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, 85 insertions, 0 deletions

diff --git a/drivers/scsi/scsi_logging.h b/drivers/scsi/scsi_logging.h
new file mode 100644
index 000000000..3df877886
--- /dev/null
+++ b/drivers/scsi/scsi_logging.h
@@ -0,0 +1,85 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _SCSI_LOGGING_H
+#define _SCSI_LOGGING_H
+
+
+/*
+ * This defines the scsi logging feature.  It is a means by which the user can
+ * select how much information they get about various goings on, and it can be
+ * really useful for fault tracing.  The logging word is divided into 10 3-bit
+ * bitfields, each of which describes a loglevel.  The division of things is
+ * somewhat arbitrary, and the division of the word could be changed if it
+ * were really needed for any reason.  The numbers below are the only place
+ * where these are specified.  For a first go-around, 3 bits is more than
+ * enough, since this gives 8 levels of logging (really 7, since 0 is always
+ * off).  Cutting to 2 bits might be wise at some point.
+ */
+
+#define SCSI_LOG_ERROR_SHIFT              0
+#define SCSI_LOG_TIMEOUT_SHIFT            3
+#define SCSI_LOG_SCAN_SHIFT               6
+#define SCSI_LOG_MLQUEUE_SHIFT            9
+#define SCSI_LOG_MLCOMPLETE_SHIFT         12
+#define SCSI_LOG_LLQUEUE_SHIFT            15
+#define SCSI_LOG_LLCOMPLETE_SHIFT         18
+#define SCSI_LOG_HLQUEUE_SHIFT            21
+#define SCSI_LOG_HLCOMPLETE_SHIFT         24
+#define SCSI_LOG_IOCTL_SHIFT              27
+
+#define SCSI_LOG_ERROR_BITS               3
+#define SCSI_LOG_TIMEOUT_BITS             3
+#define SCSI_LOG_SCAN_BITS                3
+#define SCSI_LOG_MLQUEUE_BITS             3
+#define SCSI_LOG_MLCOMPLETE_BITS          3
+#define SCSI_LOG_LLQUEUE_BITS             3
+#define SCSI_LOG_LLCOMPLETE_BITS          3
+#define SCSI_LOG_HLQUEUE_BITS             3
+#define SCSI_LOG_HLCOMPLETE_BITS          3
+#define SCSI_LOG_IOCTL_BITS               3
+
+extern unsigned int scsi_logging_level;
+
+#ifdef CONFIG_SCSI_LOGGING
+
+#define SCSI_LOG_LEVEL(SHIFT, BITS)				\
+        ((scsi_logging_level >> (SHIFT)) & ((1 << (BITS)) - 1))
+
+#define SCSI_CHECK_LOGGING(SHIFT, BITS, LEVEL, CMD)		\
+do {								\
+        if (unlikely((SCSI_LOG_LEVEL(SHIFT, BITS)) > (LEVEL)))	\
+		do {						\
+			CMD;					\
+		} while (0);					\
+} while (0)
+#else
+#define SCSI_LOG_LEVEL(SHIFT, BITS) 0
+#define SCSI_CHECK_LOGGING(SHIFT, BITS, LEVEL, CMD) do { } while (0)
+#endif /* CONFIG_SCSI_LOGGING */
+
+/*
+ * These are the macros that are actually used throughout the code to
+ * log events.  If logging isn't enabled, they are no-ops and will be
+ * completely absent from the user's code.
+ */
+#define SCSI_LOG_ERROR_RECOVERY(LEVEL,CMD)  \
+        SCSI_CHECK_LOGGING(SCSI_LOG_ERROR_SHIFT, SCSI_LOG_ERROR_BITS, LEVEL,CMD);
+#define SCSI_LOG_TIMEOUT(LEVEL,CMD)  \
+        SCSI_CHECK_LOGGING(SCSI_LOG_TIMEOUT_SHIFT, SCSI_LOG_TIMEOUT_BITS, LEVEL,CMD);
+#define SCSI_LOG_SCAN_BUS(LEVEL,CMD)  \
+        SCSI_CHECK_LOGGING(SCSI_LOG_SCAN_SHIFT, SCSI_LOG_SCAN_BITS, LEVEL,CMD);
+#define SCSI_LOG_MLQUEUE(LEVEL,CMD)  \
+        SCSI_CHECK_LOGGING(SCSI_LOG_MLQUEUE_SHIFT, SCSI_LOG_MLQUEUE_BITS, LEVEL,CMD);
+#define SCSI_LOG_MLCOMPLETE(LEVEL,CMD)  \
+        SCSI_CHECK_LOGGING(SCSI_LOG_MLCOMPLETE_SHIFT, SCSI_LOG_MLCOMPLETE_BITS, LEVEL,CMD);
+#define SCSI_LOG_LLQUEUE(LEVEL,CMD)  \
+        SCSI_CHECK_LOGGING(SCSI_LOG_LLQUEUE_SHIFT, SCSI_LOG_LLQUEUE_BITS, LEVEL,CMD);
+#define SCSI_LOG_LLCOMPLETE(LEVEL,CMD)  \
+        SCSI_CHECK_LOGGING(SCSI_LOG_LLCOMPLETE_SHIFT, SCSI_LOG_LLCOMPLETE_BITS, LEVEL,CMD);
+#define SCSI_LOG_HLQUEUE(LEVEL,CMD)  \
+        SCSI_CHECK_LOGGING(SCSI_LOG_HLQUEUE_SHIFT, SCSI_LOG_HLQUEUE_BITS, LEVEL,CMD);
+#define SCSI_LOG_HLCOMPLETE(LEVEL,CMD)  \
+        SCSI_CHECK_LOGGING(SCSI_LOG_HLCOMPLETE_SHIFT, SCSI_LOG_HLCOMPLETE_BITS, LEVEL,CMD);
+#define SCSI_LOG_IOCTL(LEVEL,CMD)  \
+        SCSI_CHECK_LOGGING(SCSI_LOG_IOCTL_SHIFT, SCSI_LOG_IOCTL_BITS, LEVEL,CMD);
+
+#endif /* _SCSI_LOGGING_H */