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, 183 insertions, 0 deletions

diff --git a/drivers/rtc/rtc-ds1347.c b/drivers/rtc/rtc-ds1347.c
new file mode 100644
index 000000000..a40c1a52d
--- /dev/null
+++ b/drivers/rtc/rtc-ds1347.c
@@ -0,0 +1,183 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* rtc-ds1347.c
+ *
+ * Driver for Dallas Semiconductor DS1347 Low Current, SPI Compatible
+ * Real Time Clock
+ *
+ * Author : Raghavendra Chandra Ganiga <ravi23ganiga@gmail.com>
+ */
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/platform_device.h>
+#include <linux/rtc.h>
+#include <linux/spi/spi.h>
+#include <linux/bcd.h>
+#include <linux/regmap.h>
+
+/* Registers in ds1347 rtc */
+
+#define DS1347_SECONDS_REG	0x01
+#define DS1347_MINUTES_REG	0x03
+#define DS1347_HOURS_REG	0x05
+#define DS1347_DATE_REG		0x07
+#define DS1347_MONTH_REG	0x09
+#define DS1347_DAY_REG		0x0B
+#define DS1347_YEAR_REG		0x0D
+#define DS1347_CONTROL_REG	0x0F
+#define DS1347_CENTURY_REG	0x13
+#define DS1347_STATUS_REG	0x17
+#define DS1347_CLOCK_BURST	0x3F
+
+#define DS1347_WP_BIT		BIT(7)
+
+#define DS1347_NEOSC_BIT	BIT(7)
+#define DS1347_OSF_BIT		BIT(2)
+
+static const struct regmap_range ds1347_ranges[] = {
+	{
+		.range_min = DS1347_SECONDS_REG,
+		.range_max = DS1347_STATUS_REG,
+	},
+};
+
+static const struct regmap_access_table ds1347_access_table = {
+	.yes_ranges = ds1347_ranges,
+	.n_yes_ranges = ARRAY_SIZE(ds1347_ranges),
+};
+
+static int ds1347_read_time(struct device *dev, struct rtc_time *dt)
+{
+	struct regmap *map = dev_get_drvdata(dev);
+	unsigned int status, century, secs;
+	unsigned char buf[8];
+	int err;
+
+	err = regmap_read(map, DS1347_STATUS_REG, &status);
+	if (err)
+		return err;
+
+	if (status & DS1347_OSF_BIT)
+		return -EINVAL;
+
+	do {
+		err = regmap_bulk_read(map, DS1347_CLOCK_BURST, buf, 8);
+		if (err)
+			return err;
+
+		err = regmap_read(map, DS1347_CENTURY_REG, &century);
+		if (err)
+			return err;
+
+		err = regmap_read(map, DS1347_SECONDS_REG, &secs);
+		if (err)
+			return err;
+	} while (buf[0] != secs);
+
+	dt->tm_sec = bcd2bin(buf[0]);
+	dt->tm_min = bcd2bin(buf[1] & 0x7f);
+	dt->tm_hour = bcd2bin(buf[2] & 0x3F);
+	dt->tm_mday = bcd2bin(buf[3]);
+	dt->tm_mon = bcd2bin(buf[4]) - 1;
+	dt->tm_wday = bcd2bin(buf[5]) - 1;
+	dt->tm_year = (bcd2bin(century) * 100) + bcd2bin(buf[6]) - 1900;
+
+	return 0;
+}
+
+static int ds1347_set_time(struct device *dev, struct rtc_time *dt)
+{
+	struct regmap *map = dev_get_drvdata(dev);
+	unsigned int century;
+	unsigned char buf[8];
+	int err;
+
+	err = regmap_update_bits(map, DS1347_STATUS_REG,
+				 DS1347_NEOSC_BIT, DS1347_NEOSC_BIT);
+	if (err)
+		return err;
+
+	buf[0] = bin2bcd(dt->tm_sec);
+	buf[1] = bin2bcd(dt->tm_min);
+	buf[2] = (bin2bcd(dt->tm_hour) & 0x3F);
+	buf[3] = bin2bcd(dt->tm_mday);
+	buf[4] = bin2bcd(dt->tm_mon + 1);
+	buf[5] = bin2bcd(dt->tm_wday + 1);
+	buf[6] = bin2bcd(dt->tm_year % 100);
+	buf[7] = bin2bcd(0x00);
+
+	err = regmap_bulk_write(map, DS1347_CLOCK_BURST, buf, 8);
+	if (err)
+		return err;
+
+	century = (dt->tm_year / 100) + 19;
+	err = regmap_write(map, DS1347_CENTURY_REG, bin2bcd(century));
+	if (err)
+		return err;
+
+	return regmap_update_bits(map, DS1347_STATUS_REG,
+				  DS1347_NEOSC_BIT | DS1347_OSF_BIT, 0);
+}
+
+static const struct rtc_class_ops ds1347_rtc_ops = {
+	.read_time = ds1347_read_time,
+	.set_time = ds1347_set_time,
+};
+
+static int ds1347_probe(struct spi_device *spi)
+{
+	struct rtc_device *rtc;
+	struct regmap_config config;
+	struct regmap *map;
+	int err;
+
+	memset(&config, 0, sizeof(config));
+	config.reg_bits = 8;
+	config.val_bits = 8;
+	config.read_flag_mask = 0x80;
+	config.max_register = 0x3F;
+	config.wr_table = &ds1347_access_table;
+
+	/* spi setup with ds1347 in mode 3 and bits per word as 8 */
+	spi->mode = SPI_MODE_3;
+	spi->bits_per_word = 8;
+	spi_setup(spi);
+
+	map = devm_regmap_init_spi(spi, &config);
+
+	if (IS_ERR(map)) {
+		dev_err(&spi->dev, "ds1347 regmap init spi failed\n");
+		return PTR_ERR(map);
+	}
+
+	spi_set_drvdata(spi, map);
+
+	/* Disable the write protect of rtc */
+	err = regmap_update_bits(map, DS1347_CONTROL_REG, DS1347_WP_BIT, 0);
+	if (err)
+		return err;
+
+	rtc = devm_rtc_allocate_device(&spi->dev);
+	if (IS_ERR(rtc))
+		return PTR_ERR(rtc);
+
+	rtc->ops = &ds1347_rtc_ops;
+	rtc->range_min = RTC_TIMESTAMP_BEGIN_0000;
+	rtc->range_max = RTC_TIMESTAMP_END_9999;
+
+	return devm_rtc_register_device(rtc);
+}
+
+static struct spi_driver ds1347_driver = {
+	.driver = {
+		.name = "ds1347",
+	},
+	.probe = ds1347_probe,
+};
+
+module_spi_driver(ds1347_driver);
+
+MODULE_DESCRIPTION("DS1347 SPI RTC DRIVER");
+MODULE_AUTHOR("Raghavendra C Ganiga <ravi23ganiga@gmail.com>");
+MODULE_LICENSE("GPL v2");