From 8086cd451f08f4c0f9693fc66d87754bbd18cfba Mon Sep 17 00:00:00 2001 From: Adrian Bunk Date: Tue, 22 Jul 2008 14:19:19 -0700 Subject: netns: make get_proc_net() static get_proc_net() can now become static. Signed-off-by: Adrian Bunk Acked-by: Pavel Emelyanov Signed-off-by: David S. Miller --- include/linux/proc_fs.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include/linux/proc_fs.h') diff --git a/include/linux/proc_fs.h b/include/linux/proc_fs.h index fff1d27ddb4c..15a9eaf4a802 100644 --- a/include/linux/proc_fs.h +++ b/include/linux/proc_fs.h @@ -305,8 +305,6 @@ static inline struct net *PDE_NET(struct proc_dir_entry *pde) return pde->parent->data; } -struct net *get_proc_net(const struct inode *inode); - struct proc_maps_private { struct pid *pid; struct task_struct *task; -- cgit v1.2.3 From 6e644c3126149b65460610fe5a00d8a162092abe Mon Sep 17 00:00:00 2001 From: Adrian Bunk Date: Fri, 25 Jul 2008 01:48:28 -0700 Subject: move proc_kmsg_operations to fs/proc/internal.h This patch moves the extern of struct proc_kmsg_operations to fs/proc/internal.h and adds an #include "internal.h" to fs/proc/kmsg.c so that the latter sees the former. Signed-off-by: Adrian Bunk Cc: Alexey Dobriyan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/internal.h | 1 + fs/proc/kmsg.c | 2 ++ include/linux/proc_fs.h | 1 - 3 files changed, 3 insertions(+), 1 deletion(-) (limited to 'include/linux/proc_fs.h') diff --git a/fs/proc/internal.h b/fs/proc/internal.h index 28cbca805905..8d67616e7bb0 100644 --- a/fs/proc/internal.h +++ b/fs/proc/internal.h @@ -63,6 +63,7 @@ extern const struct file_operations proc_smaps_operations; extern const struct file_operations proc_clear_refs_operations; extern const struct file_operations proc_pagemap_operations; extern const struct file_operations proc_net_operations; +extern const struct file_operations proc_kmsg_operations; extern const struct inode_operations proc_net_inode_operations; void free_proc_entry(struct proc_dir_entry *de); diff --git a/fs/proc/kmsg.c b/fs/proc/kmsg.c index ff3b90b56e9d..9fd5df3f40ce 100644 --- a/fs/proc/kmsg.c +++ b/fs/proc/kmsg.c @@ -15,6 +15,8 @@ #include #include +#include "internal.h" + extern wait_queue_head_t log_wait; extern int do_syslog(int type, char __user *bug, int count); diff --git a/include/linux/proc_fs.h b/include/linux/proc_fs.h index 15a9eaf4a802..cdabc2fc02f7 100644 --- a/include/linux/proc_fs.h +++ b/include/linux/proc_fs.h @@ -138,7 +138,6 @@ extern int proc_readdir(struct file *, void *, filldir_t); extern struct dentry *proc_lookup(struct inode *, struct dentry *, struct nameidata *); extern const struct file_operations proc_kcore_operations; -extern const struct file_operations proc_kmsg_operations; extern const struct file_operations ppc_htab_operations; extern int pid_ns_prepare_proc(struct pid_namespace *ns); -- cgit v1.2.3 From 881adb85358309ea9c6f707394002719982ec607 Mon Sep 17 00:00:00 2001 From: Alexey Dobriyan Date: Fri, 25 Jul 2008 01:48:29 -0700 Subject: proc: always do ->release Current two-stage scheme of removing PDE emphasizes one bug in proc: open rmmod remove_proc_entry close ->release won't be called because ->proc_fops were cleared. In simple cases it's small memory leak. For every ->open, ->release has to be done. List of openers is introduced which is traversed at remove_proc_entry() if neeeded. Discussions with Al long ago (sigh). Signed-off-by: Alexey Dobriyan Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/proc/generic.c | 14 ++++++++++ fs/proc/inode.c | 74 ++++++++++++++++++++++++++++++++++++++++++++++--- fs/proc/internal.h | 7 +++++ include/linux/proc_fs.h | 1 + 4 files changed, 92 insertions(+), 4 deletions(-) (limited to 'include/linux/proc_fs.h') diff --git a/fs/proc/generic.c b/fs/proc/generic.c index 43e54e86cefd..bc0a0dd2d844 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -597,6 +597,7 @@ static struct proc_dir_entry *__proc_create(struct proc_dir_entry **parent, ent->pde_users = 0; spin_lock_init(&ent->pde_unload_lock); ent->pde_unload_completion = NULL; + INIT_LIST_HEAD(&ent->pde_openers); out: return ent; } @@ -789,6 +790,19 @@ void remove_proc_entry(const char *name, struct proc_dir_entry *parent) spin_unlock(&de->pde_unload_lock); continue_removing: + spin_lock(&de->pde_unload_lock); + while (!list_empty(&de->pde_openers)) { + struct pde_opener *pdeo; + + pdeo = list_first_entry(&de->pde_openers, struct pde_opener, lh); + list_del(&pdeo->lh); + spin_unlock(&de->pde_unload_lock); + pdeo->release(pdeo->inode, pdeo->file); + kfree(pdeo); + spin_lock(&de->pde_unload_lock); + } + spin_unlock(&de->pde_unload_lock); + if (S_ISDIR(de->mode)) parent->nlink--; de->nlink = 0; diff --git a/fs/proc/inode.c b/fs/proc/inode.c index b08d10017911..354c08485825 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -126,12 +126,17 @@ static const struct super_operations proc_sops = { .remount_fs = proc_remount, }; -static void pde_users_dec(struct proc_dir_entry *pde) +static void __pde_users_dec(struct proc_dir_entry *pde) { - spin_lock(&pde->pde_unload_lock); pde->pde_users--; if (pde->pde_unload_completion && pde->pde_users == 0) complete(pde->pde_unload_completion); +} + +static void pde_users_dec(struct proc_dir_entry *pde) +{ + spin_lock(&pde->pde_unload_lock); + __pde_users_dec(pde); spin_unlock(&pde->pde_unload_lock); } @@ -318,36 +323,97 @@ static int proc_reg_open(struct inode *inode, struct file *file) struct proc_dir_entry *pde = PDE(inode); int rv = 0; int (*open)(struct inode *, struct file *); + int (*release)(struct inode *, struct file *); + struct pde_opener *pdeo; + + /* + * What for, you ask? Well, we can have open, rmmod, remove_proc_entry + * sequence. ->release won't be called because ->proc_fops will be + * cleared. Depending on complexity of ->release, consequences vary. + * + * We can't wait for mercy when close will be done for real, it's + * deadlockable: rmmod foo release + * by hand in remove_proc_entry(). For this, save opener's credentials + * for later. + */ + pdeo = kmalloc(sizeof(struct pde_opener), GFP_KERNEL); + if (!pdeo) + return -ENOMEM; spin_lock(&pde->pde_unload_lock); if (!pde->proc_fops) { spin_unlock(&pde->pde_unload_lock); + kfree(pdeo); return rv; } pde->pde_users++; open = pde->proc_fops->open; + release = pde->proc_fops->release; spin_unlock(&pde->pde_unload_lock); if (open) rv = open(inode, file); - pde_users_dec(pde); + spin_lock(&pde->pde_unload_lock); + if (rv == 0 && release) { + /* To know what to release. */ + pdeo->inode = inode; + pdeo->file = file; + /* Strictly for "too late" ->release in proc_reg_release(). */ + pdeo->release = release; + list_add(&pdeo->lh, &pde->pde_openers); + } else + kfree(pdeo); + __pde_users_dec(pde); + spin_unlock(&pde->pde_unload_lock); return rv; } +static struct pde_opener *find_pde_opener(struct proc_dir_entry *pde, + struct inode *inode, struct file *file) +{ + struct pde_opener *pdeo; + + list_for_each_entry(pdeo, &pde->pde_openers, lh) { + if (pdeo->inode == inode && pdeo->file == file) + return pdeo; + } + return NULL; +} + static int proc_reg_release(struct inode *inode, struct file *file) { struct proc_dir_entry *pde = PDE(inode); int rv = 0; int (*release)(struct inode *, struct file *); + struct pde_opener *pdeo; spin_lock(&pde->pde_unload_lock); + pdeo = find_pde_opener(pde, inode, file); if (!pde->proc_fops) { - spin_unlock(&pde->pde_unload_lock); + /* + * Can't simply exit, __fput() will think that everything is OK, + * and move on to freeing struct file. remove_proc_entry() will + * find slacker in opener's list and will try to do non-trivial + * things with struct file. Therefore, remove opener from list. + * + * But if opener is removed from list, who will ->release it? + */ + if (pdeo) { + list_del(&pdeo->lh); + spin_unlock(&pde->pde_unload_lock); + rv = pdeo->release(inode, file); + kfree(pdeo); + } else + spin_unlock(&pde->pde_unload_lock); return rv; } pde->pde_users++; release = pde->proc_fops->release; + if (pdeo) { + list_del(&pdeo->lh); + kfree(pdeo); + } spin_unlock(&pde->pde_unload_lock); if (release) diff --git a/fs/proc/internal.h b/fs/proc/internal.h index 8d67616e7bb0..442202314d53 100644 --- a/fs/proc/internal.h +++ b/fs/proc/internal.h @@ -89,3 +89,10 @@ struct dentry *proc_lookup_de(struct proc_dir_entry *de, struct inode *ino, struct dentry *dentry); int proc_readdir_de(struct proc_dir_entry *de, struct file *filp, void *dirent, filldir_t filldir); + +struct pde_opener { + struct inode *inode; + struct file *file; + int (*release)(struct inode *, struct file *); + struct list_head lh; +}; diff --git a/include/linux/proc_fs.h b/include/linux/proc_fs.h index cdabc2fc02f7..f560d1705afe 100644 --- a/include/linux/proc_fs.h +++ b/include/linux/proc_fs.h @@ -79,6 +79,7 @@ struct proc_dir_entry { int pde_users; /* number of callers into module in progress */ spinlock_t pde_unload_lock; /* proc_fops checks and pde_users bumps */ struct completion *pde_unload_completion; + struct list_head pde_openers; /* who did ->open, but not ->release */ }; struct kcore_list { -- cgit v1.2.3 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/proc_fs.h') diff --git a/fs/proc/inode.c b/fs/proc/inode.c index b37f25dc45a5..8bb03f056c28 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 5acc001d49f6..fa1ec2433e44 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 f560d1705afe..fb61850d1cfc 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 3f6599aeb0db..d0437f36921f 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 c9a0af887033..ff5abcca5ddf 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