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, 150 insertions, 0 deletions

diff --git a/drivers/clocksource/timer-realtek.c b/drivers/clocksource/timer-realtek.c
new file mode 100644
index 000000000..4f0439de9
--- /dev/null
+++ b/drivers/clocksource/timer-realtek.c
@@ -0,0 +1,150 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (c) 2025 Realtek Semiconductor Corp.
+ */
+
+#define pr_fmt(fmt)	KBUILD_MODNAME ": " fmt
+
+#include <linux/irqflags.h>
+#include <linux/interrupt.h>
+#include "timer-of.h"
+
+#define ENBL 1
+#define DSBL 0
+
+#define SYSTIMER_RATE 1000000
+#define SYSTIMER_MIN_DELTA 0x64
+#define SYSTIMER_MAX_DELTA ULONG_MAX
+
+/* SYSTIMER Register Offset (RTK Internal Use) */
+#define TS_LW_OFST 0x0
+#define TS_HW_OFST 0x4
+#define TS_CMP_VAL_LW_OFST 0x8
+#define TS_CMP_VAL_HW_OFST 0xC
+#define TS_CMP_CTRL_OFST 0x10
+#define TS_CMP_STAT_OFST 0x14
+
+/* SYSTIMER CMP CTRL REG Mask */
+#define TS_CMP_EN_MASK 0x1
+#define TS_WR_EN0_MASK 0x2
+
+static void __iomem *systimer_base;
+
+static u64 rtk_ts64_read(void)
+{
+	u32 low, high;
+	u64 ts;
+
+	/* Caution: Read LSB word (TS_LW_OFST) first then MSB (TS_HW_OFST) */
+	low = readl(systimer_base + TS_LW_OFST);
+	high = readl(systimer_base + TS_HW_OFST);
+	ts = ((u64)high << 32) | low;
+
+	return ts;
+}
+
+static void rtk_cmp_value_write(u64 value)
+{
+	u32 high, low;
+
+	low = value & 0xFFFFFFFF;
+	high = value >> 32;
+
+	writel(high, systimer_base + TS_CMP_VAL_HW_OFST);
+	writel(low, systimer_base + TS_CMP_VAL_LW_OFST);
+}
+
+static inline void rtk_cmp_en_write(bool cmp_en)
+{
+	u32 val;
+
+	val = TS_WR_EN0_MASK;
+	if (cmp_en == ENBL)
+		val |= TS_CMP_EN_MASK;
+
+	writel(val, systimer_base + TS_CMP_CTRL_OFST);
+}
+
+static int rtk_syst_clkevt_next_event(unsigned long cycles, struct clock_event_device *clkevt)
+{
+	u64 cmp_val;
+
+	rtk_cmp_en_write(DSBL);
+	cmp_val = rtk_ts64_read();
+
+	/* Set CMP value to current timestamp plus delta_us */
+	rtk_cmp_value_write(cmp_val + cycles);
+	rtk_cmp_en_write(ENBL);
+	return 0;
+}
+
+static irqreturn_t rtk_ts_match_intr_handler(int irq, void *dev_id)
+{
+	struct clock_event_device *clkevt = dev_id;
+	void __iomem *reg_base;
+	u32 val;
+
+	/* Disable TS CMP Match */
+	rtk_cmp_en_write(DSBL);
+
+	/* Clear TS CMP INTR */
+	reg_base = systimer_base + TS_CMP_STAT_OFST;
+	val = readl(reg_base) & TS_CMP_EN_MASK;
+	writel(val | TS_CMP_EN_MASK, reg_base);
+	clkevt->event_handler(clkevt);
+
+	return IRQ_HANDLED;
+}
+
+static int rtk_syst_shutdown(struct clock_event_device *clkevt)
+{
+	void __iomem *reg_base;
+	u64 cmp_val = 0;
+
+	/* Disable TS CMP Match */
+	rtk_cmp_en_write(DSBL);
+	/* Set compare value to 0 */
+	rtk_cmp_value_write(cmp_val);
+
+	/* Clear TS CMP INTR */
+	reg_base = systimer_base + TS_CMP_STAT_OFST;
+	writel(TS_CMP_EN_MASK, reg_base);
+	return 0;
+}
+
+static struct timer_of rtk_timer_to = {
+	.flags = TIMER_OF_IRQ | TIMER_OF_BASE,
+
+	.clkevt = {
+		.name			= "rtk-clkevt",
+		.rating			= 300,
+		.cpumask		= cpu_possible_mask,
+		.features		= CLOCK_EVT_FEAT_DYNIRQ |
+					  CLOCK_EVT_FEAT_ONESHOT,
+		.set_next_event		= rtk_syst_clkevt_next_event,
+		.set_state_oneshot	= rtk_syst_shutdown,
+		.set_state_shutdown	= rtk_syst_shutdown,
+	},
+
+	.of_irq = {
+		.flags			= IRQF_TIMER | IRQF_IRQPOLL,
+		.handler		= rtk_ts_match_intr_handler,
+	},
+};
+
+static int __init rtk_systimer_init(struct device_node *node)
+{
+	int ret;
+
+	ret = timer_of_init(node, &rtk_timer_to);
+	if (ret)
+		return ret;
+
+	systimer_base = timer_of_base(&rtk_timer_to);
+	clockevents_config_and_register(&rtk_timer_to.clkevt, SYSTIMER_RATE,
+					SYSTIMER_MIN_DELTA, SYSTIMER_MAX_DELTA);
+
+	return 0;
+}
+
+TIMER_OF_DECLARE(rtk_systimer, "realtek,rtd1625-systimer", rtk_systimer_init);