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, 55 insertions, 0 deletions

diff --git a/Documentation/i2c/old-module-parameters.rst b/Documentation/i2c/old-module-parameters.rst
new file mode 100644
index 000000000..b08b6daab
--- /dev/null
+++ b/Documentation/i2c/old-module-parameters.rst
@@ -0,0 +1,55 @@
+================================================================
+I2C device driver binding control from user-space in old kernels
+================================================================
+
+.. NOTE::
+   Note: this section is only relevant if you are handling some old code
+   found in kernel 2.6. If you work with more recent kernels, you can
+   safely skip this section.
+
+Up to kernel 2.6.32, many I2C drivers used helper macros provided by
+<linux/i2c.h> which created standard module parameters to let the user
+control how the driver would probe I2C buses and attach to devices. These
+parameters were known as ``probe`` (to let the driver probe for an extra
+address), ``force`` (to forcibly attach the driver to a given device) and
+``ignore`` (to prevent a driver from probing a given address).
+
+With the conversion of the I2C subsystem to the standard device driver
+binding model, it became clear that these per-module parameters were no
+longer needed, and that a centralized implementation was possible. The new,
+sysfs-based interface is described in
+Documentation/i2c/instantiating-devices.rst, section
+"Method 4: Instantiate from user-space".
+
+Below is a mapping from the old module parameters to the new interface.
+
+Attaching a driver to an I2C device
+-----------------------------------
+
+Old method (module parameters)::
+
+  # modprobe <driver> probe=1,0x2d
+  # modprobe <driver> force=1,0x2d
+  # modprobe <driver> force_<device>=1,0x2d
+
+New method (sysfs interface)::
+
+  # echo <device> 0x2d > /sys/bus/i2c/devices/i2c-1/new_device
+
+Preventing a driver from attaching to an I2C device
+---------------------------------------------------
+
+Old method (module parameters)::
+
+  # modprobe <driver> ignore=1,0x2f
+
+New method (sysfs interface)::
+
+  # echo dummy 0x2f > /sys/bus/i2c/devices/i2c-1/new_device
+  # modprobe <driver>
+
+Of course, it is important to instantiate the ``dummy`` device before loading
+the driver. The dummy device will be handled by i2c-core itself, preventing
+other drivers from binding to it later on. If there is a real device at the
+problematic address, and you want another driver to bind to it, then simply
+pass the name of the device in question instead of ``dummy``.