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, 215 insertions, 0 deletions

diff --git a/drivers/rtc/rtc-rs5c348.c b/drivers/rtc/rtc-rs5c348.c
new file mode 100644
index 000000000..fec633f80
--- /dev/null
+++ b/drivers/rtc/rtc-rs5c348.c
@@ -0,0 +1,215 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * A SPI driver for the Ricoh RS5C348 RTC
+ *
+ * Copyright (C) 2006 Atsushi Nemoto <anemo@mba.ocn.ne.jp>
+ *
+ * The board specific init code should provide characteristics of this
+ * device:
+ *     Mode 1 (High-Active, Shift-Then-Sample), High Avtive CS
+ */
+
+#include <linux/bcd.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/errno.h>
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/string.h>
+#include <linux/slab.h>
+#include <linux/rtc.h>
+#include <linux/workqueue.h>
+#include <linux/spi/spi.h>
+#include <linux/module.h>
+
+#define RS5C348_REG_SECS	0
+#define RS5C348_REG_MINS	1
+#define RS5C348_REG_HOURS	2
+#define RS5C348_REG_WDAY	3
+#define RS5C348_REG_DAY	4
+#define RS5C348_REG_MONTH	5
+#define RS5C348_REG_YEAR	6
+#define RS5C348_REG_CTL1	14
+#define RS5C348_REG_CTL2	15
+
+#define RS5C348_SECS_MASK	0x7f
+#define RS5C348_MINS_MASK	0x7f
+#define RS5C348_HOURS_MASK	0x3f
+#define RS5C348_WDAY_MASK	0x03
+#define RS5C348_DAY_MASK	0x3f
+#define RS5C348_MONTH_MASK	0x1f
+
+#define RS5C348_BIT_PM	0x20	/* REG_HOURS */
+#define RS5C348_BIT_Y2K	0x80	/* REG_MONTH */
+#define RS5C348_BIT_24H	0x20	/* REG_CTL1 */
+#define RS5C348_BIT_XSTP	0x10	/* REG_CTL2 */
+#define RS5C348_BIT_VDET	0x40	/* REG_CTL2 */
+
+#define RS5C348_CMD_W(addr)	(((addr) << 4) | 0x08)	/* single write */
+#define RS5C348_CMD_R(addr)	(((addr) << 4) | 0x0c)	/* single read */
+#define RS5C348_CMD_MW(addr)	(((addr) << 4) | 0x00)	/* burst write */
+#define RS5C348_CMD_MR(addr)	(((addr) << 4) | 0x04)	/* burst read */
+
+struct rs5c348_plat_data {
+	struct rtc_device *rtc;
+	int rtc_24h;
+};
+
+static int
+rs5c348_rtc_set_time(struct device *dev, struct rtc_time *tm)
+{
+	struct spi_device *spi = to_spi_device(dev);
+	struct rs5c348_plat_data *pdata = dev_get_platdata(&spi->dev);
+	u8 txbuf[5+7], *txp;
+	int ret;
+
+	ret = spi_w8r8(spi, RS5C348_CMD_R(RS5C348_REG_CTL2));
+	if (ret < 0)
+		return ret;
+	if (ret & RS5C348_BIT_XSTP) {
+		txbuf[0] = RS5C348_CMD_W(RS5C348_REG_CTL2);
+		txbuf[1] = 0;
+		ret = spi_write_then_read(spi, txbuf, 2, NULL, 0);
+		if (ret < 0)
+			return ret;
+	}
+
+	/* Transfer 5 bytes before writing SEC.  This gives 31us for carry. */
+	txp = txbuf;
+	txbuf[0] = RS5C348_CMD_R(RS5C348_REG_CTL2); /* cmd, ctl2 */
+	txbuf[1] = 0;	/* dummy */
+	txbuf[2] = RS5C348_CMD_R(RS5C348_REG_CTL2); /* cmd, ctl2 */
+	txbuf[3] = 0;	/* dummy */
+	txbuf[4] = RS5C348_CMD_MW(RS5C348_REG_SECS); /* cmd, sec, ... */
+	txp = &txbuf[5];
+	txp[RS5C348_REG_SECS] = bin2bcd(tm->tm_sec);
+	txp[RS5C348_REG_MINS] = bin2bcd(tm->tm_min);
+	if (pdata->rtc_24h) {
+		txp[RS5C348_REG_HOURS] = bin2bcd(tm->tm_hour);
+	} else {
+		/* hour 0 is AM12, noon is PM12 */
+		txp[RS5C348_REG_HOURS] = bin2bcd((tm->tm_hour + 11) % 12 + 1) |
+			(tm->tm_hour >= 12 ? RS5C348_BIT_PM : 0);
+	}
+	txp[RS5C348_REG_WDAY] = bin2bcd(tm->tm_wday);
+	txp[RS5C348_REG_DAY] = bin2bcd(tm->tm_mday);
+	txp[RS5C348_REG_MONTH] = bin2bcd(tm->tm_mon + 1) |
+		(tm->tm_year >= 100 ? RS5C348_BIT_Y2K : 0);
+	txp[RS5C348_REG_YEAR] = bin2bcd(tm->tm_year % 100);
+	/* write in one transfer to avoid data inconsistency */
+	ret = spi_write_then_read(spi, txbuf, sizeof(txbuf), NULL, 0);
+	udelay(62);	/* Tcsr 62us */
+	return ret;
+}
+
+static int
+rs5c348_rtc_read_time(struct device *dev, struct rtc_time *tm)
+{
+	struct spi_device *spi = to_spi_device(dev);
+	struct rs5c348_plat_data *pdata = dev_get_platdata(&spi->dev);
+	u8 txbuf[5], rxbuf[7];
+	int ret;
+
+	ret = spi_w8r8(spi, RS5C348_CMD_R(RS5C348_REG_CTL2));
+	if (ret < 0)
+		return ret;
+	if (ret & RS5C348_BIT_VDET)
+		dev_warn(&spi->dev, "voltage-low detected.\n");
+	if (ret & RS5C348_BIT_XSTP) {
+		dev_warn(&spi->dev, "oscillator-stop detected.\n");
+		return -EINVAL;
+	}
+
+	/* Transfer 5 byte befores reading SEC.  This gives 31us for carry. */
+	txbuf[0] = RS5C348_CMD_R(RS5C348_REG_CTL2); /* cmd, ctl2 */
+	txbuf[1] = 0;	/* dummy */
+	txbuf[2] = RS5C348_CMD_R(RS5C348_REG_CTL2); /* cmd, ctl2 */
+	txbuf[3] = 0;	/* dummy */
+	txbuf[4] = RS5C348_CMD_MR(RS5C348_REG_SECS); /* cmd, sec, ... */
+
+	/* read in one transfer to avoid data inconsistency */
+	ret = spi_write_then_read(spi, txbuf, sizeof(txbuf),
+				  rxbuf, sizeof(rxbuf));
+	udelay(62);	/* Tcsr 62us */
+	if (ret < 0)
+		return ret;
+
+	tm->tm_sec = bcd2bin(rxbuf[RS5C348_REG_SECS] & RS5C348_SECS_MASK);
+	tm->tm_min = bcd2bin(rxbuf[RS5C348_REG_MINS] & RS5C348_MINS_MASK);
+	tm->tm_hour = bcd2bin(rxbuf[RS5C348_REG_HOURS] & RS5C348_HOURS_MASK);
+	if (!pdata->rtc_24h) {
+		if (rxbuf[RS5C348_REG_HOURS] & RS5C348_BIT_PM) {
+			tm->tm_hour -= 20;
+			tm->tm_hour %= 12;
+			tm->tm_hour += 12;
+		} else
+			tm->tm_hour %= 12;
+	}
+	tm->tm_wday = bcd2bin(rxbuf[RS5C348_REG_WDAY] & RS5C348_WDAY_MASK);
+	tm->tm_mday = bcd2bin(rxbuf[RS5C348_REG_DAY] & RS5C348_DAY_MASK);
+	tm->tm_mon =
+		bcd2bin(rxbuf[RS5C348_REG_MONTH] & RS5C348_MONTH_MASK) - 1;
+	/* year is 1900 + tm->tm_year */
+	tm->tm_year = bcd2bin(rxbuf[RS5C348_REG_YEAR]) +
+		((rxbuf[RS5C348_REG_MONTH] & RS5C348_BIT_Y2K) ? 100 : 0);
+
+	return 0;
+}
+
+static const struct rtc_class_ops rs5c348_rtc_ops = {
+	.read_time	= rs5c348_rtc_read_time,
+	.set_time	= rs5c348_rtc_set_time,
+};
+
+static int rs5c348_probe(struct spi_device *spi)
+{
+	int ret;
+	struct rtc_device *rtc;
+	struct rs5c348_plat_data *pdata;
+
+	pdata = devm_kzalloc(&spi->dev, sizeof(struct rs5c348_plat_data),
+				GFP_KERNEL);
+	if (!pdata)
+		return -ENOMEM;
+	spi->dev.platform_data = pdata;
+
+	/* Check D7 of SECOND register */
+	ret = spi_w8r8(spi, RS5C348_CMD_R(RS5C348_REG_SECS));
+	if (ret < 0 || (ret & 0x80)) {
+		dev_err(&spi->dev, "not found.\n");
+		return ret;
+	}
+
+	dev_info(&spi->dev, "spiclk %u KHz.\n",
+		 (spi->max_speed_hz + 500) / 1000);
+
+	ret = spi_w8r8(spi, RS5C348_CMD_R(RS5C348_REG_CTL1));
+	if (ret < 0)
+		return ret;
+	if (ret & RS5C348_BIT_24H)
+		pdata->rtc_24h = 1;
+
+	rtc = devm_rtc_allocate_device(&spi->dev);
+	if (IS_ERR(rtc))
+		return PTR_ERR(rtc);
+
+	pdata->rtc = rtc;
+
+	rtc->ops = &rs5c348_rtc_ops;
+
+	return devm_rtc_register_device(rtc);
+}
+
+static struct spi_driver rs5c348_driver = {
+	.driver = {
+		.name	= "rtc-rs5c348",
+	},
+	.probe	= rs5c348_probe,
+};
+
+module_spi_driver(rs5c348_driver);
+
+MODULE_AUTHOR("Atsushi Nemoto <anemo@mba.ocn.ne.jp>");
+MODULE_DESCRIPTION("Ricoh RS5C348 RTC driver");
+MODULE_LICENSE("GPL");
+MODULE_ALIAS("spi:rtc-rs5c348");