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, 120 insertions, 0 deletions

diff --git a/drivers/reset/reset-socfpga.c b/drivers/reset/reset-socfpga.c
new file mode 100644
index 000000000..da208270b
--- /dev/null
+++ b/drivers/reset/reset-socfpga.c
@@ -0,0 +1,120 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2018, Intel Corporation
+ * Copied from reset-sunxi.c
+ */
+
+#include <linux/err.h>
+#include <linux/io.h>
+#include <linux/init.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/platform_device.h>
+#include <linux/reset-controller.h>
+#include <linux/reset/reset-simple.h>
+#include <linux/reset/socfpga.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <linux/types.h>
+
+#define SOCFPGA_NR_BANKS	8
+
+static int a10_reset_init(struct device_node *np)
+{
+	struct reset_simple_data *data;
+	struct resource res;
+	resource_size_t size;
+	int ret;
+	u32 reg_offset = 0x10;
+
+	data = kzalloc_obj(*data);
+	if (!data)
+		return -ENOMEM;
+
+	ret = of_address_to_resource(np, 0, &res);
+	if (ret)
+		goto err_alloc;
+
+	size = resource_size(&res);
+	if (!request_mem_region(res.start, size, np->name)) {
+		ret = -EBUSY;
+		goto err_alloc;
+	}
+
+	data->membase = ioremap(res.start, size);
+	if (!data->membase) {
+		ret = -ENOMEM;
+		goto release_region;
+	}
+
+	if (of_property_read_u32(np, "altr,modrst-offset", &reg_offset))
+		pr_warn("missing altr,modrst-offset property, assuming 0x10\n");
+	data->membase += reg_offset;
+
+	spin_lock_init(&data->lock);
+
+	data->rcdev.owner = THIS_MODULE;
+	data->rcdev.nr_resets = SOCFPGA_NR_BANKS * 32;
+	data->rcdev.ops = &reset_simple_ops;
+	data->rcdev.of_node = np;
+	data->status_active_low = true;
+
+	ret = reset_controller_register(&data->rcdev);
+	if (ret)
+		pr_err("unable to register device\n");
+
+	return ret;
+
+release_region:
+	release_mem_region(res.start, size);
+
+err_alloc:
+	kfree(data);
+	return ret;
+};
+
+/*
+ * These are the reset controller we need to initialize early on in
+ * our system, before we can even think of using a regular device
+ * driver for it.
+ * The controllers that we can register through the regular device
+ * model are handled by the simple reset driver directly.
+ */
+static const struct of_device_id socfpga_early_reset_dt_ids[] __initconst = {
+	{ .compatible = "altr,rst-mgr", },
+	{ /* sentinel */ },
+};
+
+void __init socfpga_reset_init(void)
+{
+	struct device_node *np;
+
+	for_each_matching_node(np, socfpga_early_reset_dt_ids)
+		a10_reset_init(np);
+}
+
+/*
+ * The early driver is problematic, because it doesn't register
+ * itself as a driver. This causes certain device links to prevent
+ * consumer devices from probing. The hacky solution is to register
+ * an empty driver, whose only job is to attach itself to the reset
+ * manager and call probe.
+ */
+static const struct of_device_id socfpga_reset_dt_ids[] = {
+	{ .compatible = "altr,rst-mgr", },
+	{ /* sentinel */ },
+};
+
+static int reset_simple_probe(struct platform_device *pdev)
+{
+	return 0;
+}
+
+static struct platform_driver reset_socfpga_driver = {
+	.probe	= reset_simple_probe,
+	.driver = {
+		.name		= "socfpga-reset",
+		.of_match_table	= socfpga_reset_dt_ids,
+	},
+};
+builtin_platform_driver(reset_socfpga_driver);