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, 135 insertions, 0 deletions

diff --git a/arch/sh/drivers/push-switch.c b/arch/sh/drivers/push-switch.c
new file mode 100644
index 000000000..a39e2edd4
--- /dev/null
+++ b/arch/sh/drivers/push-switch.c
@@ -0,0 +1,135 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Generic push-switch framework
+ *
+ * Copyright (C) 2006  Paul Mundt
+ */
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/module.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <asm/push-switch.h>
+
+#define DRV_NAME "push-switch"
+#define DRV_VERSION "0.1.1"
+
+static ssize_t switch_show(struct device *dev,
+			   struct device_attribute *attr,
+			   char *buf)
+{
+	struct push_switch_platform_info *psw_info = dev->platform_data;
+	return sprintf(buf, "%s\n", psw_info->name);
+}
+static DEVICE_ATTR_RO(switch);
+
+static void switch_timer(struct timer_list *t)
+{
+	struct push_switch *psw = timer_container_of(psw, t, debounce);
+
+	schedule_work(&psw->work);
+}
+
+static void switch_work_handler(struct work_struct *work)
+{
+	struct push_switch *psw = container_of(work, struct push_switch, work);
+	struct platform_device *pdev = psw->pdev;
+
+	psw->state = 0;
+
+	kobject_uevent(&pdev->dev.kobj, KOBJ_CHANGE);
+}
+
+static int switch_drv_probe(struct platform_device *pdev)
+{
+	struct push_switch_platform_info *psw_info;
+	struct push_switch *psw;
+	int ret, irq;
+
+	psw = kzalloc_obj(struct push_switch);
+	if (unlikely(!psw))
+		return -ENOMEM;
+
+	irq = platform_get_irq(pdev, 0);
+	if (unlikely(irq < 0)) {
+		ret = -ENODEV;
+		goto err;
+	}
+
+	psw_info = pdev->dev.platform_data;
+	BUG_ON(!psw_info);
+
+	ret = request_irq(irq, psw_info->irq_handler,
+			  psw_info->irq_flags,
+			  psw_info->name ? psw_info->name : DRV_NAME, pdev);
+	if (unlikely(ret < 0))
+		goto err;
+
+	if (psw_info->name) {
+		ret = device_create_file(&pdev->dev, &dev_attr_switch);
+		if (unlikely(ret)) {
+			dev_err(&pdev->dev, "Failed creating device attrs\n");
+			ret = -EINVAL;
+			goto err_irq;
+		}
+	}
+
+	INIT_WORK(&psw->work, switch_work_handler);
+	timer_setup(&psw->debounce, switch_timer, 0);
+
+	/* Workqueue API brain-damage */
+	psw->pdev = pdev;
+
+	platform_set_drvdata(pdev, psw);
+
+	return 0;
+
+err_irq:
+	free_irq(irq, pdev);
+err:
+	kfree(psw);
+	return ret;
+}
+
+static void switch_drv_remove(struct platform_device *pdev)
+{
+	struct push_switch *psw = platform_get_drvdata(pdev);
+	struct push_switch_platform_info *psw_info = pdev->dev.platform_data;
+	int irq = platform_get_irq(pdev, 0);
+
+	if (psw_info->name)
+		device_remove_file(&pdev->dev, &dev_attr_switch);
+
+	platform_set_drvdata(pdev, NULL);
+	timer_shutdown_sync(&psw->debounce);
+	flush_work(&psw->work);
+	free_irq(irq, pdev);
+
+	kfree(psw);
+}
+
+static struct platform_driver switch_driver = {
+	.probe		= switch_drv_probe,
+	.remove		= switch_drv_remove,
+	.driver		= {
+		.name	= DRV_NAME,
+	},
+};
+
+static int __init switch_init(void)
+{
+	printk(KERN_NOTICE DRV_NAME ": version %s loaded\n", DRV_VERSION);
+	return platform_driver_register(&switch_driver);
+}
+
+static void __exit switch_exit(void)
+{
+	platform_driver_unregister(&switch_driver);
+}
+module_init(switch_init);
+module_exit(switch_exit);
+
+MODULE_VERSION(DRV_VERSION);
+MODULE_AUTHOR("Paul Mundt");
+MODULE_DESCRIPTION("Generic push-switch framework");
+MODULE_LICENSE("GPL v2");