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, 75 insertions, 0 deletions

diff --git a/drivers/eisa/Kconfig b/drivers/eisa/Kconfig
new file mode 100644
index 000000000..a66b3be50
--- /dev/null
+++ b/drivers/eisa/Kconfig
@@ -0,0 +1,75 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# EISA configuration
+#
+
+config HAVE_EISA
+	bool
+
+menuconfig EISA
+	bool "EISA support"
+	depends on HAVE_EISA
+	help
+	  The Extended Industry Standard Architecture (EISA) bus was
+	  developed as an open alternative to the IBM MicroChannel bus.
+
+	  The EISA bus provided some of the features of the IBM MicroChannel
+	  bus while maintaining backward compatibility with cards made for
+	  the older ISA bus.  The EISA bus saw limited use between 1988 and
+	  1995 when it was made obsolete by the PCI bus.
+
+	  Say Y here if you are building a kernel for an EISA-based machine.
+
+	  Otherwise, say N.
+
+config EISA_VLB_PRIMING
+	bool "Vesa Local Bus priming"
+	depends on X86 && EISA
+	default n
+	help
+	  Activate this option if your system contains a Vesa Local
+	  Bus (VLB) card that identify itself as an EISA card (such as
+	  the Adaptec AHA-284x).
+
+	  When in doubt, say N.
+
+config EISA_PCI_EISA
+	bool "Generic PCI/EISA bridge"
+	depends on !PARISC && PCI && EISA
+	default y
+	help
+	  Activate this option if your system contains a PCI to EISA
+	  bridge. If your system have both PCI and EISA slots, you
+	  certainly need this option.
+
+	  When in doubt, say Y.
+
+# Using EISA_VIRTUAL_ROOT on something other than an X86 may lead
+# to crashes...
+
+config EISA_VIRTUAL_ROOT
+	bool "EISA virtual root device"
+	depends on EISA && X86
+	default y
+	help
+	  Activate this option if your system only have EISA bus
+	  (no PCI slots).
+
+	  When in doubt, say Y.
+
+config EISA_NAMES
+	bool "EISA device name database"
+	depends on EISA
+	default y
+	help
+	  By default, the kernel contains a database of all known EISA
+	  device names to make the information in sysfs comprehensible
+	  to the user. This database increases size of the kernel
+	  image by about 40KB, but it gets freed after the system
+	  boots up, so it doesn't take up kernel memory. Anyway, if
+	  you are building an installation floppy or kernel for an
+	  embedded system where kernel image size really matters, you
+	  can disable this feature and you'll get device ID instead of
+	  names.
+
+	  When in doubt, say Y.