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, 207 insertions, 0 deletions

diff --git a/drivers/reset/reset-mpfs.c b/drivers/reset/reset-mpfs.c
new file mode 100644
index 000000000..8ffcc54ee
--- /dev/null
+++ b/drivers/reset/reset-mpfs.c
@@ -0,0 +1,207 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * PolarFire SoC (MPFS) Peripheral Clock Reset Controller
+ *
+ * Author: Conor Dooley <conor.dooley@microchip.com>
+ * Copyright (c) 2022 Microchip Technology Inc. and its subsidiaries.
+ *
+ */
+#include <linux/auxiliary_bus.h>
+#include <linux/delay.h>
+#include <linux/io.h>
+#include <linux/mfd/syscon.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/regmap.h>
+#include <linux/reset-controller.h>
+#include <linux/slab.h>
+#include <dt-bindings/clock/microchip,mpfs-clock.h>
+#include <soc/microchip/mpfs.h>
+
+/*
+ * The ENVM reset is the lowest bit in the register & I am using the CLK_FOO
+ * defines in the dt to make things easier to configure - so this is accounting
+ * for the offset of 3 there.
+ */
+#define MPFS_PERIPH_OFFSET	CLK_ENVM
+#define MPFS_NUM_RESETS		30u
+#define MPFS_SLEEP_MIN_US	100
+#define MPFS_SLEEP_MAX_US	200
+
+#define REG_SUBBLK_RESET_CR	0x88u
+
+struct mpfs_reset {
+	struct regmap *regmap;
+	struct reset_controller_dev rcdev;
+};
+
+static inline struct mpfs_reset *to_mpfs_reset(struct reset_controller_dev *rcdev)
+{
+	return container_of(rcdev, struct mpfs_reset, rcdev);
+}
+
+/*
+ * Peripheral clock resets
+ */
+static int mpfs_assert(struct reset_controller_dev *rcdev, unsigned long id)
+{
+	struct mpfs_reset *rst = to_mpfs_reset(rcdev);
+
+	return regmap_set_bits(rst->regmap, REG_SUBBLK_RESET_CR, BIT(id));
+
+}
+
+static int mpfs_deassert(struct reset_controller_dev *rcdev, unsigned long id)
+{
+	struct mpfs_reset *rst = to_mpfs_reset(rcdev);
+
+	return regmap_clear_bits(rst->regmap, REG_SUBBLK_RESET_CR, BIT(id));
+
+}
+
+static int mpfs_status(struct reset_controller_dev *rcdev, unsigned long id)
+{
+	struct mpfs_reset *rst = to_mpfs_reset(rcdev);
+	u32 reg;
+
+	regmap_read(rst->regmap, REG_SUBBLK_RESET_CR, &reg);
+
+	/*
+	 * It is safe to return here as MPFS_NUM_RESETS makes sure the sign bit
+	 * is never hit.
+	 */
+	return (reg & BIT(id));
+}
+
+static int mpfs_reset(struct reset_controller_dev *rcdev, unsigned long id)
+{
+	mpfs_assert(rcdev, id);
+
+	usleep_range(MPFS_SLEEP_MIN_US, MPFS_SLEEP_MAX_US);
+
+	mpfs_deassert(rcdev, id);
+
+	return 0;
+}
+
+static const struct reset_control_ops mpfs_reset_ops = {
+	.reset = mpfs_reset,
+	.assert = mpfs_assert,
+	.deassert = mpfs_deassert,
+	.status = mpfs_status,
+};
+
+static int mpfs_reset_xlate(struct reset_controller_dev *rcdev,
+			    const struct of_phandle_args *reset_spec)
+{
+	unsigned int index = reset_spec->args[0];
+
+	/*
+	 * CLK_RESERVED does not map to a clock, but it does map to a reset,
+	 * so it has to be accounted for here. It is the reset for the fabric,
+	 * so if this reset gets called - do not reset it.
+	 */
+	if (index == CLK_RESERVED) {
+		dev_err(rcdev->dev, "Resetting the fabric is not supported\n");
+		return -EINVAL;
+	}
+
+	if (index < MPFS_PERIPH_OFFSET || index >= (MPFS_PERIPH_OFFSET + rcdev->nr_resets)) {
+		dev_err(rcdev->dev, "Invalid reset index %u\n", index);
+		return -EINVAL;
+	}
+
+	return index - MPFS_PERIPH_OFFSET;
+}
+
+static int mpfs_reset_mfd_probe(struct platform_device *pdev)
+{
+	struct reset_controller_dev *rcdev;
+	struct device *dev = &pdev->dev;
+	struct mpfs_reset *rst;
+
+	rst = devm_kzalloc(dev, sizeof(*rst), GFP_KERNEL);
+	if (!rst)
+		return -ENOMEM;
+
+	rcdev = &rst->rcdev;
+	rcdev->dev = dev;
+	rcdev->ops = &mpfs_reset_ops;
+
+	rcdev->of_node = pdev->dev.parent->of_node;
+	rcdev->of_reset_n_cells = 1;
+	rcdev->of_xlate = mpfs_reset_xlate;
+	rcdev->nr_resets = MPFS_NUM_RESETS;
+
+	rst->regmap = device_node_to_regmap(pdev->dev.parent->of_node);
+	if (IS_ERR(rst->regmap))
+		return dev_err_probe(dev, PTR_ERR(rst->regmap),
+				     "Failed to find syscon regmap\n");
+
+	return devm_reset_controller_register(dev, rcdev);
+}
+
+static struct platform_driver mpfs_reset_mfd_driver = {
+	.probe = mpfs_reset_mfd_probe,
+	.driver = {
+		.name = "mpfs-reset",
+	},
+};
+module_platform_driver(mpfs_reset_mfd_driver);
+
+static int mpfs_reset_adev_probe(struct auxiliary_device *adev,
+				 const struct auxiliary_device_id *id)
+{
+	struct reset_controller_dev *rcdev;
+	struct device *dev = &adev->dev;
+	struct mpfs_reset *rst;
+
+	rst = devm_kzalloc(dev, sizeof(*rst), GFP_KERNEL);
+	if (!rst)
+		return -ENOMEM;
+
+	rst->regmap = (struct regmap *)adev->dev.platform_data;
+
+	rcdev = &rst->rcdev;
+	rcdev->dev = dev;
+	rcdev->ops = &mpfs_reset_ops;
+
+	rcdev->of_node = dev->parent->of_node;
+	rcdev->of_reset_n_cells = 1;
+	rcdev->of_xlate = mpfs_reset_xlate;
+	rcdev->nr_resets = MPFS_NUM_RESETS;
+
+	return devm_reset_controller_register(dev, rcdev);
+}
+
+int mpfs_reset_controller_register(struct device *clk_dev, struct regmap *map)
+{
+	struct auxiliary_device *adev;
+
+	adev = devm_auxiliary_device_create(clk_dev, "reset-mpfs", (void *)map);
+	if (!adev)
+		return -ENODEV;
+
+	return 0;
+}
+EXPORT_SYMBOL_NS_GPL(mpfs_reset_controller_register, "MCHP_CLK_MPFS");
+
+static const struct auxiliary_device_id mpfs_reset_ids[] = {
+	{
+		.name = "reset_mpfs.reset-mpfs",
+	},
+	{ }
+};
+MODULE_DEVICE_TABLE(auxiliary, mpfs_reset_ids);
+
+static struct auxiliary_driver mpfs_reset_aux_driver = {
+	.probe		= mpfs_reset_adev_probe,
+	.id_table	= mpfs_reset_ids,
+};
+
+module_auxiliary_driver(mpfs_reset_aux_driver);
+
+MODULE_DESCRIPTION("Microchip PolarFire SoC Reset Driver");
+MODULE_AUTHOR("Conor Dooley <conor.dooley@microchip.com>");
+MODULE_IMPORT_NS("MCHP_CLK_MPFS");