Skip to contentSkip to conversation

x86/ioperm: fix a memory leak bug

PATCH Applied Jay Lang 3 participants 3 messages May 24, 2020 – May 28, 2020 x86

lore
Share & tools
lore

Jay Lang

author
#0 lore raw

PATCH 1/1 x86/ioperm: fix a memory leak bug

In the copy_process() routine called by _do_fork(), failure to allocate a PID (or further along in the function) will trigger an invocation to exit_thread(). This is done to clean up from an earlier call to copy_thread_tls(). Naturally, the child task is passed into exit_thread(), however during the process, io_bitmap_exit() nullifies the parent's io_bitmap rather than the child's.

As copy_thread_tls() has been called ahead of the failure, the reference count on the calling thread's io_bitmap is incremented as we would expect. However, io_bitmap_exit() doesn't accept any arguments, and thus assumes it should trash the current thread's io_bitmap reference rather than the child's. This is pretty sneaky in practice, because in all instances but this one, exit_thread() is called with respect to the current task and everything works out.

A determined attacker can issue an appropriate ioctl (i.e. KDENABIO) to get a bitmap allocated, and force a clone3() syscall to fail by passing in a zeroed clone_args structure. The kernel handles the erroneous struct and the buggy code path is followed, and even though the parent's reference to the io_bitmap is trashed, the child still holds a reference and thus the structure will never be freed.

Fix this by tweaking io_bitmap_exit() and its subroutines to accept a task_struct argument which to operate on. This may not be the most elegant solution, but it mitigates the trigger described above on an x86_64 kernel.

---
 arch/x86/include/asm/io_bitmap.h |  4 ++--
 arch/x86/kernel/ioport.c         | 22 +++++++++++-----------
 arch/x86/kernel/process.c        |  4 ++--
 3 files changed, 15 insertions(+), 15 deletions(-)
arch/x86/include/asm/io_bitmap.h +2 -2
index 07344d82e88e..ac1a99ffbd8d 100644
@@ -17,7 +17,7 @@ struct task_struct;
 
 #ifdef CONFIG_X86_IOPL_IOPERM
 void io_bitmap_share(struct task_struct *tsk);
-void io_bitmap_exit(void);
+void io_bitmap_exit(struct task_struct *tsk);
 
 void native_tss_update_io_bitmap(void);
 
@@ -29,7 +29,7 @@ void native_tss_update_io_bitmap(void);
 
 #else
 static inline void io_bitmap_share(struct task_struct *tsk) { }
-static inline void io_bitmap_exit(void) { }
+static inline void io_bitmap_exit(struct task_struct *tsk) { }
 static inline void tss_update_io_bitmap(void) { }
 #endif
 
arch/x86/kernel/ioport.c +11 -11
index a53e7b4a7419..e2fab3ceb09f 100644
@@ -33,15 +33,15 @@ void io_bitmap_share(struct task_struct *tsk)
 	set_tsk_thread_flag(tsk, TIF_IO_BITMAP);
 }
 
-static void task_update_io_bitmap(void)
+static void task_update_io_bitmap(struct task_struct *tsk)
 {
-	struct thread_struct *t = &current->thread;
+	struct thread_struct *t = &tsk->thread;
 
 	if (t->iopl_emul == 3 || t->io_bitmap) {
 		/* TSS update is handled on exit to user space */
-		set_thread_flag(TIF_IO_BITMAP);
+		set_tsk_thread_flag(tsk, TIF_IO_BITMAP);
 	} else {
-		clear_thread_flag(TIF_IO_BITMAP);
+		clear_tsk_thread_flag(tsk, TIF_IO_BITMAP);
 		/* Invalidate TSS */
 		preempt_disable();
 		tss_update_io_bitmap();
@@ -49,12 +49,12 @@ static void task_update_io_bitmap(void)
 	}
 }
 
-void io_bitmap_exit(void)
+void io_bitmap_exit(struct task_struct *tsk)
 {
-	struct io_bitmap *iobm = current->thread.io_bitmap;
+	struct io_bitmap *iobm = tsk->thread.io_bitmap;
 
-	current->thread.io_bitmap = NULL;
-	task_update_io_bitmap();
+	tsk->thread.io_bitmap = NULL;
+	task_update_io_bitmap(tsk);
 	if (iobm && refcount_dec_and_test(&iobm->refcnt))
 		kfree(iobm);
 }
@@ -102,7 +102,7 @@ long ksys_ioperm(unsigned long from, unsigned long num, int turn_on)
 		if (!iobm)
 			return -ENOMEM;
 		refcount_set(&iobm->refcnt, 1);
-		io_bitmap_exit();
+		io_bitmap_exit(current);
 	}
 
 	/*
@@ -134,7 +134,7 @@ long ksys_ioperm(unsigned long from, unsigned long num, int turn_on)
 	}
 	/* All permissions dropped? */
 	if (max_long == UINT_MAX) {
-		io_bitmap_exit();
+		io_bitmap_exit(current);
 		return 0;
 	}
 
@@ -192,7 +192,7 @@ SYSCALL_DEFINE1(iopl, unsigned int, level)
 	}
 
 	t->iopl_emul = level;
-	task_update_io_bitmap();
+	task_update_io_bitmap(current);
 
 	return 0;
 }
arch/x86/kernel/process.c +2 -2
index 9da70b279dad..35638f1c5791 100644
@@ -96,7 +96,7 @@ int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src)
 }
 
 /*
- * Free current thread data structures etc..
+ * Free thread data structures etc..
  */
 void exit_thread(struct task_struct *tsk)
 {
@@ -104,7 +104,7 @@ void exit_thread(struct task_struct *tsk)
 	struct fpu *fpu = &t->fpu;
 
 	if (test_thread_flag(TIF_IO_BITMAP))
-		io_bitmap_exit();
+		io_bitmap_exit(tsk);
 
 	free_vm86(t);
 
Signature · 1 line
2.25.1
#1 lore raw

Jay,

Quoted 15 lines · Jay Lang

In the copy_process() routine called by _do_fork(), failure to allocate a PID (or further along in the function) will trigger an invocation to exit_thread(). This is done to clean up from an earlier call to copy_thread_tls(). Naturally, the child task is passed into exit_thread(), however during the process, io_bitmap_exit() nullifies the parent's io_bitmap rather than the child's.

As copy_thread_tls() has been called ahead of the failure, the reference count on the calling thread's io_bitmap is incremented as we would expect. However, io_bitmap_exit() doesn't accept any arguments, and thus assumes it should trash the current thread's io_bitmap reference rather than the child's. This is pretty sneaky in practice, because in all instances but this one, exit_thread() is called with respect to the current task and everything works out.

Quoted 6 lines

A determined attacker can issue an appropriate ioctl (i.e. KDENABIO) to get a bitmap allocated, and force a clone3() syscall to fail by passing in a zeroed clone_args structure. The kernel handles the erroneous struct and the buggy code path is followed, and even though the parent's reference to the io_bitmap is trashed, the child still holds a reference and thus the structure will never be freed.

Nice catch! I'm sure I thought about that at some point and then forgot again.

Quoted 5 lines

Fix this by tweaking io_bitmap_exit() and its subroutines to accept a task_struct argument which to operate on.

This may not be the most elegant solution, but it mitigates the trigger described above on an x86_64 kernel.

It's well done and straight forward and comes with a well done change log. Appreciated!

Thanks,

tglx

merge bot Applied
#2 lore raw

[tip: x86/urgent] x86/ioperm: Prevent a memory leak when fork fails

In reply to Jay Lang
The following commit has been merged into the x86/urgent branch of tip:

The following commit has been merged into the x86/urgent branch of tip:

Commit-ID:     4bfe6cce133cad82cea04490c308795275857782
Gitweb:        https://git.kernel.org/tip/4bfe6cce133cad82cea04490c308795275857782
Author:        Jay Lang <[email protected]>
AuthorDate:    Sun, 24 May 2020 12:27:39 -04:00
Committer:     Thomas Gleixner <[email protected]>
CommitterDate: Thu, 28 May 2020 21:36:20 +02:00

x86/ioperm: Prevent a memory leak when fork fails

In the copy_process() routine called by _do_fork(), failure to allocate a PID (or further along in the function) will trigger an invocation to exit_thread(). This is done to clean up from an earlier call to copy_thread_tls(). Naturally, the child task is passed into exit_thread(), however during the process, io_bitmap_exit() nullifies the parent's io_bitmap rather than the child's.

As copy_thread_tls() has been called ahead of the failure, the reference count on the calling thread's io_bitmap is incremented as we would expect. However, io_bitmap_exit() doesn't accept any arguments, and thus assumes it should trash the current thread's io_bitmap reference rather than the child's. This is pretty sneaky in practice, because in all instances but this one, exit_thread() is called with respect to the current task and everything works out.

A determined attacker can issue an appropriate ioctl (i.e. KDENABIO) to get a bitmap allocated, and force a clone3() syscall to fail by passing in a zeroed clone_args structure. The kernel handles the erroneous struct and the buggy code path is followed, and even though the parent's reference to the io_bitmap is trashed, the child still holds a reference and thus the structure will never be freed.

Fix this by tweaking io_bitmap_exit() and its subroutines to accept a task_struct argument which to operate on.

Fixes ea5f1cd7ab49 ("x86/ioperm: Remove bitmap if all permissions dropped")
Cc stable#@vger.kernel.org
---
 arch/x86/include/asm/io_bitmap.h |  4 ++--
 arch/x86/kernel/ioport.c         | 22 +++++++++++-----------
 arch/x86/kernel/process.c        |  4 ++--
 3 files changed, 15 insertions(+), 15 deletions(-)
arch/x86/include/asm/io_bitmap.h +2 -2
index 07344d8..ac1a99f 100644
@@ -17,7 +17,7 @@ struct task_struct;
 
 #ifdef CONFIG_X86_IOPL_IOPERM
 void io_bitmap_share(struct task_struct *tsk);
-void io_bitmap_exit(void);
+void io_bitmap_exit(struct task_struct *tsk);
 
 void native_tss_update_io_bitmap(void);
 
@@ -29,7 +29,7 @@ void native_tss_update_io_bitmap(void);
 
 #else
 static inline void io_bitmap_share(struct task_struct *tsk) { }
-static inline void io_bitmap_exit(void) { }
+static inline void io_bitmap_exit(struct task_struct *tsk) { }
 static inline void tss_update_io_bitmap(void) { }
 #endif
 
arch/x86/kernel/ioport.c +11 -11
index a53e7b4..e2fab3c 100644
@@ -33,15 +33,15 @@ void io_bitmap_share(struct task_struct *tsk)
 	set_tsk_thread_flag(tsk, TIF_IO_BITMAP);
 }
 
-static void task_update_io_bitmap(void)
+static void task_update_io_bitmap(struct task_struct *tsk)
 {
-	struct thread_struct *t = &current->thread;
+	struct thread_struct *t = &tsk->thread;
 
 	if (t->iopl_emul == 3 || t->io_bitmap) {
 		/* TSS update is handled on exit to user space */
-		set_thread_flag(TIF_IO_BITMAP);
+		set_tsk_thread_flag(tsk, TIF_IO_BITMAP);
 	} else {
-		clear_thread_flag(TIF_IO_BITMAP);
+		clear_tsk_thread_flag(tsk, TIF_IO_BITMAP);
 		/* Invalidate TSS */
 		preempt_disable();
 		tss_update_io_bitmap();
@@ -49,12 +49,12 @@ static void task_update_io_bitmap(void)
 	}
 }
 
-void io_bitmap_exit(void)
+void io_bitmap_exit(struct task_struct *tsk)
 {
-	struct io_bitmap *iobm = current->thread.io_bitmap;
+	struct io_bitmap *iobm = tsk->thread.io_bitmap;
 
-	current->thread.io_bitmap = NULL;
-	task_update_io_bitmap();
+	tsk->thread.io_bitmap = NULL;
+	task_update_io_bitmap(tsk);
 	if (iobm && refcount_dec_and_test(&iobm->refcnt))
 		kfree(iobm);
 }
@@ -102,7 +102,7 @@ long ksys_ioperm(unsigned long from, unsigned long num, int turn_on)
 		if (!iobm)
 			return -ENOMEM;
 		refcount_set(&iobm->refcnt, 1);
-		io_bitmap_exit();
+		io_bitmap_exit(current);
 	}
 
 	/*
@@ -134,7 +134,7 @@ long ksys_ioperm(unsigned long from, unsigned long num, int turn_on)
 	}
 	/* All permissions dropped? */
 	if (max_long == UINT_MAX) {
-		io_bitmap_exit();
+		io_bitmap_exit(current);
 		return 0;
 	}
 
@@ -192,7 +192,7 @@ SYSCALL_DEFINE1(iopl, unsigned int, level)
 	}
 
 	t->iopl_emul = level;
-	task_update_io_bitmap();
+	task_update_io_bitmap(current);
 
 	return 0;
 }
arch/x86/kernel/process.c +2 -2
index 9da70b2..35638f1 100644
@@ -96,7 +96,7 @@ int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src)
 }
 
 /*
- * Free current thread data structures etc..
+ * Free thread data structures etc..
  */
 void exit_thread(struct task_struct *tsk)
 {
@@ -104,7 +104,7 @@ void exit_thread(struct task_struct *tsk)
 	struct fpu *fpu = &t->fpu;
 
 	if (test_thread_flag(TIF_IO_BITMAP))
-		io_bitmap_exit();
+		io_bitmap_exit(tsk);
 
 	free_vm86(t);