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, 88 insertions, 0 deletions

diff --git a/include/net/cls_cgroup.h b/include/net/cls_cgroup.h
new file mode 100644
index 000000000..668aeee9b
--- /dev/null
+++ b/include/net/cls_cgroup.h
@@ -0,0 +1,88 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * cls_cgroup.h			Control Group Classifier
+ *
+ * Authors:	Thomas Graf <tgraf@suug.ch>
+ */
+
+#ifndef _NET_CLS_CGROUP_H
+#define _NET_CLS_CGROUP_H
+
+#include <linux/cgroup.h>
+#include <linux/hardirq.h>
+#include <linux/rcupdate.h>
+#include <net/sock.h>
+#include <net/inet_sock.h>
+
+#ifdef CONFIG_CGROUP_NET_CLASSID
+struct cgroup_cls_state {
+	struct cgroup_subsys_state css;
+	u32 classid;
+};
+
+struct cgroup_cls_state *task_cls_state(struct task_struct *p);
+
+static inline u32 task_cls_classid(struct task_struct *p)
+{
+	u32 classid;
+
+	if (in_interrupt())
+		return 0;
+
+	rcu_read_lock();
+	classid = container_of(task_css(p, net_cls_cgrp_id),
+			       struct cgroup_cls_state, css)->classid;
+	rcu_read_unlock();
+
+	return classid;
+}
+
+static inline void sock_update_classid(struct sock_cgroup_data *skcd)
+{
+	u32 classid;
+
+	classid = task_cls_classid(current);
+	sock_cgroup_set_classid(skcd, classid);
+}
+
+static inline u32 __task_get_classid(struct task_struct *task)
+{
+	return task_cls_state(task)->classid;
+}
+
+static inline u32 task_get_classid(const struct sk_buff *skb)
+{
+	u32 classid = __task_get_classid(current);
+
+	/* Due to the nature of the classifier it is required to ignore all
+	 * packets originating from softirq context as accessing `current'
+	 * would lead to false results.
+	 *
+	 * This test assumes that all callers of dev_queue_xmit() explicitly
+	 * disable bh. Knowing this, it is possible to detect softirq based
+	 * calls by looking at the number of nested bh disable calls because
+	 * softirqs always disables bh.
+	 */
+	if (softirq_count()) {
+		struct sock *sk = skb_to_full_sk(skb);
+
+		/* If there is an sock_cgroup_classid we'll use that. */
+		if (!sk || !sk_fullsock(sk))
+			return 0;
+
+		classid = sock_cgroup_classid(&sk->sk_cgrp_data);
+	}
+
+	return classid;
+}
+#else /* !CONFIG_CGROUP_NET_CLASSID */
+static inline void sock_update_classid(struct sock_cgroup_data *skcd)
+{
+}
+
+static inline u32 task_get_classid(const struct sk_buff *skb)
+{
+	return 0;
+}
+#endif /* CONFIG_CGROUP_NET_CLASSID */
+#endif  /* _NET_CLS_CGROUP_H */