[PATCH] nommu build error fix
[deliverable/linux.git] / fs / proc / base.c
index 520978e49e92474f428054785c6c68a1978138a5..a170450aadb1adc0ef3f81b595b0eacd9978ede7 100644 (file)
@@ -62,6 +62,7 @@
 #include <linux/namespace.h>
 #include <linux/mm.h>
 #include <linux/smp_lock.h>
+#include <linux/rcupdate.h>
 #include <linux/kallsyms.h>
 #include <linux/mount.h>
 #include <linux/security.h>
@@ -102,7 +103,9 @@ enum pid_directory_inos {
        PROC_TGID_NUMA_MAPS,
        PROC_TGID_MOUNTS,
        PROC_TGID_WCHAN,
+#ifdef CONFIG_MMU
        PROC_TGID_SMAPS,
+#endif
 #ifdef CONFIG_SCHEDSTATS
        PROC_TGID_SCHEDSTAT,
 #endif
@@ -119,7 +122,6 @@ enum pid_directory_inos {
 #ifdef CONFIG_AUDITSYSCALL
        PROC_TGID_LOGINUID,
 #endif
-       PROC_TGID_FD_DIR,
        PROC_TGID_OOM_SCORE,
        PROC_TGID_OOM_ADJUST,
        PROC_TID_INO,
@@ -141,7 +143,9 @@ enum pid_directory_inos {
        PROC_TID_NUMA_MAPS,
        PROC_TID_MOUNTS,
        PROC_TID_WCHAN,
+#ifdef CONFIG_MMU
        PROC_TID_SMAPS,
+#endif
 #ifdef CONFIG_SCHEDSTATS
        PROC_TID_SCHEDSTAT,
 #endif
@@ -158,9 +162,11 @@ enum pid_directory_inos {
 #ifdef CONFIG_AUDITSYSCALL
        PROC_TID_LOGINUID,
 #endif
-       PROC_TID_FD_DIR = 0x8000,       /* 0x8000-0xffff */
        PROC_TID_OOM_SCORE,
        PROC_TID_OOM_ADJUST,
+
+       /* Add new entries before this */
+       PROC_TID_FD_DIR = 0x8000,       /* 0x8000-0xffff */
 };
 
 struct pid_entry {
@@ -193,7 +199,9 @@ static struct pid_entry tgid_base_stuff[] = {
        E(PROC_TGID_ROOT,      "root",    S_IFLNK|S_IRWXUGO),
        E(PROC_TGID_EXE,       "exe",     S_IFLNK|S_IRWXUGO),
        E(PROC_TGID_MOUNTS,    "mounts",  S_IFREG|S_IRUGO),
+#ifdef CONFIG_MMU
        E(PROC_TGID_SMAPS,     "smaps",   S_IFREG|S_IRUGO),
+#endif
 #ifdef CONFIG_SECURITY
        E(PROC_TGID_ATTR,      "attr",    S_IFDIR|S_IRUGO|S_IXUGO),
 #endif
@@ -233,7 +241,9 @@ static struct pid_entry tid_base_stuff[] = {
        E(PROC_TID_ROOT,       "root",    S_IFLNK|S_IRWXUGO),
        E(PROC_TID_EXE,        "exe",     S_IFLNK|S_IRWXUGO),
        E(PROC_TID_MOUNTS,     "mounts",  S_IFREG|S_IRUGO),
+#ifdef CONFIG_MMU
        E(PROC_TID_SMAPS,      "smaps",   S_IFREG|S_IRUGO),
+#endif
 #ifdef CONFIG_SECURITY
        E(PROC_TID_ATTR,       "attr",    S_IFDIR|S_IRUGO|S_IXUGO),
 #endif
@@ -282,30 +292,36 @@ static int proc_fd_link(struct inode *inode, struct dentry **dentry, struct vfsm
 
        files = get_files_struct(task);
        if (files) {
-               spin_lock(&files->file_lock);
+               rcu_read_lock();
                file = fcheck_files(files, fd);
                if (file) {
                        *mnt = mntget(file->f_vfsmnt);
                        *dentry = dget(file->f_dentry);
-                       spin_unlock(&files->file_lock);
+                       rcu_read_unlock();
                        put_files_struct(files);
                        return 0;
                }
-               spin_unlock(&files->file_lock);
+               rcu_read_unlock();
                put_files_struct(files);
        }
        return -ENOENT;
 }
 
-static int proc_cwd_link(struct inode *inode, struct dentry **dentry, struct vfsmount **mnt)
+static struct fs_struct *get_fs_struct(struct task_struct *task)
 {
        struct fs_struct *fs;
-       int result = -ENOENT;
-       task_lock(proc_task(inode));
-       fs = proc_task(inode)->fs;
+       task_lock(task);
+       fs = task->fs;
        if(fs)
                atomic_inc(&fs->count);
-       task_unlock(proc_task(inode));
+       task_unlock(task);
+       return fs;
+}
+
+static int proc_cwd_link(struct inode *inode, struct dentry **dentry, struct vfsmount **mnt)
+{
+       struct fs_struct *fs = get_fs_struct(proc_task(inode));
+       int result = -ENOENT;
        if (fs) {
                read_lock(&fs->lock);
                *mnt = mntget(fs->pwdmnt);
@@ -318,14 +334,56 @@ static int proc_cwd_link(struct inode *inode, struct dentry **dentry, struct vfs
 }
 
 static int proc_root_link(struct inode *inode, struct dentry **dentry, struct vfsmount **mnt)
+{
+       struct fs_struct *fs = get_fs_struct(proc_task(inode));
+       int result = -ENOENT;
+       if (fs) {
+               read_lock(&fs->lock);
+               *mnt = mntget(fs->rootmnt);
+               *dentry = dget(fs->root);
+               read_unlock(&fs->lock);
+               result = 0;
+               put_fs_struct(fs);
+       }
+       return result;
+}
+
+
+/* Same as proc_root_link, but this addionally tries to get fs from other
+ * threads in the group */
+static int proc_task_root_link(struct inode *inode, struct dentry **dentry,
+                               struct vfsmount **mnt)
 {
        struct fs_struct *fs;
        int result = -ENOENT;
-       task_lock(proc_task(inode));
-       fs = proc_task(inode)->fs;
-       if(fs)
+       struct task_struct *leader = proc_task(inode);
+
+       task_lock(leader);
+       fs = leader->fs;
+       if (fs) {
                atomic_inc(&fs->count);
-       task_unlock(proc_task(inode));
+               task_unlock(leader);
+       } else {
+               /* Try to get fs from other threads */
+               task_unlock(leader);
+               read_lock(&tasklist_lock);
+               if (pid_alive(leader)) {
+                       struct task_struct *task = leader;
+
+                       while ((task = next_thread(task)) != leader) {
+                               task_lock(task);
+                               fs = task->fs;
+                               if (fs) {
+                                       atomic_inc(&fs->count);
+                                       task_unlock(task);
+                                       break;
+                               }
+                               task_unlock(task);
+                       }
+               }
+               read_unlock(&tasklist_lock);
+       }
+
        if (fs) {
                read_lock(&fs->lock);
                *mnt = mntget(fs->rootmnt);
@@ -337,6 +395,7 @@ static int proc_root_link(struct inode *inode, struct dentry **dentry, struct vf
        return result;
 }
 
+
 #define MAY_PTRACE(task) \
        (task == current || \
        (task->parent == current && \
@@ -344,33 +403,6 @@ static int proc_root_link(struct inode *inode, struct dentry **dentry, struct vf
         (task->state == TASK_STOPPED || task->state == TASK_TRACED) && \
         security_ptrace(current,task) == 0))
 
-static int may_ptrace_attach(struct task_struct *task)
-{
-       int retval = 0;
-
-       task_lock(task);
-
-       if (!task->mm)
-               goto out;
-       if (((current->uid != task->euid) ||
-            (current->uid != task->suid) ||
-            (current->uid != task->uid) ||
-            (current->gid != task->egid) ||
-            (current->gid != task->sgid) ||
-            (current->gid != task->gid)) && !capable(CAP_SYS_PTRACE))
-               goto out;
-       rmb();
-       if (task->mm->dumpable != 1 && !capable(CAP_SYS_PTRACE))
-               goto out;
-       if (security_ptrace(current, task))
-               goto out;
-
-       retval = 1;
-out:
-       task_unlock(task);
-       return retval;
-}
-
 static int proc_pid_environ(struct task_struct *task, char * buffer)
 {
        int res = 0;
@@ -380,7 +412,7 @@ static int proc_pid_environ(struct task_struct *task, char * buffer)
                if (len > PAGE_SIZE)
                        len = PAGE_SIZE;
                res = access_process_vm(task, mm->env_start, buffer, len, 0);
-               if (!may_ptrace_attach(task))
+               if (!ptrace_may_attach(task))
                        res = -ESRCH;
                mmput(mm);
        }
@@ -495,14 +527,14 @@ static int proc_oom_score(struct task_struct *task, char *buffer)
 
 /* permission checks */
 
-static int proc_check_root(struct inode *inode)
+/* If the process being read is separated by chroot from the reading process,
+ * don't let the reader access the threads.
+ */
+static int proc_check_chroot(struct dentry *root, struct vfsmount *vfsmnt)
 {
-       struct dentry *de, *base, *root;
-       struct vfsmount *our_vfsmnt, *vfsmnt, *mnt;
+       struct dentry *de, *base;
+       struct vfsmount *our_vfsmnt, *mnt;
        int res = 0;
-
-       if (proc_root_link(inode, &root, &vfsmnt)) /* Ewww... */
-               return -ENOENT;
        read_lock(&current->fs->lock);
        our_vfsmnt = mntget(current->fs->rootmnt);
        base = dget(current->fs->root);
@@ -535,6 +567,16 @@ out:
        goto exit;
 }
 
+static int proc_check_root(struct inode *inode)
+{
+       struct dentry *root;
+       struct vfsmount *vfsmnt;
+
+       if (proc_root_link(inode, &root, &vfsmnt)) /* Ewww... */
+               return -ENOENT;
+       return proc_check_chroot(root, vfsmnt);
+}
+
 static int proc_permission(struct inode *inode, int mask, struct nameidata *nd)
 {
        if (generic_permission(inode, mask, NULL) != 0)
@@ -542,6 +584,20 @@ static int proc_permission(struct inode *inode, int mask, struct nameidata *nd)
        return proc_check_root(inode);
 }
 
+static int proc_task_permission(struct inode *inode, int mask, struct nameidata *nd)
+{
+       struct dentry *root;
+       struct vfsmount *vfsmnt;
+
+       if (generic_permission(inode, mask, NULL) != 0)
+               return -EACCES;
+
+       if (proc_task_root_link(inode, &root, &vfsmnt))
+               return -ENOENT;
+
+       return proc_check_chroot(root, vfsmnt);
+}
+
 extern struct seq_operations proc_pid_maps_op;
 static int maps_open(struct inode *inode, struct file *file)
 {
@@ -582,6 +638,7 @@ static struct file_operations proc_numa_maps_operations = {
 };
 #endif
 
+#ifdef CONFIG_MMU
 extern struct seq_operations proc_pid_smaps_op;
 static int smaps_open(struct inode *inode, struct file *file)
 {
@@ -600,6 +657,7 @@ static struct file_operations proc_smaps_operations = {
        .llseek         = seq_lseek,
        .release        = seq_release,
 };
+#endif
 
 extern struct seq_operations mounts_op;
 static int mounts_open(struct inode *inode, struct file *file)
@@ -683,7 +741,7 @@ static ssize_t mem_read(struct file * file, char __user * buf,
        int ret = -ESRCH;
        struct mm_struct *mm;
 
-       if (!MAY_PTRACE(task) || !may_ptrace_attach(task))
+       if (!MAY_PTRACE(task) || !ptrace_may_attach(task))
                goto out;
 
        ret = -ENOMEM;
@@ -709,7 +767,7 @@ static ssize_t mem_read(struct file * file, char __user * buf,
 
                this_len = (count > PAGE_SIZE) ? PAGE_SIZE : count;
                retval = access_process_vm(task, src, page, this_len, 0);
-               if (!retval || !MAY_PTRACE(task) || !may_ptrace_attach(task)) {
+               if (!retval || !MAY_PTRACE(task) || !ptrace_may_attach(task)) {
                        if (!ret)
                                ret = -EIO;
                        break;
@@ -747,7 +805,7 @@ static ssize_t mem_write(struct file * file, const char * buf,
        struct task_struct *task = proc_task(file->f_dentry->d_inode);
        unsigned long dst = *ppos;
 
-       if (!MAY_PTRACE(task) || !may_ptrace_attach(task))
+       if (!MAY_PTRACE(task) || !ptrace_may_attach(task))
                return -ESRCH;
 
        page = (char *)__get_free_page(GFP_USER);
@@ -1064,6 +1122,7 @@ static int proc_readfd(struct file * filp, void * dirent, filldir_t filldir)
        int retval;
        char buf[NUMBUF];
        struct files_struct * files;
+       struct fdtable *fdt;
 
        retval = -ENOENT;
        if (!pid_alive(p))
@@ -1086,15 +1145,16 @@ static int proc_readfd(struct file * filp, void * dirent, filldir_t filldir)
                        files = get_files_struct(p);
                        if (!files)
                                goto out;
-                       spin_lock(&files->file_lock);
+                       rcu_read_lock();
+                       fdt = files_fdtable(files);
                        for (fd = filp->f_pos-2;
-                            fd < files->max_fds;
+                            fd < fdt->max_fds;
                             fd++, filp->f_pos++) {
                                unsigned int i,j;
 
                                if (!fcheck_files(files, fd))
                                        continue;
-                               spin_unlock(&files->file_lock);
+                               rcu_read_unlock();
 
                                j = NUMBUF;
                                i = fd;
@@ -1106,12 +1166,12 @@ static int proc_readfd(struct file * filp, void * dirent, filldir_t filldir)
 
                                ino = fake_ino(tid, PROC_TID_FD_DIR + fd);
                                if (filldir(dirent, buf+j, NUMBUF-j, fd+2, ino, DT_LNK) < 0) {
-                                       spin_lock(&files->file_lock);
+                                       rcu_read_lock();
                                        break;
                                }
-                               spin_lock(&files->file_lock);
+                               rcu_read_lock();
                        }
-                       spin_unlock(&files->file_lock);
+                       rcu_read_unlock();
                        put_files_struct(files);
        }
 out:
@@ -1286,9 +1346,9 @@ static int tid_fd_revalidate(struct dentry *dentry, struct nameidata *nd)
 
        files = get_files_struct(task);
        if (files) {
-               spin_lock(&files->file_lock);
+               rcu_read_lock();
                if (fcheck_files(files, fd)) {
-                       spin_unlock(&files->file_lock);
+                       rcu_read_unlock();
                        put_files_struct(files);
                        if (task_dumpable(task)) {
                                inode->i_uid = task->euid;
@@ -1300,7 +1360,7 @@ static int tid_fd_revalidate(struct dentry *dentry, struct nameidata *nd)
                        security_task_to_inode(task, inode);
                        return 1;
                }
-               spin_unlock(&files->file_lock);
+               rcu_read_unlock();
                put_files_struct(files);
        }
        d_drop(dentry);
@@ -1392,7 +1452,7 @@ static struct dentry *proc_lookupfd(struct inode * dir, struct dentry * dentry,
        if (!files)
                goto out_unlock;
        inode->i_mode = S_IFLNK;
-       spin_lock(&files->file_lock);
+       rcu_read_lock();
        file = fcheck_files(files, fd);
        if (!file)
                goto out_unlock2;
@@ -1400,7 +1460,7 @@ static struct dentry *proc_lookupfd(struct inode * dir, struct dentry * dentry,
                inode->i_mode |= S_IRUSR | S_IXUSR;
        if (file->f_mode & 2)
                inode->i_mode |= S_IWUSR | S_IXUSR;
-       spin_unlock(&files->file_lock);
+       rcu_read_unlock();
        put_files_struct(files);
        inode->i_op = &proc_pid_link_inode_operations;
        inode->i_size = 64;
@@ -1410,7 +1470,7 @@ static struct dentry *proc_lookupfd(struct inode * dir, struct dentry * dentry,
        return NULL;
 
 out_unlock2:
-       spin_unlock(&files->file_lock);
+       rcu_read_unlock();
        put_files_struct(files);
 out_unlock:
        iput(inode);
@@ -1441,7 +1501,7 @@ static struct inode_operations proc_fd_inode_operations = {
 
 static struct inode_operations proc_task_inode_operations = {
        .lookup         = proc_task_lookup,
-       .permission     = proc_permission,
+       .permission     = proc_task_permission,
 };
 
 #ifdef CONFIG_SECURITY
@@ -1631,10 +1691,12 @@ static struct dentry *proc_pident_lookup(struct inode *dir,
                case PROC_TGID_MOUNTS:
                        inode->i_fop = &proc_mounts_operations;
                        break;
+#ifdef CONFIG_MMU
                case PROC_TID_SMAPS:
                case PROC_TGID_SMAPS:
                        inode->i_fop = &proc_smaps_operations;
                        break;
+#endif
 #ifdef CONFIG_SECURITY
                case PROC_TID_ATTR:
                        inode->i_nlink = 2;
This page took 0.029662 seconds and 5 git commands to generate.