binder: fix possible UAF when freeing buffer
diff mbox series

Message ID 20190612202927.54518-1-tkjos@google.com
State Accepted
Commit a370003cc301d4361bae20c9ef615f89bf8d1e8a
Headers show
Series
  • binder: fix possible UAF when freeing buffer
Related show

Commit Message

Todd Kjos June 12, 2019, 8:29 p.m. UTC
There is a race between the binder driver cleaning
up a completed transaction via binder_free_transaction()
and a user calling binder_ioctl(BC_FREE_BUFFER) to
release a buffer. It doesn't matter which is first but
they need to be protected against running concurrently
which can result in a UAF.

Signed-off-by: Todd Kjos <tkjos@google.com>
---
 drivers/android/binder.c | 16 ++++++++++++++--
 1 file changed, 14 insertions(+), 2 deletions(-)

Comments

Greg Kroah-Hartman June 13, 2019, 5:41 a.m. UTC | #1
On Wed, Jun 12, 2019 at 01:29:27PM -0700, Todd Kjos wrote:
> There is a race between the binder driver cleaning
> up a completed transaction via binder_free_transaction()
> and a user calling binder_ioctl(BC_FREE_BUFFER) to
> release a buffer. It doesn't matter which is first but
> they need to be protected against running concurrently
> which can result in a UAF.
> 
> Signed-off-by: Todd Kjos <tkjos@google.com>
> ---
>  drivers/android/binder.c | 16 ++++++++++++++--
>  1 file changed, 14 insertions(+), 2 deletions(-)

Does this also need to go to the stable kernels?

thanks,

greg k-h
Todd Kjos June 13, 2019, 3:43 p.m. UTC | #2
On Wed, Jun 12, 2019 at 10:41 PM Greg KH <gregkh@linuxfoundation.org> wrote:
>
> On Wed, Jun 12, 2019 at 01:29:27PM -0700, Todd Kjos wrote:
> > There is a race between the binder driver cleaning
> > up a completed transaction via binder_free_transaction()
> > and a user calling binder_ioctl(BC_FREE_BUFFER) to
> > release a buffer. It doesn't matter which is first but
> > they need to be protected against running concurrently
> > which can result in a UAF.
> >
> > Signed-off-by: Todd Kjos <tkjos@google.com>
> > ---
> >  drivers/android/binder.c | 16 ++++++++++++++--
> >  1 file changed, 14 insertions(+), 2 deletions(-)
>
> Does this also need to go to the stable kernels?

This patch won't apply cleanly to stable kernels so I'll submit a new
patch directly to stable@ (4.14, 4.19) once this one lands.

>
> thanks,
>
> greg k-h

Patch
diff mbox series

diff --git a/drivers/android/binder.c b/drivers/android/binder.c
index 748ac489ef7eb..bc26b5511f0a9 100644
--- a/drivers/android/binder.c
+++ b/drivers/android/binder.c
@@ -1941,8 +1941,18 @@  static void binder_free_txn_fixups(struct binder_transaction *t)
 
 static void binder_free_transaction(struct binder_transaction *t)
 {
-	if (t->buffer)
-		t->buffer->transaction = NULL;
+	struct binder_proc *target_proc = t->to_proc;
+
+	if (target_proc) {
+		binder_inner_proc_lock(target_proc);
+		if (t->buffer)
+			t->buffer->transaction = NULL;
+		binder_inner_proc_unlock(target_proc);
+	}
+	/*
+	 * If the transaction has no target_proc, then
+	 * t->buffer->transaction has already been cleared.
+	 */
 	binder_free_txn_fixups(t);
 	kfree(t);
 	binder_stats_deleted(BINDER_STAT_TRANSACTION);
@@ -3551,10 +3561,12 @@  static void binder_transaction(struct binder_proc *proc,
 static void
 binder_free_buf(struct binder_proc *proc, struct binder_buffer *buffer)
 {
+	binder_inner_proc_lock(proc);
 	if (buffer->transaction) {
 		buffer->transaction->buffer = NULL;
 		buffer->transaction = NULL;
 	}
+	binder_inner_proc_unlock(proc);
 	if (buffer->async_transaction && buffer->target_node) {
 		struct binder_node *buf_node;
 		struct binder_work *w;