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, 177 insertions, 0 deletions

diff --git a/drivers/gpio/gpio-moxtet.c b/drivers/gpio/gpio-moxtet.c
new file mode 100644
index 000000000..4eb9f1a27
--- /dev/null
+++ b/drivers/gpio/gpio-moxtet.c
@@ -0,0 +1,177 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ *  Turris Mox Moxtet GPIO expander
+ *
+ *  Copyright (C) 2018 Marek Behún <kabel@kernel.org>
+ */
+
+#include <linux/bitops.h>
+#include <linux/gpio/driver.h>
+#include <linux/moxtet.h>
+#include <linux/module.h>
+
+#define MOXTET_GPIO_NGPIOS	12
+#define MOXTET_GPIO_INPUTS	4
+
+struct moxtet_gpio_desc {
+	u16 in_mask;
+	u16 out_mask;
+};
+
+static const struct moxtet_gpio_desc descs[] = {
+	[TURRIS_MOX_MODULE_SFP] = {
+		.in_mask = GENMASK(2, 0),
+		.out_mask = GENMASK(5, 4),
+	},
+};
+
+struct moxtet_gpio_chip {
+	struct device			*dev;
+	struct gpio_chip		gpio_chip;
+	const struct moxtet_gpio_desc	*desc;
+};
+
+static int moxtet_gpio_get_value(struct gpio_chip *gc, unsigned int offset)
+{
+	struct moxtet_gpio_chip *chip = gpiochip_get_data(gc);
+	int ret;
+
+	if (chip->desc->in_mask & BIT(offset)) {
+		ret = moxtet_device_read(chip->dev);
+	} else if (chip->desc->out_mask & BIT(offset)) {
+		ret = moxtet_device_written(chip->dev);
+		if (ret >= 0)
+			ret <<= MOXTET_GPIO_INPUTS;
+	} else {
+		return -EINVAL;
+	}
+
+	if (ret < 0)
+		return ret;
+
+	return !!(ret & BIT(offset));
+}
+
+static int moxtet_gpio_set_value(struct gpio_chip *gc, unsigned int offset,
+				 int val)
+{
+	struct moxtet_gpio_chip *chip = gpiochip_get_data(gc);
+	int state;
+
+	state = moxtet_device_written(chip->dev);
+	if (state < 0)
+		return state;
+
+	offset -= MOXTET_GPIO_INPUTS;
+
+	if (val)
+		state |= BIT(offset);
+	else
+		state &= ~BIT(offset);
+
+	return moxtet_device_write(chip->dev, state);
+}
+
+static int moxtet_gpio_get_direction(struct gpio_chip *gc, unsigned int offset)
+{
+	struct moxtet_gpio_chip *chip = gpiochip_get_data(gc);
+
+	/* All lines are hard wired to be either input or output, not both. */
+	if (chip->desc->in_mask & BIT(offset))
+		return GPIO_LINE_DIRECTION_IN;
+	else if (chip->desc->out_mask & BIT(offset))
+		return GPIO_LINE_DIRECTION_OUT;
+	else
+		return -EINVAL;
+}
+
+static int moxtet_gpio_direction_input(struct gpio_chip *gc,
+				       unsigned int offset)
+{
+	struct moxtet_gpio_chip *chip = gpiochip_get_data(gc);
+
+	if (chip->desc->in_mask & BIT(offset))
+		return 0;
+	else if (chip->desc->out_mask & BIT(offset))
+		return -ENOTSUPP;
+	else
+		return -EINVAL;
+}
+
+static int moxtet_gpio_direction_output(struct gpio_chip *gc,
+					unsigned int offset, int val)
+{
+	struct moxtet_gpio_chip *chip = gpiochip_get_data(gc);
+
+	if (chip->desc->out_mask & BIT(offset))
+		return moxtet_gpio_set_value(gc, offset, val);
+	else if (chip->desc->in_mask & BIT(offset))
+		return -ENOTSUPP;
+
+	return -EINVAL;
+}
+
+static int moxtet_gpio_probe(struct device *dev)
+{
+	struct moxtet_gpio_chip *chip;
+	struct device_node *nc = dev->of_node;
+	int id;
+
+	id = to_moxtet_device(dev)->id;
+
+	if (id >= ARRAY_SIZE(descs)) {
+		dev_err(dev, "%pOF Moxtet device id 0x%x is not supported by gpio-moxtet driver\n",
+			nc, id);
+		return -ENOTSUPP;
+	}
+
+	chip = devm_kzalloc(dev, sizeof(*chip), GFP_KERNEL);
+	if (!chip)
+		return -ENOMEM;
+
+	chip->dev = dev;
+	chip->gpio_chip.parent = dev;
+	chip->desc = &descs[id];
+
+	dev_set_drvdata(dev, chip);
+
+	chip->gpio_chip.label = dev_name(dev);
+	chip->gpio_chip.get_direction = moxtet_gpio_get_direction;
+	chip->gpio_chip.direction_input = moxtet_gpio_direction_input;
+	chip->gpio_chip.direction_output = moxtet_gpio_direction_output;
+	chip->gpio_chip.get = moxtet_gpio_get_value;
+	chip->gpio_chip.set = moxtet_gpio_set_value;
+	chip->gpio_chip.base = -1;
+
+	chip->gpio_chip.ngpio = MOXTET_GPIO_NGPIOS;
+
+	chip->gpio_chip.can_sleep = true;
+	chip->gpio_chip.owner = THIS_MODULE;
+
+	return devm_gpiochip_add_data(dev, &chip->gpio_chip, chip);
+}
+
+static const struct of_device_id moxtet_gpio_dt_ids[] = {
+	{ .compatible = "cznic,moxtet-gpio", },
+	{},
+};
+MODULE_DEVICE_TABLE(of, moxtet_gpio_dt_ids);
+
+static const enum turris_mox_module_id moxtet_gpio_module_table[] = {
+	TURRIS_MOX_MODULE_SFP,
+	0,
+};
+
+static struct moxtet_driver moxtet_gpio_driver = {
+	.driver = {
+		.name		= "moxtet-gpio",
+		.of_match_table	= moxtet_gpio_dt_ids,
+		.probe		= moxtet_gpio_probe,
+	},
+	.id_table = moxtet_gpio_module_table,
+};
+module_moxtet_driver(moxtet_gpio_driver);
+
+MODULE_AUTHOR("Marek Behun <kabel@kernel.org>");
+MODULE_DESCRIPTION("Turris Mox Moxtet GPIO expander");
+MODULE_LICENSE("GPL v2");