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, 129 insertions, 0 deletions

diff --git a/drivers/base/regmap/regmap-sccb.c b/drivers/base/regmap/regmap-sccb.c
new file mode 100644
index 000000000..12bbbb03e
--- /dev/null
+++ b/drivers/base/regmap/regmap-sccb.c
@@ -0,0 +1,129 @@
+// SPDX-License-Identifier: GPL-2.0
+// Register map access API - SCCB support
+
+#include <linux/i2c.h>
+#include <linux/module.h>
+#include <linux/regmap.h>
+
+#include "internal.h"
+
+/**
+ * sccb_is_available - Check if the adapter supports SCCB protocol
+ * @adap: I2C adapter
+ *
+ * Return true if the I2C adapter is capable of using SCCB helper functions,
+ * false otherwise.
+ */
+static bool sccb_is_available(struct i2c_adapter *adap)
+{
+	u32 needed_funcs = I2C_FUNC_SMBUS_BYTE | I2C_FUNC_SMBUS_WRITE_BYTE_DATA;
+
+	/*
+	 * If we ever want support for hardware doing SCCB natively, we will
+	 * introduce a sccb_xfer() callback to struct i2c_algorithm and check
+	 * for it here.
+	 */
+
+	return (i2c_get_functionality(adap) & needed_funcs) == needed_funcs;
+}
+
+/**
+ * regmap_sccb_read - Read data from SCCB slave device
+ * @context: Device that will be interacted with
+ * @reg: Register to be read from
+ * @val: Pointer to store read value
+ *
+ * This executes the 2-phase write transmission cycle that is followed by a
+ * 2-phase read transmission cycle, returning negative errno else zero on
+ * success.
+ */
+static int regmap_sccb_read(void *context, unsigned int reg, unsigned int *val)
+{
+	struct device *dev = context;
+	struct i2c_client *i2c = to_i2c_client(dev);
+	int ret;
+	union i2c_smbus_data data;
+
+	i2c_lock_bus(i2c->adapter, I2C_LOCK_SEGMENT);
+
+	ret = __i2c_smbus_xfer(i2c->adapter, i2c->addr, i2c->flags,
+			       I2C_SMBUS_WRITE, reg, I2C_SMBUS_BYTE, NULL);
+	if (ret < 0)
+		goto out;
+
+	ret = __i2c_smbus_xfer(i2c->adapter, i2c->addr, i2c->flags,
+			       I2C_SMBUS_READ, 0, I2C_SMBUS_BYTE, &data);
+	if (ret < 0)
+		goto out;
+
+	*val = data.byte;
+out:
+	i2c_unlock_bus(i2c->adapter, I2C_LOCK_SEGMENT);
+
+	return ret;
+}
+
+/**
+ * regmap_sccb_write - Write data to SCCB slave device
+ * @context: Device that will be interacted with
+ * @reg: Register to write to
+ * @val: Value to be written
+ *
+ * This executes the SCCB 3-phase write transmission cycle, returning negative
+ * errno else zero on success.
+ */
+static int regmap_sccb_write(void *context, unsigned int reg, unsigned int val)
+{
+	struct device *dev = context;
+	struct i2c_client *i2c = to_i2c_client(dev);
+
+	return i2c_smbus_write_byte_data(i2c, reg, val);
+}
+
+static const struct regmap_bus regmap_sccb_bus = {
+	.reg_write = regmap_sccb_write,
+	.reg_read = regmap_sccb_read,
+};
+
+static const struct regmap_bus *regmap_get_sccb_bus(struct i2c_client *i2c,
+					const struct regmap_config *config)
+{
+	if (config->val_bits == 8 && config->reg_bits == 8 &&
+			sccb_is_available(i2c->adapter))
+		return &regmap_sccb_bus;
+
+	return ERR_PTR(-ENOTSUPP);
+}
+
+struct regmap *__regmap_init_sccb(struct i2c_client *i2c,
+				  const struct regmap_config *config,
+				  struct lock_class_key *lock_key,
+				  const char *lock_name)
+{
+	const struct regmap_bus *bus = regmap_get_sccb_bus(i2c, config);
+
+	if (IS_ERR(bus))
+		return ERR_CAST(bus);
+
+	return __regmap_init(&i2c->dev, bus, &i2c->dev, config,
+			     lock_key, lock_name);
+}
+EXPORT_SYMBOL_GPL(__regmap_init_sccb);
+
+struct regmap *__devm_regmap_init_sccb(struct i2c_client *i2c,
+				       const struct regmap_config *config,
+				       struct lock_class_key *lock_key,
+				       const char *lock_name)
+{
+	const struct regmap_bus *bus = regmap_get_sccb_bus(i2c, config);
+
+	if (IS_ERR(bus))
+		return ERR_CAST(bus);
+
+	return __devm_regmap_init(&i2c->dev, bus, &i2c->dev, config,
+				  lock_key, lock_name);
+}
+EXPORT_SYMBOL_GPL(__devm_regmap_init_sccb);
+
+MODULE_DESCRIPTION("Register map access API - SCCB support");
+MODULE_LICENSE("GPL v2");