diff mbox series

[v2,2/6] exit: Put an upper limit on how often we can oops

Message ID 20221109200050.3400857-2-keescook@chromium.org (mailing list archive)
State New
Headers show
Series exit: Put an upper limit on how often we can oops | expand

Commit Message

Kees Cook Nov. 9, 2022, 8 p.m. UTC 
From: Jann Horn <jannh@google.com>

Many Linux systems are configured to not panic on oops; but allowing an
attacker to oops the system **really** often can make even bugs that look
completely unexploitable exploitable (like NULL dereferences and such) if
each crash elevates a refcount by one or a lock is taken in read mode, and
this causes a counter to eventually overflow.

The most interesting counters for this are 32 bits wide (like open-coded
refcounts that don't use refcount_t). (The ldsem reader count on 32-bit
platforms is just 16 bits, but probably nobody cares about 32-bit platforms
that much nowadays.)

So let's panic the system if the kernel is constantly oopsing.

The speed of oopsing 2^32 times probably depends on several factors, like
how long the stack trace is and which unwinder you're using; an empirically
important one is whether your console is showing a graphical environment or
a text console that oopses will be printed to.
In a quick single-threaded benchmark, it looks like oopsing in a vfork()
child with a very short stack trace only takes ~510 microseconds per run
when a graphical console is active; but switching to a text console that
oopses are printed to slows it down around 87x, to ~45 milliseconds per
run.
(Adding more threads makes this faster, but the actual oops printing
happens under &die_lock on x86, so you can maybe speed this up by a factor
of around 2 and then any further improvement gets eaten up by lock
contention.)

It looks like it would take around 8-12 days to overflow a 32-bit counter
with repeated oopsing on a multi-core X86 system running a graphical
environment; both me (in an X86 VM) and Seth (with a distro kernel on
normal hardware in a standard configuration) got numbers in that ballpark.

12 days aren't *that* short on a desktop system, and you'd likely need much
longer on a typical server system (assuming that people don't run graphical
desktop environments on their servers), and this is a *very* noisy and
violent approach to exploiting the kernel; and it also seems to take orders
of magnitude longer on some machines, probably because stuff like EFI
pstore will slow it down a ton if that's active.

[Moved sysctl into kernel/exit.c -kees]

Signed-off-by: Jann Horn <jannh@google.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20221107201317.324457-1-jannh@google.com
---
 Documentation/admin-guide/sysctl/kernel.rst |  8 ++++
 kernel/exit.c                               | 42 +++++++++++++++++++++
 2 files changed, 50 insertions(+)

Comments

kernel test robot Nov. 12, 2022, 12:35 p.m. UTC | #1 
Hi Kees,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on next-20221109]

url:    https://github.com/intel-lab-lkp/linux/commits/Kees-Cook/exit-Put-an-upper-limit-on-how-often-we-can-oops/20221110-040244
patch link:    https://lore.kernel.org/r/20221109200050.3400857-2-keescook%40chromium.org
patch subject: [PATCH v2 2/6] exit: Put an upper limit on how often we can oops
config: m68k-randconfig-r024-20221111
compiler: m68k-linux-gcc (GCC) 12.1.0
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # https://github.com/intel-lab-lkp/linux/commit/3db52f787eeec7924f3c2f952ac8fdf33b70b0bf
        git remote add linux-review https://github.com/intel-lab-lkp/linux
        git fetch --no-tags linux-review Kees-Cook/exit-Put-an-upper-limit-on-how-often-we-can-oops/20221110-040244
        git checkout 3db52f787eeec7924f3c2f952ac8fdf33b70b0bf
        # save the config file
        mkdir build_dir && cp config build_dir/.config
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-12.1.0 make.cross W=1 O=build_dir ARCH=m68k SHELL=/bin/bash

If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <lkp@intel.com>

All warnings (new ones prefixed by >>):

>> kernel/exit.c:82:5: warning: "CONFIG_SYSCTL" is not defined, evaluates to 0 [-Wundef]
      82 | #if CONFIG_SYSCTL
         |     ^~~~~~~~~~~~~
   kernel/exit.c:1881:13: warning: no previous prototype for 'abort' [-Wmissing-prototypes]
    1881 | __weak void abort(void)
         |             ^~~~~


vim +/CONFIG_SYSCTL +82 kernel/exit.c

    81	
  > 82	#if CONFIG_SYSCTL
    83	static struct ctl_table kern_exit_table[] = {
    84		{
    85			.procname       = "oops_limit",
    86			.data           = &oops_limit,
    87			.maxlen         = sizeof(oops_limit),
    88			.mode           = 0644,
    89			.proc_handler   = proc_douintvec,
    90		},
    91		{ }
    92	};
    93
kernel test robot Nov. 13, 2022, 7:44 p.m. UTC | #2 
Hi Kees,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on next-20221109]

url:    https://github.com/intel-lab-lkp/linux/commits/Kees-Cook/exit-Put-an-upper-limit-on-how-often-we-can-oops/20221110-040244
patch link:    https://lore.kernel.org/r/20221109200050.3400857-2-keescook%40chromium.org
patch subject: [PATCH v2 2/6] exit: Put an upper limit on how often we can oops
config: riscv-randconfig-s053-20221113
compiler: riscv32-linux-gcc (GCC) 12.1.0
reproduce:
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # apt-get install sparse
        # sparse version: v0.6.4-39-gce1a6720-dirty
        # https://github.com/intel-lab-lkp/linux/commit/3db52f787eeec7924f3c2f952ac8fdf33b70b0bf
        git remote add linux-review https://github.com/intel-lab-lkp/linux
        git fetch --no-tags linux-review Kees-Cook/exit-Put-an-upper-limit-on-how-often-we-can-oops/20221110-040244
        git checkout 3db52f787eeec7924f3c2f952ac8fdf33b70b0bf
        # save the config file
        mkdir build_dir && cp config build_dir/.config
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-12.1.0 make.cross C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' O=build_dir ARCH=riscv SHELL=/bin/bash

If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <lkp@intel.com>

sparse warnings: (new ones prefixed by >>)
   WARNING: invalid argument to '-march': '_zihintpause'
>> kernel/exit.c:82:5: sparse: sparse: undefined preprocessor identifier 'CONFIG_SYSCTL'
   kernel/exit.c:313:37: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct task_struct *tsk @@     got struct task_struct [noderef] __rcu *real_parent @@
   kernel/exit.c:313:37: sparse:     expected struct task_struct *tsk
   kernel/exit.c:313:37: sparse:     got struct task_struct [noderef] __rcu *real_parent
   kernel/exit.c:316:32: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct task_struct *task @@     got struct task_struct [noderef] __rcu *real_parent @@
   kernel/exit.c:316:32: sparse:     expected struct task_struct *task
   kernel/exit.c:316:32: sparse:     got struct task_struct [noderef] __rcu *real_parent
   kernel/exit.c:317:35: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct task_struct *task @@     got struct task_struct [noderef] __rcu *real_parent @@
   kernel/exit.c:317:35: sparse:     expected struct task_struct *task
   kernel/exit.c:317:35: sparse:     got struct task_struct [noderef] __rcu *real_parent
   kernel/exit.c:362:24: sparse: sparse: incorrect type in assignment (different address spaces) @@     expected struct task_struct *parent @@     got struct task_struct [noderef] __rcu *real_parent @@
   kernel/exit.c:362:24: sparse:     expected struct task_struct *parent
   kernel/exit.c:362:24: sparse:     got struct task_struct [noderef] __rcu *real_parent
   kernel/exit.c:389:27: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct spinlock [usertype] *lock @@     got struct spinlock [noderef] __rcu * @@
   kernel/exit.c:389:27: sparse:     expected struct spinlock [usertype] *lock
   kernel/exit.c:389:27: sparse:     got struct spinlock [noderef] __rcu *
   kernel/exit.c:392:29: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct spinlock [usertype] *lock @@     got struct spinlock [noderef] __rcu * @@
   kernel/exit.c:392:29: sparse:     expected struct spinlock [usertype] *lock
   kernel/exit.c:392:29: sparse:     got struct spinlock [noderef] __rcu *
   kernel/exit.c:616:29: sparse: sparse: incorrect type in assignment (different address spaces) @@     expected struct task_struct *reaper @@     got struct task_struct [noderef] __rcu *real_parent @@
   kernel/exit.c:616:29: sparse:     expected struct task_struct *reaper
   kernel/exit.c:616:29: sparse:     got struct task_struct [noderef] __rcu *real_parent
   kernel/exit.c:618:29: sparse: sparse: incorrect type in assignment (different address spaces) @@     expected struct task_struct *reaper @@     got struct task_struct [noderef] __rcu *real_parent @@
   kernel/exit.c:618:29: sparse:     expected struct task_struct *reaper
   kernel/exit.c:618:29: sparse:     got struct task_struct [noderef] __rcu *real_parent
   kernel/exit.c:771:45: sparse: sparse: incorrect type in initializer (different address spaces) @@     expected struct sighand_struct *sighand @@     got struct sighand_struct [noderef] __rcu *sighand @@
   kernel/exit.c:771:45: sparse:     expected struct sighand_struct *sighand
   kernel/exit.c:771:45: sparse:     got struct sighand_struct [noderef] __rcu *sighand
   kernel/exit.c:976:63: sparse: sparse: incorrect type in initializer (different address spaces) @@     expected struct sighand_struct *const sighand @@     got struct sighand_struct [noderef] __rcu *sighand @@
   kernel/exit.c:976:63: sparse:     expected struct sighand_struct *const sighand
   kernel/exit.c:976:63: sparse:     got struct sighand_struct [noderef] __rcu *sighand
   kernel/exit.c:1131:39: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct spinlock [usertype] *lock @@     got struct spinlock [noderef] __rcu * @@
   kernel/exit.c:1131:39: sparse:     expected struct spinlock [usertype] *lock
   kernel/exit.c:1131:39: sparse:     got struct spinlock [noderef] __rcu *
   kernel/exit.c:1156:41: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct spinlock [usertype] *lock @@     got struct spinlock [noderef] __rcu * @@
   kernel/exit.c:1156:41: sparse:     expected struct spinlock [usertype] *lock
   kernel/exit.c:1156:41: sparse:     got struct spinlock [noderef] __rcu *
   kernel/exit.c:1245:25: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct spinlock [usertype] *lock @@     got struct spinlock [noderef] __rcu * @@
   kernel/exit.c:1245:25: sparse:     expected struct spinlock [usertype] *lock
   kernel/exit.c:1245:25: sparse:     got struct spinlock [noderef] __rcu *
   kernel/exit.c:1260:27: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct spinlock [usertype] *lock @@     got struct spinlock [noderef] __rcu * @@
   kernel/exit.c:1260:27: sparse:     expected struct spinlock [usertype] *lock
   kernel/exit.c:1260:27: sparse:     got struct spinlock [noderef] __rcu *
   kernel/exit.c:1311:25: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct spinlock [usertype] *lock @@     got struct spinlock [noderef] __rcu * @@
   kernel/exit.c:1311:25: sparse:     expected struct spinlock [usertype] *lock
   kernel/exit.c:1311:25: sparse:     got struct spinlock [noderef] __rcu *
   kernel/exit.c:1314:35: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct spinlock [usertype] *lock @@     got struct spinlock [noderef] __rcu * @@
   kernel/exit.c:1314:35: sparse:     expected struct spinlock [usertype] *lock
   kernel/exit.c:1314:35: sparse:     got struct spinlock [noderef] __rcu *
   kernel/exit.c:1320:27: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct spinlock [usertype] *lock @@     got struct spinlock [noderef] __rcu * @@
   kernel/exit.c:1320:27: sparse:     expected struct spinlock [usertype] *lock
   kernel/exit.c:1320:27: sparse:     got struct spinlock [noderef] __rcu *
   kernel/exit.c:1501:59: sparse: sparse: incompatible types in comparison expression (different base types):
   kernel/exit.c:1501:59: sparse:    void *
   kernel/exit.c:1501:59: sparse:    struct task_struct [noderef] __rcu *
   kernel/exit.c:1517:25: sparse: sparse: incorrect type in initializer (different address spaces) @@     expected struct task_struct *parent @@     got struct task_struct [noderef] __rcu * @@
   kernel/exit.c:1517:25: sparse:     expected struct task_struct *parent
   kernel/exit.c:1517:25: sparse:     got struct task_struct [noderef] __rcu *
   kernel/exit.c: note: in included file:
   include/linux/ptrace.h:92:40: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct task_struct *p1 @@     got struct task_struct [noderef] __rcu *real_parent @@
   include/linux/ptrace.h:92:40: sparse:     expected struct task_struct *p1
   include/linux/ptrace.h:92:40: sparse:     got struct task_struct [noderef] __rcu *real_parent
   include/linux/ptrace.h:92:60: sparse: sparse: incorrect type in argument 2 (different address spaces) @@     expected struct task_struct *p2 @@     got struct task_struct [noderef] __rcu *parent @@
   include/linux/ptrace.h:92:60: sparse:     expected struct task_struct *p2
   include/linux/ptrace.h:92:60: sparse:     got struct task_struct [noderef] __rcu *parent
   include/linux/ptrace.h:92:40: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct task_struct *p1 @@     got struct task_struct [noderef] __rcu *real_parent @@
   include/linux/ptrace.h:92:40: sparse:     expected struct task_struct *p1
   include/linux/ptrace.h:92:40: sparse:     got struct task_struct [noderef] __rcu *real_parent
   include/linux/ptrace.h:92:60: sparse: sparse: incorrect type in argument 2 (different address spaces) @@     expected struct task_struct *p2 @@     got struct task_struct [noderef] __rcu *parent @@
   include/linux/ptrace.h:92:60: sparse:     expected struct task_struct *p2
   include/linux/ptrace.h:92:60: sparse:     got struct task_struct [noderef] __rcu *parent
   kernel/exit.c: note: in included file (through include/linux/sched/signal.h, include/linux/rcuwait.h, include/linux/percpu-rwsem.h, ...):
   include/linux/sched/task.h:110:21: sparse: sparse: context imbalance in 'wait_task_zombie' - unexpected unlock
   include/linux/sched/task.h:110:21: sparse: sparse: context imbalance in 'wait_task_stopped' - unexpected unlock
   include/linux/sched/task.h:110:21: sparse: sparse: context imbalance in 'wait_task_continued' - unexpected unlock
   kernel/exit.c: note: in included file:
   include/linux/ptrace.h:92:40: sparse: sparse: incorrect type in argument 1 (different address spaces) @@     expected struct task_struct *p1 @@     got struct task_struct [noderef] __rcu *real_parent @@
   include/linux/ptrace.h:92:40: sparse:     expected struct task_struct *p1
   include/linux/ptrace.h:92:40: sparse:     got struct task_struct [noderef] __rcu *real_parent
   include/linux/ptrace.h:92:60: sparse: sparse: incorrect type in argument 2 (different address spaces) @@     expected struct task_struct *p2 @@     got struct task_struct [noderef] __rcu *parent @@
   include/linux/ptrace.h:92:60: sparse:     expected struct task_struct *p2
   include/linux/ptrace.h:92:60: sparse:     got struct task_struct [noderef] __rcu *parent
   kernel/exit.c: note: in included file (through include/linux/thread_info.h, include/asm-generic/preempt.h, arch/riscv/include/generated/asm/preempt.h, ...):
   arch/riscv/include/asm/current.h:31:9: sparse: sparse: context imbalance in 'do_wait' - wrong count at exit

vim +/CONFIG_SYSCTL +82 kernel/exit.c

    81	
  > 82	#if CONFIG_SYSCTL
    83	static struct ctl_table kern_exit_table[] = {
    84		{
    85			.procname       = "oops_limit",
    86			.data           = &oops_limit,
    87			.maxlen         = sizeof(oops_limit),
    88			.mode           = 0644,
    89			.proc_handler   = proc_douintvec,
    90		},
    91		{ }
    92	};
    93
diff mbox series

Patch

diff --git a/Documentation/admin-guide/sysctl/kernel.rst b/Documentation/admin-guide/sysctl/kernel.rst
index 98d1b198b2b4..09f3fb2f8585 100644
--- a/Documentation/admin-guide/sysctl/kernel.rst
+++ b/Documentation/admin-guide/sysctl/kernel.rst
@@ -667,6 +667,14 @@  This is the default behavior.
 an oops event is detected.
 
 
+oops_limit
+==========
+
+Number of kernel oopses after which the kernel should panic when
+``panic_on_oops`` is not set. Setting this to 0 or 1 has the same effect
+as setting ``panic_on_oops=1``.
+
+
 osrelease, ostype & version
 ===========================
 
diff --git a/kernel/exit.c b/kernel/exit.c
index 35e0a31a0315..892f38aeb0a4 100644
--- a/kernel/exit.c
+++ b/kernel/exit.c
@@ -72,6 +72,33 @@ 
 #include <asm/unistd.h>
 #include <asm/mmu_context.h>
 
+/*
+ * The default value should be high enough to not crash a system that randomly
+ * crashes its kernel from time to time, but low enough to at least not permit
+ * overflowing 32-bit refcounts or the ldsem writer count.
+ */
+static unsigned int oops_limit = 10000;
+
+#if CONFIG_SYSCTL
+static struct ctl_table kern_exit_table[] = {
+	{
+		.procname       = "oops_limit",
+		.data           = &oops_limit,
+		.maxlen         = sizeof(oops_limit),
+		.mode           = 0644,
+		.proc_handler   = proc_douintvec,
+	},
+	{ }
+};
+
+static __init int kernel_exit_sysctls_init(void)
+{
+	register_sysctl_init("kernel", kern_exit_table);
+	return 0;
+}
+late_initcall(kernel_exit_sysctls_init);
+#endif
+
 static void __unhash_process(struct task_struct *p, bool group_dead)
 {
 	nr_threads--;
@@ -874,6 +901,8 @@  void __noreturn do_exit(long code)
 
 void __noreturn make_task_dead(int signr)
 {
+	static atomic_t oops_count = ATOMIC_INIT(0);
+
 	/*
 	 * Take the task off the cpu after something catastrophic has
 	 * happened.
@@ -897,6 +926,19 @@  void __noreturn make_task_dead(int signr)
 		preempt_count_set(PREEMPT_ENABLED);
 	}
 
+	/*
+	 * Every time the system oopses, if the oops happens while a reference
+	 * to an object was held, the reference leaks.
+	 * If the oops doesn't also leak memory, repeated oopsing can cause
+	 * reference counters to wrap around (if they're not using refcount_t).
+	 * This means that repeated oopsing can make unexploitable-looking bugs
+	 * exploitable through repeated oopsing.
+	 * To make sure this can't happen, place an upper bound on how often the
+	 * kernel may oops without panic().
+	 */
+	if (atomic_inc_return(&oops_count) >= READ_ONCE(oops_limit))
+		panic("Oopsed too often (oops_limit is %d)", oops_limit);
+
 	/*
 	 * We're taking recursive faults here in make_task_dead. Safest is to just
 	 * leave this task alone and wait for reboot.