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, 124 insertions, 0 deletions

diff --git a/Documentation/admin-guide/btmrvl.rst b/Documentation/admin-guide/btmrvl.rst
new file mode 100644
index 000000000..ec57740ea
--- /dev/null
+++ b/Documentation/admin-guide/btmrvl.rst
@@ -0,0 +1,124 @@
+=============
+btmrvl driver
+=============
+
+All commands are used via debugfs interface.
+
+Set/get driver configurations
+=============================
+
+Path:	/debug/btmrvl/config/
+
+gpiogap=[n], hscfgcmd
+	These commands are used to configure the host sleep parameters::
+	bit 8:0  -- Gap
+	bit 16:8 -- GPIO
+
+	where GPIO is the pin number of GPIO used to wake up the host.
+	It could be any valid GPIO pin# (e.g. 0-7) or 0xff (SDIO interface
+	wakeup will be used instead).
+
+	where Gap is the gap in milli seconds between wakeup signal and
+	wakeup event, or 0xff for special host sleep setting.
+
+	Usage::
+
+		# Use SDIO interface to wake up the host and set GAP to 0x80:
+		echo 0xff80 > /debug/btmrvl/config/gpiogap
+		echo 1 > /debug/btmrvl/config/hscfgcmd
+
+		# Use GPIO pin #3 to wake up the host and set GAP to 0xff:
+		echo 0x03ff >  /debug/btmrvl/config/gpiogap
+		echo 1 > /debug/btmrvl/config/hscfgcmd
+
+psmode=[n], pscmd
+	These commands are used to enable/disable auto sleep mode
+
+	where the option is::
+
+			1 	-- Enable auto sleep mode
+			0 	-- Disable auto sleep mode
+
+	Usage::
+
+		# Enable auto sleep mode
+		echo 1 > /debug/btmrvl/config/psmode
+		echo 1 > /debug/btmrvl/config/pscmd
+
+		# Disable auto sleep mode
+		echo 0 > /debug/btmrvl/config/psmode
+		echo 1 > /debug/btmrvl/config/pscmd
+
+
+hsmode=[n], hscmd
+	These commands are used to enable host sleep or wake up firmware
+
+	where the option is::
+
+			1	-- Enable host sleep
+			0	-- Wake up firmware
+
+	Usage::
+
+		# Enable host sleep
+		echo 1 > /debug/btmrvl/config/hsmode
+		echo 1 > /debug/btmrvl/config/hscmd
+
+		# Wake up firmware
+		echo 0 > /debug/btmrvl/config/hsmode
+		echo 1 > /debug/btmrvl/config/hscmd
+
+
+Get driver status
+=================
+
+Path:	/debug/btmrvl/status/
+
+Usage::
+
+	cat /debug/btmrvl/status/<args>
+
+where the args are:
+
+curpsmode
+	This command displays current auto sleep status.
+
+psstate
+	This command display the power save state.
+
+hsstate
+	This command display the host sleep state.
+
+txdnldrdy
+	This command displays the value of Tx download ready flag.
+
+Issuing a raw hci command
+=========================
+
+Use hcitool to issue raw hci command, refer to hcitool manual
+
+Usage::
+
+	Hcitool cmd <ogf> <ocf> [Parameters]
+
+Interface Control Command::
+
+	hcitool cmd 0x3f 0x5b 0xf5 0x01 0x00    --Enable All interface
+	hcitool cmd 0x3f 0x5b 0xf5 0x01 0x01    --Enable Wlan interface
+	hcitool cmd 0x3f 0x5b 0xf5 0x01 0x02    --Enable BT interface
+	hcitool cmd 0x3f 0x5b 0xf5 0x00 0x00    --Disable All interface
+	hcitool cmd 0x3f 0x5b 0xf5 0x00 0x01    --Disable Wlan interface
+	hcitool cmd 0x3f 0x5b 0xf5 0x00 0x02    --Disable BT interface
+
+SD8688 firmware
+===============
+
+Images:
+
+- /lib/firmware/sd8688_helper.bin
+- /lib/firmware/sd8688.bin
+
+
+The images can be downloaded from:
+
+git.infradead.org/users/dwmw2/linux-firmware.git/libertas/