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, 125 insertions, 0 deletions

diff --git a/drivers/tty/tty.h b/drivers/tty/tty.h
new file mode 100644
index 000000000..93cf5ef1e
--- /dev/null
+++ b/drivers/tty/tty.h
@@ -0,0 +1,125 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * TTY core internal functions
+ */
+
+#ifndef _TTY_INTERNAL_H
+#define _TTY_INTERNAL_H
+
+#define tty_msg(fn, tty, f, ...) \
+	fn("%s %s: " f, tty_driver_name(tty), tty_name(tty), ##__VA_ARGS__)
+
+#define tty_debug(tty, f, ...)	tty_msg(pr_debug, tty, f, ##__VA_ARGS__)
+#define tty_notice(tty, f, ...)	tty_msg(pr_notice, tty, f, ##__VA_ARGS__)
+#define tty_warn(tty, f, ...)	tty_msg(pr_warn, tty, f, ##__VA_ARGS__)
+#define tty_err(tty, f, ...)	tty_msg(pr_err, tty, f, ##__VA_ARGS__)
+
+#define tty_info_ratelimited(tty, f, ...) \
+		tty_msg(pr_info_ratelimited, tty, f, ##__VA_ARGS__)
+
+/*
+ * Lock subclasses for tty locks
+ *
+ * TTY_LOCK_NORMAL is for normal ttys and master ptys.
+ * TTY_LOCK_SLAVE is for slave ptys only.
+ *
+ * Lock subclasses are necessary for handling nested locking with pty pairs.
+ * tty locks which use nested locking:
+ *
+ * legacy_mutex - Nested tty locks are necessary for releasing pty pairs.
+ *		  The stable lock order is master pty first, then slave pty.
+ * termios_rwsem - The stable lock order is tty_buffer lock->termios_rwsem.
+ *		   Subclassing this lock enables the slave pty to hold its
+ *		   termios_rwsem when claiming the master tty_buffer lock.
+ * tty_buffer lock - slave ptys can claim nested buffer lock when handling
+ *		     signal chars. The stable lock order is slave pty, then
+ *		     master.
+ */
+enum {
+	TTY_LOCK_NORMAL = 0,
+	TTY_LOCK_SLAVE,
+};
+
+/* Values for tty->flow_change */
+enum tty_flow_change {
+	TTY_FLOW_NO_CHANGE,
+	TTY_THROTTLE_SAFE,
+	TTY_UNTHROTTLE_SAFE,
+};
+
+static inline void __tty_set_flow_change(struct tty_struct *tty,
+					 enum tty_flow_change val)
+{
+	tty->flow_change = val;
+}
+
+static inline void tty_set_flow_change(struct tty_struct *tty,
+				       enum tty_flow_change val)
+{
+	tty->flow_change = val;
+	smp_mb();
+}
+
+int tty_ldisc_lock(struct tty_struct *tty, unsigned long timeout);
+void tty_ldisc_unlock(struct tty_struct *tty);
+
+int __tty_check_change(struct tty_struct *tty, int sig);
+int tty_check_change(struct tty_struct *tty);
+void __stop_tty(struct tty_struct *tty);
+void __start_tty(struct tty_struct *tty);
+void tty_write_unlock(struct tty_struct *tty);
+int tty_write_lock(struct tty_struct *tty, bool ndelay);
+void tty_vhangup_session(struct tty_struct *tty);
+void tty_open_proc_set_tty(struct file *filp, struct tty_struct *tty);
+int tty_signal_session_leader(struct tty_struct *tty, int exit_session);
+void session_clear_tty(struct pid *session);
+void tty_buffer_free_all(struct tty_port *port);
+void tty_buffer_flush(struct tty_struct *tty, struct tty_ldisc *ld);
+void tty_buffer_init(struct tty_port *port);
+void tty_buffer_set_lock_subclass(struct tty_port *port);
+bool tty_buffer_restart_work(struct tty_port *port);
+bool tty_buffer_cancel_work(struct tty_port *port);
+void tty_buffer_flush_work(struct tty_port *port);
+speed_t tty_termios_input_baud_rate(const struct ktermios *termios);
+void tty_ldisc_hangup(struct tty_struct *tty, bool reset);
+int tty_ldisc_reinit(struct tty_struct *tty, int disc);
+long tty_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
+long tty_jobctrl_ioctl(struct tty_struct *tty, struct tty_struct *real_tty,
+		       struct file *file, unsigned int cmd, unsigned long arg);
+void tty_default_fops(struct file_operations *fops);
+struct tty_struct *alloc_tty_struct(struct tty_driver *driver, int idx);
+int tty_alloc_file(struct file *file);
+void tty_add_file(struct tty_struct *tty, struct file *file);
+void tty_free_file(struct file *file);
+int tty_release(struct inode *inode, struct file *filp);
+
+#define tty_is_writelocked(tty)  (mutex_is_locked(&tty->atomic_write_lock))
+
+int tty_ldisc_setup(struct tty_struct *tty, struct tty_struct *o_tty);
+void tty_ldisc_release(struct tty_struct *tty);
+int __must_check tty_ldisc_init(struct tty_struct *tty);
+void tty_ldisc_deinit(struct tty_struct *tty);
+
+extern int tty_ldisc_autoload;
+
+/* tty_audit.c */
+#ifdef CONFIG_AUDIT
+void tty_audit_add_data(const struct tty_struct *tty, const void *data,
+			size_t size);
+void tty_audit_tiocsti(const struct tty_struct *tty, u8 ch);
+#else
+static inline void tty_audit_add_data(const struct tty_struct *tty,
+				      const void *data, size_t size)
+{
+}
+static inline void tty_audit_tiocsti(const struct tty_struct *tty, u8 ch)
+{
+}
+#endif
+
+ssize_t redirected_tty_write(struct kiocb *, struct iov_iter *);
+
+int tty_insert_flip_string_and_push_buffer(struct tty_port *port,
+					   const u8 *chars, size_t cnt);
+
+#endif