Merge branch 'perf/core' into perf/uprobes

Merge in latest upstream (and the latest perf development tree),
to prepare for tooling changes, and also to pick up v3.4 MM
changes that the uprobes code needs to take care of.

Signed-off-by: Ingo Molnar <mingo@kernel.org>

1 files changed, 82 insertions, 6 deletions

diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c
index c4e2a58a2e8..10f5e0b484d 100644
--- a/fs/devpts/inode.c
+++ b/fs/devpts/inode.c
@@ -36,7 +36,61 @@
 #define DEVPTS_DEFAULT_PTMX_MODE 0000
 #define PTMX_MINOR	2
 
-extern int pty_limit;			/* Config limit on Unix98 ptys */
+/*
+ * sysctl support for setting limits on the number of Unix98 ptys allocated.
+ * Otherwise one can eat up all kernel memory by opening /dev/ptmx repeatedly.
+ */
+static int pty_limit = NR_UNIX98_PTY_DEFAULT;
+static int pty_reserve = NR_UNIX98_PTY_RESERVE;
+static int pty_limit_min;
+static int pty_limit_max = INT_MAX;
+static int pty_count;
+
+static struct ctl_table pty_table[] = {
+	{
+		.procname	= "max",
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.data		= &pty_limit,
+		.proc_handler	= proc_dointvec_minmax,
+		.extra1		= &pty_limit_min,
+		.extra2		= &pty_limit_max,
+	}, {
+		.procname	= "reserve",
+		.maxlen		= sizeof(int),
+		.mode		= 0644,
+		.data		= &pty_reserve,
+		.proc_handler	= proc_dointvec_minmax,
+		.extra1		= &pty_limit_min,
+		.extra2		= &pty_limit_max,
+	}, {
+		.procname	= "nr",
+		.maxlen		= sizeof(int),
+		.mode		= 0444,
+		.data		= &pty_count,
+		.proc_handler	= proc_dointvec,
+	},
+	{}
+};
+
+static struct ctl_table pty_kern_table[] = {
+	{
+		.procname	= "pty",
+		.mode		= 0555,
+		.child		= pty_table,
+	},
+	{}
+};
+
+static struct ctl_table pty_root_table[] = {
+	{
+		.procname	= "kernel",
+		.mode		= 0555,
+		.child		= pty_kern_table,
+	},
+	{}
+};
+
 static DEFINE_MUTEX(allocated_ptys_lock);
 
 static struct vfsmount *devpts_mnt;
@@ -49,10 +103,11 @@ struct pts_mount_opts {
 	umode_t mode;
 	umode_t ptmxmode;
 	int newinstance;
+	int max;
 };
 
 enum {
-	Opt_uid, Opt_gid, Opt_mode, Opt_ptmxmode, Opt_newinstance,
+	Opt_uid, Opt_gid, Opt_mode, Opt_ptmxmode, Opt_newinstance,  Opt_max,
 	Opt_err
 };
 
@@ -63,6 +118,7 @@ static const match_table_t tokens = {
 #ifdef CONFIG_DEVPTS_MULTIPLE_INSTANCES
 	{Opt_ptmxmode, "ptmxmode=%o"},
 	{Opt_newinstance, "newinstance"},
+	{Opt_max, "max=%d"},
 #endif
 	{Opt_err, NULL}
 };
@@ -109,6 +165,7 @@ static int parse_mount_options(char *data, int op, struct pts_mount_opts *opts)
 	opts->gid     = 0;
 	opts->mode    = DEVPTS_DEFAULT_MODE;
 	opts->ptmxmode = DEVPTS_DEFAULT_PTMX_MODE;
+	opts->max     = NR_UNIX98_PTY_MAX;
 
 	/* newinstance makes sense only on initial mount */
 	if (op == PARSE_MOUNT)
@@ -152,6 +209,12 @@ static int parse_mount_options(char *data, int op, struct pts_mount_opts *opts)
 			if (op == PARSE_MOUNT)
 				opts->newinstance = 1;
 			break;
+		case Opt_max:
+			if (match_int(&args[0], &option) ||
+			    option < 0 || option > NR_UNIX98_PTY_MAX)
+				return -EINVAL;
+			opts->max = option;
+			break;
 #endif
 		default:
 			printk(KERN_ERR "devpts: called with bogus options\n");
@@ -258,6 +321,8 @@ static int devpts_show_options(struct seq_file *seq, struct dentry *root)
 	seq_printf(seq, ",mode=%03o", opts->mode);
 #ifdef CONFIG_DEVPTS_MULTIPLE_INSTANCES
 	seq_printf(seq, ",ptmxmode=%03o", opts->ptmxmode);
+	if (opts->max < NR_UNIX98_PTY_MAX)
+		seq_printf(seq, ",max=%d", opts->max);
 #endif
 
 	return 0;
@@ -309,12 +374,11 @@ devpts_fill_super(struct super_block *s, void *data, int silent)
 	inode->i_fop = &simple_dir_operations;
 	set_nlink(inode, 2);
 
-	s->s_root = d_alloc_root(inode);
+	s->s_root = d_make_root(inode);
 	if (s->s_root)
 		return 0;
 
 	printk(KERN_ERR "devpts: get root dentry failed\n");
-	iput(inode);
 
 fail:
 	return -ENOMEM;
@@ -438,6 +502,12 @@ retry:
 		return -ENOMEM;
 
 	mutex_lock(&allocated_ptys_lock);
+	if (pty_count >= pty_limit -
+			(fsi->mount_opts.newinstance ? pty_reserve : 0)) {
+		mutex_unlock(&allocated_ptys_lock);
+		return -ENOSPC;
+	}
+
 	ida_ret = ida_get_new(&fsi->allocated_ptys, &index);
 	if (ida_ret < 0) {
 		mutex_unlock(&allocated_ptys_lock);
@@ -446,11 +516,12 @@ retry:
 		return -EIO;
 	}
 
-	if (index >= pty_limit) {
+	if (index >= fsi->mount_opts.max) {
 		ida_remove(&fsi->allocated_ptys, index);
 		mutex_unlock(&allocated_ptys_lock);
-		return -EIO;
+		return -ENOSPC;
 	}
+	pty_count++;
 	mutex_unlock(&allocated_ptys_lock);
 	return index;
 }
@@ -462,6 +533,7 @@ void devpts_kill_index(struct inode *ptmx_inode, int idx)
 
 	mutex_lock(&allocated_ptys_lock);
 	ida_remove(&fsi->allocated_ptys, idx);
+	pty_count--;
 	mutex_unlock(&allocated_ptys_lock);
 }
 
@@ -558,11 +630,15 @@ void devpts_pty_kill(struct tty_struct *tty)
 static int __init init_devpts_fs(void)
 {
 	int err = register_filesystem(&devpts_fs_type);
+	struct ctl_table_header *table;
+
 	if (!err) {
+		table = register_sysctl_table(pty_root_table);
 		devpts_mnt = kern_mount(&devpts_fs_type);
 		if (IS_ERR(devpts_mnt)) {
 			err = PTR_ERR(devpts_mnt);
 			unregister_filesystem(&devpts_fs_type);
+			unregister_sysctl_table(table);
 		}
 	}
 	return err;