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, 162 insertions, 0 deletions

diff --git a/drivers/rtc/rtc-imx-sm-bbm.c b/drivers/rtc/rtc-imx-sm-bbm.c
new file mode 100644
index 000000000..daa472be7
--- /dev/null
+++ b/drivers/rtc/rtc-imx-sm-bbm.c
@@ -0,0 +1,162 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * Copyright 2024 NXP.
+ */
+
+#include <linux/jiffies.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/rtc.h>
+#include <linux/scmi_protocol.h>
+#include <linux/scmi_imx_protocol.h>
+
+struct scmi_imx_bbm {
+	const struct scmi_imx_bbm_proto_ops *ops;
+	struct rtc_device *rtc_dev;
+	struct scmi_protocol_handle *ph;
+	struct notifier_block nb;
+};
+
+static int scmi_imx_bbm_read_time(struct device *dev, struct rtc_time *tm)
+{
+	struct scmi_imx_bbm *bbnsm = dev_get_drvdata(dev);
+	struct scmi_protocol_handle *ph = bbnsm->ph;
+	u64 val;
+	int ret;
+
+	ret = bbnsm->ops->rtc_time_get(ph, 0, &val);
+	if (ret)
+		return ret;
+
+	rtc_time64_to_tm(val, tm);
+
+	return 0;
+}
+
+static int scmi_imx_bbm_set_time(struct device *dev, struct rtc_time *tm)
+{
+	struct scmi_imx_bbm *bbnsm = dev_get_drvdata(dev);
+	struct scmi_protocol_handle *ph = bbnsm->ph;
+	u64 val;
+
+	val = rtc_tm_to_time64(tm);
+
+	return bbnsm->ops->rtc_time_set(ph, 0, val);
+}
+
+static int scmi_imx_bbm_alarm_irq_enable(struct device *dev, unsigned int enable)
+{
+	struct scmi_imx_bbm *bbnsm = dev_get_drvdata(dev);
+	struct scmi_protocol_handle *ph = bbnsm->ph;
+
+	/* scmi_imx_bbm_set_alarm enables the irq, just handle disable here */
+	if (!enable)
+		return bbnsm->ops->rtc_alarm_set(ph, 0, false, 0);
+
+	return 0;
+}
+
+static int scmi_imx_bbm_set_alarm(struct device *dev, struct rtc_wkalrm *alrm)
+{
+	struct scmi_imx_bbm *bbnsm = dev_get_drvdata(dev);
+	struct scmi_protocol_handle *ph = bbnsm->ph;
+	struct rtc_time *alrm_tm = &alrm->time;
+	u64 val;
+
+	val = rtc_tm_to_time64(alrm_tm);
+
+	return bbnsm->ops->rtc_alarm_set(ph, 0, true, val);
+}
+
+static const struct rtc_class_ops smci_imx_bbm_rtc_ops = {
+	.read_time = scmi_imx_bbm_read_time,
+	.set_time = scmi_imx_bbm_set_time,
+	.set_alarm = scmi_imx_bbm_set_alarm,
+	.alarm_irq_enable = scmi_imx_bbm_alarm_irq_enable,
+};
+
+static int scmi_imx_bbm_rtc_notifier(struct notifier_block *nb, unsigned long event, void *data)
+{
+	struct scmi_imx_bbm *bbnsm = container_of(nb, struct scmi_imx_bbm, nb);
+	struct scmi_imx_bbm_notif_report *r = data;
+
+	if (r->is_rtc)
+		rtc_update_irq(bbnsm->rtc_dev, 1, RTC_AF | RTC_IRQF);
+	else
+		pr_err("Unexpected bbm event: %s\n", __func__);
+
+	return 0;
+}
+
+static int scmi_imx_bbm_rtc_init(struct scmi_device *sdev)
+{
+	const struct scmi_handle *handle = sdev->handle;
+	struct device *dev = &sdev->dev;
+	struct scmi_imx_bbm *bbnsm = dev_get_drvdata(dev);
+	int ret;
+
+	bbnsm->rtc_dev = devm_rtc_allocate_device(dev);
+	if (IS_ERR(bbnsm->rtc_dev))
+		return PTR_ERR(bbnsm->rtc_dev);
+
+	bbnsm->rtc_dev->ops = &smci_imx_bbm_rtc_ops;
+	bbnsm->rtc_dev->range_max = U32_MAX;
+
+	bbnsm->nb.notifier_call = &scmi_imx_bbm_rtc_notifier;
+	ret = handle->notify_ops->devm_event_notifier_register(sdev, SCMI_PROTOCOL_IMX_BBM,
+							       SCMI_EVENT_IMX_BBM_RTC,
+							       NULL, &bbnsm->nb);
+	if (ret)
+		return ret;
+
+	return devm_rtc_register_device(bbnsm->rtc_dev);
+}
+
+static int scmi_imx_bbm_rtc_probe(struct scmi_device *sdev)
+{
+	const struct scmi_handle *handle = sdev->handle;
+	struct device *dev = &sdev->dev;
+	struct scmi_protocol_handle *ph;
+	struct scmi_imx_bbm *bbnsm;
+	int ret;
+
+	if (!handle)
+		return -ENODEV;
+
+	bbnsm = devm_kzalloc(dev, sizeof(*bbnsm), GFP_KERNEL);
+	if (!bbnsm)
+		return -ENOMEM;
+
+	bbnsm->ops = handle->devm_protocol_get(sdev, SCMI_PROTOCOL_IMX_BBM, &ph);
+	if (IS_ERR(bbnsm->ops))
+		return PTR_ERR(bbnsm->ops);
+
+	bbnsm->ph = ph;
+
+	device_init_wakeup(dev, true);
+
+	dev_set_drvdata(dev, bbnsm);
+
+	ret = scmi_imx_bbm_rtc_init(sdev);
+	if (ret)
+		device_init_wakeup(dev, false);
+
+	return ret;
+}
+
+static const struct scmi_device_id scmi_id_table[] = {
+	{ SCMI_PROTOCOL_IMX_BBM, "imx-bbm-rtc" },
+	{ },
+};
+MODULE_DEVICE_TABLE(scmi, scmi_id_table);
+
+static struct scmi_driver scmi_imx_bbm_rtc_driver = {
+	.name = "scmi-imx-bbm-rtc",
+	.probe = scmi_imx_bbm_rtc_probe,
+	.id_table = scmi_id_table,
+};
+module_scmi_driver(scmi_imx_bbm_rtc_driver);
+
+MODULE_AUTHOR("Peng Fan <peng.fan@nxp.com>");
+MODULE_DESCRIPTION("IMX SM BBM RTC driver");
+MODULE_LICENSE("GPL");