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, 53 insertions, 0 deletions

diff --git a/Documentation/usb/dwc3.rst b/Documentation/usb/dwc3.rst
new file mode 100644
index 000000000..f94a7ba16
--- /dev/null
+++ b/Documentation/usb/dwc3.rst
@@ -0,0 +1,53 @@
+===========
+DWC3 driver
+===========
+
+
+TODO
+~~~~
+
+Please pick something while reading :)
+
+- Convert interrupt handler to per-ep-thread-irq
+
+  As it turns out some DWC3-commands ~1ms to complete. Currently we spin
+  until the command completes which is bad.
+
+  Implementation idea:
+
+  - dwc core implements a demultiplexing irq chip for interrupts per
+    endpoint. The interrupt numbers are allocated during probe and belong
+    to the device. If MSI provides per-endpoint interrupt this dummy
+    interrupt chip can be replaced with "real" interrupts.
+  - interrupts are requested / allocated on usb_ep_enable() and removed on
+    usb_ep_disable(). Worst case are 32 interrupts, the lower limit is two
+    for ep0/1.
+  - dwc3_send_gadget_ep_cmd() will sleep in wait_for_completion_timeout()
+    until the command completes.
+  - the interrupt handler is split into the following pieces:
+
+    - primary handler of the device
+      goes through every event and calls generic_handle_irq() for event
+      it. On return from generic_handle_irq() in acknowledges the event
+      counter so interrupt goes away (eventually).
+
+    - threaded handler of the device
+      none
+
+    - primary handler of the EP-interrupt
+      reads the event and tries to process it. Everything that requires
+      sleeping is handed over to the Thread. The event is saved in an
+      per-endpoint data-structure.
+      We probably have to pay attention not to process events once we
+      handed something to thread so we don't process event X prio Y
+      where X > Y.
+
+    - threaded handler of the EP-interrupt
+      handles the remaining EP work which might sleep such as waiting
+      for command completion.
+
+  Latency:
+
+   There should be no increase in latency since the interrupt-thread has a
+   high priority and will be run before an average task in user land
+   (except the user changed priorities).