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, 200 insertions, 0 deletions

diff --git a/drivers/net/dummy.c b/drivers/net/dummy.c
new file mode 100644
index 000000000..36a4e85c6
--- /dev/null
+++ b/drivers/net/dummy.c
@@ -0,0 +1,200 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* dummy.c: a dummy net driver
+
+	The purpose of this driver is to provide a device to point a
+	route through, but not to actually transmit packets.
+
+	Why?  If you have a machine whose only connection is an occasional
+	PPP/SLIP/PLIP link, you can only connect to your own hostname
+	when the link is up.  Otherwise you have to use localhost.
+	This isn't very consistent.
+
+	One solution is to set up a dummy link using PPP/SLIP/PLIP,
+	but this seems (to me) too much overhead for too little gain.
+	This driver provides a small alternative. Thus you can do
+
+	[when not running slip]
+		ifconfig dummy slip.addr.ess.here up
+	[to go to slip]
+		ifconfig dummy down
+		dip whatever
+
+	This was written by looking at Donald Becker's skeleton driver
+	and the loopback driver.  I then threw away anything that didn't
+	apply!	Thanks to Alan Cox for the key clue on what to do with
+	misguided packets.
+
+			Nick Holloway, 27th May 1994
+	[I tweaked this explanation a little but that's all]
+			Alan Cox, 30th May 1994
+*/
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/ethtool.h>
+#include <linux/init.h>
+#include <linux/moduleparam.h>
+#include <linux/rtnetlink.h>
+#include <linux/net_tstamp.h>
+#include <net/netdev_lock.h>
+#include <net/rtnetlink.h>
+#include <linux/u64_stats_sync.h>
+
+#define DRV_NAME	"dummy"
+
+static int numdummies = 1;
+
+/* fake multicast ability */
+static int set_multicast_list(struct net_device *dev,
+			      struct netdev_hw_addr_list *uc,
+			      struct netdev_hw_addr_list *mc)
+{
+	return 0;
+}
+
+static void dummy_get_stats64(struct net_device *dev,
+			      struct rtnl_link_stats64 *stats)
+{
+	dev_lstats_read(dev, &stats->tx_packets, &stats->tx_bytes);
+}
+
+static netdev_tx_t dummy_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	dev_lstats_add(dev, skb->len);
+
+	skb_tx_timestamp(skb);
+	dev_kfree_skb(skb);
+	return NETDEV_TX_OK;
+}
+
+static int dummy_dev_init(struct net_device *dev)
+{
+	dev->pcpu_stat_type = NETDEV_PCPU_STAT_LSTATS;
+
+	netdev_lockdep_set_classes(dev);
+	return 0;
+}
+
+static int dummy_change_carrier(struct net_device *dev, bool new_carrier)
+{
+	if (new_carrier)
+		netif_carrier_on(dev);
+	else
+		netif_carrier_off(dev);
+	return 0;
+}
+
+static const struct net_device_ops dummy_netdev_ops = {
+	.ndo_init		= dummy_dev_init,
+	.ndo_start_xmit		= dummy_xmit,
+	.ndo_validate_addr	= eth_validate_addr,
+	.ndo_set_rx_mode_async	= set_multicast_list,
+	.ndo_set_mac_address	= eth_mac_addr,
+	.ndo_get_stats64	= dummy_get_stats64,
+	.ndo_change_carrier	= dummy_change_carrier,
+};
+
+static const struct ethtool_ops dummy_ethtool_ops = {
+	.get_ts_info		= ethtool_op_get_ts_info,
+};
+
+static void dummy_setup(struct net_device *dev)
+{
+	ether_setup(dev);
+
+	/* Initialize the device structure. */
+	dev->netdev_ops = &dummy_netdev_ops;
+	dev->ethtool_ops = &dummy_ethtool_ops;
+	dev->needs_free_netdev = true;
+	dev->request_ops_lock = true;
+
+	/* Fill in device structure with ethernet-generic values. */
+	dev->flags |= IFF_NOARP;
+	dev->flags &= ~IFF_MULTICAST;
+	dev->priv_flags |= IFF_LIVE_ADDR_CHANGE | IFF_NO_QUEUE;
+	dev->lltx = true;
+	dev->features	|= NETIF_F_SG | NETIF_F_FRAGLIST;
+	dev->features	|= NETIF_F_GSO_SOFTWARE;
+	dev->features	|= NETIF_F_HW_CSUM | NETIF_F_HIGHDMA;
+	dev->features	|= NETIF_F_GSO_ENCAP_ALL;
+	dev->hw_features |= dev->features;
+	dev->hw_enc_features |= dev->features;
+	eth_hw_addr_random(dev);
+
+	dev->min_mtu = 0;
+	dev->max_mtu = 0;
+}
+
+static int dummy_validate(struct nlattr *tb[], struct nlattr *data[],
+			  struct netlink_ext_ack *extack)
+{
+	if (tb[IFLA_ADDRESS]) {
+		if (nla_len(tb[IFLA_ADDRESS]) != ETH_ALEN)
+			return -EINVAL;
+		if (!is_valid_ether_addr(nla_data(tb[IFLA_ADDRESS])))
+			return -EADDRNOTAVAIL;
+	}
+	return 0;
+}
+
+static struct rtnl_link_ops dummy_link_ops __read_mostly = {
+	.kind		= DRV_NAME,
+	.setup		= dummy_setup,
+	.validate	= dummy_validate,
+};
+
+/* Number of dummy devices to be set up by this module. */
+module_param(numdummies, int, 0);
+MODULE_PARM_DESC(numdummies, "Number of dummy pseudo devices");
+
+static int __init dummy_init_one(void)
+{
+	struct net_device *dev_dummy;
+	int err;
+
+	dev_dummy = alloc_netdev(0, "dummy%d", NET_NAME_ENUM, dummy_setup);
+	if (!dev_dummy)
+		return -ENOMEM;
+
+	dev_dummy->rtnl_link_ops = &dummy_link_ops;
+	err = register_netdev(dev_dummy);
+	if (err < 0)
+		goto err;
+	return 0;
+
+err:
+	free_netdev(dev_dummy);
+	return err;
+}
+
+static int __init dummy_init_module(void)
+{
+	int i, err = 0;
+
+	err = rtnl_link_register(&dummy_link_ops);
+	if (err < 0)
+		return err;
+
+	for (i = 0; i < numdummies && !err; i++) {
+		err = dummy_init_one();
+		cond_resched();
+	}
+
+	if (err < 0)
+		rtnl_link_unregister(&dummy_link_ops);
+
+	return err;
+}
+
+static void __exit dummy_cleanup_module(void)
+{
+	rtnl_link_unregister(&dummy_link_ops);
+}
+
+module_init(dummy_init_module);
+module_exit(dummy_cleanup_module);
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Dummy netdevice driver which discards all packets sent to it");
+MODULE_ALIAS_RTNL_LINK(DRV_NAME);