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, 212 insertions, 0 deletions

diff --git a/drivers/clocksource/timer-meson6.c b/drivers/clocksource/timer-meson6.c
new file mode 100644
index 000000000..99f5510a2
--- /dev/null
+++ b/drivers/clocksource/timer-meson6.c
@@ -0,0 +1,212 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Amlogic Meson6 SoCs timer handling.
+ *
+ * Copyright (C) 2014 Carlo Caione <carlo@caione.org>
+ *
+ * Based on code from Amlogic, Inc
+ */
+
+#include <linux/bitfield.h>
+#include <linux/bitops.h>
+#include <linux/clk.h>
+#include <linux/clockchips.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/irqreturn.h>
+#include <linux/sched_clock.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+
+#ifdef CONFIG_ARM
+#include <linux/delay.h>
+#endif
+
+#define MESON_ISA_TIMER_MUX					0x00
+#define MESON_ISA_TIMER_MUX_TIMERD_EN				BIT(19)
+#define MESON_ISA_TIMER_MUX_TIMERC_EN				BIT(18)
+#define MESON_ISA_TIMER_MUX_TIMERB_EN				BIT(17)
+#define MESON_ISA_TIMER_MUX_TIMERA_EN				BIT(16)
+#define MESON_ISA_TIMER_MUX_TIMERD_MODE				BIT(15)
+#define MESON_ISA_TIMER_MUX_TIMERC_MODE				BIT(14)
+#define MESON_ISA_TIMER_MUX_TIMERB_MODE				BIT(13)
+#define MESON_ISA_TIMER_MUX_TIMERA_MODE				BIT(12)
+#define MESON_ISA_TIMER_MUX_TIMERE_INPUT_CLOCK_MASK		GENMASK(10, 8)
+#define MESON_ISA_TIMER_MUX_TIMERE_INPUT_CLOCK_SYSTEM_CLOCK	0x0
+#define MESON_ISA_TIMER_MUX_TIMERE_INPUT_CLOCK_1US		0x1
+#define MESON_ISA_TIMER_MUX_TIMERE_INPUT_CLOCK_10US		0x2
+#define MESON_ISA_TIMER_MUX_TIMERE_INPUT_CLOCK_100US		0x3
+#define MESON_ISA_TIMER_MUX_TIMERE_INPUT_CLOCK_1MS		0x4
+#define MESON_ISA_TIMER_MUX_TIMERD_INPUT_CLOCK_MASK		GENMASK(7, 6)
+#define MESON_ISA_TIMER_MUX_TIMERC_INPUT_CLOCK_MASK		GENMASK(5, 4)
+#define MESON_ISA_TIMER_MUX_TIMERB_INPUT_CLOCK_MASK		GENMASK(3, 2)
+#define MESON_ISA_TIMER_MUX_TIMERA_INPUT_CLOCK_MASK		GENMASK(1, 0)
+#define MESON_ISA_TIMER_MUX_TIMERABCD_INPUT_CLOCK_1US		0x0
+#define MESON_ISA_TIMER_MUX_TIMERABCD_INPUT_CLOCK_10US		0x1
+#define MESON_ISA_TIMER_MUX_TIMERABCD_INPUT_CLOCK_100US		0x0
+#define MESON_ISA_TIMER_MUX_TIMERABCD_INPUT_CLOCK_1MS		0x3
+
+#define MESON_ISA_TIMERA					0x04
+#define MESON_ISA_TIMERB					0x08
+#define MESON_ISA_TIMERC					0x0c
+#define MESON_ISA_TIMERD					0x10
+#define MESON_ISA_TIMERE					0x14
+
+static void __iomem *timer_base;
+
+#ifdef CONFIG_ARM
+static unsigned long meson6_read_current_timer(void)
+{
+	return readl_relaxed(timer_base + MESON_ISA_TIMERE);
+}
+
+static struct delay_timer meson6_delay_timer = {
+	.read_current_timer = meson6_read_current_timer,
+	.freq = 1000 * 1000,
+};
+#endif
+
+static u64 notrace meson6_timer_sched_read(void)
+{
+	return (u64)readl(timer_base + MESON_ISA_TIMERE);
+}
+
+static void meson6_clkevt_time_stop(void)
+{
+	u32 val = readl(timer_base + MESON_ISA_TIMER_MUX);
+
+	writel(val & ~MESON_ISA_TIMER_MUX_TIMERA_EN,
+	       timer_base + MESON_ISA_TIMER_MUX);
+}
+
+static void meson6_clkevt_time_setup(unsigned long delay)
+{
+	writel(delay, timer_base + MESON_ISA_TIMERA);
+}
+
+static void meson6_clkevt_time_start(bool periodic)
+{
+	u32 val = readl(timer_base + MESON_ISA_TIMER_MUX);
+
+	if (periodic)
+		val |= MESON_ISA_TIMER_MUX_TIMERA_MODE;
+	else
+		val &= ~MESON_ISA_TIMER_MUX_TIMERA_MODE;
+
+	writel(val | MESON_ISA_TIMER_MUX_TIMERA_EN,
+	       timer_base + MESON_ISA_TIMER_MUX);
+}
+
+static int meson6_shutdown(struct clock_event_device *evt)
+{
+	meson6_clkevt_time_stop();
+	return 0;
+}
+
+static int meson6_set_oneshot(struct clock_event_device *evt)
+{
+	meson6_clkevt_time_stop();
+	meson6_clkevt_time_start(false);
+	return 0;
+}
+
+static int meson6_set_periodic(struct clock_event_device *evt)
+{
+	meson6_clkevt_time_stop();
+	meson6_clkevt_time_setup(USEC_PER_SEC / HZ - 1);
+	meson6_clkevt_time_start(true);
+	return 0;
+}
+
+static int meson6_clkevt_next_event(unsigned long evt,
+				    struct clock_event_device *unused)
+{
+	meson6_clkevt_time_stop();
+	meson6_clkevt_time_setup(evt);
+	meson6_clkevt_time_start(false);
+
+	return 0;
+}
+
+static struct clock_event_device meson6_clockevent = {
+	.name			= "meson6_tick",
+	.rating			= 400,
+	.features		= CLOCK_EVT_FEAT_PERIODIC |
+				  CLOCK_EVT_FEAT_ONESHOT,
+	.set_state_shutdown	= meson6_shutdown,
+	.set_state_periodic	= meson6_set_periodic,
+	.set_state_oneshot	= meson6_set_oneshot,
+	.tick_resume		= meson6_shutdown,
+	.set_next_event		= meson6_clkevt_next_event,
+};
+
+static irqreturn_t meson6_timer_interrupt(int irq, void *dev_id)
+{
+	struct clock_event_device *evt = (struct clock_event_device *)dev_id;
+
+	evt->event_handler(evt);
+
+	return IRQ_HANDLED;
+}
+
+static int __init meson6_timer_init(struct device_node *node)
+{
+	u32 val;
+	int ret, irq;
+
+	timer_base = of_io_request_and_map(node, 0, "meson6-timer");
+	if (IS_ERR(timer_base)) {
+		pr_err("Can't map registers\n");
+		return -ENXIO;
+	}
+
+	irq = irq_of_parse_and_map(node, 0);
+	if (irq <= 0) {
+		pr_err("Can't parse IRQ\n");
+		return -EINVAL;
+	}
+
+	/* Set 1us for timer E */
+	val = readl(timer_base + MESON_ISA_TIMER_MUX);
+	val &= ~MESON_ISA_TIMER_MUX_TIMERE_INPUT_CLOCK_MASK;
+	val |= FIELD_PREP(MESON_ISA_TIMER_MUX_TIMERE_INPUT_CLOCK_MASK,
+			  MESON_ISA_TIMER_MUX_TIMERE_INPUT_CLOCK_1US);
+	writel(val, timer_base + MESON_ISA_TIMER_MUX);
+
+	sched_clock_register(meson6_timer_sched_read, 32, USEC_PER_SEC);
+	clocksource_mmio_init(timer_base + MESON_ISA_TIMERE, node->name,
+			      1000 * 1000, 300, 32, clocksource_mmio_readl_up);
+
+	/* Timer A base 1us */
+	val &= ~MESON_ISA_TIMER_MUX_TIMERA_INPUT_CLOCK_MASK;
+	val |= FIELD_PREP(MESON_ISA_TIMER_MUX_TIMERA_INPUT_CLOCK_MASK,
+			  MESON_ISA_TIMER_MUX_TIMERABCD_INPUT_CLOCK_1US);
+	writel(val, timer_base + MESON_ISA_TIMER_MUX);
+
+	/* Stop the timer A */
+	meson6_clkevt_time_stop();
+
+	ret = request_irq(irq, meson6_timer_interrupt,
+			  IRQF_TIMER | IRQF_IRQPOLL, "meson6_timer",
+			  &meson6_clockevent);
+	if (ret) {
+		pr_warn("failed to setup irq %d\n", irq);
+		return ret;
+	}
+
+	meson6_clockevent.cpumask = cpu_possible_mask;
+	meson6_clockevent.irq = irq;
+
+	clockevents_config_and_register(&meson6_clockevent, USEC_PER_SEC,
+					1, 0xfffe);
+
+#ifdef CONFIG_ARM
+	/* Also use MESON_ISA_TIMERE for delays */
+	register_current_timer_delay(&meson6_delay_timer);
+#endif
+
+	return 0;
+}
+TIMER_OF_DECLARE(meson6, "amlogic,meson6-timer",
+		       meson6_timer_init);