x86/kprobes: Fix crash when probing CS CALL instructionsgrafted

When using eBPF to probe CS CALL instructions within a function,
a crash can be triggered.

The eBPF tool probes offset 257 of the __hrtimer_run_queues()
function:

<__hrtimer_run_queues+249>:  nopl   0x0(%rax,%rax,1)
<__hrtimer_run_queues+254>:  mov    %r14,%rdi
<__hrtimer_run_queues+257>:  cs call <__x86_indirect_thunk_r12>
<__hrtimer_run_queues+263>:  mov    %eax,%r12d
<__hrtimer_run_queues+266>:  xchg   %ax,%ax
<__hrtimer_run_queues+268>:  mov    %r13,%rdi

Which triggers this crash:

  BUG: unable to handle page fault for address: 00000000000f41c9
  #PF: supervisor write access in kernel mode
  #PF: error_code(0x0002) - not-present page
  PGD 0 P4D 0
  Oops: 0002 [#1] SMP NOPTI
  CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P
  RIP: 0010:__hrtimer_run_queues+0x106/0x230

Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is
at the 6th byte of the above CS CALL instruction. Since the CS CALL
instruction occupies 6 bytes, the exception occurred in the middle of that
call instruction.

The root cause is that when using eBPF tools to probe in the middle of a
function, a kprobe with INT3 is used as the underlying implementation.

During single-step emulation of the original CALL instruction,
int3_emulate_call() assumes that the probed CALL instruction is 5 bytes
long. However, the actual CS-prefixed CALL instruction occupies 6 bytes,
so it constructs an incorrect exception return address. When the CPU
returns from the kprobe handler, the next instruction to be executed is at
the address of the last byte of that CS CALL instruction. Coincidentally,
starting from that address, the CPU fetches and decodes a completely
different instruction, which ultimately triggers a kernel crash.

Fix the issue by using the actual instruction length obtained from
the instruction decoder when constructing the exception return
address, rather than relying on the hardcoded CALL_INSN_SIZE macro.

[ mingo: Refined the changelog ]

Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step")
Suggested-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Jinke Han <jinkehan@didiglobal.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Acked-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000

1 files changed, 221 insertions, 0 deletions

diff --git a/drivers/input/input-poller.c b/drivers/input/input-poller.c
new file mode 100644
index 000000000..54dc07fca
--- /dev/null
+++ b/drivers/input/input-poller.c
@@ -0,0 +1,221 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Support for polling mode for input devices.
+ */
+
+#include <linux/device.h>
+#include <linux/export.h>
+#include <linux/input.h>
+#include <linux/jiffies.h>
+#include <linux/mutex.h>
+#include <linux/slab.h>
+#include <linux/types.h>
+#include <linux/workqueue.h>
+#include "input-poller.h"
+
+struct input_dev_poller {
+	void (*poll)(struct input_dev *dev);
+
+	unsigned int poll_interval; /* msec */
+	unsigned int poll_interval_max; /* msec */
+	unsigned int poll_interval_min; /* msec */
+
+	struct input_dev *input;
+	struct delayed_work work;
+};
+
+static void input_dev_poller_queue_work(struct input_dev_poller *poller)
+{
+	unsigned long delay;
+
+	delay = msecs_to_jiffies(poller->poll_interval);
+	if (delay >= HZ)
+		delay = round_jiffies_relative(delay);
+
+	queue_delayed_work(system_freezable_wq, &poller->work, delay);
+}
+
+static void input_dev_poller_work(struct work_struct *work)
+{
+	struct input_dev_poller *poller =
+		container_of(work, struct input_dev_poller, work.work);
+
+	poller->poll(poller->input);
+	input_dev_poller_queue_work(poller);
+}
+
+void input_dev_poller_finalize(struct input_dev_poller *poller)
+{
+	if (!poller->poll_interval)
+		poller->poll_interval = 500;
+	if (!poller->poll_interval_max)
+		poller->poll_interval_max = poller->poll_interval;
+}
+
+void input_dev_poller_start(struct input_dev_poller *poller)
+{
+	/* Only start polling if polling is enabled */
+	if (poller->poll_interval > 0) {
+		poller->poll(poller->input);
+		input_dev_poller_queue_work(poller);
+	}
+}
+
+void input_dev_poller_stop(struct input_dev_poller *poller)
+{
+	cancel_delayed_work_sync(&poller->work);
+}
+
+int input_setup_polling(struct input_dev *dev,
+			void (*poll_fn)(struct input_dev *dev))
+{
+	struct input_dev_poller *poller;
+
+	poller = kzalloc_obj(*poller);
+	if (!poller) {
+		/*
+		 * We want to show message even though kzalloc() may have
+		 * printed backtrace as knowing what instance of input
+		 * device we were dealing with is helpful.
+		 */
+		dev_err(dev->dev.parent ?: &dev->dev,
+			"%s: unable to allocate poller structure\n", __func__);
+		return -ENOMEM;
+	}
+
+	INIT_DELAYED_WORK(&poller->work, input_dev_poller_work);
+	poller->input = dev;
+	poller->poll = poll_fn;
+
+	dev->poller = poller;
+	return 0;
+}
+EXPORT_SYMBOL(input_setup_polling);
+
+static bool input_dev_ensure_poller(struct input_dev *dev)
+{
+	if (!dev->poller) {
+		dev_err(dev->dev.parent ?: &dev->dev,
+			"poller structure has not been set up\n");
+		return false;
+	}
+
+	return true;
+}
+
+void input_set_poll_interval(struct input_dev *dev, unsigned int interval)
+{
+	if (input_dev_ensure_poller(dev))
+		dev->poller->poll_interval = interval;
+}
+EXPORT_SYMBOL(input_set_poll_interval);
+
+void input_set_min_poll_interval(struct input_dev *dev, unsigned int interval)
+{
+	if (input_dev_ensure_poller(dev))
+		dev->poller->poll_interval_min = interval;
+}
+EXPORT_SYMBOL(input_set_min_poll_interval);
+
+void input_set_max_poll_interval(struct input_dev *dev, unsigned int interval)
+{
+	if (input_dev_ensure_poller(dev))
+		dev->poller->poll_interval_max = interval;
+}
+EXPORT_SYMBOL(input_set_max_poll_interval);
+
+int input_get_poll_interval(struct input_dev *dev)
+{
+	if (!dev->poller)
+		return -EINVAL;
+
+	return dev->poller->poll_interval;
+}
+EXPORT_SYMBOL(input_get_poll_interval);
+
+/* SYSFS interface */
+
+static ssize_t input_dev_get_poll_interval(struct device *dev,
+					   struct device_attribute *attr,
+					   char *buf)
+{
+	struct input_dev *input = to_input_dev(dev);
+
+	return sprintf(buf, "%d\n", input->poller->poll_interval);
+}
+
+static ssize_t input_dev_set_poll_interval(struct device *dev,
+					   struct device_attribute *attr,
+					   const char *buf, size_t count)
+{
+	struct input_dev *input = to_input_dev(dev);
+	struct input_dev_poller *poller = input->poller;
+	unsigned int interval;
+	int err;
+
+	err = kstrtouint(buf, 0, &interval);
+	if (err)
+		return err;
+
+	if (interval < poller->poll_interval_min)
+		return -EINVAL;
+
+	if (interval > poller->poll_interval_max)
+		return -EINVAL;
+
+	guard(mutex)(&input->mutex);
+
+	poller->poll_interval = interval;
+
+	if (input_device_enabled(input)) {
+		cancel_delayed_work_sync(&poller->work);
+		if (poller->poll_interval > 0)
+			input_dev_poller_queue_work(poller);
+	}
+
+	return count;
+}
+
+static DEVICE_ATTR(poll, 0644,
+		   input_dev_get_poll_interval, input_dev_set_poll_interval);
+
+static ssize_t input_dev_get_poll_max(struct device *dev,
+				      struct device_attribute *attr, char *buf)
+{
+	struct input_dev *input = to_input_dev(dev);
+
+	return sprintf(buf, "%d\n", input->poller->poll_interval_max);
+}
+
+static DEVICE_ATTR(max, 0444, input_dev_get_poll_max, NULL);
+
+static ssize_t input_dev_get_poll_min(struct device *dev,
+				     struct device_attribute *attr, char *buf)
+{
+	struct input_dev *input = to_input_dev(dev);
+
+	return sprintf(buf, "%d\n", input->poller->poll_interval_min);
+}
+
+static DEVICE_ATTR(min, 0444, input_dev_get_poll_min, NULL);
+
+static umode_t input_poller_attrs_visible(struct kobject *kobj,
+					  struct attribute *attr, int n)
+{
+	struct device *dev = kobj_to_dev(kobj);
+	struct input_dev *input = to_input_dev(dev);
+
+	return input->poller ? attr->mode : 0;
+}
+
+static struct attribute *input_poller_attrs[] = {
+	&dev_attr_poll.attr,
+	&dev_attr_max.attr,
+	&dev_attr_min.attr,
+	NULL
+};
+
+struct attribute_group input_poller_attribute_group = {
+	.is_visible	= input_poller_attrs_visible,
+	.attrs		= input_poller_attrs,
+};