[PATCH] binderfs: implement sysctls

From: Christian Brauner
Date: Fri Dec 21 2018 - 08:39:25 EST


This implements three sysctls that have very specific goals:

1. /proc/sys/fs/binder/max:
Allow global root to globally limit the number of allocatable binder
devices.
2. /proc/sys/fs/binder/nr:
Allow global root to easily detect how many binder devices are currently
in use across all binderfs mounts.
3. /proc/sys/fs/binder/reserved:
Ensure that global root can reserve binder devices for the initial
binderfs mount in the initial ipc namespace to prevent DOS attacks.

This is equivalent to sysctls of devpts.

Signed-off-by: Christian Brauner <christian.brauner@xxxxxxxxxx>
---
drivers/android/binderfs.c | 81 +++++++++++++++++++++++++++++++++++++-
1 file changed, 79 insertions(+), 2 deletions(-)

diff --git a/drivers/android/binderfs.c b/drivers/android/binderfs.c
index 7496b10532aa..5ff015f82314 100644
--- a/drivers/android/binderfs.c
+++ b/drivers/android/binderfs.c
@@ -64,6 +64,71 @@ struct binderfs_info {

};

+/* Global default limit on the number of binder devices. */
+static int device_limit = 4096;
+
+/*
+ * Number of binder devices reserved for the initial binderfs mount in the
+ * initial ipc namespace.
+ */
+static int device_reserve = 1024;
+
+/* Dummy sysctl minimum. */
+static int device_limit_min;
+
+/* Cap sysctl at BINDERFS_MAX_MINOR. */
+static int device_limit_max = BINDERFS_MAX_MINOR;
+
+/* Current number of allocated binder devices. */
+static atomic_t device_count = ATOMIC_INIT(0);
+
+static struct ctl_table binderfs_table[] = {
+ {
+ .procname = "max",
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .data = &device_limit,
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = &device_limit_min,
+ .extra2 = &device_limit_max,
+ },
+ {
+ .procname = "reserve",
+ .maxlen = sizeof(int),
+ .mode = 0644,
+ .data = &device_reserve,
+ .proc_handler = proc_dointvec_minmax,
+ .extra1 = &device_limit_min,
+ .extra2 = &device_limit_max,
+ },
+ {
+ .procname = "nr",
+ .maxlen = sizeof(int),
+ .mode = 0444,
+ .data = &device_count,
+ .proc_handler = proc_dointvec,
+ },
+ {}
+};
+
+static struct ctl_table binderfs_fs_table[] = {
+ {
+ .procname = "binder",
+ .mode = 0555,
+ .child = binderfs_table,
+ },
+ {}
+};
+
+static struct ctl_table binderfs_root_table[] = {
+ {
+ .procname = "fs",
+ .mode = 0555,
+ .child = binderfs_fs_table,
+ },
+ {}
+};
+
static inline struct binderfs_info *BINDERFS_I(const struct inode *inode)
{
return inode->i_sb->s_fs_info;
@@ -107,13 +172,21 @@ static int binderfs_binder_device_create(struct inode *ref_inode,
struct inode *inode = NULL;
struct super_block *sb = ref_inode->i_sb;
struct binderfs_info *info = sb->s_fs_info;
+ bool use_reserved = (info->ipc_ns == &init_ipc_ns);

/* Reserve new minor number for the new device. */
mutex_lock(&binderfs_minors_mutex);
- minor = ida_alloc_max(&binderfs_minors, BINDERFS_MAX_MINOR, GFP_KERNEL);
+ if (atomic_inc_return(&device_count) <
+ (device_limit - (use_reserved ? 0 : device_reserve)))
+ minor = ida_alloc_max(&binderfs_minors, BINDERFS_MAX_MINOR,
+ GFP_KERNEL);
+ else
+ minor = -ENOSPC;
mutex_unlock(&binderfs_minors_mutex);
- if (minor < 0)
+ if (minor < 0) {
+ atomic_dec(&device_count);
return minor;
+ }

ret = -ENOMEM;
device = kzalloc(sizeof(*device), GFP_KERNEL);
@@ -187,6 +260,7 @@ static int binderfs_binder_device_create(struct inode *ref_inode,
kfree(name);
kfree(device);
mutex_lock(&binderfs_minors_mutex);
+ atomic_dec(&device_count);
ida_free(&binderfs_minors, minor);
mutex_unlock(&binderfs_minors_mutex);
iput(inode);
@@ -239,6 +313,7 @@ static void binderfs_evict_inode(struct inode *inode)
return;

mutex_lock(&binderfs_minors_mutex);
+ atomic_dec(&device_count);
ida_free(&binderfs_minors, device->miscdev.minor);
mutex_unlock(&binderfs_minors_mutex);

@@ -536,6 +611,8 @@ static int __init init_binderfs(void)
binderfs_mnt = NULL;
unregister_filesystem(&binder_fs_type);
unregister_chrdev_region(binderfs_dev, BINDERFS_MAX_MINOR);
+ } else {
+ register_sysctl_table(binderfs_root_table);
}

return ret;
--
2.19.1