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, 287 insertions, 0 deletions

diff --git a/drivers/acpi/sbshc.c b/drivers/acpi/sbshc.c
new file mode 100644
index 000000000..c0ffa267f
--- /dev/null
+++ b/drivers/acpi/sbshc.c
@@ -0,0 +1,287 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * SMBus driver for ACPI Embedded Controller (v0.1)
+ *
+ * Copyright (c) 2007 Alexey Starikovskiy
+ */
+
+#define pr_fmt(fmt) "ACPI: " fmt
+
+#include <linux/acpi.h>
+#include <linux/wait.h>
+#include <linux/slab.h>
+#include <linux/delay.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+
+#include "sbshc.h"
+#include "internal.h"
+
+struct acpi_smb_hc {
+	struct acpi_ec *ec;
+	struct mutex lock;
+	wait_queue_head_t wait;
+	u8 offset;
+	u8 query_bit;
+	smbus_alarm_callback callback;
+	void *context;
+	bool done;
+};
+
+static int acpi_smbus_hc_probe(struct platform_device *pdev);
+static void acpi_smbus_hc_remove(struct platform_device *pdev);
+
+static const struct acpi_device_id sbs_device_ids[] = {
+	{"ACPI0001", 0},
+	{"ACPI0005", 0},
+	{"", 0},
+};
+
+MODULE_DEVICE_TABLE(acpi, sbs_device_ids);
+
+static struct platform_driver acpi_smb_hc_driver = {
+	.probe = acpi_smbus_hc_probe,
+	.remove = acpi_smbus_hc_remove,
+	.driver = {
+		.name = "acpi-smbus-hc",
+		.acpi_match_table = sbs_device_ids,
+	},
+};
+
+union acpi_smb_status {
+	u8 raw;
+	struct {
+		u8 status:5;
+		u8 reserved:1;
+		u8 alarm:1;
+		u8 done:1;
+	} fields;
+};
+
+enum acpi_smb_status_codes {
+	SMBUS_OK = 0,
+	SMBUS_UNKNOWN_FAILURE = 0x07,
+	SMBUS_DEVICE_ADDRESS_NACK = 0x10,
+	SMBUS_DEVICE_ERROR = 0x11,
+	SMBUS_DEVICE_COMMAND_ACCESS_DENIED = 0x12,
+	SMBUS_UNKNOWN_ERROR = 0x13,
+	SMBUS_DEVICE_ACCESS_DENIED = 0x17,
+	SMBUS_TIMEOUT = 0x18,
+	SMBUS_HOST_UNSUPPORTED_PROTOCOL = 0x19,
+	SMBUS_BUSY = 0x1a,
+	SMBUS_PEC_ERROR = 0x1f,
+};
+
+enum acpi_smb_offset {
+	ACPI_SMB_PROTOCOL = 0,	/* protocol, PEC */
+	ACPI_SMB_STATUS = 1,	/* status */
+	ACPI_SMB_ADDRESS = 2,	/* address */
+	ACPI_SMB_COMMAND = 3,	/* command */
+	ACPI_SMB_DATA = 4,	/* 32 data registers */
+	ACPI_SMB_BLOCK_COUNT = 0x24,	/* number of data bytes */
+	ACPI_SMB_ALARM_ADDRESS = 0x25,	/* alarm address */
+	ACPI_SMB_ALARM_DATA = 0x26,	/* 2 bytes alarm data */
+};
+
+static inline int smb_hc_read(struct acpi_smb_hc *hc, u8 address, u8 *data)
+{
+	return ec_read(hc->offset + address, data);
+}
+
+static inline int smb_hc_write(struct acpi_smb_hc *hc, u8 address, u8 data)
+{
+	return ec_write(hc->offset + address, data);
+}
+
+static int wait_transaction_complete(struct acpi_smb_hc *hc, int timeout)
+{
+	if (wait_event_timeout(hc->wait, hc->done, msecs_to_jiffies(timeout)))
+		return 0;
+	return -ETIME;
+}
+
+static int acpi_smbus_transaction(struct acpi_smb_hc *hc, u8 protocol,
+				  u8 address, u8 command, u8 *data, u8 length)
+{
+	int ret = -EFAULT, i;
+	u8 temp, sz = 0;
+
+	if (!hc) {
+		pr_err("host controller is not configured\n");
+		return ret;
+	}
+
+	mutex_lock(&hc->lock);
+	hc->done = false;
+	if (smb_hc_read(hc, ACPI_SMB_PROTOCOL, &temp))
+		goto end;
+	if (temp) {
+		ret = -EBUSY;
+		goto end;
+	}
+	smb_hc_write(hc, ACPI_SMB_COMMAND, command);
+	if (!(protocol & 0x01)) {
+		smb_hc_write(hc, ACPI_SMB_BLOCK_COUNT, length);
+		for (i = 0; i < length; ++i)
+			smb_hc_write(hc, ACPI_SMB_DATA + i, data[i]);
+	}
+	smb_hc_write(hc, ACPI_SMB_ADDRESS, address << 1);
+	smb_hc_write(hc, ACPI_SMB_PROTOCOL, protocol);
+	/*
+	 * Wait for completion. Save the status code, data size,
+	 * and data into the return package (if required by the protocol).
+	 */
+	ret = wait_transaction_complete(hc, 1000);
+	if (ret || !(protocol & 0x01))
+		goto end;
+	switch (protocol) {
+	case SMBUS_RECEIVE_BYTE:
+	case SMBUS_READ_BYTE:
+		sz = 1;
+		break;
+	case SMBUS_READ_WORD:
+		sz = 2;
+		break;
+	case SMBUS_READ_BLOCK:
+		if (smb_hc_read(hc, ACPI_SMB_BLOCK_COUNT, &sz)) {
+			ret = -EFAULT;
+			goto end;
+		}
+		sz &= 0x1f;
+		break;
+	}
+	for (i = 0; i < sz; ++i)
+		smb_hc_read(hc, ACPI_SMB_DATA + i, &data[i]);
+      end:
+	mutex_unlock(&hc->lock);
+	return ret;
+}
+
+int acpi_smbus_read(struct acpi_smb_hc *hc, u8 protocol, u8 address,
+		    u8 command, u8 *data)
+{
+	return acpi_smbus_transaction(hc, protocol, address, command, data, 0);
+}
+
+EXPORT_SYMBOL_GPL(acpi_smbus_read);
+
+int acpi_smbus_write(struct acpi_smb_hc *hc, u8 protocol, u8 address,
+		     u8 command, u8 *data, u8 length)
+{
+	return acpi_smbus_transaction(hc, protocol, address, command, data, length);
+}
+
+EXPORT_SYMBOL_GPL(acpi_smbus_write);
+
+int acpi_smbus_register_callback(struct acpi_smb_hc *hc,
+				 smbus_alarm_callback callback, void *context)
+{
+	mutex_lock(&hc->lock);
+	hc->callback = callback;
+	hc->context = context;
+	mutex_unlock(&hc->lock);
+	return 0;
+}
+
+EXPORT_SYMBOL_GPL(acpi_smbus_register_callback);
+
+int acpi_smbus_unregister_callback(struct acpi_smb_hc *hc)
+{
+	mutex_lock(&hc->lock);
+	hc->callback = NULL;
+	hc->context = NULL;
+	mutex_unlock(&hc->lock);
+	acpi_os_wait_events_complete();
+	return 0;
+}
+
+EXPORT_SYMBOL_GPL(acpi_smbus_unregister_callback);
+
+static inline void acpi_smbus_callback(void *context)
+{
+	struct acpi_smb_hc *hc = context;
+	if (hc->callback)
+		hc->callback(hc->context);
+}
+
+static int smbus_alarm(void *context)
+{
+	struct acpi_smb_hc *hc = context;
+	union acpi_smb_status status;
+	u8 address;
+	if (smb_hc_read(hc, ACPI_SMB_STATUS, &status.raw))
+		return 0;
+	/* Check if it is only a completion notify */
+	if (status.fields.done && status.fields.status == SMBUS_OK) {
+		hc->done = true;
+		wake_up(&hc->wait);
+	}
+	if (!status.fields.alarm)
+		return 0;
+	mutex_lock(&hc->lock);
+	smb_hc_read(hc, ACPI_SMB_ALARM_ADDRESS, &address);
+	status.fields.alarm = 0;
+	smb_hc_write(hc, ACPI_SMB_STATUS, status.raw);
+	/* We are only interested in events coming from known devices */
+	switch (address >> 1) {
+		case ACPI_SBS_CHARGER:
+		case ACPI_SBS_MANAGER:
+		case ACPI_SBS_BATTERY:
+			acpi_os_execute(OSL_NOTIFY_HANDLER,
+					acpi_smbus_callback, hc);
+	}
+	mutex_unlock(&hc->lock);
+	return 0;
+}
+
+static int acpi_smbus_hc_probe(struct platform_device *pdev)
+{
+	struct acpi_device *device;
+	int status;
+	unsigned long long val;
+	struct acpi_smb_hc *hc;
+
+	device = ACPI_COMPANION(&pdev->dev);
+	if (!device)
+		return -ENODEV;
+
+	status = acpi_evaluate_integer(device->handle, "_EC", NULL, &val);
+	if (ACPI_FAILURE(status)) {
+		pr_err("error obtaining _EC.\n");
+		return -EIO;
+	}
+
+	hc = kzalloc_obj(struct acpi_smb_hc);
+	if (!hc)
+		return -ENOMEM;
+	mutex_init(&hc->lock);
+	init_waitqueue_head(&hc->wait);
+
+	platform_set_drvdata(pdev, hc);
+
+	hc->ec = dev_get_drvdata(pdev->dev.parent);
+	hc->offset = (val >> 8) & 0xff;
+	hc->query_bit = val & 0xff;
+
+	acpi_ec_add_query_handler(hc->ec, hc->query_bit, NULL, smbus_alarm, hc);
+	dev_info(&device->dev, "SBS HC: offset = 0x%0x, query_bit = 0x%0x\n",
+		 hc->offset, hc->query_bit);
+
+	return 0;
+}
+
+static void acpi_smbus_hc_remove(struct platform_device *pdev)
+{
+	struct acpi_smb_hc *hc = platform_get_drvdata(pdev);
+
+	acpi_ec_remove_query_handler(hc->ec, hc->query_bit);
+	acpi_os_wait_events_complete();
+	kfree(hc);
+}
+
+module_platform_driver(acpi_smb_hc_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Alexey Starikovskiy");
+MODULE_DESCRIPTION("ACPI SMBus HC driver");