namespaces: mqueue ns: move mqueue_mnt into struct ipc_namespace

Move mqueue vfsmount plus a few tunables into the ipc_namespace struct.
The CONFIG_IPC_NS boolean and the ipc_namespace struct will serve both the
posix message queue namespaces and the SYSV ipc namespaces.

The sysctl code will be fixed separately in patch 3.  After just this
patch, making a change to posix mqueue tunables always changes the values
in the initial ipc namespace.

Signed-off-by: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Serge E. Hallyn <serue@us.ibm.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

7 files changed, 145 insertions, 71 deletions

diff --git a/include/linux/ipc_namespace.h b/include/linux/ipc_namespace.h
index ea330f9e7100..3e6fcacebe8a 100644
--- a/include/linux/ipc_namespace.h
+++ b/include/linux/ipc_namespace.h
@@ -44,24 +44,55 @@ struct ipc_namespace {
         int             shm_tot;
 
         struct notifier_block ipcns_nb;
+
+        /* The kern_mount of the mqueuefs sb.  We take a ref on it */
+        struct vfsmount *mq_mnt;
+
+        /* # queues in this ns, protected by mq_lock */
+        unsigned int    mq_queues_count;
+
+        /* next fields are set through sysctl */
+        unsigned int    mq_queues_max;   /* initialized to DFLT_QUEUESMAX */
+        unsigned int    mq_msg_max;      /* initialized to DFLT_MSGMAX */
+        unsigned int    mq_msgsize_max;  /* initialized to DFLT_MSGSIZEMAX */
+
 };
 
 extern struct ipc_namespace init_ipc_ns;
 extern atomic_t nr_ipc_ns;
 
-#ifdef CONFIG_SYSVIPC
+#if defined(CONFIG_POSIX_MQUEUE) || defined(CONFIG_SYSVIPC)
 #define INIT_IPC_NS(ns)         .ns             = &init_ipc_ns,
+#else
+#define INIT_IPC_NS(ns)
+#endif
 
+#ifdef CONFIG_SYSVIPC
 extern int register_ipcns_notifier(struct ipc_namespace *);
 extern int cond_register_ipcns_notifier(struct ipc_namespace *);
 extern void unregister_ipcns_notifier(struct ipc_namespace *);
 extern int ipcns_notify(unsigned long);
-
 #else /* CONFIG_SYSVIPC */
-#define INIT_IPC_NS(ns)
+static inline int register_ipcns_notifier(struct ipc_namespace *ns)
+{ return 0; }
+static inline int cond_register_ipcns_notifier(struct ipc_namespace *ns)
+{ return 0; }
+static inline void unregister_ipcns_notifier(struct ipc_namespace *ns) { }
+static inline int ipcns_notify(unsigned long l) { return 0; }
 #endif /* CONFIG_SYSVIPC */
 
-#if defined(CONFIG_SYSVIPC) && defined(CONFIG_IPC_NS)
+#ifdef CONFIG_POSIX_MQUEUE
+extern void mq_init_ns(struct ipc_namespace *ns);
+/* default values */
+#define DFLT_QUEUESMAX 256     /* max number of message queues */
+#define DFLT_MSGMAX    10      /* max number of messages in each queue */
+#define HARD_MSGMAX    (131072/sizeof(void *))
+#define DFLT_MSGSIZEMAX 8192   /* max message size */
+#else
+#define mq_init_ns(ns) ((void) 0)
+#endif
+
+#if defined(CONFIG_IPC_NS)
 extern void free_ipc_ns(struct kref *kref);
 extern struct ipc_namespace *copy_ipcs(unsigned long flags,
                                        struct ipc_namespace *ns);
diff --git a/init/Kconfig b/init/Kconfig
index c52d1d48272a..a0807ba91644 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -670,10 +670,10 @@ config UTS_NS
 
 config IPC_NS
         bool "IPC namespace"
-        depends on NAMESPACES && SYSVIPC
+        depends on NAMESPACES && (SYSVIPC || POSIX_MQUEUE)
         help
           In this namespace tasks work with IPC ids which correspond to
-          different IPC objects in different namespaces
+          different IPC objects in different namespaces.
 
 config USER_NS
         bool "User namespace (EXPERIMENTAL)"
diff --git a/ipc/mqueue.c b/ipc/mqueue.c
index 916785363f0f..a3673a09069a 100644
--- a/ipc/mqueue.c
+++ b/ipc/mqueue.c
@@ -31,6 +31,7 @@
 #include <linux/mutex.h>
 #include <linux/nsproxy.h>
 #include <linux/pid.h>
+#include <linux/ipc_namespace.h>
 
 #include <net/sock.h>
 #include "util.h"
@@ -46,12 +47,6 @@
 #define STATE_PENDING   1
 #define STATE_READY     2
 
-/* default values */
-#define DFLT_QUEUESMAX  256     /* max number of message queues */
-#define DFLT_MSGMAX     10      /* max number of messages in each queue */
-#define HARD_MSGMAX     (131072/sizeof(void*))
-#define DFLT_MSGSIZEMAX 8192    /* max message size */
-
 /*
  * Define the ranges various user-specified maximum values can
  * be set to.
@@ -95,12 +90,6 @@ static void remove_notification(struct mqueue_inode_info *info);
 
 static spinlock_t mq_lock;
 static struct kmem_cache *mqueue_inode_cachep;
-static struct vfsmount *mqueue_mnt;
-
-static unsigned int queues_count;
-static unsigned int queues_max  = DFLT_QUEUESMAX;
-static unsigned int msg_max     = DFLT_MSGMAX;
-static unsigned int msgsize_max = DFLT_MSGSIZEMAX;
 
 static struct ctl_table_header * mq_sysctl_table;
 
@@ -109,11 +98,27 @@ static inline struct mqueue_inode_info *MQUEUE_I(struct inode *inode)
         return container_of(inode, struct mqueue_inode_info, vfs_inode);
 }
 
+void mq_init_ns(struct ipc_namespace *ns)
+{
+        ns->mq_queues_count  = 0;
+        ns->mq_queues_max    = DFLT_QUEUESMAX;
+        ns->mq_msg_max       = DFLT_MSGMAX;
+        ns->mq_msgsize_max   = DFLT_MSGSIZEMAX;
+        ns->mq_mnt           = mntget(init_ipc_ns.mq_mnt);
+}
+
+void mq_exit_ns(struct ipc_namespace *ns)
+{
+        /* will need to clear out ns->mq_mnt->mnt_sb->s_fs_info here */
+        mntput(ns->mq_mnt);
+}
+
 static struct inode *mqueue_get_inode(struct super_block *sb, int mode,
                                                         struct mq_attr *attr)
 {
         struct user_struct *u = current_user();
         struct inode *inode;
+        struct ipc_namespace *ipc_ns = &init_ipc_ns;
 
         inode = new_inode(sb);
         if (inode) {
@@ -141,8 +146,8 @@ static struct inode *mqueue_get_inode(struct super_block *sb, int mode,
                         info->qsize = 0;
                         info->user = NULL;      /* set when all is ok */
                         memset(&info->attr, 0, sizeof(info->attr));
-                        info->attr.mq_maxmsg = msg_max;
-                        info->attr.mq_msgsize = msgsize_max;
+                        info->attr.mq_maxmsg = ipc_ns->mq_msg_max;
+                        info->attr.mq_msgsize = ipc_ns->mq_msgsize_max;
                         if (attr) {
                                 info->attr.mq_maxmsg = attr->mq_maxmsg;
                                 info->attr.mq_msgsize = attr->mq_msgsize;
@@ -242,6 +247,7 @@ static void mqueue_delete_inode(struct inode *inode)
         struct user_struct *user;
         unsigned long mq_bytes;
         int i;
+        struct ipc_namespace *ipc_ns = &init_ipc_ns;
 
         if (S_ISDIR(inode->i_mode)) {
                 clear_inode(inode);
@@ -262,7 +268,7 @@ static void mqueue_delete_inode(struct inode *inode)
         if (user) {
                 spin_lock(&mq_lock);
                 user->mq_bytes -= mq_bytes;
-                queues_count--;
+                ipc_ns->mq_queues_count--;
                 spin_unlock(&mq_lock);
                 free_uid(user);
         }
@@ -274,21 +280,23 @@ static int mqueue_create(struct inode *dir, struct dentry *dentry,
         struct inode *inode;
         struct mq_attr *attr = dentry->d_fsdata;
         int error;
+        struct ipc_namespace *ipc_ns = &init_ipc_ns;
 
         spin_lock(&mq_lock);
-        if (queues_count >= queues_max && !capable(CAP_SYS_RESOURCE)) {
+        if (ipc_ns->mq_queues_count >= ipc_ns->mq_queues_max &&
+                        !capable(CAP_SYS_RESOURCE)) {
                 error = -ENOSPC;
-                goto out_lock;
+                goto out_unlock;
         }
-        queues_count++;
+        ipc_ns->mq_queues_count++;
         spin_unlock(&mq_lock);
 
         inode = mqueue_get_inode(dir->i_sb, mode, attr);
         if (!inode) {
                 error = -ENOMEM;
                 spin_lock(&mq_lock);
-                queues_count--;
-                goto out_lock;
+                ipc_ns->mq_queues_count--;
+                goto out_unlock;
         }
 
         dir->i_size += DIRENT_SIZE;
@@ -297,7 +305,7 @@ static int mqueue_create(struct inode *dir, struct dentry *dentry,
         d_instantiate(dentry, inode);
         dget(dentry);
         return 0;
-out_lock:
+out_unlock:
         spin_unlock(&mq_lock);
         return error;
 }
@@ -562,7 +570,7 @@ static void remove_notification(struct mqueue_inode_info *info)
         info->notify_owner = NULL;
 }
 
-static int mq_attr_ok(struct mq_attr *attr)
+static int mq_attr_ok(struct ipc_namespace *ipc_ns, struct mq_attr *attr)
 {
         if (attr->mq_maxmsg <= 0 || attr->mq_msgsize <= 0)
                 return 0;
@@ -570,8 +578,8 @@ static int mq_attr_ok(struct mq_attr *attr)
                 if (attr->mq_maxmsg > HARD_MSGMAX)
                         return 0;
         } else {
-                if (attr->mq_maxmsg > msg_max ||
-                                attr->mq_msgsize > msgsize_max)
+                if (attr->mq_maxmsg > ipc_ns->mq_msg_max ||
+                                attr->mq_msgsize > ipc_ns->mq_msgsize_max)
                         return 0;
         }
         /* check for overflow */
@@ -587,8 +595,9 @@ static int mq_attr_ok(struct mq_attr *attr)
 /*
  * Invoked when creating a new queue via sys_mq_open
  */
-static struct file *do_create(struct dentry *dir, struct dentry *dentry,
-                        int oflag, mode_t mode, struct mq_attr *attr)
+static struct file *do_create(struct ipc_namespace *ipc_ns, struct dentry *dir,
+                        struct dentry *dentry, int oflag, mode_t mode,
+                        struct mq_attr *attr)
 {
         const struct cred *cred = current_cred();
         struct file *result;
@@ -596,14 +605,14 @@ static struct file *do_create(struct dentry *dir, struct dentry *dentry,
 
         if (attr) {
                 ret = -EINVAL;
-                if (!mq_attr_ok(attr))
+                if (!mq_attr_ok(ipc_ns, attr))
                         goto out;
                 /* store for use during create */
                 dentry->d_fsdata = attr;
         }
 
         mode &= ~current_umask();
-        ret = mnt_want_write(mqueue_mnt);
+        ret = mnt_want_write(ipc_ns->mq_mnt);
         if (ret)
                 goto out;
         ret = vfs_create(dir->d_inode, dentry, mode, NULL);
@@ -611,24 +620,25 @@ static struct file *do_create(struct dentry *dir, struct dentry *dentry,
         if (ret)
                 goto out_drop_write;
 
-        result = dentry_open(dentry, mqueue_mnt, oflag, cred);
+        result = dentry_open(dentry, ipc_ns->mq_mnt, oflag, cred);
         /*
          * dentry_open() took a persistent mnt_want_write(),
          * so we can now drop this one.
          */
-        mnt_drop_write(mqueue_mnt);
+        mnt_drop_write(ipc_ns->mq_mnt);
         return result;
 
 out_drop_write:
-        mnt_drop_write(mqueue_mnt);
+        mnt_drop_write(ipc_ns->mq_mnt);
 out:
         dput(dentry);
-        mntput(mqueue_mnt);
+        mntput(ipc_ns->mq_mnt);
         return ERR_PTR(ret);
 }
 
 /* Opens existing queue */
-static struct file *do_open(struct dentry *dentry, int oflag)
+static struct file *do_open(struct ipc_namespace *ipc_ns,
+                                struct dentry *dentry, int oflag)
 {
         const struct cred *cred = current_cred();
 
@@ -637,17 +647,17 @@ static struct file *do_open(struct dentry *dentry, int oflag)
 
         if ((oflag & O_ACCMODE) == (O_RDWR | O_WRONLY)) {
                 dput(dentry);
-                mntput(mqueue_mnt);
+                mntput(ipc_ns->mq_mnt);
                 return ERR_PTR(-EINVAL);
         }
 
         if (inode_permission(dentry->d_inode, oflag2acc[oflag & O_ACCMODE])) {
                 dput(dentry);
-                mntput(mqueue_mnt);
+                mntput(ipc_ns->mq_mnt);
                 return ERR_PTR(-EACCES);
         }
 
-        return dentry_open(dentry, mqueue_mnt, oflag, cred);
+        return dentry_open(dentry, ipc_ns->mq_mnt, oflag, cred);
 }
 
 SYSCALL_DEFINE4(mq_open, const char __user *, u_name, int, oflag, mode_t, mode,
@@ -658,6 +668,7 @@ SYSCALL_DEFINE4(mq_open, const char __user *, u_name, int, oflag, mode_t, mode,
         char *name;
         struct mq_attr attr;
         int fd, error;
+        struct ipc_namespace *ipc_ns = &init_ipc_ns;
 
         if (u_attr && copy_from_user(&attr, u_attr, sizeof(struct mq_attr)))
                 return -EFAULT;
@@ -671,13 +682,13 @@ SYSCALL_DEFINE4(mq_open, const char __user *, u_name, int, oflag, mode_t, mode,
         if (fd < 0)
                 goto out_putname;
 
-        mutex_lock(&mqueue_mnt->mnt_root->d_inode->i_mutex);
-        dentry = lookup_one_len(name, mqueue_mnt->mnt_root, strlen(name));
+        mutex_lock(&ipc_ns->mq_mnt->mnt_root->d_inode->i_mutex);
+        dentry = lookup_one_len(name, ipc_ns->mq_mnt->mnt_root, strlen(name));
         if (IS_ERR(dentry)) {
                 error = PTR_ERR(dentry);
                 goto out_err;
         }
-        mntget(mqueue_mnt);
+        mntget(ipc_ns->mq_mnt);
 
         if (oflag & O_CREAT) {
                 if (dentry->d_inode) {  /* entry already exists */
@@ -685,10 +696,10 @@ SYSCALL_DEFINE4(mq_open, const char __user *, u_name, int, oflag, mode_t, mode,
                         error = -EEXIST;
                         if (oflag & O_EXCL)
                                 goto out;
-                        filp = do_open(dentry, oflag);
+                        filp = do_open(ipc_ns, dentry, oflag);
                 } else {
-                        filp = do_create(mqueue_mnt->mnt_root, dentry,
-                                                oflag, mode,
+                        filp = do_create(ipc_ns, ipc_ns->mq_mnt->mnt_root,
+                                                dentry, oflag, mode,
                                                 u_attr ? &attr : NULL);
                 }
         } else {
@@ -696,7 +707,7 @@ SYSCALL_DEFINE4(mq_open, const char __user *, u_name, int, oflag, mode_t, mode,
                 if (!dentry->d_inode)
                         goto out;
                 audit_inode(name, dentry);
-                filp = do_open(dentry, oflag);
+                filp = do_open(ipc_ns, dentry, oflag);
         }
 
         if (IS_ERR(filp)) {
@@ -709,13 +720,13 @@ SYSCALL_DEFINE4(mq_open, const char __user *, u_name, int, oflag, mode_t, mode,
 
 out:
         dput(dentry);
-        mntput(mqueue_mnt);
+        mntput(ipc_ns->mq_mnt);
 out_putfd:
         put_unused_fd(fd);
 out_err:
         fd = error;
 out_upsem:
-        mutex_unlock(&mqueue_mnt->mnt_root->d_inode->i_mutex);
+        mutex_unlock(&ipc_ns->mq_mnt->mnt_root->d_inode->i_mutex);
 out_putname:
         putname(name);
         return fd;
@@ -727,14 +738,15 @@ SYSCALL_DEFINE1(mq_unlink, const char __user *, u_name)
         char *name;
         struct dentry *dentry;
         struct inode *inode = NULL;
+        struct ipc_namespace *ipc_ns = &init_ipc_ns;
 
         name = getname(u_name);
         if (IS_ERR(name))
                 return PTR_ERR(name);
 
-        mutex_lock_nested(&mqueue_mnt->mnt_root->d_inode->i_mutex,
+        mutex_lock_nested(&ipc_ns->mq_mnt->mnt_root->d_inode->i_mutex,
                         I_MUTEX_PARENT);
-        dentry = lookup_one_len(name, mqueue_mnt->mnt_root, strlen(name));
+        dentry = lookup_one_len(name, ipc_ns->mq_mnt->mnt_root, strlen(name));
         if (IS_ERR(dentry)) {
                 err = PTR_ERR(dentry);
                 goto out_unlock;
@@ -748,16 +760,16 @@ SYSCALL_DEFINE1(mq_unlink, const char __user *, u_name)
         inode = dentry->d_inode;
         if (inode)
                 atomic_inc(&inode->i_count);
-        err = mnt_want_write(mqueue_mnt);
+        err = mnt_want_write(ipc_ns->mq_mnt);
         if (err)
                 goto out_err;
         err = vfs_unlink(dentry->d_parent->d_inode, dentry);
-        mnt_drop_write(mqueue_mnt);
+        mnt_drop_write(ipc_ns->mq_mnt);
 out_err:
         dput(dentry);
 
 out_unlock:
-        mutex_unlock(&mqueue_mnt->mnt_root->d_inode->i_mutex);
+        mutex_unlock(&ipc_ns->mq_mnt->mnt_root->d_inode->i_mutex);
         putname(name);
         if (inode)
                 iput(inode);
@@ -1214,14 +1226,14 @@ static int msg_maxsize_limit_max = MAX_MSGSIZEMAX;
 static ctl_table mq_sysctls[] = {
         {
                 .procname       = "queues_max",
-                .data           = &queues_max,
+                .data           = &init_ipc_ns.mq_queues_max,
                 .maxlen         = sizeof(int),
                 .mode           = 0644,
                 .proc_handler   = &proc_dointvec,
         },
         {
                 .procname       = "msg_max",
-                .data           = &msg_max,
+                .data           = &init_ipc_ns.mq_msg_max,
                 .maxlen         = sizeof(int),
                 .mode           = 0644,
                 .proc_handler   = &proc_dointvec_minmax,
@@ -1230,7 +1242,7 @@ static ctl_table mq_sysctls[] = {
         },
         {
                 .procname       = "msgsize_max",
-                .data           = &msgsize_max,
+                .data           = &init_ipc_ns.mq_msgsize_max,
                 .maxlen         = sizeof(int),
                 .mode           = 0644,
                 .proc_handler   = &proc_dointvec_minmax,
@@ -1276,13 +1288,13 @@ static int __init init_mqueue_fs(void)
         if (error)
                 goto out_sysctl;
 
-        if (IS_ERR(mqueue_mnt = kern_mount(&mqueue_fs_type))) {
-                error = PTR_ERR(mqueue_mnt);
+        init_ipc_ns.mq_mnt = kern_mount(&mqueue_fs_type);
+        if (IS_ERR(init_ipc_ns.mq_mnt)) {
+                error = PTR_ERR(init_ipc_ns.mq_mnt);
                 goto out_filesystem;
         }
 
         /* internal initialization - not common for vfs */
-        queues_count = 0;
         spin_lock_init(&mq_lock);
 
         return 0;
diff --git a/ipc/msgutil.c b/ipc/msgutil.c
index c82c215693d7..73c316cb8613 100644
--- a/ipc/msgutil.c
+++ b/ipc/msgutil.c
@@ -13,10 +13,32 @@
 #include <linux/security.h>
 #include <linux/slab.h>
 #include <linux/ipc.h>
+#include <linux/ipc_namespace.h>
 #include <asm/uaccess.h>
 
 #include "util.h"
 
+/*
+ * The next 2 defines are here bc this is the only file
+ * compiled when either CONFIG_SYSVIPC and CONFIG_POSIX_MQUEUE
+ * and not CONFIG_IPC_NS.
+ */
+struct ipc_namespace init_ipc_ns = {
+        .kref = {
+                /* It's not for this patch to change, but should this be 1? */
+                .refcount       = ATOMIC_INIT(2),
+        },
+#ifdef CONFIG_POSIX_MQUEUE
+        .mq_mnt          = NULL,
+        .mq_queues_count = 0,
+        .mq_queues_max   = DFLT_QUEUESMAX,
+        .mq_msg_max      = DFLT_MSGMAX,
+        .mq_msgsize_max  = DFLT_MSGSIZEMAX,
+#endif
+};
+
+atomic_t nr_ipc_ns = ATOMIC_INIT(1);
+
 struct msg_msgseg {
         struct msg_msgseg* next;
         /* the next part of the message follows immediately */
diff --git a/ipc/namespace.c b/ipc/namespace.c
index 9171d948751e..4b4dc6d847f1 100644
--- a/ipc/namespace.c
+++ b/ipc/namespace.c
@@ -25,6 +25,7 @@ static struct ipc_namespace *clone_ipc_ns(struct ipc_namespace *old_ns)
         sem_init_ns(ns);
         msg_init_ns(ns);
         shm_init_ns(ns);
+        mq_init_ns(ns);
 
         /*
          * msgmni has already been computed for the new ipc ns.
@@ -101,6 +102,7 @@ void free_ipc_ns(struct kref *kref)
         sem_exit_ns(ns);
         msg_exit_ns(ns);
         shm_exit_ns(ns);
+        mq_exit_ns(ns);
         kfree(ns);
         atomic_dec(&nr_ipc_ns);
 
diff --git a/ipc/util.c b/ipc/util.c
index 7585a72e259b..b8e4ba92f6d1 100644
--- a/ipc/util.c
+++ b/ipc/util.c
@@ -47,15 +47,6 @@ struct ipc_proc_iface {
         int (*show)(struct seq_file *, void *);
 };
 
-struct ipc_namespace init_ipc_ns = {
-        .kref = {
-                .refcount       = ATOMIC_INIT(2),
-        },
-};
-
-atomic_t nr_ipc_ns = ATOMIC_INIT(1);
-
-
 #ifdef CONFIG_MEMORY_HOTPLUG
 
 static void ipc_memory_notifier(struct work_struct *work)
diff --git a/ipc/util.h b/ipc/util.h
index 3646b45a03c9..0e7d9223acc1 100644
--- a/ipc/util.h
+++ b/ipc/util.h
@@ -20,6 +20,13 @@ void shm_init (void);
 
 struct ipc_namespace;
 
+#ifdef CONFIG_POSIX_MQUEUE
+void mq_exit_ns(struct ipc_namespace *ns);
+#else
+static inline void mq_exit_ns(struct ipc_namespace *ns) { }
+#endif
+
+#ifdef CONFIG_SYSVIPC
 void sem_init_ns(struct ipc_namespace *ns);
 void msg_init_ns(struct ipc_namespace *ns);
 void shm_init_ns(struct ipc_namespace *ns);
@@ -27,6 +34,15 @@ void shm_init_ns(struct ipc_namespace *ns);
 void sem_exit_ns(struct ipc_namespace *ns);
 void msg_exit_ns(struct ipc_namespace *ns);
 void shm_exit_ns(struct ipc_namespace *ns);
+#else
+static inline void sem_init_ns(struct ipc_namespace *ns) { }
+static inline void msg_init_ns(struct ipc_namespace *ns) { }
+static inline void shm_init_ns(struct ipc_namespace *ns) { }
+
+static inline void sem_exit_ns(struct ipc_namespace *ns) { }
+static inline void msg_exit_ns(struct ipc_namespace *ns) { }
+static inline void shm_exit_ns(struct ipc_namespace *ns) { }
+#endif
 
 /*
  * Structure that holds the parameters needed by the ipc operations