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, 168 insertions, 0 deletions

diff --git a/drivers/scsi/arm/msgqueue.c b/drivers/scsi/arm/msgqueue.c
new file mode 100644
index 000000000..581158313
--- /dev/null
+++ b/drivers/scsi/arm/msgqueue.c
@@ -0,0 +1,168 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  linux/drivers/acorn/scsi/msgqueue.c
+ *
+ *  Copyright (C) 1997-1998 Russell King
+ *
+ *  message queue handling
+ */
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/stddef.h>
+#include <linux/init.h>
+
+#include "msgqueue.h"
+
+/*
+ * Function: struct msgqueue_entry *mqe_alloc(MsgQueue_t *msgq)
+ * Purpose : Allocate a message queue entry
+ * Params  : msgq - message queue to claim entry for
+ * Returns : message queue entry or NULL.
+ */
+static struct msgqueue_entry *mqe_alloc(MsgQueue_t *msgq)
+{
+	struct msgqueue_entry *mq;
+
+	if ((mq = msgq->free) != NULL)
+		msgq->free = mq->next;
+
+	return mq;
+}
+
+/*
+ * Function: void mqe_free(MsgQueue_t *msgq, struct msgqueue_entry *mq)
+ * Purpose : free a message queue entry
+ * Params  : msgq - message queue to free entry from
+ *	     mq   - message queue entry to free
+ */
+static void mqe_free(MsgQueue_t *msgq, struct msgqueue_entry *mq)
+{
+	if (mq) {
+		mq->next = msgq->free;
+		msgq->free = mq;
+	}
+}
+
+/*
+ * Function: void msgqueue_initialise(MsgQueue_t *msgq)
+ * Purpose : initialise a message queue
+ * Params  : msgq - queue to initialise
+ */
+void msgqueue_initialise(MsgQueue_t *msgq)
+{
+	int i;
+
+	msgq->qe = NULL;
+	msgq->free = &msgq->entries[0];
+
+	for (i = 0; i < NR_MESSAGES; i++)
+		msgq->entries[i].next = &msgq->entries[i + 1];
+
+	msgq->entries[NR_MESSAGES - 1].next = NULL;
+}
+
+
+/*
+ * Function: void msgqueue_free(MsgQueue_t *msgq)
+ * Purpose : free a queue
+ * Params  : msgq - queue to free
+ */
+void msgqueue_free(MsgQueue_t *msgq)
+{
+}
+
+/*
+ * Function: int msgqueue_msglength(MsgQueue_t *msgq)
+ * Purpose : calculate the total length of all messages on the message queue
+ * Params  : msgq - queue to examine
+ * Returns : number of bytes of messages in queue
+ */
+int msgqueue_msglength(MsgQueue_t *msgq)
+{
+	struct msgqueue_entry *mq = msgq->qe;
+	int length = 0;
+
+	for (mq = msgq->qe; mq; mq = mq->next)
+		length += mq->msg.length;
+
+	return length;
+}
+
+/*
+ * Function: struct message *msgqueue_getmsg(MsgQueue_t *msgq, int msgno)
+ * Purpose : return a message
+ * Params  : msgq   - queue to obtain message from
+ *	   : msgno  - message number
+ * Returns : pointer to message string, or NULL
+ */
+struct message *msgqueue_getmsg(MsgQueue_t *msgq, int msgno)
+{
+	struct msgqueue_entry *mq;
+
+	for (mq = msgq->qe; mq && msgno; mq = mq->next, msgno--);
+
+	return mq ? &mq->msg : NULL;
+}
+
+/*
+ * Function: int msgqueue_addmsg(MsgQueue_t *msgq, int length, ...)
+ * Purpose : add a message onto a message queue
+ * Params  : msgq   - queue to add message on
+ *	     length - length of message
+ *	     ...    - message bytes
+ * Returns : != 0 if successful
+ */
+int msgqueue_addmsg(MsgQueue_t *msgq, int length, ...)
+{
+	struct msgqueue_entry *mq = mqe_alloc(msgq);
+	va_list ap;
+
+	if (mq) {
+		struct msgqueue_entry **mqp;
+		int i;
+
+		va_start(ap, length);
+		for (i = 0; i < length; i++)
+			mq->msg.msg[i] = va_arg(ap, unsigned int);
+		va_end(ap);
+
+		mq->msg.length = length;
+		mq->msg.fifo = 0;
+		mq->next = NULL;
+
+		mqp = &msgq->qe;
+		while (*mqp)
+			mqp = &(*mqp)->next;
+
+		*mqp = mq;
+	}
+
+	return mq != NULL;
+}
+
+/*
+ * Function: void msgqueue_flush(MsgQueue_t *msgq)
+ * Purpose : flush all messages from message queue
+ * Params  : msgq - queue to flush
+ */
+void msgqueue_flush(MsgQueue_t *msgq)
+{
+	struct msgqueue_entry *mq, *mqnext;
+
+	for (mq = msgq->qe; mq; mq = mqnext) {
+		mqnext = mq->next;
+		mqe_free(msgq, mq);
+	}
+	msgq->qe = NULL;
+}
+
+EXPORT_SYMBOL(msgqueue_initialise);
+EXPORT_SYMBOL(msgqueue_free);
+EXPORT_SYMBOL(msgqueue_msglength);
+EXPORT_SYMBOL(msgqueue_getmsg);
+EXPORT_SYMBOL(msgqueue_addmsg);
+EXPORT_SYMBOL(msgqueue_flush);
+
+MODULE_AUTHOR("Russell King");
+MODULE_DESCRIPTION("SCSI message queue handling");
+MODULE_LICENSE("GPL");