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, 306 insertions, 0 deletions

diff --git a/net/ipv6/tunnel6.c b/net/ipv6/tunnel6.c
new file mode 100644
index 000000000..dc4ea9b11
--- /dev/null
+++ b/net/ipv6/tunnel6.c
@@ -0,0 +1,306 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (C)2003,2004 USAGI/WIDE Project
+ *
+ * Authors	Mitsuru KANDA  <mk@linux-ipv6.org>
+ *		YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
+ */
+
+#define pr_fmt(fmt) "IPv6: " fmt
+
+#include <linux/icmpv6.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/netdevice.h>
+#include <linux/skbuff.h>
+#include <linux/slab.h>
+#include <net/ipv6.h>
+#include <net/protocol.h>
+#include <net/xfrm.h>
+
+static struct xfrm6_tunnel __rcu *tunnel6_handlers __read_mostly;
+static struct xfrm6_tunnel __rcu *tunnel46_handlers __read_mostly;
+static struct xfrm6_tunnel __rcu *tunnelmpls6_handlers __read_mostly;
+static DEFINE_MUTEX(tunnel6_mutex);
+
+static inline int xfrm6_tunnel_mpls_supported(void)
+{
+	return IS_ENABLED(CONFIG_MPLS);
+}
+
+int xfrm6_tunnel_register(struct xfrm6_tunnel *handler, unsigned short family)
+{
+	struct xfrm6_tunnel __rcu **pprev;
+	struct xfrm6_tunnel *t;
+	int ret = -EEXIST;
+	int priority = handler->priority;
+
+	mutex_lock(&tunnel6_mutex);
+
+	switch (family) {
+	case AF_INET6:
+		pprev = &tunnel6_handlers;
+		break;
+	case AF_INET:
+		pprev = &tunnel46_handlers;
+		break;
+	case AF_MPLS:
+		pprev = &tunnelmpls6_handlers;
+		break;
+	default:
+		goto err;
+	}
+
+	for (; (t = rcu_dereference_protected(*pprev,
+			lockdep_is_held(&tunnel6_mutex))) != NULL;
+	     pprev = &t->next) {
+		if (t->priority > priority)
+			break;
+		if (t->priority == priority)
+			goto err;
+	}
+
+	handler->next = *pprev;
+	rcu_assign_pointer(*pprev, handler);
+
+	ret = 0;
+
+err:
+	mutex_unlock(&tunnel6_mutex);
+
+	return ret;
+}
+EXPORT_SYMBOL(xfrm6_tunnel_register);
+
+int xfrm6_tunnel_deregister(struct xfrm6_tunnel *handler, unsigned short family)
+{
+	struct xfrm6_tunnel __rcu **pprev;
+	struct xfrm6_tunnel *t;
+	int ret = -ENOENT;
+
+	mutex_lock(&tunnel6_mutex);
+
+	switch (family) {
+	case AF_INET6:
+		pprev = &tunnel6_handlers;
+		break;
+	case AF_INET:
+		pprev = &tunnel46_handlers;
+		break;
+	case AF_MPLS:
+		pprev = &tunnelmpls6_handlers;
+		break;
+	default:
+		goto err;
+	}
+
+	for (; (t = rcu_dereference_protected(*pprev,
+			lockdep_is_held(&tunnel6_mutex))) != NULL;
+	     pprev = &t->next) {
+		if (t == handler) {
+			*pprev = handler->next;
+			ret = 0;
+			break;
+		}
+	}
+
+err:
+	mutex_unlock(&tunnel6_mutex);
+
+	synchronize_net();
+
+	return ret;
+}
+EXPORT_SYMBOL(xfrm6_tunnel_deregister);
+
+#define for_each_tunnel_rcu(head, handler)		\
+	for (handler = rcu_dereference(head);		\
+	     handler != NULL;				\
+	     handler = rcu_dereference(handler->next))	\
+
+static int tunnelmpls6_rcv(struct sk_buff *skb)
+{
+	struct xfrm6_tunnel *handler;
+
+	if (!pskb_may_pull(skb, sizeof(struct ipv6hdr)))
+		goto drop;
+
+	for_each_tunnel_rcu(tunnelmpls6_handlers, handler)
+		if (!handler->handler(skb))
+			return 0;
+
+	icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);
+
+drop:
+	kfree_skb(skb);
+	return 0;
+}
+
+static int tunnel6_rcv(struct sk_buff *skb)
+{
+	struct xfrm6_tunnel *handler;
+
+	if (!pskb_may_pull(skb, sizeof(struct ipv6hdr)))
+		goto drop;
+
+	for_each_tunnel_rcu(tunnel6_handlers, handler)
+		if (!handler->handler(skb))
+			return 0;
+
+	icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);
+
+drop:
+	kfree_skb(skb);
+	return 0;
+}
+
+#if IS_ENABLED(CONFIG_INET6_XFRM_TUNNEL)
+static int tunnel6_rcv_cb(struct sk_buff *skb, u8 proto, int err)
+{
+	struct xfrm6_tunnel __rcu *head;
+	struct xfrm6_tunnel *handler;
+	int ret;
+
+	head = (proto == IPPROTO_IPV6) ? tunnel6_handlers : tunnel46_handlers;
+
+	for_each_tunnel_rcu(head, handler) {
+		if (handler->cb_handler) {
+			ret = handler->cb_handler(skb, err);
+			if (ret <= 0)
+				return ret;
+		}
+	}
+
+	return 0;
+}
+
+static const struct xfrm_input_afinfo tunnel6_input_afinfo = {
+	.family		=	AF_INET6,
+	.is_ipip	=	true,
+	.callback	=	tunnel6_rcv_cb,
+};
+#endif
+
+static int tunnel46_rcv(struct sk_buff *skb)
+{
+	struct xfrm6_tunnel *handler;
+
+	if (!pskb_may_pull(skb, sizeof(struct iphdr)))
+		goto drop;
+
+	for_each_tunnel_rcu(tunnel46_handlers, handler)
+		if (!handler->handler(skb))
+			return 0;
+
+	icmpv6_send(skb, ICMPV6_DEST_UNREACH, ICMPV6_PORT_UNREACH, 0);
+
+drop:
+	kfree_skb(skb);
+	return 0;
+}
+
+static int tunnel6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
+			u8 type, u8 code, int offset, __be32 info)
+{
+	struct xfrm6_tunnel *handler;
+
+	for_each_tunnel_rcu(tunnel6_handlers, handler)
+		if (!handler->err_handler(skb, opt, type, code, offset, info))
+			return 0;
+
+	return -ENOENT;
+}
+
+static int tunnel46_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
+			 u8 type, u8 code, int offset, __be32 info)
+{
+	struct xfrm6_tunnel *handler;
+
+	for_each_tunnel_rcu(tunnel46_handlers, handler)
+		if (!handler->err_handler(skb, opt, type, code, offset, info))
+			return 0;
+
+	return -ENOENT;
+}
+
+static int tunnelmpls6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
+			   u8 type, u8 code, int offset, __be32 info)
+{
+	struct xfrm6_tunnel *handler;
+
+	for_each_tunnel_rcu(tunnelmpls6_handlers, handler)
+		if (!handler->err_handler(skb, opt, type, code, offset, info))
+			return 0;
+
+	return -ENOENT;
+}
+
+static const struct inet6_protocol tunnel6_protocol = {
+	.handler	= tunnel6_rcv,
+	.err_handler	= tunnel6_err,
+	.flags          = INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
+};
+
+static const struct inet6_protocol tunnel46_protocol = {
+	.handler	= tunnel46_rcv,
+	.err_handler	= tunnel46_err,
+	.flags          = INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
+};
+
+static const struct inet6_protocol tunnelmpls6_protocol = {
+	.handler	= tunnelmpls6_rcv,
+	.err_handler	= tunnelmpls6_err,
+	.flags          = INET6_PROTO_NOPOLICY|INET6_PROTO_FINAL,
+};
+
+static int __init tunnel6_init(void)
+{
+	if (inet6_add_protocol(&tunnel6_protocol, IPPROTO_IPV6)) {
+		pr_err("%s: can't add protocol\n", __func__);
+		return -EAGAIN;
+	}
+	if (inet6_add_protocol(&tunnel46_protocol, IPPROTO_IPIP)) {
+		pr_err("%s: can't add protocol\n", __func__);
+		inet6_del_protocol(&tunnel6_protocol, IPPROTO_IPV6);
+		return -EAGAIN;
+	}
+	if (xfrm6_tunnel_mpls_supported() &&
+	    inet6_add_protocol(&tunnelmpls6_protocol, IPPROTO_MPLS)) {
+		pr_err("%s: can't add protocol\n", __func__);
+		inet6_del_protocol(&tunnel6_protocol, IPPROTO_IPV6);
+		inet6_del_protocol(&tunnel46_protocol, IPPROTO_IPIP);
+		return -EAGAIN;
+	}
+#if IS_ENABLED(CONFIG_INET6_XFRM_TUNNEL)
+	if (xfrm_input_register_afinfo(&tunnel6_input_afinfo)) {
+		pr_err("%s: can't add input afinfo\n", __func__);
+		inet6_del_protocol(&tunnel6_protocol, IPPROTO_IPV6);
+		inet6_del_protocol(&tunnel46_protocol, IPPROTO_IPIP);
+		if (xfrm6_tunnel_mpls_supported())
+			inet6_del_protocol(&tunnelmpls6_protocol, IPPROTO_MPLS);
+		return -EAGAIN;
+	}
+#endif
+	return 0;
+}
+
+static void __exit tunnel6_fini(void)
+{
+#if IS_ENABLED(CONFIG_INET6_XFRM_TUNNEL)
+	if (xfrm_input_unregister_afinfo(&tunnel6_input_afinfo))
+		pr_err("%s: can't remove input afinfo\n", __func__);
+#endif
+	if (inet6_del_protocol(&tunnel46_protocol, IPPROTO_IPIP))
+		pr_err("%s: can't remove protocol\n", __func__);
+	if (inet6_del_protocol(&tunnel6_protocol, IPPROTO_IPV6))
+		pr_err("%s: can't remove protocol\n", __func__);
+	if (xfrm6_tunnel_mpls_supported() &&
+	    inet6_del_protocol(&tunnelmpls6_protocol, IPPROTO_MPLS))
+		pr_err("%s: can't remove protocol\n", __func__);
+}
+
+module_init(tunnel6_init);
+module_exit(tunnel6_fini);
+MODULE_DESCRIPTION("IP-in-IPv6 tunnel driver");
+MODULE_LICENSE("GPL");