From 4e4d6d860b9393c5395ba5920edb5b4c5d43a3a3 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sun, 18 Dec 2011 20:05:43 -0800 Subject: sysfs: Add s_hash to sysfs_dirent and order directory entries by hash Compute a 31 bit hash of directory entries (that can fit in a signed 32bit off_t) and index the sysfs directory entries by that hash, replacing the per directory indexes by name and by inode. Because we now only use a single rbtree this reduces the size of sysfs_dirent by 2 pointers. Because we have fewer cases to deal with the code is now simpler. For now I use the simple hash that the dcache uses as that is easy to use and seems simple enough. In addition to makeing the code simpler using a hash for the file position in readdir brings sysfs in line with other filesystems that have non-trivial directory structures. Signed-off-by: Eric W. Biederman Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/dir.c | 219 +++++++++++++++++++++++++++++-------------------------- fs/sysfs/sysfs.h | 9 +-- 2 files changed, 120 insertions(+), 108 deletions(-) (limited to 'fs') diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index 7fdf6a7b7436..0daf255b7bf9 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -22,76 +22,103 @@ #include #include #include +#include #include "sysfs.h" DEFINE_MUTEX(sysfs_mutex); DEFINE_SPINLOCK(sysfs_assoc_lock); +#define to_sysfs_dirent(X) rb_entry((X), struct sysfs_dirent, s_rb); + static DEFINE_SPINLOCK(sysfs_ino_lock); static DEFINE_IDA(sysfs_ino_ida); /** - * sysfs_link_sibling - link sysfs_dirent into sibling list + * sysfs_name_hash + * @ns: Namespace tag to hash + * @name: Null terminated string to hash + * + * Returns 31 bit hash of ns + name (so it fits in an off_t ) + */ +static unsigned int sysfs_name_hash(const void *ns, const char *name) +{ + unsigned long hash = init_name_hash(); + unsigned int len = strlen(name); + while (len--) + hash = partial_name_hash(*name++, hash); + hash = ( end_name_hash(hash) ^ hash_ptr( (void *)ns, 31 ) ); + hash &= 0x7fffffffU; + /* Reserve hash numbers 0, 1 and INT_MAX for magic directory entries */ + if (hash < 1) + hash += 2; + if (hash >= INT_MAX) + hash = INT_MAX - 1; + return hash; +} + +static int sysfs_name_compare(unsigned int hash, const void *ns, + const char *name, const struct sysfs_dirent *sd) +{ + if (hash != sd->s_hash) + return hash - sd->s_hash; + if (ns != sd->s_ns) + return ns - sd->s_ns; + return strcmp(name, sd->s_name); +} + +static int sysfs_sd_compare(const struct sysfs_dirent *left, + const struct sysfs_dirent *right) +{ + return sysfs_name_compare(left->s_hash, left->s_ns, left->s_name, + right); +} + +/** + * sysfs_link_subling - link sysfs_dirent into sibling rbtree * @sd: sysfs_dirent of interest * - * Link @sd into its sibling list which starts from + * Link @sd into its sibling rbtree which starts from * sd->s_parent->s_dir.children. * * Locking: * mutex_lock(sysfs_mutex) + * + * RETURNS: + * 0 on susccess -EEXIST on failure. */ -static void sysfs_link_sibling(struct sysfs_dirent *sd) +static int sysfs_link_sibling(struct sysfs_dirent *sd) { - struct sysfs_dirent *parent_sd = sd->s_parent; - - struct rb_node **p; - struct rb_node *parent; + struct rb_node **node = &sd->s_parent->s_dir.children.rb_node; + struct rb_node *parent = NULL; if (sysfs_type(sd) == SYSFS_DIR) - parent_sd->s_dir.subdirs++; - - p = &parent_sd->s_dir.inode_tree.rb_node; - parent = NULL; - while (*p) { - parent = *p; -#define node rb_entry(parent, struct sysfs_dirent, inode_node) - if (sd->s_ino < node->s_ino) { - p = &node->inode_node.rb_left; - } else if (sd->s_ino > node->s_ino) { - p = &node->inode_node.rb_right; - } else { - printk(KERN_CRIT "sysfs: inserting duplicate inode '%lx'\n", - (unsigned long) sd->s_ino); - BUG(); - } -#undef node - } - rb_link_node(&sd->inode_node, parent, p); - rb_insert_color(&sd->inode_node, &parent_sd->s_dir.inode_tree); - - p = &parent_sd->s_dir.name_tree.rb_node; - parent = NULL; - while (*p) { - int c; - parent = *p; -#define node rb_entry(parent, struct sysfs_dirent, name_node) - c = strcmp(sd->s_name, node->s_name); - if (c < 0) { - p = &node->name_node.rb_left; - } else { - p = &node->name_node.rb_right; - } -#undef node + sd->s_parent->s_dir.subdirs++; + + while (*node) { + struct sysfs_dirent *pos; + int result; + + pos = to_sysfs_dirent(*node); + parent = *node; + result = sysfs_sd_compare(sd, pos); + if (result < 0) + node = &pos->s_rb.rb_left; + else if (result > 0) + node = &pos->s_rb.rb_right; + else + return -EEXIST; } - rb_link_node(&sd->name_node, parent, p); - rb_insert_color(&sd->name_node, &parent_sd->s_dir.name_tree); + /* add new node and rebalance the tree */ + rb_link_node(&sd->s_rb, parent, node); + rb_insert_color(&sd->s_rb, &sd->s_parent->s_dir.children); + return 0; } /** - * sysfs_unlink_sibling - unlink sysfs_dirent from sibling list + * sysfs_unlink_sibling - unlink sysfs_dirent from sibling rbtree * @sd: sysfs_dirent of interest * - * Unlink @sd from its sibling list which starts from + * Unlink @sd from its sibling rbtree which starts from * sd->s_parent->s_dir.children. * * Locking: @@ -102,8 +129,7 @@ static void sysfs_unlink_sibling(struct sysfs_dirent *sd) if (sysfs_type(sd) == SYSFS_DIR) sd->s_parent->s_dir.subdirs--; - rb_erase(&sd->inode_node, &sd->s_parent->s_dir.inode_tree); - rb_erase(&sd->name_node, &sd->s_parent->s_dir.name_tree); + rb_erase(&sd->s_rb, &sd->s_parent->s_dir.children); } /** @@ -402,6 +428,7 @@ void sysfs_addrm_start(struct sysfs_addrm_cxt *acxt, int __sysfs_add_one(struct sysfs_addrm_cxt *acxt, struct sysfs_dirent *sd) { struct sysfs_inode_attrs *ps_iattr; + int ret; if (!!sysfs_ns_type(acxt->parent_sd) != !!sd->s_ns) { WARN(1, KERN_WARNING "sysfs: ns %s in '%s' for '%s'\n", @@ -410,12 +437,12 @@ int __sysfs_add_one(struct sysfs_addrm_cxt *acxt, struct sysfs_dirent *sd) return -EINVAL; } - if (sysfs_find_dirent(acxt->parent_sd, sd->s_ns, sd->s_name)) - return -EEXIST; - + sd->s_hash = sysfs_name_hash(sd->s_ns, sd->s_name); sd->s_parent = sysfs_get(acxt->parent_sd); - sysfs_link_sibling(sd); + ret = sysfs_link_sibling(sd); + if (ret) + return ret; /* Update timestamps on the parent */ ps_iattr = acxt->parent_sd->s_iattr; @@ -565,8 +592,8 @@ struct sysfs_dirent *sysfs_find_dirent(struct sysfs_dirent *parent_sd, const void *ns, const unsigned char *name) { - struct rb_node *p = parent_sd->s_dir.name_tree.rb_node; - struct sysfs_dirent *found = NULL; + struct rb_node *node = parent_sd->s_dir.children.rb_node; + unsigned int hash; if (!!sysfs_ns_type(parent_sd) != !!ns) { WARN(1, KERN_WARNING "sysfs: ns %s in '%s' for '%s'\n", @@ -575,33 +602,21 @@ struct sysfs_dirent *sysfs_find_dirent(struct sysfs_dirent *parent_sd, return NULL; } - while (p) { - int c; -#define node rb_entry(p, struct sysfs_dirent, name_node) - c = strcmp(name, node->s_name); - if (c < 0) { - p = node->name_node.rb_left; - } else if (c > 0) { - p = node->name_node.rb_right; - } else { - found = node; - p = node->name_node.rb_left; - } -#undef node - } - - if (found) { - while (found->s_ns != ns) { - p = rb_next(&found->name_node); - if (!p) - return NULL; - found = rb_entry(p, struct sysfs_dirent, name_node); - if (strcmp(name, found->s_name)) - return NULL; - } + hash = sysfs_name_hash(ns, name); + while (node) { + struct sysfs_dirent *sd; + int result; + + sd = to_sysfs_dirent(node); + result = sysfs_name_compare(hash, ns, name, sd); + if (result < 0) + node = node->rb_left; + else if (result > 0) + node = node->rb_right; + else + return sd; } - - return found; + return NULL; } /** @@ -804,9 +819,9 @@ static void __sysfs_remove_dir(struct sysfs_dirent *dir_sd) pr_debug("sysfs %s: removing dir\n", dir_sd->s_name); sysfs_addrm_start(&acxt, dir_sd); - pos = rb_first(&dir_sd->s_dir.inode_tree); + pos = rb_first(&dir_sd->s_dir.children); while (pos) { - struct sysfs_dirent *sd = rb_entry(pos, struct sysfs_dirent, inode_node); + struct sysfs_dirent *sd = to_sysfs_dirent(pos); pos = rb_next(pos); if (sysfs_type(sd) != SYSFS_DIR) sysfs_remove_one(&acxt, sd); @@ -919,38 +934,36 @@ static int sysfs_dir_release(struct inode *inode, struct file *filp) } static struct sysfs_dirent *sysfs_dir_pos(const void *ns, - struct sysfs_dirent *parent_sd, ino_t ino, struct sysfs_dirent *pos) + struct sysfs_dirent *parent_sd, loff_t hash, struct sysfs_dirent *pos) { if (pos) { int valid = !(pos->s_flags & SYSFS_FLAG_REMOVED) && pos->s_parent == parent_sd && - ino == pos->s_ino; + hash == pos->s_hash; sysfs_put(pos); if (!valid) pos = NULL; } - if (!pos && (ino > 1) && (ino < INT_MAX)) { - struct rb_node *p = parent_sd->s_dir.inode_tree.rb_node; - while (p) { -#define node rb_entry(p, struct sysfs_dirent, inode_node) - if (ino < node->s_ino) { - pos = node; - p = node->inode_node.rb_left; - } else if (ino > node->s_ino) { - p = node->inode_node.rb_right; - } else { - pos = node; + if (!pos && (hash > 1) && (hash < INT_MAX)) { + struct rb_node *node = parent_sd->s_dir.children.rb_node; + while (node) { + pos = to_sysfs_dirent(node); + + if (hash < pos->s_hash) + node = node->rb_left; + else if (hash > pos->s_hash) + node = node->rb_right; + else break; - } -#undef node } } + /* Skip over entries in the wrong namespace */ while (pos && pos->s_ns != ns) { - struct rb_node *p = rb_next(&pos->inode_node); - if (!p) + struct rb_node *node = rb_next(&pos->s_rb); + if (!node) pos = NULL; else - pos = rb_entry(p, struct sysfs_dirent, inode_node); + pos = to_sysfs_dirent(node); } return pos; } @@ -960,11 +973,11 @@ static struct sysfs_dirent *sysfs_dir_next_pos(const void *ns, { pos = sysfs_dir_pos(ns, parent_sd, ino, pos); if (pos) do { - struct rb_node *p = rb_next(&pos->inode_node); - if (!p) + struct rb_node *node = rb_next(&pos->s_rb); + if (!node) pos = NULL; else - pos = rb_entry(p, struct sysfs_dirent, inode_node); + pos = to_sysfs_dirent(node); } while (pos && pos->s_ns != ns); return pos; } @@ -1006,7 +1019,7 @@ static int sysfs_readdir(struct file * filp, void * dirent, filldir_t filldir) len = strlen(name); ino = pos->s_ino; type = dt_type(pos); - filp->f_pos = ino; + filp->f_pos = pos->s_hash; filp->private_data = sysfs_get(pos); mutex_unlock(&sysfs_mutex); diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index 7484a36ee678..2b5c923b4b90 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -20,9 +20,8 @@ struct sysfs_elem_dir { struct kobject *kobj; unsigned long subdirs; - - struct rb_root inode_tree; - struct rb_root name_tree; + /* children rbtree starts here and goes through sd->s_rb */ + struct rb_root children; }; struct sysfs_elem_symlink { @@ -62,8 +61,7 @@ struct sysfs_dirent { struct sysfs_dirent *s_parent; const char *s_name; - struct rb_node inode_node; - struct rb_node name_node; + struct rb_node s_rb; union { struct completion *completion; @@ -71,6 +69,7 @@ struct sysfs_dirent { } u; const void *s_ns; /* namespace tag */ + unsigned int s_hash; /* ns + name hash */ union { struct sysfs_elem_dir s_dir; struct sysfs_elem_symlink s_symlink; -- cgit v1.2.3 From 15a3382451e51925facfe430deeca63d90137f5d Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sun, 18 Dec 2011 20:07:23 -0800 Subject: sysfs: Reduce s_flags to an unsinged short so it packs well with s_mode On 32bit this reduces sizeof(struct sysfs_dirent) by 2 bytes. Signed-off-by: Eric W. Biederman Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/sysfs.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index 2b5c923b4b90..19994948ac5c 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -77,7 +77,7 @@ struct sysfs_dirent { struct sysfs_elem_bin_attr s_bin_attr; }; - unsigned int s_flags; + unsigned short s_flags; umode_t s_mode; ino_t s_ino; struct sysfs_inode_attrs *s_iattr; @@ -94,11 +94,11 @@ struct sysfs_dirent { #define SYSFS_ACTIVE_REF (SYSFS_KOBJ_ATTR | SYSFS_KOBJ_BIN_ATTR) /* identify any namespace tag on sysfs_dirents */ -#define SYSFS_NS_TYPE_MASK 0xff00 +#define SYSFS_NS_TYPE_MASK 0xf00 #define SYSFS_NS_TYPE_SHIFT 8 #define SYSFS_FLAG_MASK ~(SYSFS_NS_TYPE_MASK|SYSFS_TYPE_MASK) -#define SYSFS_FLAG_REMOVED 0x020000 +#define SYSFS_FLAG_REMOVED 0x02000 static inline unsigned int sysfs_type(struct sysfs_dirent *sd) { -- cgit v1.2.3 From cafa6b5dd7ce4f0e0a30be301be4efed587a7808 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sun, 18 Dec 2011 20:08:16 -0800 Subject: sysfs: Store the sysfs inode in an unsigned int. Store the sysfs inode number in an unsided int because ida inode allocator can return at most a 31 bit number, reducing the size of struct sysfs_dirent by 8 bytes on 64bit platforms. Signed-off-by: Eric W. Biederman Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/dir.c | 4 ++-- fs/sysfs/sysfs.h | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) (limited to 'fs') diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index 0daf255b7bf9..0589c9a694bf 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -224,7 +224,7 @@ static void sysfs_deactivate(struct sysfs_dirent *sd) rwsem_release(&sd->dep_map, 1, _RET_IP_); } -static int sysfs_alloc_ino(ino_t *pino) +static int sysfs_alloc_ino(unsigned int *pino) { int ino, rc; @@ -243,7 +243,7 @@ static int sysfs_alloc_ino(ino_t *pino) return rc; } -static void sysfs_free_ino(ino_t ino) +static void sysfs_free_ino(unsigned int ino) { spin_lock(&sysfs_ino_lock); ida_remove(&sysfs_ino_ida, ino); diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index 19994948ac5c..661a9639570b 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -79,7 +79,7 @@ struct sysfs_dirent { unsigned short s_flags; umode_t s_mode; - ino_t s_ino; + unsigned int s_ino; struct sysfs_inode_attrs *s_iattr; }; -- cgit v1.2.3 From 524b6c5b39b931311dfe5a2f5abae2f5c9731676 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sun, 18 Dec 2011 20:09:31 -0800 Subject: sysfs: Kill nlink counting. Tracking the number of subdirectories requires an extra field that increases the size of sysfs_dirent. nlinks are not particularly interesting for sysfs and the nlink counts are wrong when network namespaces are involved so stop counting them, and always return nlink == 1. Userspace already knows that directories with nlink == 1 have an nlink count they can't use to count subdirectories. This reduces the size of sysfs_dirent by 8 bytes on 64bit platforms. Signed-off-by: Eric W. Biederman Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/dir.c | 6 ------ fs/sysfs/inode.c | 3 --- fs/sysfs/sysfs.h | 1 - 3 files changed, 10 deletions(-) (limited to 'fs') diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index 0589c9a694bf..ea64d01400ac 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -91,9 +91,6 @@ static int sysfs_link_sibling(struct sysfs_dirent *sd) struct rb_node **node = &sd->s_parent->s_dir.children.rb_node; struct rb_node *parent = NULL; - if (sysfs_type(sd) == SYSFS_DIR) - sd->s_parent->s_dir.subdirs++; - while (*node) { struct sysfs_dirent *pos; int result; @@ -126,9 +123,6 @@ static int sysfs_link_sibling(struct sysfs_dirent *sd) */ static void sysfs_unlink_sibling(struct sysfs_dirent *sd) { - if (sysfs_type(sd) == SYSFS_DIR) - sd->s_parent->s_dir.subdirs--; - rb_erase(&sd->s_rb, &sd->s_parent->s_dir.children); } diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c index 4a802b4a9056..0ac3e1c1a7d8 100644 --- a/fs/sysfs/inode.c +++ b/fs/sysfs/inode.c @@ -216,9 +216,6 @@ static void sysfs_refresh_inode(struct sysfs_dirent *sd, struct inode *inode) iattrs->ia_secdata, iattrs->ia_secdata_len); } - - if (sysfs_type(sd) == SYSFS_DIR) - set_nlink(inode, sd->s_dir.subdirs + 2); } int sysfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index 661a9639570b..6289a00287db 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -19,7 +19,6 @@ struct sysfs_open_dirent; struct sysfs_elem_dir { struct kobject *kobj; - unsigned long subdirs; /* children rbtree starts here and goes through sd->s_rb */ struct rb_root children; }; -- cgit v1.2.3 From c56d8a7362665d165ba992b6b7a8d6c13a26eafc Mon Sep 17 00:00:00 2001 From: Vitaly Kuznetsov Date: Tue, 17 Jan 2012 12:17:22 +0000 Subject: sysfs: change permissions for /sys from 0755 to 0555 There is a misleading difference between /proc and /sys permissions, /proc is 0555 and /sys is 0755. But as it is impossible to create or unlink something in /sys it would be nice to have same permissions. Signed-off-by: Vitaly Kuznetsov Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/mount.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs') diff --git a/fs/sysfs/mount.c b/fs/sysfs/mount.c index e34f0d99ea4e..140f26a34288 100644 --- a/fs/sysfs/mount.c +++ b/fs/sysfs/mount.c @@ -36,7 +36,7 @@ struct sysfs_dirent sysfs_root = { .s_name = "", .s_count = ATOMIC_INIT(1), .s_flags = SYSFS_DIR | (KOBJ_NS_TYPE_NONE << SYSFS_NS_TYPE_SHIFT), - .s_mode = S_IFDIR | S_IRWXU | S_IRUGO | S_IXUGO, + .s_mode = S_IFDIR | S_IRUGO | S_IXUGO, .s_ino = 1, }; -- cgit v1.2.3 From d6e486868cde585842d55ba3b6ec57af090fc343 Mon Sep 17 00:00:00 2001 From: Ludwig Nussel Date: Wed, 25 Jan 2012 11:52:28 +0100 Subject: debugfs: add mode, uid and gid options Cautious admins may want to restrict access to debugfs. Currently a manual chown/chmod e.g. in an init script is needed to achieve that. Distributions that want to make the mount options configurable need to add extra config files. By allowing to set the root inode's uid, gid and mode via mount options no such hacks are needed anymore. Instead configuration becomes straight forward via fstab. Signed-off-by: Ludwig Nussel Signed-off-by: Greg Kroah-Hartman --- Documentation/filesystems/debugfs.txt | 5 +- fs/debugfs/inode.c | 149 +++++++++++++++++++++++++++++++++- 2 files changed, 152 insertions(+), 2 deletions(-) (limited to 'fs') diff --git a/Documentation/filesystems/debugfs.txt b/Documentation/filesystems/debugfs.txt index 6872c91bce35..4e2575873187 100644 --- a/Documentation/filesystems/debugfs.txt +++ b/Documentation/filesystems/debugfs.txt @@ -14,7 +14,10 @@ Debugfs is typically mounted with a command like: mount -t debugfs none /sys/kernel/debug -(Or an equivalent /etc/fstab line). +(Or an equivalent /etc/fstab line). +The debugfs root directory is accessible by anyone by default. To +restrict access to the tree the "uid", "gid" and "mode" mount +options can be used. Note that the debugfs API is exported GPL-only to modules. diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index 956d5ddddf6e..b80bc846a15a 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -23,9 +23,13 @@ #include #include #include +#include +#include #include #include +#define DEBUGFS_DEFAULT_MODE 0755 + static struct vfsmount *debugfs_mount; static int debugfs_mount_count; static bool debugfs_registered; @@ -125,11 +129,154 @@ static inline int debugfs_positive(struct dentry *dentry) return dentry->d_inode && !d_unhashed(dentry); } +struct debugfs_mount_opts { + uid_t uid; + gid_t gid; + umode_t mode; +}; + +enum { + Opt_uid, + Opt_gid, + Opt_mode, + Opt_err +}; + +static const match_table_t tokens = { + {Opt_uid, "uid=%u"}, + {Opt_gid, "gid=%u"}, + {Opt_mode, "mode=%o"}, + {Opt_err, NULL} +}; + +struct debugfs_fs_info { + struct debugfs_mount_opts mount_opts; +}; + +static int debugfs_parse_options(char *data, struct debugfs_mount_opts *opts) +{ + substring_t args[MAX_OPT_ARGS]; + int option; + int token; + char *p; + + opts->mode = DEBUGFS_DEFAULT_MODE; + + while ((p = strsep(&data, ",")) != NULL) { + if (!*p) + continue; + + token = match_token(p, tokens, args); + switch (token) { + case Opt_uid: + if (match_int(&args[0], &option)) + return -EINVAL; + opts->uid = option; + break; + case Opt_gid: + if (match_octal(&args[0], &option)) + return -EINVAL; + opts->gid = option; + break; + case Opt_mode: + if (match_octal(&args[0], &option)) + return -EINVAL; + opts->mode = option & S_IALLUGO; + break; + /* + * We might like to report bad mount options here; + * but traditionally debugfs has ignored all mount options + */ + } + } + + return 0; +} + +static int debugfs_apply_options(struct super_block *sb) +{ + struct debugfs_fs_info *fsi = sb->s_fs_info; + struct inode *inode = sb->s_root->d_inode; + struct debugfs_mount_opts *opts = &fsi->mount_opts; + + inode->i_mode &= ~S_IALLUGO; + inode->i_mode |= opts->mode; + + inode->i_uid = opts->uid; + inode->i_gid = opts->gid; + + return 0; +} + +static int debugfs_remount(struct super_block *sb, int *flags, char *data) +{ + int err; + struct debugfs_fs_info *fsi = sb->s_fs_info; + + err = debugfs_parse_options(data, &fsi->mount_opts); + if (err) + goto fail; + + debugfs_apply_options(sb); + +fail: + return err; +} + +static int debugfs_show_options(struct seq_file *m, struct dentry *root) +{ + struct debugfs_fs_info *fsi = root->d_sb->s_fs_info; + struct debugfs_mount_opts *opts = &fsi->mount_opts; + + if (opts->uid != 0) + seq_printf(m, ",uid=%u", opts->uid); + if (opts->gid != 0) + seq_printf(m, ",gid=%u", opts->gid); + if (opts->mode != DEBUGFS_DEFAULT_MODE) + seq_printf(m, ",mode=%o", opts->mode); + + return 0; +} + +static const struct super_operations debugfs_super_operations = { + .statfs = simple_statfs, + .remount_fs = debugfs_remount, + .show_options = debugfs_show_options, +}; + static int debug_fill_super(struct super_block *sb, void *data, int silent) { static struct tree_descr debug_files[] = {{""}}; + struct debugfs_fs_info *fsi; + int err; + + save_mount_options(sb, data); + + fsi = kzalloc(sizeof(struct debugfs_fs_info), GFP_KERNEL); + sb->s_fs_info = fsi; + if (!fsi) { + err = -ENOMEM; + goto fail; + } + + err = debugfs_parse_options(data, &fsi->mount_opts); + if (err) + goto fail; + + err = simple_fill_super(sb, DEBUGFS_MAGIC, debug_files); + if (err) + goto fail; + + sb->s_op = &debugfs_super_operations; + + debugfs_apply_options(sb); + + return 0; - return simple_fill_super(sb, DEBUGFS_MAGIC, debug_files); +fail: + kfree(fsi); + sb->s_fs_info = NULL; + return err; } static struct dentry *debug_mount(struct file_system_type *fs_type, -- cgit v1.2.3 From d5c38b137ac8a6e3dbed13bc494d60df5b69dfc4 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Tue, 31 Jan 2012 06:40:26 -0800 Subject: sysfs: Update the name hash when renaming sysfs entries This fixes a bug introduced with sysfs name hashes where renaming a network device appears to succeed but silently makes the sysfs files for that network device inaccessible. In at least one configuration this bug has stopped networking from coming up during boot. Signed-off-by: Eric W. Biederman Tested-by: Jiri Slaby Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/dir.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs') diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index ea64d01400ac..dd3779cf3a3b 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -872,6 +872,7 @@ int sysfs_rename(struct sysfs_dirent *sd, dup_name = sd->s_name; sd->s_name = new_name; + sd->s_hash = sysfs_name_hash(sd->s_ns, sd->s_name); } /* Move to the appropriate place in the appropriate directories rbtree. */ -- cgit v1.2.3 From 93518dd2ebafcc761a8637b2877008cfd748c202 Mon Sep 17 00:00:00 2001 From: Masami Ichikawa Date: Tue, 21 Feb 2012 07:43:50 +0900 Subject: sysfs: Fix memory leak in sysfs_sd_setsecdata(). This patch fixies follwing two memory leak patterns that reported by kmemleak. sysfs_sd_setsecdata() is called during sys_lsetxattr() operation. It checks sd->s_iattr is NULL or not. Then if it is NULL, it calls sysfs_init_inode_attrs() to allocate memory. That code is this. iattrs = sd->s_iattr; if (!iattrs) iattrs = sysfs_init_inode_attrs(sd); The iattrs recieves sysfs_init_inode_attrs()'s result, but sd->s_iattr doesn't know the address. so it needs to set correct address to sd->s_iattr to free memory in other function. unreferenced object 0xffff880250b73e60 (size 32): comm "systemd", pid 1, jiffies 4294683888 (age 94.553s) hex dump (first 32 bytes): 73 79 73 74 65 6d 5f 75 3a 6f 62 6a 65 63 74 5f system_u:object_ 72 3a 73 79 73 66 73 5f 74 3a 73 30 00 00 00 00 r:sysfs_t:s0.... backtrace: [] kmemleak_alloc+0x73/0x98 [] __kmalloc+0x100/0x12c [] context_struct_to_string+0x106/0x210 [] security_sid_to_context_core+0x10b/0x129 [] security_sid_to_context+0x10/0x12 [] selinux_inode_getsecurity+0x7d/0xa8 [] selinux_inode_getsecctx+0x22/0x2e [] security_inode_getsecctx+0x16/0x18 [] sysfs_setxattr+0x96/0x117 [] __vfs_setxattr_noperm+0x73/0xd9 [] vfs_setxattr+0x83/0xa1 [] setxattr+0xcf/0x101 [] sys_lsetxattr+0x6a/0x8f [] system_call_fastpath+0x16/0x1b [] 0xffffffffffffffff unreferenced object 0xffff88024163c5a0 (size 96): comm "systemd", pid 1, jiffies 4294683888 (age 94.553s) hex dump (first 32 bytes): 00 00 00 00 ed 41 00 00 00 00 00 00 00 00 00 00 .....A.......... 00 00 00 00 00 00 00 00 0c 64 42 4f 00 00 00 00 .........dBO.... backtrace: [] kmemleak_alloc+0x73/0x98 [] kmem_cache_alloc_trace+0xc4/0xee [] sysfs_init_inode_attrs+0x2a/0x83 [] sysfs_setxattr+0xbf/0x117 [] __vfs_setxattr_noperm+0x73/0xd9 [] vfs_setxattr+0x83/0xa1 [] setxattr+0xcf/0x101 [] sys_lsetxattr+0x6a/0x8f [] system_call_fastpath+0x16/0x1b [] 0xffffffffffffffff ` Signed-off-by: Masami Ichikawa Cc: stable Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/inode.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) (limited to 'fs') diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c index 4291fd1617ab..cc7ea5de2fdd 100644 --- a/fs/sysfs/inode.c +++ b/fs/sysfs/inode.c @@ -136,12 +136,13 @@ static int sysfs_sd_setsecdata(struct sysfs_dirent *sd, void **secdata, u32 *sec void *old_secdata; size_t old_secdata_len; - iattrs = sd->s_iattr; - if (!iattrs) - iattrs = sysfs_init_inode_attrs(sd); - if (!iattrs) - return -ENOMEM; + if (!sd->s_iattr) { + sd->s_iattr = sysfs_init_inode_attrs(sd); + if (!sd->s_iattr) + return -ENOMEM; + } + iattrs = sd->s_iattr; old_secdata = iattrs->ia_secdata; old_secdata_len = iattrs->ia_secdata_len; -- cgit v1.2.3 From 54d20f006ceff1f2f1e69d0e54049b6c0765c039 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Thu, 8 Mar 2012 13:03:10 -0800 Subject: Revert "sysfs: Kill nlink counting." This reverts commit 524b6c5b39b931311dfe5a2f5abae2f5c9731676. It has shown to break userspace tools, which is not acceptable. Reported-by: Jiri Slaby Cc: Eric W. Biederman Cc: Linus Torvalds Signed-off-by: Greg Kroah-Hartman --- fs/sysfs/dir.c | 6 ++++++ fs/sysfs/inode.c | 3 +++ fs/sysfs/sysfs.h | 1 + 3 files changed, 10 insertions(+) (limited to 'fs') diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c index dd3779cf3a3b..2a7a3f5d1ca6 100644 --- a/fs/sysfs/dir.c +++ b/fs/sysfs/dir.c @@ -91,6 +91,9 @@ static int sysfs_link_sibling(struct sysfs_dirent *sd) struct rb_node **node = &sd->s_parent->s_dir.children.rb_node; struct rb_node *parent = NULL; + if (sysfs_type(sd) == SYSFS_DIR) + sd->s_parent->s_dir.subdirs++; + while (*node) { struct sysfs_dirent *pos; int result; @@ -123,6 +126,9 @@ static int sysfs_link_sibling(struct sysfs_dirent *sd) */ static void sysfs_unlink_sibling(struct sysfs_dirent *sd) { + if (sysfs_type(sd) == SYSFS_DIR) + sd->s_parent->s_dir.subdirs--; + rb_erase(&sd->s_rb, &sd->s_parent->s_dir.children); } diff --git a/fs/sysfs/inode.c b/fs/sysfs/inode.c index cc7ea5de2fdd..feb2d69396cf 100644 --- a/fs/sysfs/inode.c +++ b/fs/sysfs/inode.c @@ -217,6 +217,9 @@ static void sysfs_refresh_inode(struct sysfs_dirent *sd, struct inode *inode) iattrs->ia_secdata, iattrs->ia_secdata_len); } + + if (sysfs_type(sd) == SYSFS_DIR) + set_nlink(inode, sd->s_dir.subdirs + 2); } int sysfs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) diff --git a/fs/sysfs/sysfs.h b/fs/sysfs/sysfs.h index 6289a00287db..661a9639570b 100644 --- a/fs/sysfs/sysfs.h +++ b/fs/sysfs/sysfs.h @@ -19,6 +19,7 @@ struct sysfs_open_dirent; struct sysfs_elem_dir { struct kobject *kobj; + unsigned long subdirs; /* children rbtree starts here and goes through sd->s_rb */ struct rb_root children; }; -- cgit v1.2.3