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, 197 insertions, 0 deletions

diff --git a/drivers/input/misc/pf1550-onkey.c b/drivers/input/misc/pf1550-onkey.c
new file mode 100644
index 000000000..0d1b570bb
--- /dev/null
+++ b/drivers/input/misc/pf1550-onkey.c
@@ -0,0 +1,197 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Driver for the PF1550 ONKEY
+ * Copyright (C) 2016 Freescale Semiconductor, Inc. All Rights Reserved.
+ *
+ * Portions Copyright (c) 2025 Savoir-faire Linux Inc.
+ * Samuel Kayode <samuel.kayode@savoirfairelinux.com>
+ */
+
+#include <linux/err.h>
+#include <linux/input.h>
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/mfd/pf1550.h>
+#include <linux/platform_device.h>
+
+#define PF1550_ONKEY_IRQ_NR	6
+
+struct onkey_drv_data {
+	struct device *dev;
+	const struct pf1550_ddata *pf1550;
+	bool wakeup;
+	struct input_dev *input;
+};
+
+static irqreturn_t pf1550_onkey_irq_handler(int irq, void *data)
+{
+	struct onkey_drv_data *onkey = data;
+	struct platform_device *pdev = to_platform_device(onkey->dev);
+	int i, state, irq_type = -1;
+
+	for (i = 0; i < PF1550_ONKEY_IRQ_NR; i++)
+		if (irq == platform_get_irq(pdev, i))
+			irq_type = i;
+
+	switch (irq_type) {
+	case PF1550_ONKEY_IRQ_PUSHI:
+		state = 0;
+		break;
+	case PF1550_ONKEY_IRQ_1SI:
+	case PF1550_ONKEY_IRQ_2SI:
+	case PF1550_ONKEY_IRQ_3SI:
+	case PF1550_ONKEY_IRQ_4SI:
+	case PF1550_ONKEY_IRQ_8SI:
+		state = 1;
+		break;
+	default:
+		dev_err(onkey->dev, "onkey interrupt: irq %d occurred\n",
+			irq_type);
+		return IRQ_HANDLED;
+	}
+
+	input_event(onkey->input, EV_KEY, KEY_POWER, state);
+	input_sync(onkey->input);
+
+	return IRQ_HANDLED;
+}
+
+static int pf1550_onkey_probe(struct platform_device *pdev)
+{
+	struct onkey_drv_data *onkey;
+	struct input_dev *input;
+	bool key_power = false;
+	int i, irq, error;
+
+	onkey = devm_kzalloc(&pdev->dev, sizeof(*onkey), GFP_KERNEL);
+	if (!onkey)
+		return -ENOMEM;
+
+	onkey->dev = &pdev->dev;
+
+	onkey->pf1550 = dev_get_drvdata(pdev->dev.parent);
+	if (!onkey->pf1550->regmap)
+		return dev_err_probe(&pdev->dev, -ENODEV,
+				     "failed to get regmap\n");
+
+	onkey->wakeup = device_property_read_bool(pdev->dev.parent,
+						  "wakeup-source");
+
+	if (device_property_read_bool(pdev->dev.parent,
+				      "nxp,disable-key-power")) {
+		error = regmap_clear_bits(onkey->pf1550->regmap,
+					  PF1550_PMIC_REG_PWRCTRL1,
+					  PF1550_ONKEY_RST_EN);
+		if (error)
+			return dev_err_probe(&pdev->dev, error,
+					     "failed: disable turn system off");
+	} else {
+		key_power = true;
+	}
+
+	input = devm_input_allocate_device(&pdev->dev);
+	if (!input)
+		return dev_err_probe(&pdev->dev, -ENOMEM,
+				     "failed to allocate the input device\n");
+
+	input->name = pdev->name;
+	input->phys = "pf1550-onkey/input0";
+	input->id.bustype = BUS_HOST;
+
+	if (key_power)
+		input_set_capability(input, EV_KEY, KEY_POWER);
+
+	onkey->input = input;
+	platform_set_drvdata(pdev, onkey);
+
+	for (i = 0; i < PF1550_ONKEY_IRQ_NR; i++) {
+		irq = platform_get_irq(pdev, i);
+		if (irq < 0)
+			return irq;
+
+		error = devm_request_threaded_irq(&pdev->dev, irq, NULL,
+						  pf1550_onkey_irq_handler,
+						  IRQF_NO_SUSPEND,
+						  "pf1550-onkey", onkey);
+		if (error)
+			return dev_err_probe(&pdev->dev, error,
+					     "failed: irq request (IRQ: %d)\n",
+					     i);
+	}
+
+	error = input_register_device(input);
+	if (error)
+		return dev_err_probe(&pdev->dev, error,
+				     "failed to register input device\n");
+
+	device_init_wakeup(&pdev->dev, onkey->wakeup);
+
+	return 0;
+}
+
+static int pf1550_onkey_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct onkey_drv_data *onkey = platform_get_drvdata(pdev);
+	int i, irq;
+
+	if (!device_may_wakeup(&pdev->dev))
+		regmap_write(onkey->pf1550->regmap,
+			     PF1550_PMIC_REG_ONKEY_INT_MASK0,
+			     ONKEY_IRQ_PUSHI | ONKEY_IRQ_1SI | ONKEY_IRQ_2SI |
+			     ONKEY_IRQ_3SI | ONKEY_IRQ_4SI | ONKEY_IRQ_8SI);
+	else
+		for (i = 0; i < PF1550_ONKEY_IRQ_NR; i++) {
+			irq = platform_get_irq(pdev, i);
+			if (irq > 0)
+				enable_irq_wake(irq);
+		}
+
+	return 0;
+}
+
+static int pf1550_onkey_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct onkey_drv_data *onkey = platform_get_drvdata(pdev);
+	int i, irq;
+
+	if (!device_may_wakeup(&pdev->dev))
+		regmap_write(onkey->pf1550->regmap,
+			     PF1550_PMIC_REG_ONKEY_INT_MASK0,
+			     ~((u8)(ONKEY_IRQ_PUSHI | ONKEY_IRQ_1SI |
+			     ONKEY_IRQ_2SI | ONKEY_IRQ_3SI | ONKEY_IRQ_4SI |
+			     ONKEY_IRQ_8SI)));
+	else
+		for (i = 0; i < PF1550_ONKEY_IRQ_NR; i++) {
+			irq = platform_get_irq(pdev, i);
+			if (irq > 0)
+				disable_irq_wake(irq);
+		}
+
+	return 0;
+}
+
+static DEFINE_SIMPLE_DEV_PM_OPS(pf1550_onkey_pm_ops, pf1550_onkey_suspend,
+			 pf1550_onkey_resume);
+
+static const struct platform_device_id pf1550_onkey_id[] = {
+	{ "pf1550-onkey", },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(platform, pf1550_onkey_id);
+
+static struct platform_driver pf1550_onkey_driver = {
+	.driver = {
+		.name = "pf1550-onkey",
+		.pm   = pm_sleep_ptr(&pf1550_onkey_pm_ops),
+	},
+	.probe = pf1550_onkey_probe,
+	.id_table = pf1550_onkey_id,
+};
+module_platform_driver(pf1550_onkey_driver);
+
+MODULE_AUTHOR("Freescale Semiconductor");
+MODULE_DESCRIPTION("PF1550 onkey Driver");
+MODULE_LICENSE("GPL");