From 734550921e9b7ab924a43aa3d0bd4239dac4fbf1 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Mon, 14 Jul 2008 21:22:20 -0400 Subject: [PATCH] beginning of sysctl cleanup - ctl_table_set New object: set of sysctls [currently - root and per-net-ns]. Contains: pointer to parent set, list of tables and "should I see this set?" method (->is_seen(set)). Current lists of tables are subsumed by that; net-ns contains such a beast. ->lookup() for ctl_table_root returns pointer to ctl_table_set instead of that to ->list of that ctl_table_set. [folded compile fixes by rdd for configs without sysctl] Signed-off-by: Al Viro --- include/linux/sysctl.h | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) (limited to 'include/linux/sysctl.h') diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h index 24141b4d1a1..c1e0cf408af 100644 --- a/include/linux/sysctl.h +++ b/include/linux/sysctl.h @@ -947,6 +947,16 @@ struct ctl_table; struct nsproxy; struct ctl_table_root; +struct ctl_table_set { + struct list_head list; + struct ctl_table_set *parent; + int (*is_seen)(struct ctl_table_set *); +}; + +extern void setup_sysctl_set(struct ctl_table_set *p, + struct ctl_table_set *parent, + int (*is_seen)(struct ctl_table_set *)); + extern struct ctl_table_header *sysctl_head_next(struct ctl_table_header *prev); extern struct ctl_table_header *__sysctl_head_next(struct nsproxy *namespaces, struct ctl_table_header *prev); @@ -1049,8 +1059,8 @@ struct ctl_table struct ctl_table_root { struct list_head root_list; - struct list_head header_list; - struct list_head *(*lookup)(struct ctl_table_root *root, + struct ctl_table_set default_set; + struct ctl_table_set *(*lookup)(struct ctl_table_root *root, struct nsproxy *namespaces); int (*permissions)(struct ctl_table_root *root, struct nsproxy *namespaces, struct ctl_table *table); @@ -1066,6 +1076,7 @@ struct ctl_table_header struct completion *unregistering; struct ctl_table *ctl_table_arg; struct ctl_table_root *root; + struct ctl_table_set *set; }; /* struct ctl_path describes where in the hierarchy a table is added */ -- cgit v1.2.3-70-g09d2 From f7e6ced4061da509f737541ca4dbd44d83a6e82f Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 15 Jul 2008 01:44:23 -0400 Subject: [PATCH] allow delayed freeing of ctl_table_header Refcount the sucker; instead of freeing it by the end of unregistration just drop the refcount and free only when it hits zero. Make sure that we _always_ make ->unregistering non-NULL in start_unregistering(). That allows anybody to get a reference to such puppy, preventing its freeing and reuse. It does *not* block unregistration. Anybody who holds such a reference can * try to grab a "use" reference (ctl_head_grab()); that will succeeds if and only if it hadn't entered unregistration yet. If it succeeds, we can use it in all normal ways until we release the "use" reference (with ctl_head_finish()). Note that this relies on having ->unregistering become non-NULL in all cases when one starts to unregister the sucker. * keep pointers to ctl_table entries; they *can* be freed if the entire thing is unregistered. However, if ctl_head_grab() succeeds, we know that unregistration had not happened (and will not happen until ctl_head_finish()) and such pointers can be used safely. IOW, now we can have inodes under /proc/sys keep references to ctl_table entries, protecting them with references to ctl_table_header and grabbing the latter for the duration of operations that require access to ctl_table. That won't cause deadlocks, since unregistration will not be stopped by mere keeping a reference to ctl_table_header. Signed-off-by: Al Viro --- include/linux/sysctl.h | 6 ++++++ kernel/sysctl.c | 37 ++++++++++++++++++++++++++++++++++++- 2 files changed, 42 insertions(+), 1 deletion(-) (limited to 'include/linux/sysctl.h') diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h index c1e0cf408af..956264d09ba 100644 --- a/include/linux/sysctl.h +++ b/include/linux/sysctl.h @@ -957,6 +957,11 @@ extern void setup_sysctl_set(struct ctl_table_set *p, struct ctl_table_set *parent, int (*is_seen)(struct ctl_table_set *)); +struct ctl_table_header; + +extern void sysctl_head_get(struct ctl_table_header *); +extern void sysctl_head_put(struct ctl_table_header *); +extern struct ctl_table_header *sysctl_head_grab(struct ctl_table_header *); extern struct ctl_table_header *sysctl_head_next(struct ctl_table_header *prev); extern struct ctl_table_header *__sysctl_head_next(struct nsproxy *namespaces, struct ctl_table_header *prev); @@ -1073,6 +1078,7 @@ struct ctl_table_header struct ctl_table *ctl_table; struct list_head ctl_entry; int used; + int count; struct completion *unregistering; struct ctl_table *ctl_table_arg; struct ctl_table_root *root; diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 8ee4a0619fb..60d9357e717 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -1387,6 +1387,9 @@ static void start_unregistering(struct ctl_table_header *p) spin_unlock(&sysctl_lock); wait_for_completion(&wait); spin_lock(&sysctl_lock); + } else { + /* anything non-NULL; we'll never dereference it */ + p->unregistering = ERR_PTR(-EINVAL); } /* * do not remove from the list until nobody holds it; walking the @@ -1395,6 +1398,32 @@ static void start_unregistering(struct ctl_table_header *p) list_del_init(&p->ctl_entry); } +void sysctl_head_get(struct ctl_table_header *head) +{ + spin_lock(&sysctl_lock); + head->count++; + spin_unlock(&sysctl_lock); +} + +void sysctl_head_put(struct ctl_table_header *head) +{ + spin_lock(&sysctl_lock); + if (!--head->count) + kfree(head); + spin_unlock(&sysctl_lock); +} + +struct ctl_table_header *sysctl_head_grab(struct ctl_table_header *head) +{ + if (!head) + BUG(); + spin_lock(&sysctl_lock); + if (!use_table(head)) + head = ERR_PTR(-ENOENT); + spin_unlock(&sysctl_lock); + return head; +} + void sysctl_head_finish(struct ctl_table_header *head) { if (!head) @@ -1771,6 +1800,7 @@ struct ctl_table_header *__register_sysctl_paths( header->unregistering = NULL; header->root = root; sysctl_set_parent(NULL, header->ctl_table); + header->count = 1; #ifdef CONFIG_SYSCTL_SYSCALL_CHECK if (sysctl_check_table(namespaces, header->ctl_table)) { kfree(header); @@ -1834,8 +1864,9 @@ void unregister_sysctl_table(struct ctl_table_header * header) spin_lock(&sysctl_lock); start_unregistering(header); + if (!--header->count) + kfree(header); spin_unlock(&sysctl_lock); - kfree(header); } void setup_sysctl_set(struct ctl_table_set *p, @@ -1869,6 +1900,10 @@ void setup_sysctl_set(struct ctl_table_set *p, { } +void sysctl_head_put(struct ctl_table_header *head) +{ +} + #endif /* CONFIG_SYSCTL */ /* -- cgit v1.2.3-70-g09d2 From ae7edecc9b8810770a8e5cb9a466ea4bdcfa8401 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 15 Jul 2008 06:33:31 -0400 Subject: [PATCH] sysctl: keep track of tree relationships In a sense, that's the heart of the series. It's based on the following property of the trees we are actually asked to add: they can be split into stem that is already covered by registered trees and crown that is entirely new. IOW, if a/b and a/c/d are introduced by our tree, then a/c is also introduced by it. That allows to associate tree and table entry with each node in the union; while directory nodes might be covered by many trees, only one will cover the node by its crown. And that will allow much saner logics for /proc/sys in the next patches. This patch introduces the data structures needed to keep track of that. When adding a sysctl table, we find a "parent" one. Which is to say, find the deepest node on its stem that already is present in one of the tables from our table set or its ancestor sets. That table will be our parent and that node in it - attachment point. Add our table to list anchored in parent, have it refer the parent and contents of attachment point. Also remember where its crown lives. Signed-off-by: Al Viro --- include/linux/sysctl.h | 3 +++ kernel/sysctl.c | 63 ++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 66 insertions(+) (limited to 'include/linux/sysctl.h') diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h index 956264d09ba..3f6599aeb0d 100644 --- a/include/linux/sysctl.h +++ b/include/linux/sysctl.h @@ -1083,6 +1083,9 @@ struct ctl_table_header struct ctl_table *ctl_table_arg; struct ctl_table_root *root; struct ctl_table_set *set; + struct ctl_table *attached_by; + struct ctl_table *attached_to; + struct ctl_table_header *parent; }; /* struct ctl_path describes where in the hierarchy a table is added */ diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 60d9357e717..c9a0af88703 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -1680,6 +1680,52 @@ static __init int sysctl_init(void) core_initcall(sysctl_init); +static int is_branch_in(struct ctl_table *branch, struct ctl_table *table) +{ + struct ctl_table *p; + const char *s = branch->procname; + + /* branch should have named subdirectory as its first element */ + if (!s || !branch->child) + return 0; + + /* ... and nothing else */ + if (branch[1].procname || branch[1].ctl_name) + return 0; + + /* table should contain subdirectory with the same name */ + for (p = table; p->procname || p->ctl_name; p++) { + if (!p->child) + continue; + if (p->procname && strcmp(p->procname, s) == 0) + return 1; + } + return 0; +} + +/* see if attaching q to p would be an improvement */ +static void try_attach(struct ctl_table_header *p, struct ctl_table_header *q) +{ + struct ctl_table *to = p->ctl_table, *by = q->ctl_table; + int is_better = 0; + int not_in_parent = !p->attached_by; + + while (is_branch_in(by, to)) { + if (by == q->attached_by) + is_better = 1; + if (to == p->attached_by) + not_in_parent = 1; + by = by->child; + to = to->child; + } + + if (is_better && not_in_parent) { + q->attached_by = by; + q->attached_to = to; + q->parent = p; + } +} + /** * __register_sysctl_paths - register a sysctl hierarchy * @root: List of sysctl headers to register on @@ -1759,6 +1805,7 @@ struct ctl_table_header *__register_sysctl_paths( struct ctl_table_header *header; struct ctl_table *new, **prevp; unsigned int n, npath; + struct ctl_table_set *set; /* Count the path components */ for (npath = 0; path[npath].ctl_name || path[npath].procname; ++npath) @@ -1809,6 +1856,18 @@ struct ctl_table_header *__register_sysctl_paths( #endif spin_lock(&sysctl_lock); header->set = lookup_header_set(root, namespaces); + header->attached_by = header->ctl_table; + header->attached_to = root_table; + header->parent = &root_table_header; + for (set = header->set; set; set = set->parent) { + struct ctl_table_header *p; + list_for_each_entry(p, &set->list, ctl_entry) { + if (p->unregistering) + continue; + try_attach(p, header); + } + } + header->parent->count++; list_add_tail(&header->ctl_entry, &header->set->list); spin_unlock(&sysctl_lock); @@ -1864,6 +1923,10 @@ void unregister_sysctl_table(struct ctl_table_header * header) spin_lock(&sysctl_lock); start_unregistering(header); + if (!--header->parent->count) { + WARN_ON(1); + kfree(header->parent); + } if (!--header->count) kfree(header); spin_unlock(&sysctl_lock); -- cgit v1.2.3-70-g09d2 From 9043476f726802f4b00c96d0c4f418dde48d1304 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 15 Jul 2008 08:54:06 -0400 Subject: [PATCH] sanitize proc_sysctl * keep references to ctl_table_head and ctl_table in /proc/sys inodes * grab the former during operations, use the latter for access to entry if that succeeds * have ->d_compare() check if table should be seen for one who does lookup; that allows us to avoid flipping inodes - if we have the same name resolve to different things, we'll just keep several dentries and ->d_compare() will reject the wrong ones. * have ->lookup() and ->readdir() scan the table of our inode first, then walk all ctl_table_header and scan ->attached_by for those that are attached to our directory. * implement ->getattr(). * get rid of insane amounts of tree-walking * get rid of the need to know dentry in ->permission() and of the contortions induced by that. Signed-off-by: Al Viro --- fs/proc/inode.c | 5 + fs/proc/proc_sysctl.c | 427 ++++++++++++++++++++++-------------------------- include/linux/proc_fs.h | 5 + include/linux/sysctl.h | 1 + kernel/sysctl.c | 15 ++ 5 files changed, 218 insertions(+), 235 deletions(-) (limited to 'include/linux/sysctl.h') diff --git a/fs/proc/inode.c b/fs/proc/inode.c index b37f25dc45a..8bb03f056c2 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include @@ -65,6 +66,8 @@ static void proc_delete_inode(struct inode *inode) module_put(de->owner); de_put(de); } + if (PROC_I(inode)->sysctl) + sysctl_head_put(PROC_I(inode)->sysctl); clear_inode(inode); } @@ -84,6 +87,8 @@ static struct inode *proc_alloc_inode(struct super_block *sb) ei->fd = 0; ei->op.proc_get_link = NULL; ei->pde = NULL; + ei->sysctl = NULL; + ei->sysctl_entry = NULL; inode = &ei->vfs_inode; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; return inode; diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index 5acc001d49f..fa1ec2433e4 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -10,149 +10,110 @@ static struct dentry_operations proc_sys_dentry_operations; static const struct file_operations proc_sys_file_operations; static const struct inode_operations proc_sys_inode_operations; +static const struct file_operations proc_sys_dir_file_operations; +static const struct inode_operations proc_sys_dir_operations; -static void proc_sys_refresh_inode(struct inode *inode, struct ctl_table *table) -{ - /* Refresh the cached information bits in the inode */ - if (table) { - inode->i_uid = 0; - inode->i_gid = 0; - inode->i_mode = table->mode; - if (table->proc_handler) { - inode->i_mode |= S_IFREG; - inode->i_nlink = 1; - } else { - inode->i_mode |= S_IFDIR; - inode->i_nlink = 0; /* It is too hard to figure out */ - } - } -} - -static struct inode *proc_sys_make_inode(struct inode *dir, struct ctl_table *table) +static struct inode *proc_sys_make_inode(struct super_block *sb, + struct ctl_table_header *head, struct ctl_table *table) { struct inode *inode; - struct proc_inode *dir_ei, *ei; - int depth; + struct proc_inode *ei; - inode = new_inode(dir->i_sb); + inode = new_inode(sb); if (!inode) goto out; - /* A directory is always one deeper than it's parent */ - dir_ei = PROC_I(dir); - depth = dir_ei->fd + 1; - + sysctl_head_get(head); ei = PROC_I(inode); - ei->fd = depth; + ei->sysctl = head; + ei->sysctl_entry = table; + inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; - inode->i_op = &proc_sys_inode_operations; - inode->i_fop = &proc_sys_file_operations; inode->i_flags |= S_PRIVATE; /* tell selinux to ignore this inode */ - proc_sys_refresh_inode(inode, table); + inode->i_mode = table->mode; + if (!table->child) { + inode->i_mode |= S_IFREG; + inode->i_op = &proc_sys_inode_operations; + inode->i_fop = &proc_sys_file_operations; + } else { + inode->i_mode |= S_IFDIR; + inode->i_nlink = 0; + inode->i_op = &proc_sys_dir_operations; + inode->i_fop = &proc_sys_dir_file_operations; + } out: return inode; } -static struct dentry *proc_sys_ancestor(struct dentry *dentry, int depth) -{ - for (;;) { - struct proc_inode *ei; - - ei = PROC_I(dentry->d_inode); - if (ei->fd == depth) - break; /* found */ - - dentry = dentry->d_parent; - } - return dentry; -} - -static struct ctl_table *proc_sys_lookup_table_one(struct ctl_table *table, - struct qstr *name) +static struct ctl_table *find_in_table(struct ctl_table *p, struct qstr *name) { int len; - for ( ; table->ctl_name || table->procname; table++) { + for ( ; p->ctl_name || p->procname; p++) { - if (!table->procname) + if (!p->procname) continue; - len = strlen(table->procname); + len = strlen(p->procname); if (len != name->len) continue; - if (memcmp(table->procname, name->name, len) != 0) + if (memcmp(p->procname, name->name, len) != 0) continue; /* I have a match */ - return table; + return p; } return NULL; } -static struct ctl_table *proc_sys_lookup_table(struct dentry *dentry, - struct ctl_table *table) +struct ctl_table_header *grab_header(struct inode *inode) { - struct dentry *ancestor; - struct proc_inode *ei; - int depth, i; + if (PROC_I(inode)->sysctl) + return sysctl_head_grab(PROC_I(inode)->sysctl); + else + return sysctl_head_next(NULL); +} - ei = PROC_I(dentry->d_inode); - depth = ei->fd; +static struct dentry *proc_sys_lookup(struct inode *dir, struct dentry *dentry, + struct nameidata *nd) +{ + struct ctl_table_header *head = grab_header(dir); + struct ctl_table *table = PROC_I(dir)->sysctl_entry; + struct ctl_table_header *h = NULL; + struct qstr *name = &dentry->d_name; + struct ctl_table *p; + struct inode *inode; + struct dentry *err = ERR_PTR(-ENOENT); - if (depth == 0) - return table; + if (IS_ERR(head)) + return ERR_CAST(head); - for (i = 1; table && (i <= depth); i++) { - ancestor = proc_sys_ancestor(dentry, i); - table = proc_sys_lookup_table_one(table, &ancestor->d_name); - if (table) - table = table->child; + if (table && !table->child) { + WARN_ON(1); + goto out; } - return table; - -} -static struct ctl_table *proc_sys_lookup_entry(struct dentry *dparent, - struct qstr *name, - struct ctl_table *table) -{ - table = proc_sys_lookup_table(dparent, table); - if (table) - table = proc_sys_lookup_table_one(table, name); - return table; -} -static struct ctl_table *do_proc_sys_lookup(struct dentry *parent, - struct qstr *name, - struct ctl_table_header **ptr) -{ - struct ctl_table_header *head; - struct ctl_table *table = NULL; + table = table ? table->child : head->ctl_table; - for (head = sysctl_head_next(NULL); head; - head = sysctl_head_next(head)) { - table = proc_sys_lookup_entry(parent, name, head->ctl_table); - if (table) - break; + p = find_in_table(table, name); + if (!p) { + for (h = sysctl_head_next(NULL); h; h = sysctl_head_next(h)) { + if (h->attached_to != table) + continue; + p = find_in_table(h->attached_by, name); + if (p) + break; + } } - *ptr = head; - return table; -} - -static struct dentry *proc_sys_lookup(struct inode *dir, struct dentry *dentry, - struct nameidata *nd) -{ - struct ctl_table_header *head; - struct inode *inode; - struct dentry *err; - struct ctl_table *table; - err = ERR_PTR(-ENOENT); - table = do_proc_sys_lookup(dentry->d_parent, &dentry->d_name, &head); - if (!table) + if (!p) goto out; err = ERR_PTR(-ENOMEM); - inode = proc_sys_make_inode(dir, table); + inode = proc_sys_make_inode(dir->i_sb, h ? h : head, p); + if (h) + sysctl_head_finish(h); + if (!inode) goto out; @@ -168,22 +129,14 @@ out: static ssize_t proc_sys_call_handler(struct file *filp, void __user *buf, size_t count, loff_t *ppos, int write) { - struct dentry *dentry = filp->f_dentry; - struct ctl_table_header *head; - struct ctl_table *table; + struct inode *inode = filp->f_path.dentry->d_inode; + struct ctl_table_header *head = grab_header(inode); + struct ctl_table *table = PROC_I(inode)->sysctl_entry; ssize_t error; size_t res; - table = do_proc_sys_lookup(dentry->d_parent, &dentry->d_name, &head); - /* Has the sysctl entry disappeared on us? */ - error = -ENOENT; - if (!table) - goto out; - - /* Has the sysctl entry been replaced by a directory? */ - error = -EISDIR; - if (!table->proc_handler) - goto out; + if (IS_ERR(head)) + return PTR_ERR(head); /* * At this point we know that the sysctl was not unregistered @@ -193,6 +146,11 @@ static ssize_t proc_sys_call_handler(struct file *filp, void __user *buf, if (sysctl_perm(head->root, table, write ? MAY_WRITE : MAY_READ)) goto out; + /* if that can happen at all, it should be -EINVAL, not -EISDIR */ + error = -EINVAL; + if (!table->proc_handler) + goto out; + /* careful: calling conventions are nasty here */ res = count; error = table->proc_handler(table, write, filp, buf, &res, ppos); @@ -218,82 +176,86 @@ static ssize_t proc_sys_write(struct file *filp, const char __user *buf, static int proc_sys_fill_cache(struct file *filp, void *dirent, - filldir_t filldir, struct ctl_table *table) + filldir_t filldir, + struct ctl_table_header *head, + struct ctl_table *table) { - struct ctl_table_header *head; - struct ctl_table *child_table = NULL; struct dentry *child, *dir = filp->f_path.dentry; struct inode *inode; struct qstr qname; ino_t ino = 0; unsigned type = DT_UNKNOWN; - int ret; qname.name = table->procname; qname.len = strlen(table->procname); qname.hash = full_name_hash(qname.name, qname.len); - /* Suppress duplicates. - * Only fill a directory entry if it is the value that - * an ordinary lookup of that name returns. Hide all - * others. - * - * If we ever cache this translation in the dcache - * I should do a dcache lookup first. But for now - * it is just simpler not to. - */ - ret = 0; - child_table = do_proc_sys_lookup(dir, &qname, &head); - sysctl_head_finish(head); - if (child_table != table) - return 0; - child = d_lookup(dir, &qname); if (!child) { - struct dentry *new; - new = d_alloc(dir, &qname); - if (new) { - inode = proc_sys_make_inode(dir->d_inode, table); - if (!inode) - child = ERR_PTR(-ENOMEM); - else { - new->d_op = &proc_sys_dentry_operations; - d_add(new, inode); + child = d_alloc(dir, &qname); + if (child) { + inode = proc_sys_make_inode(dir->d_sb, head, table); + if (!inode) { + dput(child); + return -ENOMEM; + } else { + child->d_op = &proc_sys_dentry_operations; + d_add(child, inode); } - if (child) - dput(new); - else - child = new; + } else { + return -ENOMEM; } } - if (!child || IS_ERR(child) || !child->d_inode) - goto end_instantiate; inode = child->d_inode; - if (inode) { - ino = inode->i_ino; - type = inode->i_mode >> 12; - } + ino = inode->i_ino; + type = inode->i_mode >> 12; dput(child); -end_instantiate: - if (!ino) - ino= find_inode_number(dir, &qname); - if (!ino) - ino = 1; - return filldir(dirent, qname.name, qname.len, filp->f_pos, ino, type); + return !!filldir(dirent, qname.name, qname.len, filp->f_pos, ino, type); +} + +static int scan(struct ctl_table_header *head, ctl_table *table, + unsigned long *pos, struct file *file, + void *dirent, filldir_t filldir) +{ + + for (; table->ctl_name || table->procname; table++, (*pos)++) { + int res; + + /* Can't do anything without a proc name */ + if (!table->procname) + continue; + + if (*pos < file->f_pos) + continue; + + res = proc_sys_fill_cache(file, dirent, filldir, head, table); + if (res) + return res; + + file->f_pos = *pos + 1; + } + return 0; } static int proc_sys_readdir(struct file *filp, void *dirent, filldir_t filldir) { - struct dentry *dentry = filp->f_dentry; + struct dentry *dentry = filp->f_path.dentry; struct inode *inode = dentry->d_inode; - struct ctl_table_header *head = NULL; - struct ctl_table *table; + struct ctl_table_header *head = grab_header(inode); + struct ctl_table *table = PROC_I(inode)->sysctl_entry; + struct ctl_table_header *h = NULL; unsigned long pos; - int ret; + int ret = -EINVAL; + + if (IS_ERR(head)) + return PTR_ERR(head); - ret = -ENOTDIR; - if (!S_ISDIR(inode->i_mode)) + if (table && !table->child) { + WARN_ON(1); goto out; + } + + table = table ? table->child : head->ctl_table; ret = 0; /* Avoid a switch here: arm builds fail with missing __cmpdi2 */ @@ -311,30 +273,17 @@ static int proc_sys_readdir(struct file *filp, void *dirent, filldir_t filldir) } pos = 2; - /* - Find each instance of the directory - * - Read all entries in each instance - * - Before returning an entry to user space lookup the entry - * by name and if I find a different entry don't return - * this one because it means it is a buried dup. - * For sysctl this should only happen for directory entries. - */ - for (head = sysctl_head_next(NULL); head; head = sysctl_head_next(head)) { - table = proc_sys_lookup_table(dentry, head->ctl_table); + ret = scan(head, table, &pos, filp, dirent, filldir); + if (ret) + goto out; - if (!table) + for (h = sysctl_head_next(NULL); h; h = sysctl_head_next(h)) { + if (h->attached_to != table) continue; - - for (; table->ctl_name || table->procname; table++, pos++) { - /* Can't do anything without a proc name */ - if (!table->procname) - continue; - - if (pos < filp->f_pos) - continue; - - if (proc_sys_fill_cache(filp, dirent, filldir, table) < 0) - goto out; - filp->f_pos = pos + 1; + ret = scan(h, h->attached_by, &pos, filp, dirent, filldir); + if (ret) { + sysctl_head_finish(h); + break; } } ret = 1; @@ -349,47 +298,18 @@ static int proc_sys_permission(struct inode *inode, int mask, struct nameidata * * sysctl entries that are not writeable, * are _NOT_ writeable, capabilities or not. */ - struct ctl_table_header *head; - struct ctl_table *table; - struct dentry *dentry; - int mode; - int depth; + struct ctl_table_header *head = grab_header(inode); + struct ctl_table *table = PROC_I(inode)->sysctl_entry; int error; - head = NULL; - depth = PROC_I(inode)->fd; - - /* First check the cached permissions, in case we don't have - * enough information to lookup the sysctl table entry. - */ - error = -EACCES; - mode = inode->i_mode; - - if (current->euid == 0) - mode >>= 6; - else if (in_group_p(0)) - mode >>= 3; - - if ((mode & mask & (MAY_READ|MAY_WRITE|MAY_EXEC)) == mask) - error = 0; - - /* If we can't get a sysctl table entry the permission - * checks on the cached mode will have to be enough. - */ - if (!nd || !depth) - goto out; + if (IS_ERR(head)) + return PTR_ERR(head); - dentry = nd->path.dentry; - table = do_proc_sys_lookup(dentry->d_parent, &dentry->d_name, &head); + if (!table) /* global root - r-xr-xr-x */ + error = mask & MAY_WRITE ? -EACCES : 0; + else /* Use the permissions on the sysctl table entry */ + error = sysctl_perm(head->root, table, mask); - /* If the entry does not exist deny permission */ - error = -EACCES; - if (!table) - goto out; - - /* Use the permissions on the sysctl table entry */ - error = sysctl_perm(head->root, table, mask); -out: sysctl_head_finish(head); return error; } @@ -409,33 +329,70 @@ static int proc_sys_setattr(struct dentry *dentry, struct iattr *attr) return error; } -/* I'm lazy and don't distinguish between files and directories, - * until access time. - */ +static int proc_sys_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) +{ + struct inode *inode = dentry->d_inode; + struct ctl_table_header *head = grab_header(inode); + struct ctl_table *table = PROC_I(inode)->sysctl_entry; + + if (IS_ERR(head)) + return PTR_ERR(head); + + generic_fillattr(inode, stat); + if (table) + stat->mode = (stat->mode & S_IFMT) | table->mode; + + sysctl_head_finish(head); + return 0; +} + static const struct file_operations proc_sys_file_operations = { .read = proc_sys_read, .write = proc_sys_write, +}; + +static const struct file_operations proc_sys_dir_file_operations = { .readdir = proc_sys_readdir, }; static const struct inode_operations proc_sys_inode_operations = { + .permission = proc_sys_permission, + .setattr = proc_sys_setattr, + .getattr = proc_sys_getattr, +}; + +static const struct inode_operations proc_sys_dir_operations = { .lookup = proc_sys_lookup, .permission = proc_sys_permission, .setattr = proc_sys_setattr, + .getattr = proc_sys_getattr, }; static int proc_sys_revalidate(struct dentry *dentry, struct nameidata *nd) { - struct ctl_table_header *head; - struct ctl_table *table; - table = do_proc_sys_lookup(dentry->d_parent, &dentry->d_name, &head); - proc_sys_refresh_inode(dentry->d_inode, table); - sysctl_head_finish(head); - return !!table; + return !PROC_I(dentry->d_inode)->sysctl->unregistering; +} + +static int proc_sys_delete(struct dentry *dentry) +{ + return !!PROC_I(dentry->d_inode)->sysctl->unregistering; +} + +static int proc_sys_compare(struct dentry *dir, struct qstr *qstr, + struct qstr *name) +{ + struct dentry *dentry = container_of(qstr, struct dentry, d_name); + if (qstr->len != name->len) + return 1; + if (memcmp(qstr->name, name->name, name->len)) + return 1; + return !sysctl_is_seen(PROC_I(dentry->d_inode)->sysctl); } static struct dentry_operations proc_sys_dentry_operations = { .d_revalidate = proc_sys_revalidate, + .d_delete = proc_sys_delete, + .d_compare = proc_sys_compare, }; static struct proc_dir_entry *proc_sys_root; @@ -443,8 +400,8 @@ static struct proc_dir_entry *proc_sys_root; int proc_sys_init(void) { proc_sys_root = proc_mkdir("sys", NULL); - proc_sys_root->proc_iops = &proc_sys_inode_operations; - proc_sys_root->proc_fops = &proc_sys_file_operations; + proc_sys_root->proc_iops = &proc_sys_dir_operations; + proc_sys_root->proc_fops = &proc_sys_dir_file_operations; proc_sys_root->nlink = 0; return 0; } diff --git a/include/linux/proc_fs.h b/include/linux/proc_fs.h index f560d1705af..fb61850d1cf 100644 --- a/include/linux/proc_fs.h +++ b/include/linux/proc_fs.h @@ -282,11 +282,16 @@ union proc_op { struct task_struct *task); }; +struct ctl_table_header; +struct ctl_table; + struct proc_inode { struct pid *pid; int fd; union proc_op op; struct proc_dir_entry *pde; + struct ctl_table_header *sysctl; + struct ctl_table *sysctl_entry; struct inode vfs_inode; }; diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h index 3f6599aeb0d..d0437f36921 100644 --- a/include/linux/sysctl.h +++ b/include/linux/sysctl.h @@ -961,6 +961,7 @@ struct ctl_table_header; extern void sysctl_head_get(struct ctl_table_header *); extern void sysctl_head_put(struct ctl_table_header *); +extern int sysctl_is_seen(struct ctl_table_header *); extern struct ctl_table_header *sysctl_head_grab(struct ctl_table_header *); extern struct ctl_table_header *sysctl_head_next(struct ctl_table_header *prev); extern struct ctl_table_header *__sysctl_head_next(struct nsproxy *namespaces, diff --git a/kernel/sysctl.c b/kernel/sysctl.c index c9a0af88703..ff5abcca5dd 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -1932,6 +1932,21 @@ void unregister_sysctl_table(struct ctl_table_header * header) spin_unlock(&sysctl_lock); } +int sysctl_is_seen(struct ctl_table_header *p) +{ + struct ctl_table_set *set = p->set; + int res; + spin_lock(&sysctl_lock); + if (p->unregistering) + res = 0; + else if (!set->is_seen) + res = 1; + else + res = set->is_seen(set); + spin_unlock(&sysctl_lock); + return res; +} + void setup_sysctl_set(struct ctl_table_set *p, struct ctl_table_set *parent, int (*is_seen)(struct ctl_table_set *)) -- cgit v1.2.3-70-g09d2