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, 213 insertions, 0 deletions

diff --git a/drivers/input/keyboard/lm8333.c b/drivers/input/keyboard/lm8333.c
new file mode 100644
index 000000000..34be5b2e4
--- /dev/null
+++ b/drivers/input/keyboard/lm8333.c
@@ -0,0 +1,213 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * LM8333 keypad driver
+ * Copyright (C) 2012 Wolfram Sang, Pengutronix <kernel@pengutronix.de>
+ */
+
+#include <linux/i2c.h>
+#include <linux/input.h>
+#include <linux/input/matrix_keypad.h>
+#include <linux/input/lm8333.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+
+#define LM8333_FIFO_READ		0x20
+#define LM8333_DEBOUNCE			0x22
+#define LM8333_READ_INT			0xD0
+#define LM8333_ACTIVE			0xE4
+#define LM8333_READ_ERROR		0xF0
+
+#define LM8333_KEYPAD_IRQ		(1 << 0)
+#define LM8333_ERROR_IRQ		(1 << 3)
+
+#define LM8333_ERROR_KEYOVR		0x04
+#define LM8333_ERROR_FIFOOVR		0x40
+
+#define LM8333_FIFO_TRANSFER_SIZE	16
+
+#define LM8333_NUM_ROWS		8
+#define LM8333_NUM_COLS		16
+#define LM8333_ROW_SHIFT	4
+
+struct lm8333 {
+	struct i2c_client *client;
+	struct input_dev *input;
+	unsigned short keycodes[LM8333_NUM_ROWS << LM8333_ROW_SHIFT];
+};
+
+/* The accessors try twice because the first access may be needed for wakeup */
+#define LM8333_READ_RETRIES 2
+
+int lm8333_read8(struct lm8333 *lm8333, u8 cmd)
+{
+	int retries = 0, ret;
+
+	do {
+		ret = i2c_smbus_read_byte_data(lm8333->client, cmd);
+	} while (ret < 0 && retries++ < LM8333_READ_RETRIES);
+
+	return ret;
+}
+
+int lm8333_write8(struct lm8333 *lm8333, u8 cmd, u8 val)
+{
+	int retries = 0, ret;
+
+	do {
+		ret = i2c_smbus_write_byte_data(lm8333->client, cmd, val);
+	} while (ret < 0 && retries++ < LM8333_READ_RETRIES);
+
+	return ret;
+}
+
+int lm8333_read_block(struct lm8333 *lm8333, u8 cmd, u8 len, u8 *buf)
+{
+	int retries = 0, ret;
+
+	do {
+		ret = i2c_smbus_read_i2c_block_data(lm8333->client,
+						    cmd, len, buf);
+	} while (ret < 0 && retries++ < LM8333_READ_RETRIES);
+
+	return ret;
+}
+
+static void lm8333_key_handler(struct lm8333 *lm8333)
+{
+	struct input_dev *input = lm8333->input;
+	u8 keys[LM8333_FIFO_TRANSFER_SIZE];
+	u8 code, pressed;
+	int i, ret;
+
+	ret = lm8333_read_block(lm8333, LM8333_FIFO_READ,
+				LM8333_FIFO_TRANSFER_SIZE, keys);
+	if (ret != LM8333_FIFO_TRANSFER_SIZE) {
+		dev_err(&lm8333->client->dev,
+			"Error %d while reading FIFO\n", ret);
+		return;
+	}
+
+	for (i = 0; i < LM8333_FIFO_TRANSFER_SIZE && keys[i]; i++) {
+		pressed = keys[i] & 0x80;
+		code = keys[i] & 0x7f;
+
+		input_event(input, EV_MSC, MSC_SCAN, code);
+		input_report_key(input, lm8333->keycodes[code], pressed);
+	}
+
+	input_sync(input);
+}
+
+static irqreturn_t lm8333_irq_thread(int irq, void *data)
+{
+	struct lm8333 *lm8333 = data;
+	u8 status = lm8333_read8(lm8333, LM8333_READ_INT);
+
+	if (!status)
+		return IRQ_NONE;
+
+	if (status & LM8333_ERROR_IRQ) {
+		u8 err = lm8333_read8(lm8333, LM8333_READ_ERROR);
+
+		if (err & (LM8333_ERROR_KEYOVR | LM8333_ERROR_FIFOOVR)) {
+			u8 dummy[LM8333_FIFO_TRANSFER_SIZE];
+
+			lm8333_read_block(lm8333, LM8333_FIFO_READ,
+					LM8333_FIFO_TRANSFER_SIZE, dummy);
+		}
+		dev_err(&lm8333->client->dev, "Got error %02x\n", err);
+	}
+
+	if (status & LM8333_KEYPAD_IRQ)
+		lm8333_key_handler(lm8333);
+
+	return IRQ_HANDLED;
+}
+
+static int lm8333_probe(struct i2c_client *client)
+{
+	const struct lm8333_platform_data *pdata =
+			dev_get_platdata(&client->dev);
+	struct lm8333 *lm8333;
+	struct input_dev *input;
+	int err, active_time;
+
+	if (!pdata)
+		return -EINVAL;
+
+	active_time = pdata->active_time ?: 500;
+	if (active_time / 3 <= pdata->debounce_time / 3) {
+		dev_err(&client->dev, "Active time not big enough!\n");
+		return -EINVAL;
+	}
+
+	lm8333 = devm_kzalloc(&client->dev, sizeof(*lm8333), GFP_KERNEL);
+	if (!lm8333)
+		return -ENOMEM;
+
+	input = devm_input_allocate_device(&client->dev);
+	if (!input)
+		return -ENOMEM;
+
+	lm8333->client = client;
+	lm8333->input = input;
+
+	input->name = client->name;
+	input->id.bustype = BUS_I2C;
+
+	input_set_capability(input, EV_MSC, MSC_SCAN);
+
+	err = matrix_keypad_build_keymap(pdata->matrix_data, NULL,
+					 LM8333_NUM_ROWS, LM8333_NUM_COLS,
+					 lm8333->keycodes, input);
+	if (err)
+		return err;
+
+	if (pdata->debounce_time) {
+		err = lm8333_write8(lm8333, LM8333_DEBOUNCE,
+				    pdata->debounce_time / 3);
+		if (err)
+			dev_warn(&client->dev, "Unable to set debounce time\n");
+	}
+
+	if (pdata->active_time) {
+		err = lm8333_write8(lm8333, LM8333_ACTIVE,
+				    pdata->active_time / 3);
+		if (err)
+			dev_warn(&client->dev, "Unable to set active time\n");
+	}
+
+	err = devm_request_threaded_irq(&client->dev, client->irq,
+					NULL, lm8333_irq_thread,
+					IRQF_TRIGGER_FALLING | IRQF_ONESHOT,
+					"lm8333", lm8333);
+	if (err)
+		return err;
+
+	err = input_register_device(input);
+	if (err)
+		return err;
+
+	i2c_set_clientdata(client, lm8333);
+	return 0;
+}
+
+static const struct i2c_device_id lm8333_id[] = {
+	{ .name = "lm8333" },
+	{ }
+};
+MODULE_DEVICE_TABLE(i2c, lm8333_id);
+
+static struct i2c_driver lm8333_driver = {
+	.driver = {
+		.name		= "lm8333",
+	},
+	.probe		= lm8333_probe,
+	.id_table	= lm8333_id,
+};
+module_i2c_driver(lm8333_driver);
+
+MODULE_AUTHOR("Wolfram Sang <kernel@pengutronix.de>");
+MODULE_DESCRIPTION("LM8333 keyboard driver");
+MODULE_LICENSE("GPL v2");