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, 107 insertions, 0 deletions

diff --git a/drivers/base/pinctrl.c b/drivers/base/pinctrl.c
new file mode 100644
index 000000000..0bbc83231
--- /dev/null
+++ b/drivers/base/pinctrl.c
@@ -0,0 +1,107 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Driver core interface to the pinctrl subsystem.
+ *
+ * Copyright (C) 2012 ST-Ericsson SA
+ * Written on behalf of Linaro for ST-Ericsson
+ * Based on bits of regulator core, gpio core and clk core
+ *
+ * Author: Linus Walleij <linus.walleij@linaro.org>
+ */
+
+#include <linux/device.h>
+#include <linux/pinctrl/devinfo.h>
+#include <linux/pinctrl/consumer.h>
+#include <linux/slab.h>
+
+#include "base.h"
+
+/**
+ * pinctrl_bind_pins() - called by the device core before probe
+ * @dev: the device that is just about to probe
+ */
+int pinctrl_bind_pins(struct device *dev)
+{
+	int ret;
+
+	if (dev_of_node_reused(dev))
+		return 0;
+
+	dev->pins = devm_kzalloc(dev, sizeof(*(dev->pins)), GFP_KERNEL);
+	if (!dev->pins)
+		return -ENOMEM;
+
+	dev->pins->p = devm_pinctrl_get(dev);
+	if (IS_ERR(dev->pins->p)) {
+		dev_dbg(dev, "no pinctrl handle\n");
+		ret = PTR_ERR(dev->pins->p);
+		goto cleanup_alloc;
+	}
+
+	dev->pins->default_state = pinctrl_lookup_state(dev->pins->p,
+					PINCTRL_STATE_DEFAULT);
+	if (IS_ERR(dev->pins->default_state)) {
+		dev_dbg(dev, "no default pinctrl state\n");
+		ret = 0;
+		goto cleanup_get;
+	}
+
+	dev->pins->init_state = pinctrl_lookup_state(dev->pins->p,
+					PINCTRL_STATE_INIT);
+	if (IS_ERR(dev->pins->init_state)) {
+		/* Not supplying this state is perfectly legal */
+		dev_dbg(dev, "no init pinctrl state\n");
+
+		ret = pinctrl_select_state(dev->pins->p,
+					   dev->pins->default_state);
+	} else {
+		ret = pinctrl_select_state(dev->pins->p, dev->pins->init_state);
+	}
+
+	if (ret) {
+		dev_dbg(dev, "failed to activate initial pinctrl state\n");
+		goto cleanup_get;
+	}
+
+#ifdef CONFIG_PM
+	/*
+	 * If power management is enabled, we also look for the optional
+	 * sleep and idle pin states, with semantics as defined in
+	 * <linux/pinctrl/pinctrl-state.h>
+	 */
+	dev->pins->sleep_state = pinctrl_lookup_state(dev->pins->p,
+					PINCTRL_STATE_SLEEP);
+	if (IS_ERR(dev->pins->sleep_state))
+		/* Not supplying this state is perfectly legal */
+		dev_dbg(dev, "no sleep pinctrl state\n");
+
+	dev->pins->idle_state = pinctrl_lookup_state(dev->pins->p,
+					PINCTRL_STATE_IDLE);
+	if (IS_ERR(dev->pins->idle_state))
+		/* Not supplying this state is perfectly legal */
+		dev_dbg(dev, "no idle pinctrl state\n");
+#endif
+
+	return 0;
+
+	/*
+	 * If no pinctrl handle or default state was found for this device,
+	 * let's explicitly free the pin container in the device, there is
+	 * no point in keeping it around.
+	 */
+cleanup_get:
+	devm_pinctrl_put(dev->pins->p);
+cleanup_alloc:
+	devm_kfree(dev, dev->pins);
+	dev->pins = NULL;
+
+	/* Return deferrals */
+	if (ret == -EPROBE_DEFER)
+		return ret;
+	/* Return serious errors */
+	if (ret == -EINVAL)
+		return ret;
+	/* We ignore errors like -ENOENT meaning no pinctrl state */
+
+	return 0;
+}