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, 352 insertions, 0 deletions

diff --git a/drivers/input/misc/sparcspkr.c b/drivers/input/misc/sparcspkr.c
new file mode 100644
index 000000000..1cfadd738
--- /dev/null
+++ b/drivers/input/misc/sparcspkr.c
@@ -0,0 +1,352 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  Driver for PC-speaker like devices found on various Sparc systems.
+ *
+ *  Copyright (c) 2002 Vojtech Pavlik
+ *  Copyright (c) 2002, 2006, 2008 David S. Miller (davem@davemloft.net)
+ */
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/input.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+
+#include <asm/io.h>
+
+MODULE_AUTHOR("David S. Miller <davem@davemloft.net>");
+MODULE_DESCRIPTION("Sparc Speaker beeper driver");
+MODULE_LICENSE("GPL");
+
+struct grover_beep_info {
+	void __iomem	*freq_regs;
+	void __iomem	*enable_reg;
+};
+
+struct bbc_beep_info {
+	u32		clock_freq;
+	void __iomem	*regs;
+};
+
+struct sparcspkr_state {
+	const char		*name;
+	int (*event)(struct input_dev *dev, unsigned int type, unsigned int code, int value);
+	spinlock_t		lock;
+	struct input_dev	*input_dev;
+	union {
+		struct grover_beep_info grover;
+		struct bbc_beep_info bbc;
+	} u;
+};
+
+static u32 bbc_count_to_reg(struct bbc_beep_info *info, unsigned int count)
+{
+	u32 val, clock_freq = info->clock_freq;
+	int i;
+
+	if (!count)
+		return 0;
+
+	if (count <= clock_freq >> 20)
+		return 1 << 18;
+
+	if (count >= clock_freq >> 12)
+		return 1 << 10;
+
+	val = 1 << 18;
+	for (i = 19; i >= 11; i--) {
+		val >>= 1;
+		if (count <= clock_freq >> i)
+			break;
+	}
+
+	return val;
+}
+
+static int bbc_spkr_event(struct input_dev *dev, unsigned int type, unsigned int code, int value)
+{
+	struct sparcspkr_state *state = dev_get_drvdata(dev->dev.parent);
+	struct bbc_beep_info *info = &state->u.bbc;
+	unsigned int count = 0;
+
+	if (type != EV_SND)
+		return -1;
+
+	switch (code) {
+	case SND_BELL:
+		if (value)
+			value = 1000;
+		break;
+	case SND_TONE:
+		break;
+	default:
+		return -1;
+	}
+
+	if (value > 20 && value < 32767)
+		count = 1193182 / value;
+
+	count = bbc_count_to_reg(info, count);
+
+	guard(spinlock_irqsave)(&state->lock);
+
+	if (count) {
+		sbus_writeb(0x01,                 info->regs + 0);
+		sbus_writeb(0x00,                 info->regs + 2);
+		sbus_writeb((count >> 16) & 0xff, info->regs + 3);
+		sbus_writeb((count >>  8) & 0xff, info->regs + 4);
+		sbus_writeb(0x00,                 info->regs + 5);
+	} else {
+		sbus_writeb(0x00,                 info->regs + 0);
+	}
+
+	return 0;
+}
+
+static int grover_spkr_event(struct input_dev *dev, unsigned int type, unsigned int code, int value)
+{
+	struct sparcspkr_state *state = dev_get_drvdata(dev->dev.parent);
+	struct grover_beep_info *info = &state->u.grover;
+	unsigned int count = 0;
+
+	if (type != EV_SND)
+		return -1;
+
+	switch (code) {
+	case SND_BELL:
+		if (value)
+			value = 1000;
+		break;
+	case SND_TONE:
+		break;
+	default:
+		return -1;
+	}
+
+	if (value > 20 && value < 32767)
+		count = 1193182 / value;
+
+	guard(spinlock_irqsave)(&state->lock);
+
+	if (count) {
+		/* enable counter 2 */
+		sbus_writeb(sbus_readb(info->enable_reg) | 3, info->enable_reg);
+		/* set command for counter 2, 2 byte write */
+		sbus_writeb(0xB6, info->freq_regs + 1);
+		/* select desired HZ */
+		sbus_writeb(count & 0xff, info->freq_regs + 0);
+		sbus_writeb((count >> 8) & 0xff, info->freq_regs + 0);
+	} else {
+		/* disable counter 2 */
+		sbus_writeb(sbus_readb(info->enable_reg) & 0xFC, info->enable_reg);
+	}
+
+	return 0;
+}
+
+static int sparcspkr_probe(struct device *dev)
+{
+	struct sparcspkr_state *state = dev_get_drvdata(dev);
+	struct input_dev *input_dev;
+	int error;
+
+	input_dev = input_allocate_device();
+	if (!input_dev)
+		return -ENOMEM;
+
+	input_dev->name = state->name;
+	input_dev->phys = "sparc/input0";
+	input_dev->id.bustype = BUS_ISA;
+	input_dev->id.vendor = 0x001f;
+	input_dev->id.product = 0x0001;
+	input_dev->id.version = 0x0100;
+	input_dev->dev.parent = dev;
+
+	input_dev->evbit[0] = BIT_MASK(EV_SND);
+	input_dev->sndbit[0] = BIT_MASK(SND_BELL) | BIT_MASK(SND_TONE);
+
+	input_dev->event = state->event;
+
+	error = input_register_device(input_dev);
+	if (error) {
+		input_free_device(input_dev);
+		return error;
+	}
+
+	state->input_dev = input_dev;
+
+	return 0;
+}
+
+static void sparcspkr_shutdown(struct platform_device *dev)
+{
+	struct sparcspkr_state *state = platform_get_drvdata(dev);
+	struct input_dev *input_dev = state->input_dev;
+
+	/* turn off the speaker */
+	state->event(input_dev, EV_SND, SND_BELL, 0);
+}
+
+static int bbc_beep_probe(struct platform_device *op)
+{
+	struct sparcspkr_state *state;
+	struct bbc_beep_info *info;
+	int err;
+
+	state = devm_kzalloc(&op->dev, sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return -ENOMEM;
+
+	state->name = "Sparc BBC Speaker";
+	state->event = bbc_spkr_event;
+	spin_lock_init(&state->lock);
+
+	struct device_node *dp __free(device_node) = of_find_node_by_path("/");
+	if (!dp)
+		return -ENODEV;
+
+	info = &state->u.bbc;
+	info->clock_freq = of_getintprop_default(dp, "clock-frequency", 0);
+	if (!info->clock_freq)
+		return -ENODEV;
+
+	info->regs = of_ioremap(&op->resource[0], 0, 6, "bbc beep");
+	if (!info->regs)
+		return -ENODEV;
+
+	platform_set_drvdata(op, state);
+
+	err = sparcspkr_probe(&op->dev);
+	if (err) {
+		of_iounmap(&op->resource[0], info->regs, 6);
+		return err;
+	}
+
+	return 0;
+}
+
+static void bbc_remove(struct platform_device *op)
+{
+	struct sparcspkr_state *state = platform_get_drvdata(op);
+	struct input_dev *input_dev = state->input_dev;
+	struct bbc_beep_info *info = &state->u.bbc;
+
+	/* turn off the speaker */
+	state->event(input_dev, EV_SND, SND_BELL, 0);
+
+	input_unregister_device(input_dev);
+
+	of_iounmap(&op->resource[0], info->regs, 6);
+}
+
+static const struct of_device_id bbc_beep_match[] = {
+	{
+		.name = "beep",
+		.compatible = "SUNW,bbc-beep",
+	},
+	{},
+};
+MODULE_DEVICE_TABLE(of, bbc_beep_match);
+
+static struct platform_driver bbc_beep_driver = {
+	.driver = {
+		.name = "bbcbeep",
+		.of_match_table = bbc_beep_match,
+	},
+	.probe		= bbc_beep_probe,
+	.remove		= bbc_remove,
+	.shutdown	= sparcspkr_shutdown,
+};
+
+static int grover_beep_probe(struct platform_device *op)
+{
+	struct sparcspkr_state *state;
+	struct grover_beep_info *info;
+	int err = -ENOMEM;
+
+	state = devm_kzalloc(&op->dev, sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return err;
+
+	state->name = "Sparc Grover Speaker";
+	state->event = grover_spkr_event;
+	spin_lock_init(&state->lock);
+
+	info = &state->u.grover;
+	info->freq_regs = of_ioremap(&op->resource[2], 0, 2, "grover beep freq");
+	if (!info->freq_regs)
+		return err;
+
+	info->enable_reg = of_ioremap(&op->resource[3], 0, 1, "grover beep enable");
+	if (!info->enable_reg)
+		goto out_unmap_freq_regs;
+
+	platform_set_drvdata(op, state);
+
+	err = sparcspkr_probe(&op->dev);
+	if (err)
+		goto out_clear_drvdata;
+
+	return 0;
+
+out_clear_drvdata:
+	of_iounmap(&op->resource[3], info->enable_reg, 1);
+
+out_unmap_freq_regs:
+	of_iounmap(&op->resource[2], info->freq_regs, 2);
+
+	return err;
+}
+
+static void grover_remove(struct platform_device *op)
+{
+	struct sparcspkr_state *state = platform_get_drvdata(op);
+	struct grover_beep_info *info = &state->u.grover;
+	struct input_dev *input_dev = state->input_dev;
+
+	/* turn off the speaker */
+	state->event(input_dev, EV_SND, SND_BELL, 0);
+
+	input_unregister_device(input_dev);
+
+	of_iounmap(&op->resource[3], info->enable_reg, 1);
+	of_iounmap(&op->resource[2], info->freq_regs, 2);
+}
+
+static const struct of_device_id grover_beep_match[] = {
+	{
+		.name = "beep",
+		.compatible = "SUNW,smbus-beep",
+	},
+	{},
+};
+MODULE_DEVICE_TABLE(of, grover_beep_match);
+
+static struct platform_driver grover_beep_driver = {
+	.driver = {
+		.name = "groverbeep",
+		.of_match_table = grover_beep_match,
+	},
+	.probe		= grover_beep_probe,
+	.remove		= grover_remove,
+	.shutdown	= sparcspkr_shutdown,
+};
+
+static struct platform_driver * const drivers[] = {
+	&bbc_beep_driver,
+	&grover_beep_driver,
+};
+
+static int __init sparcspkr_init(void)
+{
+	return platform_register_drivers(drivers, ARRAY_SIZE(drivers));
+}
+
+static void __exit sparcspkr_exit(void)
+{
+	platform_unregister_drivers(drivers, ARRAY_SIZE(drivers));
+}
+
+module_init(sparcspkr_init);
+module_exit(sparcspkr_exit);