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, 130 insertions, 0 deletions

diff --git a/drivers/rtc/rtc-aspeed.c b/drivers/rtc/rtc-aspeed.c
new file mode 100644
index 000000000..8f5b440f8
--- /dev/null
+++ b/drivers/rtc/rtc-aspeed.c
@@ -0,0 +1,130 @@
+// SPDX-License-Identifier: GPL-2.0+
+// Copyright 2015 IBM Corp.
+
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/rtc.h>
+#include <linux/io.h>
+
+struct aspeed_rtc {
+	void __iomem *base;
+};
+
+#define RTC_TIME	0x00
+#define RTC_YEAR	0x04
+#define RTC_CTRL	0x10
+
+#define RTC_UNLOCK	BIT(1)
+#define RTC_ENABLE	BIT(0)
+
+static int aspeed_rtc_read_time(struct device *dev, struct rtc_time *tm)
+{
+	struct aspeed_rtc *rtc = dev_get_drvdata(dev);
+	unsigned int cent, year;
+	u32 reg1, reg2;
+
+	if (!(readl(rtc->base + RTC_CTRL) & RTC_ENABLE)) {
+		dev_dbg(dev, "%s failing as rtc disabled\n", __func__);
+		return -EINVAL;
+	}
+
+	do {
+		reg2 = readl(rtc->base + RTC_YEAR);
+		reg1 = readl(rtc->base + RTC_TIME);
+	} while (reg2 != readl(rtc->base + RTC_YEAR));
+
+	tm->tm_mday = (reg1 >> 24) & 0x1f;
+	tm->tm_hour = (reg1 >> 16) & 0x1f;
+	tm->tm_min = (reg1 >> 8) & 0x3f;
+	tm->tm_sec = (reg1 >> 0) & 0x3f;
+
+	cent = (reg2 >> 16) & 0x1f;
+	year = (reg2 >> 8) & 0x7f;
+	tm->tm_mon = ((reg2 >>  0) & 0x0f) - 1;
+	tm->tm_year = year + (cent * 100) - 1900;
+
+	dev_dbg(dev, "%s %ptR", __func__, tm);
+
+	return 0;
+}
+
+static int aspeed_rtc_set_time(struct device *dev, struct rtc_time *tm)
+{
+	struct aspeed_rtc *rtc = dev_get_drvdata(dev);
+	u32 reg1, reg2, ctrl;
+	int year, cent;
+
+	cent = (tm->tm_year + 1900) / 100;
+	year = tm->tm_year % 100;
+
+	reg1 = (tm->tm_mday << 24) | (tm->tm_hour << 16) | (tm->tm_min << 8) |
+		tm->tm_sec;
+
+	reg2 = ((cent & 0x1f) << 16) | ((year & 0x7f) << 8) |
+		((tm->tm_mon + 1) & 0xf);
+
+	ctrl = readl(rtc->base + RTC_CTRL);
+	writel(ctrl | RTC_UNLOCK, rtc->base + RTC_CTRL);
+
+	writel(reg1, rtc->base + RTC_TIME);
+	writel(reg2, rtc->base + RTC_YEAR);
+
+	/* Re-lock and ensure enable is set now that a time is programmed */
+	writel(ctrl | RTC_ENABLE, rtc->base + RTC_CTRL);
+
+	return 0;
+}
+
+static const struct rtc_class_ops aspeed_rtc_ops = {
+	.read_time = aspeed_rtc_read_time,
+	.set_time = aspeed_rtc_set_time,
+};
+
+static int aspeed_rtc_probe(struct platform_device *pdev)
+{
+	struct aspeed_rtc *rtc;
+	struct rtc_device *rtc_dev;
+
+	rtc = devm_kzalloc(&pdev->dev, sizeof(*rtc), GFP_KERNEL);
+	if (!rtc)
+		return -ENOMEM;
+
+	rtc->base = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(rtc->base))
+		return PTR_ERR(rtc->base);
+
+	rtc_dev = devm_rtc_allocate_device(&pdev->dev);
+	if (IS_ERR(rtc_dev))
+		return PTR_ERR(rtc_dev);
+
+	platform_set_drvdata(pdev, rtc);
+
+	rtc_dev->ops = &aspeed_rtc_ops;
+	rtc_dev->range_min = RTC_TIMESTAMP_BEGIN_1900;
+	rtc_dev->range_max = 38814989399LL; /* 3199-12-31 23:59:59 */
+
+	return devm_rtc_register_device(rtc_dev);
+}
+
+static const struct of_device_id aspeed_rtc_match[] = {
+	{ .compatible = "aspeed,ast2400-rtc", },
+	{ .compatible = "aspeed,ast2500-rtc", },
+	{ .compatible = "aspeed,ast2600-rtc", },
+	{ .compatible = "aspeed,ast2700-rtc", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, aspeed_rtc_match);
+
+static struct platform_driver aspeed_rtc_driver = {
+	.driver = {
+		.name = "aspeed-rtc",
+		.of_match_table = aspeed_rtc_match,
+	},
+};
+
+module_platform_driver_probe(aspeed_rtc_driver, aspeed_rtc_probe);
+
+MODULE_DESCRIPTION("ASPEED RTC driver");
+MODULE_AUTHOR("Joel Stanley <joel@jms.id.au>");
+MODULE_LICENSE("GPL");