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/net/ipv4/tcp_ulp.c b/net/ipv4/tcp_ulp.c
new file mode 100644
index 000000000..2aa442128
--- /dev/null
+++ b/net/ipv4/tcp_ulp.c
@@ -0,0 +1,168 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Pluggable TCP upper layer protocol support.
+ *
+ * Copyright (c) 2016-2017, Mellanox Technologies. All rights reserved.
+ * Copyright (c) 2016-2017, Dave Watson <davejwatson@fb.com>. All rights reserved.
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/mm.h>
+#include <linux/types.h>
+#include <linux/list.h>
+#include <linux/gfp.h>
+#include <net/tcp.h>
+
+static DEFINE_SPINLOCK(tcp_ulp_list_lock);
+static LIST_HEAD(tcp_ulp_list);
+
+/* Simple linear search, don't expect many entries! */
+static struct tcp_ulp_ops *tcp_ulp_find(const char *name)
+{
+	struct tcp_ulp_ops *e;
+
+	list_for_each_entry_rcu(e, &tcp_ulp_list, list,
+				lockdep_is_held(&tcp_ulp_list_lock)) {
+		if (strcmp(e->name, name) == 0)
+			return e;
+	}
+
+	return NULL;
+}
+
+static const struct tcp_ulp_ops *__tcp_ulp_find_autoload(const char *name)
+{
+	const struct tcp_ulp_ops *ulp = NULL;
+
+	rcu_read_lock();
+	ulp = tcp_ulp_find(name);
+
+#ifdef CONFIG_MODULES
+	if (!ulp && capable(CAP_NET_ADMIN)) {
+		rcu_read_unlock();
+		request_module("tcp-ulp-%s", name);
+		rcu_read_lock();
+		ulp = tcp_ulp_find(name);
+	}
+#endif
+	if (!ulp || !try_module_get(ulp->owner))
+		ulp = NULL;
+
+	rcu_read_unlock();
+	return ulp;
+}
+
+/* Attach new upper layer protocol to the list
+ * of available protocols.
+ */
+int tcp_register_ulp(struct tcp_ulp_ops *ulp)
+{
+	int ret = 0;
+
+	spin_lock(&tcp_ulp_list_lock);
+	if (tcp_ulp_find(ulp->name))
+		ret = -EEXIST;
+	else
+		list_add_tail_rcu(&ulp->list, &tcp_ulp_list);
+	spin_unlock(&tcp_ulp_list_lock);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(tcp_register_ulp);
+
+void tcp_unregister_ulp(struct tcp_ulp_ops *ulp)
+{
+	spin_lock(&tcp_ulp_list_lock);
+	list_del_rcu(&ulp->list);
+	spin_unlock(&tcp_ulp_list_lock);
+
+	synchronize_rcu();
+}
+EXPORT_SYMBOL_GPL(tcp_unregister_ulp);
+
+/* Build string with list of available upper layer protocl values */
+void tcp_get_available_ulp(char *buf, size_t maxlen)
+{
+	struct tcp_ulp_ops *ulp_ops;
+	size_t offs = 0;
+
+	*buf = '\0';
+	rcu_read_lock();
+	list_for_each_entry_rcu(ulp_ops, &tcp_ulp_list, list) {
+		offs += snprintf(buf + offs, maxlen - offs,
+				 "%s%s",
+				 offs == 0 ? "" : " ", ulp_ops->name);
+
+		if (WARN_ON_ONCE(offs >= maxlen))
+			break;
+	}
+	rcu_read_unlock();
+}
+
+void tcp_update_ulp(struct sock *sk, struct proto *proto,
+		    void (*write_space)(struct sock *sk))
+{
+	struct inet_connection_sock *icsk = inet_csk(sk);
+
+	if (icsk->icsk_ulp_ops->update)
+		icsk->icsk_ulp_ops->update(sk, proto, write_space);
+}
+
+void tcp_cleanup_ulp(struct sock *sk)
+{
+	struct inet_connection_sock *icsk = inet_csk(sk);
+
+	/* No sock_owned_by_me() check here as at the time the
+	 * stack calls this function, the socket is dead and
+	 * about to be destroyed.
+	 */
+	if (!icsk->icsk_ulp_ops)
+		return;
+
+	if (icsk->icsk_ulp_ops->release)
+		icsk->icsk_ulp_ops->release(sk);
+	module_put(icsk->icsk_ulp_ops->owner);
+
+	icsk->icsk_ulp_ops = NULL;
+}
+
+static int __tcp_set_ulp(struct sock *sk, const struct tcp_ulp_ops *ulp_ops)
+{
+	struct inet_connection_sock *icsk = inet_csk(sk);
+	int err;
+
+	err = -EEXIST;
+	if (icsk->icsk_ulp_ops)
+		goto out_err;
+
+	if (sk->sk_socket)
+		clear_bit(SOCK_SUPPORT_ZC, &sk->sk_socket->flags);
+
+	err = -ENOTCONN;
+	if (!ulp_ops->clone && sk->sk_state == TCP_LISTEN)
+		goto out_err;
+
+	err = ulp_ops->init(sk);
+	if (err)
+		goto out_err;
+
+	icsk->icsk_ulp_ops = ulp_ops;
+	return 0;
+out_err:
+	module_put(ulp_ops->owner);
+	return err;
+}
+
+int tcp_set_ulp(struct sock *sk, const char *name)
+{
+	const struct tcp_ulp_ops *ulp_ops;
+
+	sock_owned_by_me(sk);
+
+	ulp_ops = __tcp_ulp_find_autoload(name);
+	if (!ulp_ops)
+		return -ENOENT;
+
+	return __tcp_set_ulp(sk, ulp_ops);
+}