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, 27 insertions, 0 deletions

diff --git a/Documentation/admin-guide/edid.rst b/Documentation/admin-guide/edid.rst
new file mode 100644
index 000000000..1a9b965aa
--- /dev/null
+++ b/Documentation/admin-guide/edid.rst
@@ -0,0 +1,27 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+====
+EDID
+====
+
+In the good old days when graphics parameters were configured explicitly
+in a file called xorg.conf, even broken hardware could be managed.
+
+Today, with the advent of Kernel Mode Setting, a graphics board is
+either correctly working because all components follow the standards -
+or the computer is unusable, because the screen remains dark after
+booting or it displays the wrong area. Cases when this happens are:
+
+- The graphics board does not recognize the monitor.
+- The graphics board is unable to detect any EDID data.
+- The graphics board incorrectly forwards EDID data to the driver.
+- The monitor sends no or bogus EDID data.
+- A KVM sends its own EDID data instead of querying the connected monitor.
+
+Adding the kernel parameter "nomodeset" helps in most cases, but causes
+restrictions later on.
+
+As a remedy for such situations, the kernel configuration item
+CONFIG_DRM_LOAD_EDID_FIRMWARE was introduced. It allows to provide an
+individually prepared or corrected EDID data set in the /lib/firmware
+directory from where it is loaded via the firmware interface.