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, 293 insertions, 0 deletions

diff --git a/drivers/hwmon/max1111.c b/drivers/hwmon/max1111.c
new file mode 100644
index 000000000..5cc08c720
--- /dev/null
+++ b/drivers/hwmon/max1111.c
@@ -0,0 +1,293 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * max1111.c - +2.7V, Low-Power, Multichannel, Serial 8-bit ADCs
+ *
+ * Based on arch/arm/mach-pxa/corgi_ssp.c
+ *
+ * Copyright (C) 2004-2005 Richard Purdie
+ *
+ * Copyright (C) 2008 Marvell International Ltd.
+ *	Eric Miao <eric.miao@marvell.com>
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/err.h>
+#include <linux/hwmon.h>
+#include <linux/hwmon-sysfs.h>
+#include <linux/spi/spi.h>
+#include <linux/slab.h>
+
+enum chips { max1110, max1111, max1112, max1113 };
+
+#define MAX1111_TX_BUF_SIZE	1
+#define MAX1111_RX_BUF_SIZE	2
+
+/* MAX1111 Commands */
+#define MAX1111_CTRL_PD0      (1u << 0)
+#define MAX1111_CTRL_PD1      (1u << 1)
+#define MAX1111_CTRL_SGL      (1u << 2)
+#define MAX1111_CTRL_UNI      (1u << 3)
+#define MAX1110_CTRL_SEL_SH   (4)
+#define MAX1111_CTRL_SEL_SH   (5)	/* NOTE: bit 4 is ignored */
+#define MAX1111_CTRL_STR      (1u << 7)
+
+struct max1111_data {
+	struct spi_device	*spi;
+	struct device		*hwmon_dev;
+	struct spi_message	msg;
+	struct spi_transfer	xfer[2];
+	uint8_t tx_buf[MAX1111_TX_BUF_SIZE];
+	uint8_t rx_buf[MAX1111_RX_BUF_SIZE];
+	struct mutex		drvdata_lock;
+	/* protect msg, xfer and buffers from multiple access */
+	int			sel_sh;
+	int			lsb;
+};
+
+static int max1111_read(struct device *dev, int channel)
+{
+	struct max1111_data *data = dev_get_drvdata(dev);
+	uint8_t v1, v2;
+	int err;
+
+	/* writing to drvdata struct is not thread safe, wait on mutex */
+	mutex_lock(&data->drvdata_lock);
+
+	data->tx_buf[0] = (channel << data->sel_sh) |
+		MAX1111_CTRL_PD0 | MAX1111_CTRL_PD1 |
+		MAX1111_CTRL_SGL | MAX1111_CTRL_UNI | MAX1111_CTRL_STR;
+
+	err = spi_sync(data->spi, &data->msg);
+	if (err < 0) {
+		dev_err(dev, "spi_sync failed with %d\n", err);
+		mutex_unlock(&data->drvdata_lock);
+		return err;
+	}
+
+	v1 = data->rx_buf[0];
+	v2 = data->rx_buf[1];
+
+	mutex_unlock(&data->drvdata_lock);
+
+	if ((v1 & 0xc0) || (v2 & 0x3f))
+		return -EINVAL;
+
+	return (v1 << 2) | (v2 >> 6);
+}
+
+#ifdef CONFIG_SHARPSL_PM
+static struct max1111_data *the_max1111;
+
+int max1111_read_channel(int channel);
+int max1111_read_channel(int channel)
+{
+	if (!the_max1111 || !the_max1111->spi)
+		return -ENODEV;
+
+	return max1111_read(&the_max1111->spi->dev, channel);
+}
+EXPORT_SYMBOL(max1111_read_channel);
+#endif
+
+/*
+ * NOTE: SPI devices do not have a default 'name' attribute, which is
+ * likely to be used by hwmon applications to distinguish between
+ * different devices, explicitly add a name attribute here.
+ */
+static ssize_t name_show(struct device *dev,
+			 struct device_attribute *attr, char *buf)
+{
+	return sprintf(buf, "%s\n", to_spi_device(dev)->modalias);
+}
+
+static ssize_t show_adc(struct device *dev,
+			struct device_attribute *attr, char *buf)
+{
+	struct max1111_data *data = dev_get_drvdata(dev);
+	int channel = to_sensor_dev_attr(attr)->index;
+	int ret;
+
+	ret = max1111_read(dev, channel);
+	if (ret < 0)
+		return ret;
+
+	/*
+	 * Assume the reference voltage to be 2.048V or 4.096V, with an 8-bit
+	 * sample. The LSB weight is 8mV or 16mV depending on the chip type.
+	 */
+	return sprintf(buf, "%d\n", ret * data->lsb);
+}
+
+#define MAX1111_ADC_ATTR(_id)		\
+	SENSOR_DEVICE_ATTR(in##_id##_input, S_IRUGO, show_adc, NULL, _id)
+
+static DEVICE_ATTR_RO(name);
+static MAX1111_ADC_ATTR(0);
+static MAX1111_ADC_ATTR(1);
+static MAX1111_ADC_ATTR(2);
+static MAX1111_ADC_ATTR(3);
+static MAX1111_ADC_ATTR(4);
+static MAX1111_ADC_ATTR(5);
+static MAX1111_ADC_ATTR(6);
+static MAX1111_ADC_ATTR(7);
+
+static struct attribute *max1111_attributes[] = {
+	&dev_attr_name.attr,
+	&sensor_dev_attr_in0_input.dev_attr.attr,
+	&sensor_dev_attr_in1_input.dev_attr.attr,
+	&sensor_dev_attr_in2_input.dev_attr.attr,
+	&sensor_dev_attr_in3_input.dev_attr.attr,
+	NULL,
+};
+
+static const struct attribute_group max1111_attr_group = {
+	.attrs	= max1111_attributes,
+};
+
+static struct attribute *max1110_attributes[] = {
+	&sensor_dev_attr_in4_input.dev_attr.attr,
+	&sensor_dev_attr_in5_input.dev_attr.attr,
+	&sensor_dev_attr_in6_input.dev_attr.attr,
+	&sensor_dev_attr_in7_input.dev_attr.attr,
+	NULL,
+};
+
+static const struct attribute_group max1110_attr_group = {
+	.attrs	= max1110_attributes,
+};
+
+static int setup_transfer(struct max1111_data *data)
+{
+	struct spi_message *m;
+	struct spi_transfer *x;
+
+	m = &data->msg;
+	x = &data->xfer[0];
+
+	spi_message_init(m);
+
+	x->tx_buf = &data->tx_buf[0];
+	x->len = MAX1111_TX_BUF_SIZE;
+	spi_message_add_tail(x, m);
+
+	x++;
+	x->rx_buf = &data->rx_buf[0];
+	x->len = MAX1111_RX_BUF_SIZE;
+	spi_message_add_tail(x, m);
+
+	return 0;
+}
+
+static int max1111_probe(struct spi_device *spi)
+{
+	enum chips chip = spi_get_device_id(spi)->driver_data;
+	struct max1111_data *data;
+	int err;
+
+	spi->bits_per_word = 8;
+	spi->mode = SPI_MODE_0;
+	err = spi_setup(spi);
+	if (err < 0)
+		return err;
+
+	data = devm_kzalloc(&spi->dev, sizeof(struct max1111_data), GFP_KERNEL);
+	if (data == NULL)
+		return -ENOMEM;
+
+	switch (chip) {
+	case max1110:
+		data->lsb = 8;
+		data->sel_sh = MAX1110_CTRL_SEL_SH;
+		break;
+	case max1111:
+		data->lsb = 8;
+		data->sel_sh = MAX1111_CTRL_SEL_SH;
+		break;
+	case max1112:
+		data->lsb = 16;
+		data->sel_sh = MAX1110_CTRL_SEL_SH;
+		break;
+	case max1113:
+		data->lsb = 16;
+		data->sel_sh = MAX1111_CTRL_SEL_SH;
+		break;
+	}
+	err = setup_transfer(data);
+	if (err)
+		return err;
+
+	mutex_init(&data->drvdata_lock);
+
+	data->spi = spi;
+	spi_set_drvdata(spi, data);
+
+	err = sysfs_create_group(&spi->dev.kobj, &max1111_attr_group);
+	if (err) {
+		dev_err(&spi->dev, "failed to create attribute group\n");
+		return err;
+	}
+	if (chip == max1110 || chip == max1112) {
+		err = sysfs_create_group(&spi->dev.kobj, &max1110_attr_group);
+		if (err) {
+			dev_err(&spi->dev,
+				"failed to create extended attribute group\n");
+			goto err_remove;
+		}
+	}
+
+	data->hwmon_dev = hwmon_device_register(&spi->dev);
+	if (IS_ERR(data->hwmon_dev)) {
+		dev_err(&spi->dev, "failed to create hwmon device\n");
+		err = PTR_ERR(data->hwmon_dev);
+		goto err_remove;
+	}
+
+#ifdef CONFIG_SHARPSL_PM
+	the_max1111 = data;
+#endif
+	return 0;
+
+err_remove:
+	sysfs_remove_group(&spi->dev.kobj, &max1110_attr_group);
+	sysfs_remove_group(&spi->dev.kobj, &max1111_attr_group);
+	return err;
+}
+
+static void max1111_remove(struct spi_device *spi)
+{
+	struct max1111_data *data = spi_get_drvdata(spi);
+
+#ifdef CONFIG_SHARPSL_PM
+	the_max1111 = NULL;
+#endif
+	hwmon_device_unregister(data->hwmon_dev);
+	sysfs_remove_group(&spi->dev.kobj, &max1110_attr_group);
+	sysfs_remove_group(&spi->dev.kobj, &max1111_attr_group);
+	mutex_destroy(&data->drvdata_lock);
+}
+
+static const struct spi_device_id max1111_ids[] = {
+	{ "max1110", max1110 },
+	{ "max1111", max1111 },
+	{ "max1112", max1112 },
+	{ "max1113", max1113 },
+	{ },
+};
+MODULE_DEVICE_TABLE(spi, max1111_ids);
+
+static struct spi_driver max1111_driver = {
+	.driver		= {
+		.name	= "max1111",
+	},
+	.id_table	= max1111_ids,
+	.probe		= max1111_probe,
+	.remove		= max1111_remove,
+};
+
+module_spi_driver(max1111_driver);
+
+MODULE_AUTHOR("Eric Miao <eric.miao@marvell.com>");
+MODULE_DESCRIPTION("MAX1110/MAX1111/MAX1112/MAX1113 ADC Driver");
+MODULE_LICENSE("GPL");