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, 170 insertions, 0 deletions

diff --git a/drivers/rtc/rtc-ep93xx.c b/drivers/rtc/rtc-ep93xx.c
new file mode 100644
index 000000000..507c9f220
--- /dev/null
+++ b/drivers/rtc/rtc-ep93xx.c
@@ -0,0 +1,170 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * A driver for the RTC embedded in the Cirrus Logic EP93XX processors
+ * Copyright (c) 2006 Tower Technologies
+ *
+ * Author: Alessandro Zummo <a.zummo@towertech.it>
+ */
+
+#include <linux/module.h>
+#include <linux/rtc.h>
+#include <linux/platform_device.h>
+#include <linux/io.h>
+#include <linux/gfp.h>
+
+#define EP93XX_RTC_DATA			0x000
+#define EP93XX_RTC_MATCH		0x004
+#define EP93XX_RTC_STATUS		0x008
+#define  EP93XX_RTC_STATUS_INTR		 BIT(0)
+#define EP93XX_RTC_LOAD			0x00C
+#define EP93XX_RTC_CONTROL		0x010
+#define  EP93XX_RTC_CONTROL_MIE		 BIT(0)
+#define EP93XX_RTC_SWCOMP		0x108
+#define  EP93XX_RTC_SWCOMP_DEL_MASK	 0x001f0000
+#define  EP93XX_RTC_SWCOMP_DEL_SHIFT	 16
+#define  EP93XX_RTC_SWCOMP_INT_MASK	 0x0000ffff
+#define  EP93XX_RTC_SWCOMP_INT_SHIFT	 0
+
+struct ep93xx_rtc {
+	void __iomem	*mmio_base;
+};
+
+static int ep93xx_rtc_get_swcomp(struct device *dev, unsigned short *preload,
+				 unsigned short *delete)
+{
+	struct ep93xx_rtc *ep93xx_rtc = dev_get_drvdata(dev);
+	unsigned long comp;
+
+	comp = readl(ep93xx_rtc->mmio_base + EP93XX_RTC_SWCOMP);
+
+	if (preload)
+		*preload = (comp & EP93XX_RTC_SWCOMP_INT_MASK)
+				>> EP93XX_RTC_SWCOMP_INT_SHIFT;
+
+	if (delete)
+		*delete = (comp & EP93XX_RTC_SWCOMP_DEL_MASK)
+				>> EP93XX_RTC_SWCOMP_DEL_SHIFT;
+
+	return 0;
+}
+
+static int ep93xx_rtc_read_time(struct device *dev, struct rtc_time *tm)
+{
+	struct ep93xx_rtc *ep93xx_rtc = dev_get_drvdata(dev);
+	unsigned long time;
+
+	time = readl(ep93xx_rtc->mmio_base + EP93XX_RTC_DATA);
+
+	rtc_time64_to_tm(time, tm);
+	return 0;
+}
+
+static int ep93xx_rtc_set_time(struct device *dev, struct rtc_time *tm)
+{
+	struct ep93xx_rtc *ep93xx_rtc = dev_get_drvdata(dev);
+	unsigned long secs = rtc_tm_to_time64(tm);
+
+	writel(secs + 1, ep93xx_rtc->mmio_base + EP93XX_RTC_LOAD);
+	return 0;
+}
+
+static int ep93xx_rtc_proc(struct device *dev, struct seq_file *seq)
+{
+	unsigned short preload, delete;
+
+	ep93xx_rtc_get_swcomp(dev, &preload, &delete);
+
+	seq_printf(seq, "preload\t\t: %d\n", preload);
+	seq_printf(seq, "delete\t\t: %d\n", delete);
+
+	return 0;
+}
+
+static const struct rtc_class_ops ep93xx_rtc_ops = {
+	.read_time	= ep93xx_rtc_read_time,
+	.set_time	= ep93xx_rtc_set_time,
+	.proc		= ep93xx_rtc_proc,
+};
+
+static ssize_t comp_preload_show(struct device *dev,
+				 struct device_attribute *attr, char *buf)
+{
+	unsigned short preload;
+
+	ep93xx_rtc_get_swcomp(dev->parent, &preload, NULL);
+
+	return sprintf(buf, "%d\n", preload);
+}
+static DEVICE_ATTR_RO(comp_preload);
+
+static ssize_t comp_delete_show(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	unsigned short delete;
+
+	ep93xx_rtc_get_swcomp(dev->parent, NULL, &delete);
+
+	return sprintf(buf, "%d\n", delete);
+}
+static DEVICE_ATTR_RO(comp_delete);
+
+static struct attribute *ep93xx_rtc_attrs[] = {
+	&dev_attr_comp_preload.attr,
+	&dev_attr_comp_delete.attr,
+	NULL
+};
+
+static const struct attribute_group ep93xx_rtc_sysfs_files = {
+	.attrs	= ep93xx_rtc_attrs,
+};
+
+static int ep93xx_rtc_probe(struct platform_device *pdev)
+{
+	struct ep93xx_rtc *ep93xx_rtc;
+	struct rtc_device *rtc;
+	int err;
+
+	ep93xx_rtc = devm_kzalloc(&pdev->dev, sizeof(*ep93xx_rtc), GFP_KERNEL);
+	if (!ep93xx_rtc)
+		return -ENOMEM;
+
+	ep93xx_rtc->mmio_base = devm_platform_ioremap_resource(pdev, 0);
+	if (IS_ERR(ep93xx_rtc->mmio_base))
+		return PTR_ERR(ep93xx_rtc->mmio_base);
+
+	platform_set_drvdata(pdev, ep93xx_rtc);
+
+	rtc = devm_rtc_allocate_device(&pdev->dev);
+	if (IS_ERR(rtc))
+		return PTR_ERR(rtc);
+
+	rtc->ops = &ep93xx_rtc_ops;
+	rtc->range_max = U32_MAX;
+
+	err = rtc_add_group(rtc, &ep93xx_rtc_sysfs_files);
+	if (err)
+		return err;
+
+	return devm_rtc_register_device(rtc);
+}
+
+static const struct of_device_id ep93xx_rtc_of_ids[] = {
+	{ .compatible = "cirrus,ep9301-rtc" },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, ep93xx_rtc_of_ids);
+
+static struct platform_driver ep93xx_rtc_driver = {
+	.driver		= {
+		.name	= "ep93xx-rtc",
+		.of_match_table = ep93xx_rtc_of_ids,
+	},
+	.probe		= ep93xx_rtc_probe,
+};
+
+module_platform_driver(ep93xx_rtc_driver);
+
+MODULE_AUTHOR("Alessandro Zummo <a.zummo@towertech.it>");
+MODULE_DESCRIPTION("EP93XX RTC driver");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("platform:ep93xx-rtc");