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, 140 insertions, 0 deletions

diff --git a/drivers/rtc/rtc-macsmc.c b/drivers/rtc/rtc-macsmc.c
new file mode 100644
index 000000000..8fe883066
--- /dev/null
+++ b/drivers/rtc/rtc-macsmc.c
@@ -0,0 +1,140 @@
+// SPDX-License-Identifier: GPL-2.0-only OR MIT
+/*
+ * Apple SMC RTC driver
+ * Copyright The Asahi Linux Contributors
+ */
+
+#include <linux/bitops.h>
+#include <linux/mfd/macsmc.h>
+#include <linux/module.h>
+#include <linux/nvmem-consumer.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/rtc.h>
+#include <linux/slab.h>
+
+/* 48-bit RTC */
+#define RTC_BYTES 6
+#define RTC_BITS (8 * RTC_BYTES)
+
+/* 32768 Hz clock */
+#define RTC_SEC_SHIFT 15
+
+struct macsmc_rtc {
+	struct device *dev;
+	struct apple_smc *smc;
+	struct rtc_device *rtc_dev;
+	struct nvmem_cell *rtc_offset;
+};
+
+static int macsmc_rtc_get_time(struct device *dev, struct rtc_time *tm)
+{
+	struct macsmc_rtc *rtc = dev_get_drvdata(dev);
+	u64 ctr = 0, off = 0;
+	time64_t now;
+	void *p_off;
+	size_t len;
+	int ret;
+
+	ret = apple_smc_read(rtc->smc, SMC_KEY(CLKM), &ctr, RTC_BYTES);
+	if (ret < 0)
+		return ret;
+	if (ret != RTC_BYTES)
+		return -EIO;
+
+	p_off = nvmem_cell_read(rtc->rtc_offset, &len);
+	if (IS_ERR(p_off))
+		return PTR_ERR(p_off);
+	if (len < RTC_BYTES) {
+		kfree(p_off);
+		return -EIO;
+	}
+
+	memcpy(&off, p_off, RTC_BYTES);
+	kfree(p_off);
+
+	/* Sign extend from 48 to 64 bits, then arithmetic shift right 15 bits to get seconds */
+	now = sign_extend64(ctr + off, RTC_BITS - 1) >> RTC_SEC_SHIFT;
+	rtc_time64_to_tm(now, tm);
+
+	return ret;
+}
+
+static int macsmc_rtc_set_time(struct device *dev, struct rtc_time *tm)
+{
+	struct macsmc_rtc *rtc = dev_get_drvdata(dev);
+	u64 ctr = 0, off = 0;
+	int ret;
+
+	ret = apple_smc_read(rtc->smc, SMC_KEY(CLKM), &ctr, RTC_BYTES);
+	if (ret < 0)
+		return ret;
+	if (ret != RTC_BYTES)
+		return -EIO;
+
+	/* This sets the offset such that the set second begins now */
+	off = (rtc_tm_to_time64(tm) << RTC_SEC_SHIFT) - ctr;
+	return nvmem_cell_write(rtc->rtc_offset, &off, RTC_BYTES);
+}
+
+static const struct rtc_class_ops macsmc_rtc_ops = {
+	.read_time = macsmc_rtc_get_time,
+	.set_time = macsmc_rtc_set_time,
+};
+
+static int macsmc_rtc_probe(struct platform_device *pdev)
+{
+	struct apple_smc *smc = dev_get_drvdata(pdev->dev.parent);
+	struct macsmc_rtc *rtc;
+
+	/*
+	 * MFD will probe this device even without a node in the device tree,
+	 * thus bail out early if the SMC on the current machines does not
+	 * support RTC and has no node in the device tree.
+	 */
+	if (!pdev->dev.of_node)
+		return -ENODEV;
+
+	rtc = devm_kzalloc(&pdev->dev, sizeof(*rtc), GFP_KERNEL);
+	if (!rtc)
+		return -ENOMEM;
+
+	rtc->dev = &pdev->dev;
+	rtc->smc = smc;
+
+	rtc->rtc_offset = devm_nvmem_cell_get(&pdev->dev, "rtc_offset");
+	if (IS_ERR(rtc->rtc_offset))
+		return dev_err_probe(&pdev->dev, PTR_ERR(rtc->rtc_offset),
+				     "Failed to get rtc_offset NVMEM cell\n");
+
+	rtc->rtc_dev = devm_rtc_allocate_device(&pdev->dev);
+	if (IS_ERR(rtc->rtc_dev))
+		return PTR_ERR(rtc->rtc_dev);
+
+	rtc->rtc_dev->ops = &macsmc_rtc_ops;
+	rtc->rtc_dev->range_min = S64_MIN >> (RTC_SEC_SHIFT + (64 - RTC_BITS));
+	rtc->rtc_dev->range_max = S64_MAX >> (RTC_SEC_SHIFT + (64 - RTC_BITS));
+
+	platform_set_drvdata(pdev, rtc);
+
+	return devm_rtc_register_device(rtc->rtc_dev);
+}
+
+static const struct of_device_id macsmc_rtc_of_table[] = {
+	{ .compatible = "apple,smc-rtc", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, macsmc_rtc_of_table);
+
+static struct platform_driver macsmc_rtc_driver = {
+	.driver = {
+		.name = "macsmc-rtc",
+		.of_match_table = macsmc_rtc_of_table,
+	},
+	.probe = macsmc_rtc_probe,
+};
+module_platform_driver(macsmc_rtc_driver);
+
+MODULE_LICENSE("Dual MIT/GPL");
+MODULE_DESCRIPTION("Apple SMC RTC driver");
+MODULE_AUTHOR("Hector Martin <marcan@marcan.st>");