rcu_read_(un)lock(), list_*_rcu(), and synchronize_rcu() are used for a secure
access and manipulation of the list of patches that modify the same function.
In particular, it is the variable func_stack that is accessible from the ftrace
handler via struct ftrace_ops and klp_ops.
Of course, it synchronizes also some states of the patch on the top of the
stack, e.g. func->transition in klp_ftrace_handler.
At the same time, this mechanism guards also the manipulation of
task->patch_state. It is modified according to the state of the transition and
the state of the process.
Now, all this works well as long as RCU works well. Sadly livepatching might
get into some corner cases when this is not true. For example, RCU is not
watching when rcu_read_lock() is taken in idle threads. It is because they
might sleep and prevent reaching the grace period for too long.
There are ways how to make RCU watching even in idle threads, see
rcu_irq_enter(). But there is a small location inside RCU infrastructure when
even this does not work.
This small problematic location can be detected either before calling
rcu_irq_enter() by rcu_irq_enter_disabled() or later by rcu_is_watching().
Sadly, there is no safe way how to handle it. Once we detect that RCU was not
watching, we might see inconsistent state of the function stack and the related
variables in klp_ftrace_handler(). Then we could do a wrong decision, use an
incompatible implementation of the function and break the consistency of the
system. We could warn but we could not avoid the damage.
Fortunately, ftrace has similar problems and they seem to be solved well there.
It uses a heavy weight implementation of some RCU operations. In particular, it
replaces:
+ rcu_read_lock() with preempt_disable_notrace()
+ rcu_read_unlock() with preempt_enable_notrace()
+ synchronize_rcu() with schedule_on_each_cpu(sync_work)
My understanding is that this is RCU implementation from a stone age. It meets
the core RCU requirements but it is rather ineffective. Especially, it does not
allow to batch or speed up the synchronize calls.
On the other hand, it is very trivial. It allows to safely trace and/or
livepatch even the RCU core infrastructure. And the effectiveness is a not a
big issue because using ftrace or livepatches on productive systems is a rare
operation. The safety is much more important than a negligible extra load.
Note that the alternative implementation follows the RCU principles. Therefore,
we could and actually must use list_*_rcu() variants when manipulating the
func_stack. These functions allow to access the pointers in the right
order and with the right barriers. But they do not use any other
information that would be set only by rcu_read_lock().
Also note that there are actually two problems solved in ftrace:
First, it cares about the consistency of RCU read sections. It is being solved
the way as described and used in this patch.
Second, ftrace needs to make sure that nobody is inside the dynamic trampoline
when it is being freed. For this, it also calls synchronize_rcu_tasks() in
preemptive kernel in ftrace_shutdown().
Livepatch has similar problem but it is solved by ftrace for free.
klp_ftrace_handler() is a good guy and never sleeps. In addition, it is
registered with FTRACE_OPS_FL_DYNAMIC. It causes that
unregister_ftrace_function() calls:
* schedule_on_each_cpu(ftrace_sync) - always
* synchronize_rcu_tasks() - in preemptive kernel
The effect is that nobody is neither inside the dynamic trampoline nor inside
the ftrace handler after unregister_ftrace_function() returns.
[jkosina@suse.cz: reformat changelog, fix comment]
Signed-off-by: Petr Mladek <pmladek@suse.com>
Acked-by: Josh Poimboeuf <jpoimboe@redhat.com>
Acked-by: Miroslav Benes <mbenes@suse.cz>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Change livepatch to use a basic per-task consistency model. This is the
foundation which will eventually enable us to patch those ~10% of
security patches which change function or data semantics. This is the
biggest remaining piece needed to make livepatch more generally useful.
This code stems from the design proposal made by Vojtech [1] in November
2014. It's a hybrid of kGraft and kpatch: it uses kGraft's per-task
consistency and syscall barrier switching combined with kpatch's stack
trace switching. There are also a number of fallback options which make
it quite flexible.
Patches are applied on a per-task basis, when the task is deemed safe to
switch over. When a patch is enabled, livepatch enters into a
transition state where tasks are converging to the patched state.
Usually this transition state can complete in a few seconds. The same
sequence occurs when a patch is disabled, except the tasks converge from
the patched state to the unpatched state.
An interrupt handler inherits the patched state of the task it
interrupts. The same is true for forked tasks: the child inherits the
patched state of the parent.
Livepatch uses several complementary approaches to determine when it's
safe to patch tasks:
1. The first and most effective approach is stack checking of sleeping
tasks. If no affected functions are on the stack of a given task,
the task is patched. In most cases this will patch most or all of
the tasks on the first try. Otherwise it'll keep trying
periodically. This option is only available if the architecture has
reliable stacks (HAVE_RELIABLE_STACKTRACE).
2. The second approach, if needed, is kernel exit switching. A
task is switched when it returns to user space from a system call, a
user space IRQ, or a signal. It's useful in the following cases:
a) Patching I/O-bound user tasks which are sleeping on an affected
function. In this case you have to send SIGSTOP and SIGCONT to
force it to exit the kernel and be patched.
b) Patching CPU-bound user tasks. If the task is highly CPU-bound
then it will get patched the next time it gets interrupted by an
IRQ.
c) In the future it could be useful for applying patches for
architectures which don't yet have HAVE_RELIABLE_STACKTRACE. In
this case you would have to signal most of the tasks on the
system. However this isn't supported yet because there's
currently no way to patch kthreads without
HAVE_RELIABLE_STACKTRACE.
3. For idle "swapper" tasks, since they don't ever exit the kernel, they
instead have a klp_update_patch_state() call in the idle loop which
allows them to be patched before the CPU enters the idle state.
(Note there's not yet such an approach for kthreads.)
All the above approaches may be skipped by setting the 'immediate' flag
in the 'klp_patch' struct, which will disable per-task consistency and
patch all tasks immediately. This can be useful if the patch doesn't
change any function or data semantics. Note that, even with this flag
set, it's possible that some tasks may still be running with an old
version of the function, until that function returns.
There's also an 'immediate' flag in the 'klp_func' struct which allows
you to specify that certain functions in the patch can be applied
without per-task consistency. This might be useful if you want to patch
a common function like schedule(), and the function change doesn't need
consistency but the rest of the patch does.
For architectures which don't have HAVE_RELIABLE_STACKTRACE, the user
must set patch->immediate which causes all tasks to be patched
immediately. This option should be used with care, only when the patch
doesn't change any function or data semantics.
In the future, architectures which don't have HAVE_RELIABLE_STACKTRACE
may be allowed to use per-task consistency if we can come up with
another way to patch kthreads.
The /sys/kernel/livepatch/<patch>/transition file shows whether a patch
is in transition. Only a single patch (the topmost patch on the stack)
can be in transition at a given time. A patch can remain in transition
indefinitely, if any of the tasks are stuck in the initial patch state.
A transition can be reversed and effectively canceled by writing the
opposite value to the /sys/kernel/livepatch/<patch>/enabled file while
the transition is in progress. Then all the tasks will attempt to
converge back to the original patch state.
[1] https://lkml.kernel.org/r/20141107140458.GA21774@suse.cz
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Acked-by: Miroslav Benes <mbenes@suse.cz>
Acked-by: Ingo Molnar <mingo@kernel.org> # for the scheduler changes
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Move functions related to the actual patching of functions and objects
into a new patch.c file.
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Acked-by: Miroslav Benes <mbenes@suse.cz>
Reviewed-by: Petr Mladek <pmladek@suse.com>
Reviewed-by: Kamalesh Babulal <kamalesh@linux.vnet.ibm.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Petr Mladek
Josh Poimboeuf