path) { + status = -ENOMEM; + return status; + } + + status = ocfs2_find_path(inode, path, UINT_MAX); + if (status < 0) { + mlog_errno(status); + goto out; + } + + status = ocfs2_extend_trans(handle, path_num_items(path) + + handle->h_buffer_credits); + if (status < 0) { + mlog_errno(status); + goto out; + } + + status = ocfs2_journal_access_path(inode, handle, path); + if (status < 0) { + mlog_errno(status); + goto out; + } + + el = path_leaf_el(path); + rec = &el->l_recs[le32_to_cpu(el->l_next_free_rec) - 1]; + + ocfs2_adjust_rightmost_records(inode, handle, path, rec); + +out: + ocfs2_free_path(path); + return status; +} + +/* * Add an entire tree branch to our inode. eb_bh is the extent block * to start at, if we don't want to start the branch at the dinode * structure. @@ -1038,7 +1092,7 @@ static int ocfs2_add_branch(struct ocfs2_super *osb, struct ocfs2_extent_block *eb; struct ocfs2_extent_list *eb_el; struct ocfs2_extent_list *el; - u32 new_cpos; + u32 new_cpos, root_end; mlog_entry_void(); @@ -1055,6 +1109,27 @@ static int ocfs2_add_branch(struct ocfs2_super *osb, new_blocks = le16_to_cpu(el->l_tree_depth); + eb = (struct ocfs2_extent_block *)(*last_eb_bh)->b_data; + new_cpos = ocfs2_sum_rightmost_rec(&eb->h_list); + root_end = ocfs2_sum_rightmost_rec(et->et_root_el); + + /* + * If there is a gap before the root end and the real end + * of the righmost leaf block, we need to remove the gap + * between new_cpos and root_end first so that the tree + * is consistent after we add a new branch(it will start + * from new_cpos). + */ + if (root_end > new_cpos) { + mlog(0, "adjust the cluster end from %u to %u\n", + root_end, new_cpos); + status = ocfs2_adjust_rightmost_branch(handle, inode, et); + if (status) { + mlog_errno(status); + goto bail; + } + } + /* allocate the number of new eb blocks we need */ new_eb_bhs = kcalloc(new_blocks, sizeof(struct buffer_head *), GFP_KERNEL); @@ -1071,9 +1146,6 @@ static int ocfs2_add_branch(struct ocfs2_super *osb, goto bail; } - eb = (struct ocfs2_extent_block *)(*last_eb_bh)->b_data; - new_cpos = ocfs2_sum_rightmost_rec(&eb->h_list); - /* Note: new_eb_bhs[new_blocks - 1] is the guy which will be * linked with the rest of the tree. * conversly, new_eb_bhs[0] is the new bottommost leaf. diff --git a/fs/ocfs2/blockcheck.c b/fs/ocfs2/blockcheck.c index 2a947c44e594..a1163b8b417c 100644 --- a/fs/ocfs2/blockcheck.c +++ b/fs/ocfs2/blockcheck.c @@ -22,6 +22,9 @@ #include <linux/crc32.h> #include <linux/buffer_head.h> #include <linux/bitops.h> +#include <linux/debugfs.h> +#include <linux/module.h> +#include <linux/fs.h> #include <asm/byteorder.h> #include <cluster/masklog.h> @@ -222,6 +225,155 @@ void ocfs2_hamming_fix_block(void *data, unsigned int blocksize, ocfs2_hamming_fix(data, blocksize * 8, 0, fix); } + +/* + * Debugfs handling. + */ + +#ifdef CONFIG_DEBUG_FS + +static int blockcheck_u64_get(void *data, u64 *val) +{ + *val = *(u64 *)data; + return 0; +} +DEFINE_SIMPLE_ATTRIBUTE(blockcheck_fops, blockcheck_u64_get, NULL, "%llu\n"); + +static struct dentry *blockcheck_debugfs_create(const char *name, + struct dentry *parent, + u64 *value) +{ + return debugfs_create_file(name, S_IFREG | S_IRUSR, parent, value, + &blockcheck_fops); +} + +static void ocfs2_blockcheck_debug_remove(struct ocfs2_blockcheck_stats *stats) +{ + if (stats) { + debugfs_remove(stats->b_debug_check); + stats->b_debug_check = NULL; + debugfs_remove(stats->b_debug_failure); + stats->b_debug_failure = NULL; + debugfs_remove(stats->b_debug_recover); + stats->b_debug_recover = NULL; + debugfs_remove(stats->b_debug_dir); + stats->b_debug_dir = NULL; + } +} + +static int ocfs2_blockcheck_debug_install(struct ocfs2_blockcheck_stats *stats, + struct dentry *parent) +{ + int rc = -EINVAL; + + if (!stats) + goto out; + + stats->b_debug_dir = debugfs_create_dir("blockcheck", parent); + if (!stats->b_debug_dir) + goto out; + + stats->b_debug_check = + blockcheck_debugfs_create("blocks_checked", + stats->b_debug_dir, + &stats->b_check_count); + + stats->b_debug_failure = + blockcheck_debugfs_create("checksums_failed", + stats->b_debug_dir, + &stats->b_failure_count); + + stats->b_debug_recover = + blockcheck_debugfs_create("ecc_recoveries", + stats->b_debug_dir, + &stats->b_recover_count); + if (stats->b_debug_check && stats->b_debug_failure && + stats->b_debug_recover) + rc = 0; + +out: + if (rc) + ocfs2_blockcheck_debug_remove(stats); + return rc; +} +#else +static inline int ocfs2_blockcheck_debug_install(struct ocfs2_blockcheck_stats *stats, + struct dentry *parent) +{ + return 0; +} + +static inline void ocfs2_blockcheck_debug_remove(struct ocfs2_blockcheck_stats *stats) +{ +} +#endif /* CONFIG_DEBUG_FS */ + +/* Always-called wrappers for starting and stopping the debugfs files */ +int ocfs2_blockcheck_stats_debugfs_install(struct ocfs2_blockcheck_stats *stats, + struct dentry *parent) +{ + return ocfs2_blockcheck_debug_install(stats, parent); +} + +void ocfs2_blockcheck_stats_debugfs_remove(struct ocfs2_blockcheck_stats *stats) +{ + ocfs2_blockcheck_debug_remove(stats); +} + +static void ocfs2_blockcheck_inc_check(struct ocfs2_blockcheck_stats *stats) +{ + u64 new_count; + + if (!stats) + return; + + spin_lock(&stats->b_lock); + stats->b_check_count++; + new_count = stats->b_check_count; + spin_unlock(&stats->b_lock); + + if (!new_count) + mlog(ML_NOTICE, "Block check count has wrapped\n"); +} + +static void ocfs2_blockcheck_inc_failure(struct ocfs2_blockcheck_stats *stats) +{ + u64 new_count; + + if (!stats) + return; + + spin_lock(&stats->b_lock); + stats->b_failure_count++; + new_count = stats->b_failure_count; + spin_unlock(&stats->b_lock); + + if (!new_count) + mlog(ML_NOTICE, "Checksum failure count has wrapped\n"); +} + +static void ocfs2_blockcheck_inc_recover(struct ocfs2_blockcheck_stats *stats) +{ + u64 new_count; + + if (!stats) + return; + + spin_lock(&stats->b_lock); + stats->b_recover_count++; + new_count = stats->b_recover_count; + spin_unlock(&stats->b_lock); + + if (!new_count) + mlog(ML_NOTICE, "ECC recovery count has wrapped\n"); +} + + + +/* + * These are the low-level APIs for using the ocfs2_block_check structure. + */ + /* * This function generates check information for a block. * data is the block to be checked. bc is a pointer to the @@ -266,12 +418,15 @@ void ocfs2_block_check_compute(void *data, size_t blocksize, * Again, the data passed in should be the on-disk endian. */ int ocfs2_block_check_validate(void *data, size_t blocksize, - struct ocfs2_block_check *bc) + struct ocfs2_block_check *bc, + struct ocfs2_blockcheck_stats *stats) { int rc = 0; struct ocfs2_block_check check; u32 crc, ecc; + ocfs2_blockcheck_inc_check(stats); + check.bc_crc32e = le32_to_cpu(bc->bc_crc32e); check.bc_ecc = le16_to_cpu(bc->bc_ecc); @@ -282,6 +437,7 @@ int ocfs2_block_check_validate(void *data, size_t blocksize, if (crc == check.bc_crc32e) goto out; + ocfs2_blockcheck_inc_failure(stats); mlog(ML_ERROR, "CRC32 failed: stored: %u, computed %u. Applying ECC.\n", (unsigned int)check.bc_crc32e, (unsigned int)crc); @@ -292,8 +448,10 @@ int ocfs2_block_check_validate(void *data, size_t blocksize, /* And check the crc32 again */ crc = crc32_le(~0, data, blocksize); - if (crc == check.bc_crc32e) + if (crc == check.bc_crc32e) { + ocfs2_blockcheck_inc_recover(stats); goto out; + } mlog(ML_ERROR, "Fixed CRC32 failed: stored: %u, computed %u\n", (unsigned int)check.bc_crc32e, (unsigned int)crc); @@ -366,7 +524,8 @@ void ocfs2_block_check_compute_bhs(struct buffer_head **bhs, int nr, * Again, the data passed in should be the on-disk endian. */ int ocfs2_block_check_validate_bhs(struct buffer_head **bhs, int nr, - struct ocfs2_block_check *bc) + struct ocfs2_block_check *bc, + struct ocfs2_blockcheck_stats *stats) { int i, rc = 0; struct ocfs2_block_check check; @@ -377,6 +536,8 @@ int ocfs2_block_check_validate_bhs(struct buffer_head **bhs, int nr, if (!nr) return 0; + ocfs2_blockcheck_inc_check(stats); + check.bc_crc32e = le32_to_cpu(bc->bc_crc32e); check.bc_ecc = le16_to_cpu(bc->bc_ecc); @@ -388,6 +549,7 @@ int ocfs2_block_check_validate_bhs(struct buffer_head **bhs, int nr, if (crc == check.bc_crc32e) goto out; + ocfs2_blockcheck_inc_failure(stats); mlog(ML_ERROR, "CRC32 failed: stored: %u, computed %u. Applying ECC.\n", (unsigned int)check.bc_crc32e, (unsigned int)crc); @@ -416,8 +578,10 @@ int ocfs2_block_check_validate_bhs(struct buffer_head **bhs, int nr, /* And check the crc32 again */ for (i = 0, crc = ~0; i < nr; i++) crc = crc32_le(crc, bhs[i]->b_data, bhs[i]->b_size); - if (crc == check.bc_crc32e) + if (crc == check.bc_crc32e) { + ocfs2_blockcheck_inc_recover(stats); goto out; + } mlog(ML_ERROR, "Fixed CRC32 failed: stored: %u, computed %u\n", (unsigned int)check.bc_crc32e, (unsigned int)crc); @@ -448,9 +612,11 @@ int ocfs2_validate_meta_ecc(struct super_block *sb, void *data, struct ocfs2_block_check *bc) { int rc = 0; + struct ocfs2_super *osb = OCFS2_SB(sb); - if (ocfs2_meta_ecc(OCFS2_SB(sb))) - rc = ocfs2_block_check_validate(data, sb->s_blocksize, bc); + if (ocfs2_meta_ecc(osb)) + rc = ocfs2_block_check_validate(data, sb->s_blocksize, bc, + &osb->osb_ecc_stats); return rc; } @@ -468,9 +634,11 @@ int ocfs2_validate_meta_ecc_bhs(struct super_block *sb, struct ocfs2_block_check *bc) { int rc = 0; + struct ocfs2_super *osb = OCFS2_SB(sb); - if (ocfs2_meta_ecc(OCFS2_SB(sb))) - rc = ocfs2_block_check_validate_bhs(bhs, nr, bc); + if (ocfs2_meta_ecc(osb)) + rc = ocfs2_block_check_validate_bhs(bhs, nr, bc, + &osb->osb_ecc_stats); return rc; } diff --git a/fs/ocfs2/blockcheck.h b/fs/ocfs2/blockcheck.h index 70ec3feda32f..d4b69febf70a 100644 --- a/fs/ocfs2/blockcheck.h +++ b/fs/ocfs2/blockcheck.h @@ -21,6 +21,24 @@ #define OCFS2_BLOCKCHECK_H +/* Count errors and error correction from blockcheck.c */ +struct ocfs2_blockcheck_stats { + spinlock_t b_lock; + u64 b_check_count; /* Number of blocks we've checked */ + u64 b_failure_count; /* Number of failed checksums */ + u64 b_recover_count; /* Number of blocks fixed by ecc */ + + /* + * debugfs entries, used if this is passed to + * ocfs2_blockcheck_stats_debugfs_install() + */ + struct dentry *b_debug_dir; /* Parent of the debugfs files */ + struct dentry *b_debug_check; /* Exposes b_check_count */ + struct dentry *b_debug_failure; /* Exposes b_failure_count */ + struct dentry *b_debug_recover; /* Exposes b_recover_count */ +}; + + /* High level block API */ void ocfs2_compute_meta_ecc(struct super_block *sb, void *data, struct ocfs2_block_check *bc); @@ -37,11 +55,18 @@ int ocfs2_validate_meta_ecc_bhs(struct super_block *sb, void ocfs2_block_check_compute(void *data, size_t blocksize, struct ocfs2_block_check *bc); int ocfs2_block_check_validate(void *data, size_t blocksize, - struct ocfs2_block_check *bc); + struct ocfs2_block_check *bc, + struct ocfs2_blockcheck_stats *stats); void ocfs2_block_check_compute_bhs(struct buffer_head **bhs, int nr, struct ocfs2_block_check *bc); int ocfs2_block_check_validate_bhs(struct buffer_head **bhs, int nr, - struct ocfs2_block_check *bc); + struct ocfs2_block_check *bc, + struct ocfs2_blockcheck_stats *stats); + +/* Debug Initialization */ +int ocfs2_blockcheck_stats_debugfs_install(struct ocfs2_blockcheck_stats *stats, + struct dentry *parent); +void ocfs2_blockcheck_stats_debugfs_remove(struct ocfs2_blockcheck_stats *stats); /* * Hamming code functions diff --git a/fs/ocfs2/cluster/masklog.h b/fs/ocfs2/cluster/masklog.h index 7e72a81bc2d4..696c32e50716 100644 --- a/fs/ocfs2/cluster/masklog.h +++ b/fs/ocfs2/cluster/masklog.h @@ -48,34 +48,33 @@ * only emit the appropriage printk() when the caller passes in a constant * mask, as is almost always the case. * - * All this bitmask nonsense is hidden from the /proc interface so that Joel - * doesn't have an aneurism. Reading the file gives a straight forward - * indication of which bits are on or off: - * ENTRY off - * EXIT off + * All this bitmask nonsense is managed from the files under + * /sys/fs/o2cb/logmask/. Reading the files gives a straightforward + * indication of which bits are allowed (allow) or denied (off/deny). + * ENTRY deny + * EXIT deny * TCP off * MSG off * SOCKET off - * ERROR off - * NOTICE on + * ERROR allow + * NOTICE allow * * Writing changes the state of a given bit and requires a strictly formatted * single write() call: * - * write(fd, "ENTRY on", 8); + * write(fd, "allow", 5); * - * would turn the entry bit on. "1" is also accepted in the place of "on", and - * "off" and "0" behave as expected. + * Echoing allow/deny/off string into the logmask files can flip the bits + * on or off as expected; here is the bash script for example: * - * Some trivial shell can flip all the bits on or off: + * log_mask="/sys/fs/o2cb/log_mask" + * for node in ENTRY EXIT TCP MSG SOCKET ERROR NOTICE; do + * echo allow >"$log_mask"/"$node" + * done * - * log_mask="/proc/fs/ocfs2_nodemanager/log_mask" - * cat $log_mask | ( - * while read bit status; do - * # $1 is "on" or "off", say - * echo "$bit $1" > $log_mask - * done - * ) + * The debugfs.ocfs2 tool can also flip the bits with the -l option: + * + * debugfs.ocfs2 -l TCP allow */ /* for task_struct */ diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index 9fbe849f6344..334f231a422c 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -974,7 +974,7 @@ static int o2net_tx_can_proceed(struct o2net_node *nn, int o2net_send_message_vec(u32 msg_type, u32 key, struct kvec *caller_vec, size_t caller_veclen, u8 target_node, int *status) { - int ret, error = 0; + int ret; struct o2net_msg *msg = NULL; size_t veclen, caller_bytes = 0; struct kvec *vec = NULL; @@ -1015,10 +1015,7 @@ int o2net_send_message_vec(u32 msg_type, u32 key, struct kvec *caller_vec, o2net_set_nst_sock_time(&nst); - ret = wait_event_interruptible(nn->nn_sc_wq, - o2net_tx_can_proceed(nn, &sc, &error)); - if (!ret && error) - ret = error; + wait_event(nn->nn_sc_wq, o2net_tx_can_proceed(nn, &sc, &ret)); if (ret) goto out; diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index c5752305627c..b358f3bf896d 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -2900,6 +2900,8 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh, alloc = ocfs2_clusters_for_bytes(sb, bytes); dx_alloc = 0; + down_write(&oi->ip_alloc_sem); + if (ocfs2_supports_indexed_dirs(osb)) { credits += ocfs2_add_dir_index_credits(sb); @@ -2940,8 +2942,6 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh, goto out; } - down_write(&oi->ip_alloc_sem); - /* * Prepare for worst case allocation scenario of two separate * extents in the unindexed tree. @@ -2953,7 +2953,7 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh, if (IS_ERR(handle)) { ret = PTR_ERR(handle); mlog_errno(ret); - goto out_sem; + goto out; } if (vfs_dq_alloc_space_nodirty(dir, @@ -3172,10 +3172,8 @@ out_commit: ocfs2_commit_trans(osb, handle); -out_sem: - up_write(&oi->ip_alloc_sem); - out: + up_write(&oi->ip_alloc_sem); if (data_ac) ocfs2_free_alloc_context(data_ac); if (meta_ac) @@ -3322,11 +3320,15 @@ static int ocfs2_extend_dir(struct ocfs2_super *osb, brelse(new_bh); new_bh = NULL; + down_write(&OCFS2_I(dir)->ip_alloc_sem); + drop_alloc_sem = 1; dir_i_size = i_size_read(dir); credits = OCFS2_SIMPLE_DIR_EXTEND_CREDITS; goto do_extend; } + down_write(&OCFS2_I(dir)->ip_alloc_sem); + drop_alloc_sem = 1; dir_i_size = i_size_read(dir); mlog(0, "extending dir %llu (i_size = %lld)\n", (unsigned long long)OCFS2_I(dir)->ip_blkno, dir_i_size); @@ -3370,9 +3372,6 @@ do_extend: credits++; /* For attaching the new dirent block to the * dx_root */ - down_write(&OCFS2_I(dir)->ip_alloc_sem); - drop_alloc_sem = 1; - handle = ocfs2_start_trans(osb, credits); if (IS_ERR(handle)) { status = PTR_ERR(handle); @@ -3435,10 +3434,10 @@ bail_bh: *new_de_bh = new_bh; get_bh(*new_de_bh); bail: - if (drop_alloc_sem) - up_write(&OCFS2_I(dir)->ip_alloc_sem); if (handle) ocfs2_commit_trans(osb, handle); + if (drop_alloc_sem) + up_write(&OCFS2_I(dir)->ip_alloc_sem); if (data_ac) ocfs2_free_alloc_context(data_ac); diff --git a/fs/ocfs2/dlmglue.c b/fs/ocfs2/dlmglue.c index e15fc7d50827..6cdeaa76f27f 100644 --- a/fs/ocfs2/dlmglue.c +++ b/fs/ocfs2/dlmglue.c @@ -248,6 +248,10 @@ static struct ocfs2_lock_res_ops ocfs2_nfs_sync_lops = { .flags = 0, }; +static struct ocfs2_lock_res_ops ocfs2_orphan_scan_lops = { + .flags = LOCK_TYPE_REQUIRES_REFRESH|LOCK_TYPE_USES_LVB, +}; + static struct ocfs2_lock_res_ops ocfs2_dentry_lops = { .get_osb = ocfs2_get_dentry_osb, .post_unlock = ocfs2_dentry_post_unlock, @@ -637,6 +641,19 @@ static void ocfs2_nfs_sync_lock_res_init(struct ocfs2_lock_res *res, &ocfs2_nfs_sync_lops, osb); } +static void ocfs2_orphan_scan_lock_res_init(struct ocfs2_lock_res *res, + struct ocfs2_super *osb) +{ + struct ocfs2_orphan_scan_lvb *lvb; + + ocfs2_lock_res_init_once(res); + ocfs2_build_lock_name(OCFS2_LOCK_TYPE_ORPHAN_SCAN, 0, 0, res->l_name); + ocfs2_lock_res_init_common(osb, res, OCFS2_LOCK_TYPE_ORPHAN_SCAN, + &ocfs2_orphan_scan_lops, osb); + lvb = ocfs2_dlm_lvb(&res->l_lksb); + lvb->lvb_version = OCFS2_ORPHAN_LVB_VERSION; +} + void ocfs2_file_lock_res_init(struct ocfs2_lock_res *lockres, struct ocfs2_file_private *fp) { @@ -2352,6 +2369,37 @@ void ocfs2_inode_unlock(struct inode *inode, mlog_exit_void(); } +int ocfs2_orphan_scan_lock(struct ocfs2_super *osb, u32 *seqno, int ex) +{ + struct ocfs2_lock_res *lockres; + struct ocfs2_orphan_scan_lvb *lvb; + int level = ex ? DLM_LOCK_EX : DLM_LOCK_PR; + int status = 0; + + lockres = &osb->osb_orphan_scan.os_lockres; + status = ocfs2_cluster_lock(osb, lockres, level, 0, 0); + if (status < 0) + return status; + + lvb = ocfs2_dlm_lvb(&lockres->l_lksb); + if (lvb->lvb_version == OCFS2_ORPHAN_LVB_VERSION) + *seqno = be32_to_cpu(lvb->lvb_os_seqno); + return status; +} + +void ocfs2_orphan_scan_unlock(struct ocfs2_super *osb, u32 seqno, int ex) +{ + struct ocfs2_lock_res *lockres; + struct ocfs2_orphan_scan_lvb *lvb; + int level = ex ? DLM_LOCK_EX : DLM_LOCK_PR; + + lockres = &osb->osb_orphan_scan.os_lockres; + lvb = ocfs2_dlm_lvb(&lockres->l_lksb); + lvb->lvb_version = OCFS2_ORPHAN_LVB_VERSION; + lvb->lvb_os_seqno = cpu_to_be32(seqno); + ocfs2_cluster_unlock(osb, lockres, level); +} + int ocfs2_super_lock(struct ocfs2_super *osb, int ex) { @@ -2842,6 +2890,7 @@ local: ocfs2_super_lock_res_init(&osb->osb_super_lockres, osb); ocfs2_rename_lock_res_init(&osb->osb_rename_lockres, osb); ocfs2_nfs_sync_lock_res_init(&osb->osb_nfs_sync_lockres, osb); + ocfs2_orphan_scan_lock_res_init(&osb->osb_orphan_scan.os_lockres, osb); osb->cconn = conn; @@ -2878,6 +2927,7 @@ void ocfs2_dlm_shutdown(struct ocfs2_super *osb, ocfs2_lock_res_free(&osb->osb_super_lockres); ocfs2_lock_res_free(&osb->osb_rename_lockres); ocfs2_lock_res_free(&osb->osb_nfs_sync_lockres); + ocfs2_lock_res_free(&osb->osb_orphan_scan.os_lockres); ocfs2_cluster_disconnect(osb->cconn, hangup_pending); osb->cconn = NULL; @@ -3061,6 +3111,7 @@ static void ocfs2_drop_osb_locks(struct ocfs2_super *osb) ocfs2_simple_drop_lockres(osb, &osb->osb_super_lockres); ocfs2_simple_drop_lockres(osb, &osb->osb_rename_lockres); ocfs2_simple_drop_lockres(osb, &osb->osb_nfs_sync_lockres); + ocfs2_simple_drop_lockres(osb, &osb->osb_orphan_scan.os_lockres); } int ocfs2_drop_inode_locks(struct inode *inode) diff --git a/fs/ocfs2/dlmglue.h b/fs/ocfs2/dlmglue.h index e1fd5721cd7f..31b90d7b8f51 100644 --- a/fs/ocfs2/dlmglue.h +++ b/fs/ocfs2/dlmglue.h @@ -62,6 +62,14 @@ struct ocfs2_qinfo_lvb { __be32 lvb_free_entry; }; +#define OCFS2_ORPHAN_LVB_VERSION 1 + +struct ocfs2_orphan_scan_lvb { + __u8 lvb_version; + __u8 lvb_reserved[3]; + __be32 lvb_os_seqno; +}; + /* ocfs2_inode_lock_full() 'arg_flags' flags */ /* don't wait on recovery. */ #define OCFS2_META_LOCK_RECOVERY (0x01) @@ -113,6 +121,9 @@ int ocfs2_super_lock(struct ocfs2_super *osb, int ex); void ocfs2_super_unlock(struct ocfs2_super *osb, int ex); +int ocfs2_orphan_scan_lock(struct ocfs2_super *osb, u32 *seqno, int ex); +void ocfs2_orphan_scan_unlock(struct ocfs2_super *osb, u32 seqno, int ex); + int ocfs2_rename_lock(struct ocfs2_super *osb); void ocfs2_rename_unlock(struct ocfs2_super *osb); int ocfs2_nfs_sync_lock(struct ocfs2_super *osb, int ex); diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index c2a87c885b73..07267e0da909 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -187,6 +187,9 @@ static int ocfs2_sync_file(struct file *file, if (err) goto bail; + if (datasync && !(inode->i_state & I_DIRTY_DATASYNC)) + goto bail; + journal = osb->journal->j_journal; err = jbd2_journal_force_commit(journal); @@ -894,9 +897,9 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr) struct ocfs2_super *osb = OCFS2_SB(sb); struct buffer_head *bh = NULL; handle_t *handle = NULL; - int locked[MAXQUOTAS] = {0, 0}; - int credits, qtype; - struct ocfs2_mem_dqinfo *oinfo; + int qtype; + struct dquot *transfer_from[MAXQUOTAS] = { }; + struct dquot *transfer_to[MAXQUOTAS] = { }; mlog_entry("(0x%p, '%.*s')\n", dentry, dentry->d_name.len, dentry->d_name.name); @@ -969,30 +972,37 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr) if ((attr->ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid) || (attr->ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid)) { - credits = OCFS2_INODE_UPDATE_CREDITS; + /* + * Gather pointers to quota structures so that allocation / + * freeing of quota structures happens here and not inside + * vfs_dq_transfer() where we have problems with lock ordering + */ if (attr->ia_valid & ATTR_UID && attr->ia_uid != inode->i_uid && OCFS2_HAS_RO_COMPAT_FEATURE(sb, OCFS2_FEATURE_RO_COMPAT_USRQUOTA)) { - oinfo = sb_dqinfo(sb, USRQUOTA)->dqi_priv; - status = ocfs2_lock_global_qf(oinfo, 1); - if (status < 0) + transfer_to[USRQUOTA] = dqget(sb, attr->ia_uid, + USRQUOTA); + transfer_from[USRQUOTA] = dqget(sb, inode->i_uid, + USRQUOTA); + if (!transfer_to[USRQUOTA] || !transfer_from[USRQUOTA]) { + status = -ESRCH; goto bail_unlock; - credits += ocfs2_calc_qinit_credits(sb, USRQUOTA) + - ocfs2_calc_qdel_credits(sb, USRQUOTA); - locked[USRQUOTA] = 1; + } } if (attr->ia_valid & ATTR_GID && attr->ia_gid != inode->i_gid && OCFS2_HAS_RO_COMPAT_FEATURE(sb, OCFS2_FEATURE_RO_COMPAT_GRPQUOTA)) { - oinfo = sb_dqinfo(sb, GRPQUOTA)->dqi_priv; - status = ocfs2_lock_global_qf(oinfo, 1); - if (status < 0) + transfer_to[GRPQUOTA] = dqget(sb, attr->ia_gid, + GRPQUOTA); + transfer_from[GRPQUOTA] = dqget(sb, inode->i_gid, + GRPQUOTA); + if (!transfer_to[GRPQUOTA] || !transfer_from[GRPQUOTA]) { + status = -ESRCH; goto bail_unlock; - credits += ocfs2_calc_qinit_credits(sb, GRPQUOTA) + - ocfs2_calc_qdel_credits(sb, GRPQUOTA); - locked[GRPQUOTA] = 1; + } } - handle = ocfs2_start_trans(osb, credits); + handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS + + 2 * ocfs2_quota_trans_credits(sb)); if (IS_ERR(handle)) { status = PTR_ERR(handle); mlog_errno(status); @@ -1030,12 +1040,6 @@ int ocfs2_setattr(struct dentry *dentry, struct iattr *attr) bail_commit: ocfs2_commit_trans(osb, handle); bail_unlock: - for (qtype = 0; qtype < MAXQUOTAS; qtype++) { - if (!locked[qtype]) - continue; - oinfo = sb_dqinfo(sb, qtype)->dqi_priv; - ocfs2_unlock_global_qf(oinfo, 1); - } ocfs2_inode_unlock(inode, 1); bail_unlock_rw: if (size_change) @@ -1043,6 +1047,12 @@ bail_unlock_rw: bail: brelse(bh); + /* Release quota pointers in case we acquired them */ + for (qtype = 0; qtype < MAXQUOTAS; qtype++) { + dqput(transfer_to[qtype]); + dqput(transfer_from[qtype]); + } + if (!status && attr->ia_valid & ATTR_MODE) { status = ocfs2_acl_chmod(inode); if (status < 0) diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index a20a0f1e37fd..4a3b9e6b31ad 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -28,6 +28,8 @@ #include <linux/slab.h> #include <linux/highmem.h> #include <linux/kthread.h> +#include <linux/time.h> +#include <linux/random.h> #define MLOG_MASK_PREFIX ML_JOURNAL #include <cluster/masklog.h> @@ -52,6 +54,8 @@ DEFINE_SPINLOCK(trans_inc_lock); +#define ORPHAN_SCAN_SCHEDULE_TIMEOUT 300000 + static int ocfs2_force_read_journal(struct inode *inode); static int ocfs2_recover_node(struct ocfs2_super *osb, int node_num, int slot_num); @@ -1841,6 +1845,113 @@ bail: return status; } +/* + * Scan timer should get fired every ORPHAN_SCAN_SCHEDULE_TIMEOUT. Add some + * randomness to the timeout to minimize multple nodes firing the timer at the + * same time. + */ +static inline unsigned long ocfs2_orphan_scan_timeout(void) +{ + unsigned long time; + + get_random_bytes(&time, sizeof(time)); + time = ORPHAN_SCAN_SCHEDULE_TIMEOUT + (time % 5000); + return msecs_to_jiffies(time); +} + +/* + * ocfs2_queue_orphan_scan calls ocfs2_queue_recovery_completion for + * every slot, queuing a recovery of the slot on the ocfs2_wq thread. This + * is done to catch any orphans that are left over in orphan directories. + * + * ocfs2_queue_orphan_scan gets called every ORPHAN_SCAN_SCHEDULE_TIMEOUT + * seconds. It gets an EX lock on os_lockres and checks sequence number + * stored in LVB. If the sequence number has changed, it means some other + * node has done the scan. This node skips the scan and tracks the + * sequence number. If the sequence number didn't change, it means a scan + * hasn't happened. The node queues a scan and increments the + * sequence number in the LVB. + */ +void ocfs2_queue_orphan_scan(struct ocfs2_super *osb) +{ + struct ocfs2_orphan_scan *os; + int status, i; + u32 seqno = 0; + + os = &osb->osb_orphan_scan; + + status = ocfs2_orphan_scan_lock(osb, &seqno, DLM_LOCK_EX); + if (status < 0) { + if (status != -EAGAIN) + mlog_errno(status); + goto out; + } + + if (os->os_seqno != seqno) { + os->os_seqno = seqno; + goto unlock; + } + + for (i = 0; i < osb->max_slots; i++) + ocfs2_queue_recovery_completion(osb->journal, i, NULL, NULL, + NULL); + /* + * We queued a recovery on orphan slots, increment the sequence + * number and update LVB so other node will skip the scan for a while + */ + seqno++; + os->os_count++; + os->os_scantime = CURRENT_TIME; +unlock: + ocfs2_orphan_scan_unlock(osb, seqno, DLM_LOCK_EX); +out: + return; +} + +/* Worker task that gets fired every ORPHAN_SCAN_SCHEDULE_TIMEOUT millsec */ +void ocfs2_orphan_scan_work(struct work_struct *work) +{ + struct ocfs2_orphan_scan *os; + struct ocfs2_super *osb; + + os = container_of(work, struct ocfs2_orphan_scan, + os_orphan_scan_work.work); + osb = os->os_osb; + + mutex_lock(&os->os_lock); + ocfs2_queue_orphan_scan(osb); + schedule_delayed_work(&os->os_orphan_scan_work, + ocfs2_orphan_scan_timeout()); + mutex_unlock(&os->os_lock); +} + +void ocfs2_orphan_scan_stop(struct ocfs2_super *osb) +{ + struct ocfs2_orphan_scan *os; + + os = &osb->osb_orphan_scan; + mutex_lock(&os->os_lock); + cancel_delayed_work(&os->os_orphan_scan_work); + mutex_unlock(&os->os_lock); +} + +int ocfs2_orphan_scan_init(struct ocfs2_super *osb) +{ + struct ocfs2_orphan_scan *os; + + os = &osb->osb_orphan_scan; + os->os_osb = osb; + os->os_count = 0; + os->os_scantime = CURRENT_TIME; + mutex_init(&os->os_lock); + + INIT_DELAYED_WORK(&os->os_orphan_scan_work, + ocfs2_orphan_scan_work); + schedule_delayed_work(&os->os_orphan_scan_work, + ocfs2_orphan_scan_timeout()); + return 0; +} + struct ocfs2_orphan_filldir_priv { struct inode *head; struct ocfs2_super *osb; diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h index eb7b76331eb7..61045eeb3f6e 100644 --- a/fs/ocfs2/journal.h +++ b/fs/ocfs2/journal.h @@ -144,6 +144,10 @@ static inline void ocfs2_inode_set_new(struct ocfs2_super *osb, } /* Exported only for the journal struct init code in super.c. Do not call. */ +int ocfs2_orphan_scan_init(struct ocfs2_super *osb); +void ocfs2_orphan_scan_stop(struct ocfs2_super *osb); +void ocfs2_orphan_scan_exit(struct ocfs2_super *osb); + void ocfs2_complete_recovery(struct work_struct *work); void ocfs2_wait_for_recovery(struct ocfs2_super *osb); diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index 1386281950db..18c1d9ec1c93 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -47,6 +47,9 @@ #include "ocfs2_fs.h" #include "ocfs2_lockid.h" +/* For struct ocfs2_blockcheck_stats */ +#include "blockcheck.h" + /* Most user visible OCFS2 inodes will have very few pieces of * metadata, but larger files (including bitmaps, etc) must be taken * into account when designing an access scheme. We allow a small @@ -151,6 +154,16 @@ struct ocfs2_lock_res { #endif }; +struct ocfs2_orphan_scan { + struct mutex os_lock; + struct ocfs2_super *os_osb; + struct ocfs2_lock_res os_lockres; /* lock to synchronize scans */ + struct delayed_work os_orphan_scan_work; + struct timespec os_scantime; /* time this node ran the scan */ + u32 os_count; /* tracks node specific scans */ + u32 os_seqno; /* tracks cluster wide scans */ +}; + struct ocfs2_dlm_debug { struct kref d_refcnt; struct dentry *d_locking_state; @@ -295,6 +308,7 @@ struct ocfs2_super struct ocfs2_dinode *local_alloc_copy; struct ocfs2_quota_recovery *quota_rec; + struct ocfs2_blockcheck_stats osb_ecc_stats; struct ocfs2_alloc_stats alloc_stats; char dev_str[20]; /* "major,minor" of the device */ @@ -341,6 +355,8 @@ struct ocfs2_super unsigned int *osb_orphan_wipes; wait_queue_head_t osb_wipe_event; + struct ocfs2_orphan_scan osb_orphan_scan; + /* used to protect metaecc calculation check of xattr. */ spinlock_t osb_xattr_lock; diff --git a/fs/ocfs2/ocfs2_lockid.h b/fs/ocfs2/ocfs2_lockid.h index a53ce87481bf..fcdba091af3d 100644 --- a/fs/ocfs2/ocfs2_lockid.h +++ b/fs/ocfs2/ocfs2_lockid.h @@ -48,6 +48,7 @@ enum ocfs2_lock_type { OCFS2_LOCK_TYPE_FLOCK, OCFS2_LOCK_TYPE_QINFO, OCFS2_LOCK_TYPE_NFS_SYNC, + OCFS2_LOCK_TYPE_ORPHAN_SCAN, OCFS2_NUM_LOCK_TYPES }; @@ -85,6 +86,9 @@ static inline char ocfs2_lock_type_char(enum ocfs2_lock_type type) case OCFS2_LOCK_TYPE_NFS_SYNC: c = 'Y'; break; + case OCFS2_LOCK_TYPE_ORPHAN_SCAN: + c = 'P'; + break; default: c = '\0'; } @@ -104,6 +108,7 @@ static char *ocfs2_lock_type_strings[] = { [OCFS2_LOCK_TYPE_OPEN] = "Open", [OCFS2_LOCK_TYPE_FLOCK] = "Flock", [OCFS2_LOCK_TYPE_QINFO] = "Quota", + [OCFS2_LOCK_TYPE_ORPHAN_SCAN] = "OrphanScan", }; static inline const char *ocfs2_lock_type_string(enum ocfs2_lock_type type) diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index 1ed0f7c86869..edfa60cd155c 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -421,6 +421,7 @@ int ocfs2_global_read_dquot(struct dquot *dquot) OCFS2_DQUOT(dquot)->dq_originodes = dquot->dq_dqb.dqb_curinodes; if (!dquot->dq_off) { /* No real quota entry? */ /* Upgrade to exclusive lock for allocation */ + ocfs2_qinfo_unlock(info, 0); err = ocfs2_qinfo_lock(info, 1); if (err < 0) goto out_qlock; @@ -435,7 +436,8 @@ int ocfs2_global_read_dquot(struct dquot *dquot) out_qlock: if (ex) ocfs2_qinfo_unlock(info, 1); - ocfs2_qinfo_unlock(info, 0); + else + ocfs2_qinfo_unlock(info, 0); out: if (err < 0) mlog_errno(err); diff --git a/fs/ocfs2/quota_local.c b/fs/ocfs2/quota_local.c index 07deec5e9721..5a460fa82553 100644 --- a/fs/ocfs2/quota_local.c +++ b/fs/ocfs2/quota_local.c @@ -444,10 +444,6 @@ static int ocfs2_recover_local_quota_file(struct inode *lqinode, mlog_entry("ino=%lu type=%u", (unsigned long)lqinode->i_ino, type); - status = ocfs2_lock_global_qf(oinfo, 1); - if (status < 0) - goto out; - list_for_each_entry_safe(rchunk, next, &(rec->r_list[type]), rc_list) { chunk = rchunk->rc_chunk; hbh = NULL; @@ -480,12 +476,18 @@ static int ocfs2_recover_local_quota_file(struct inode *lqinode, type); goto out_put_bh; } + status = ocfs2_lock_global_qf(oinfo, 1); + if (status < 0) { + mlog_errno(status); + goto out_put_dquot; + } + handle = ocfs2_start_trans(OCFS2_SB(sb), OCFS2_QSYNC_CREDITS); if (IS_ERR(handle)) { status = PTR_ERR(handle); mlog_errno(status); - goto out_put_dquot; + goto out_drop_lock; } mutex_lock(&sb_dqopt(sb)->dqio_mutex); spin_lock(&dq_data_lock); @@ -523,6 +525,8 @@ static int ocfs2_recover_local_quota_file(struct inode *lqinode, out_commit: mutex_unlock(&sb_dqopt(sb)->dqio_mutex); ocfs2_commit_trans(OCFS2_SB(sb), handle); +out_drop_lock: + ocfs2_unlock_global_qf(oinfo, 1); out_put_dquot: dqput(dquot); out_put_bh: @@ -537,8 +541,6 @@ out_put_bh: if (status < 0) break; } - ocfs2_unlock_global_qf(oinfo, 1); -out: if (status < 0) free_recovery_list(&(rec->r_list[type])); mlog_exit(status); @@ -655,6 +657,9 @@ static int ocfs2_local_read_info(struct super_block *sb, int type) struct ocfs2_quota_recovery *rec; int locked = 0; + /* We don't need the lock and we have to acquire quota file locks + * which will later depend on this lock */ + mutex_unlock(&sb_dqopt(sb)->dqio_mutex); info->dqi_maxblimit = 0x7fffffffffffffffLL; info->dqi_maxilimit = 0x7fffffffffffffffLL; oinfo = kmalloc(sizeof(struct ocfs2_mem_dqinfo), GFP_NOFS); @@ -733,6 +738,7 @@ static int ocfs2_local_read_info(struct super_block *sb, int type) goto out_err; } + mutex_lock(&sb_dqopt(sb)->dqio_mutex); return 0; out_err: if (oinfo) { @@ -746,6 +752,7 @@ out_err: kfree(oinfo); } brelse(bh); + mutex_lock(&sb_dqopt(sb)->dqio_mutex); return -1; } diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 201b40a441fe..d33767f17ba3 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -119,10 +119,12 @@ static void ocfs2_release_system_inodes(struct ocfs2_super *osb); static int ocfs2_check_volume(struct ocfs2_super *osb); static int ocfs2_verify_volume(struct ocfs2_dinode *di, struct buffer_head *bh, - u32 sectsize); + u32 sectsize, + struct ocfs2_blockcheck_stats *stats); static int ocfs2_initialize_super(struct super_block *sb, struct buffer_head *bh, - int sector_size); + int sector_size, + struct ocfs2_blockcheck_stats *stats); static int ocfs2_get_sector(struct super_block *sb, struct buffer_head **bh, int block, @@ -207,6 +209,7 @@ static int ocfs2_osb_dump(struct ocfs2_super *osb, char *buf, int len) int i; struct ocfs2_cluster_connection *cconn = osb->cconn; struct ocfs2_recovery_map *rm = osb->recovery_map; + struct ocfs2_orphan_scan *os; out += snprintf(buf + out, len - out, "%10s => Id: %-s Uuid: %-s Gen: 0x%X Label: %-s\n", @@ -308,6 +311,13 @@ static int ocfs2_osb_dump(struct ocfs2_super *osb, char *buf, int len) i, osb->slot_recovery_generations[i]); } + os = &osb->osb_orphan_scan; + out += snprintf(buf + out, len - out, "Orphan Scan=> "); + out += snprintf(buf + out, len - out, "Local: %u Global: %u ", + os->os_count, os->os_seqno); + out += snprintf(buf + out, len - out, " Last Scan: %lu seconds ago\n", + (get_seconds() - os->os_scantime.tv_sec)); + return out; } @@ -693,7 +703,8 @@ out: static int ocfs2_sb_probe(struct super_block *sb, struct buffer_head **bh, - int *sector_size) + int *sector_size, + struct ocfs2_blockcheck_stats *stats) { int status, tmpstat; struct ocfs1_vol_disk_hdr *hdr; @@ -759,7 +770,8 @@ static int ocfs2_sb_probe(struct super_block *sb, goto bail; } di = (struct ocfs2_dinode *) (*bh)->b_data; - status = ocfs2_verify_volume(di, *bh, blksize); + memset(stats, 0, sizeof(struct ocfs2_blockcheck_stats)); + status = ocfs2_verify_volume(di, *bh, blksize, stats); if (status >= 0) goto bail; brelse(*bh); @@ -965,6 +977,7 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent) struct ocfs2_super *osb = NULL; struct buffer_head *bh = NULL; char nodestr[8]; + struct ocfs2_blockcheck_stats stats; mlog_entry("%p, %p, %i", sb, data, silent); @@ -974,13 +987,13 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent) } /* probe for superblock */ - status = ocfs2_sb_probe(sb, &bh, §or_size); + status = ocfs2_sb_probe(sb, &bh, §or_size, &stats); if (status < 0) { mlog(ML_ERROR, "superblock probe failed!\n"); goto read_super_error; } - status = ocfs2_initialize_super(sb, bh, sector_size); + status = ocfs2_initialize_super(sb, bh, sector_size, &stats); osb = OCFS2_SB(sb); if (status < 0) { mlog_errno(status); @@ -1090,6 +1103,18 @@ static int ocfs2_fill_super(struct super_block *sb, void *data, int silent) goto read_super_error; } + if (ocfs2_meta_ecc(osb)) { + status = ocfs2_blockcheck_stats_debugfs_install( + &osb->osb_ecc_stats, + osb->osb_debug_root); + if (status) { + mlog(ML_ERROR, + "Unable to create blockcheck statistics " + "files\n"); + goto read_super_error; + } + } + status = ocfs2_mount_volume(sb); if (osb->root_inode) inode = igrab(osb->root_inode); @@ -1760,13 +1785,8 @@ static int ocfs2_mount_volume(struct super_block *sb) } status = ocfs2_truncate_log_init(osb); - if (status < 0) { + if (status < 0) mlog_errno(status); - goto leave; - } - - if (ocfs2_mount_local(osb)) - goto leave; leave: if (unlock_super) @@ -1796,6 +1816,8 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err) ocfs2_truncate_log_shutdown(osb); + ocfs2_orphan_scan_stop(osb); + /* This will disable recovery and flush any recovery work. */ ocfs2_recovery_exit(osb); @@ -1833,6 +1855,7 @@ static void ocfs2_dismount_volume(struct super_block *sb, int mnt_err) if (osb->cconn) ocfs2_dlm_shutdown(osb, hangup_needed); + ocfs2_blockcheck_stats_debugfs_remove(&osb->osb_ecc_stats); debugfs_remove(osb->osb_debug_root); if (hangup_needed) @@ -1880,7 +1903,8 @@ static int ocfs2_setup_osb_uuid(struct ocfs2_super *osb, const unsigned char *uu static int ocfs2_initialize_super(struct super_block *sb, struct buffer_head *bh, - int sector_size) + int sector_size, + struct ocfs2_blockcheck_stats *stats) { int status; int i, cbits, bbits; @@ -1939,6 +1963,9 @@ static int ocfs2_initialize_super(struct super_block *sb, atomic_set(&osb->alloc_stats.bg_allocs, 0); atomic_set(&osb->alloc_stats.bg_extends, 0); + /* Copy the blockcheck stats from the superblock probe */ + osb->osb_ecc_stats = *stats; + ocfs2_init_node_maps(osb); snprintf(osb->dev_str, sizeof(osb->dev_str), "%u,%u", @@ -1951,6 +1978,13 @@ static int ocfs2_initialize_super(struct super_block *sb, goto bail; } + status = ocfs2_orphan_scan_init(osb); + if (status) { + mlog(ML_ERROR, "Unable to initialize delayed orphan scan\n"); + mlog_errno(status); + goto bail; + } + init_waitqueue_head(&osb->checkpoint_event); atomic_set(&osb->needs_checkpoint, 0); @@ -2169,7 +2203,8 @@ bail: */ static int ocfs2_verify_volume(struct ocfs2_dinode *di, struct buffer_head *bh, - u32 blksz) + u32 blksz, + struct ocfs2_blockcheck_stats *stats) { int status = -EAGAIN; @@ -2182,7 +2217,8 @@ static int ocfs2_verify_volume(struct ocfs2_dinode *di, OCFS2_FEATURE_INCOMPAT_META_ECC) { status = ocfs2_block_check_validate(bh->b_data, bh->b_size, - &di->i_check); + &di->i_check, + stats); if (status) goto out; } diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index 15631019dc63..ba320e250747 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -3154,7 +3154,7 @@ static int ocfs2_iterate_xattr_buckets(struct inode *inode, le32_to_cpu(bucket_xh(bucket)->xh_entries[0].xe_name_hash)); if (func) { ret = func(inode, bucket, para); - if (ret) + if (ret && ret != -ERANGE) mlog_errno(ret); /* Fall through to bucket_relse() */ } @@ -3261,7 +3261,8 @@ static int ocfs2_xattr_tree_list_index_block(struct inode *inode, ocfs2_list_xattr_bucket, &xl); if (ret) { - mlog_errno(ret); + if (ret != -ERANGE) + mlog_errno(ret); goto out; } diff --git a/fs/sysfs/symlink.c b/fs/sysfs/symlink.c index a3ba217fbe74..1d897ad808e0 100644 --- a/fs/sysfs/symlink.c +++ b/fs/sysfs/symlink.c @@ -192,8 +192,11 @@ static void *sysfs_follow_link(struct dentry *dentry, struct nameidata *nd) { int error = -ENOMEM; unsigned long page = get_zeroed_page(GFP_KERNEL); - if (page) + if (page) { error = sysfs_getlink(dentry, (char *) page); + if (error < 0) + free_page((unsigned long)page); + } nd_set_link(nd, error ? ERR_PTR(error) : (char *)page); return NULL; } diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 3589eab02a2f..3260b73abe29 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -1937,6 +1937,9 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent) err = bdi_init(&c->bdi); if (err) goto out_close; + err = bdi_register(&c->bdi, NULL, "ubifs"); + if (err) + goto out_bdi; err = ubifs_parse_options(c, data, 0); if (err) |