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, 293 insertions, 0 deletions

diff --git a/net/atm/ioctl.c b/net/atm/ioctl.c
new file mode 100644
index 000000000..194587fc1
--- /dev/null
+++ b/net/atm/ioctl.c
@@ -0,0 +1,293 @@
+// SPDX-License-Identifier: GPL-2.0
+/* ATM ioctl handling */
+
+/* Written 1995-2000 by Werner Almesberger, EPFL LRC/ICA */
+/* 2003 John Levon  <levon@movementarian.org> */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ":%s: " fmt, __func__
+
+#include <linux/module.h>
+#include <linux/kmod.h>
+#include <linux/net.h>		/* struct socket, struct proto_ops */
+#include <linux/atm.h>		/* ATM stuff */
+#include <linux/atmdev.h>
+#include <linux/capability.h>
+#include <linux/mutex.h>
+#include <asm/ioctls.h>
+#include <net/compat.h>
+
+#include "resources.h"
+#include "common.h"
+
+
+static DEFINE_MUTEX(ioctl_mutex);
+static LIST_HEAD(ioctl_list);
+
+
+void register_atm_ioctl(struct atm_ioctl *ioctl)
+{
+	mutex_lock(&ioctl_mutex);
+	list_add_tail(&ioctl->list, &ioctl_list);
+	mutex_unlock(&ioctl_mutex);
+}
+EXPORT_SYMBOL(register_atm_ioctl);
+
+void deregister_atm_ioctl(struct atm_ioctl *ioctl)
+{
+	mutex_lock(&ioctl_mutex);
+	list_del(&ioctl->list);
+	mutex_unlock(&ioctl_mutex);
+}
+EXPORT_SYMBOL(deregister_atm_ioctl);
+
+static int do_vcc_ioctl(struct socket *sock, unsigned int cmd,
+			unsigned long arg, int compat)
+{
+	struct sock *sk = sock->sk;
+	struct atm_vcc *vcc;
+	int error;
+	struct list_head *pos;
+	void __user *argp = (void __user *)arg;
+	void __user *buf;
+	int __user *len;
+
+	vcc = ATM_SD(sock);
+	switch (cmd) {
+	case SIOCOUTQ:
+		if (sock->state != SS_CONNECTED ||
+		    !test_bit(ATM_VF_READY, &vcc->flags)) {
+			error =  -EINVAL;
+			goto done;
+		}
+		error = put_user(sk->sk_sndbuf - sk_wmem_alloc_get(sk),
+				 (int __user *)argp);
+		goto done;
+	case SIOCINQ:
+	{
+		struct sk_buff *skb;
+		int amount;
+
+		if (sock->state != SS_CONNECTED) {
+			error = -EINVAL;
+			goto done;
+		}
+		spin_lock_irq(&sk->sk_receive_queue.lock);
+		skb = skb_peek(&sk->sk_receive_queue);
+		amount = skb ? skb->len : 0;
+		spin_unlock_irq(&sk->sk_receive_queue.lock);
+		error = put_user(amount, (int __user *)argp);
+		goto done;
+	}
+	case ATM_SETSC:
+		net_warn_ratelimited("ATM_SETSC is obsolete; used by %s:%d\n",
+				     current->comm, task_pid_nr(current));
+		error = 0;
+		goto done;
+	case ATM_SETBACKEND:
+	case ATM_NEWBACKENDIF:
+	{
+		atm_backend_t backend;
+		error = get_user(backend, (atm_backend_t __user *)argp);
+		if (error)
+			goto done;
+		switch (backend) {
+		case ATM_BACKEND_PPP:
+			request_module("pppoatm");
+			break;
+		case ATM_BACKEND_BR2684:
+			request_module("br2684");
+			break;
+		}
+		break;
+	}
+	}
+
+	error = -ENOIOCTLCMD;
+
+	mutex_lock(&ioctl_mutex);
+	list_for_each(pos, &ioctl_list) {
+		struct atm_ioctl *ic = list_entry(pos, struct atm_ioctl, list);
+		if (try_module_get(ic->owner)) {
+			error = ic->ioctl(sock, cmd, arg);
+			module_put(ic->owner);
+			if (error != -ENOIOCTLCMD)
+				break;
+		}
+	}
+	mutex_unlock(&ioctl_mutex);
+
+	if (error != -ENOIOCTLCMD)
+		goto done;
+
+	if (cmd == ATM_GETNAMES) {
+		if (IS_ENABLED(CONFIG_COMPAT) && compat) {
+#ifdef CONFIG_COMPAT
+			struct compat_atm_iobuf __user *ciobuf = argp;
+			compat_uptr_t cbuf;
+			len = &ciobuf->length;
+			if (get_user(cbuf, &ciobuf->buffer))
+				return -EFAULT;
+			buf = compat_ptr(cbuf);
+#endif
+		} else {
+			struct atm_iobuf __user *iobuf = argp;
+			len = &iobuf->length;
+			if (get_user(buf, &iobuf->buffer))
+				return -EFAULT;
+		}
+		error = atm_getnames(buf, len);
+	} else {
+		int number;
+
+		if (IS_ENABLED(CONFIG_COMPAT) && compat) {
+#ifdef CONFIG_COMPAT
+			struct compat_atmif_sioc __user *csioc = argp;
+			compat_uptr_t carg;
+
+			len = &csioc->length;
+			if (get_user(carg, &csioc->arg))
+				return -EFAULT;
+			buf = compat_ptr(carg);
+			if (get_user(number, &csioc->number))
+				return -EFAULT;
+#endif
+		} else {
+			struct atmif_sioc __user *sioc = argp;
+
+			len = &sioc->length;
+			if (get_user(buf, &sioc->arg))
+				return -EFAULT;
+			if (get_user(number, &sioc->number))
+				return -EFAULT;
+		}
+		error = atm_dev_ioctl(cmd, buf, len, number, compat);
+	}
+
+done:
+	return error;
+}
+
+int vcc_ioctl(struct socket *sock, unsigned int cmd, unsigned long arg)
+{
+	return do_vcc_ioctl(sock, cmd, arg, 0);
+}
+
+#ifdef CONFIG_COMPAT
+/*
+ * FIXME:
+ * The compat_ioctl handling is duplicated, using both these conversion
+ * routines and the compat argument to the actual handlers. Both
+ * versions are somewhat incomplete and should be merged, e.g. by
+ * moving the ioctl number translation into the actual handlers and
+ * killing the conversion code.
+ *
+ * -arnd, November 2009
+ */
+#define ATM_GETLINKRATE32 _IOW('a', ATMIOC_ITF+1, struct compat_atmif_sioc)
+#define ATM_GETNAMES32    _IOW('a', ATMIOC_ITF+3, struct compat_atm_iobuf)
+#define ATM_GETTYPE32     _IOW('a', ATMIOC_ITF+4, struct compat_atmif_sioc)
+#define ATM_GETESI32	  _IOW('a', ATMIOC_ITF+5, struct compat_atmif_sioc)
+#define ATM_GETCIRANGE32  _IOW('a', ATMIOC_ITF+10, struct compat_atmif_sioc)
+#define ATM_SETCIRANGE32  _IOW('a', ATMIOC_ITF+11, struct compat_atmif_sioc)
+#define ATM_SETESI32      _IOW('a', ATMIOC_ITF+12, struct compat_atmif_sioc)
+#define ATM_SETESIF32     _IOW('a', ATMIOC_ITF+13, struct compat_atmif_sioc)
+#define ATM_GETSTAT32     _IOW('a', ATMIOC_SARCOM+0, struct compat_atmif_sioc)
+#define ATM_GETSTATZ32    _IOW('a', ATMIOC_SARCOM+1, struct compat_atmif_sioc)
+#define ATM_GETLOOP32	  _IOW('a', ATMIOC_SARCOM+2, struct compat_atmif_sioc)
+#define ATM_SETLOOP32	  _IOW('a', ATMIOC_SARCOM+3, struct compat_atmif_sioc)
+#define ATM_QUERYLOOP32	  _IOW('a', ATMIOC_SARCOM+4, struct compat_atmif_sioc)
+
+static struct {
+	unsigned int cmd32;
+	unsigned int cmd;
+} atm_ioctl_map[] = {
+	{ ATM_GETLINKRATE32, ATM_GETLINKRATE },
+	{ ATM_GETNAMES32,    ATM_GETNAMES },
+	{ ATM_GETTYPE32,     ATM_GETTYPE },
+	{ ATM_GETESI32,	     ATM_GETESI },
+	{ ATM_GETCIRANGE32,  ATM_GETCIRANGE },
+	{ ATM_SETCIRANGE32,  ATM_SETCIRANGE },
+	{ ATM_SETESI32,	     ATM_SETESI },
+	{ ATM_SETESIF32,     ATM_SETESIF },
+	{ ATM_GETSTAT32,     ATM_GETSTAT },
+	{ ATM_GETSTATZ32,    ATM_GETSTATZ },
+	{ ATM_GETLOOP32,     ATM_GETLOOP },
+	{ ATM_SETLOOP32,     ATM_SETLOOP },
+	{ ATM_QUERYLOOP32,   ATM_QUERYLOOP },
+};
+
+#define NR_ATM_IOCTL ARRAY_SIZE(atm_ioctl_map)
+
+static int do_atm_iobuf(struct socket *sock, unsigned int cmd,
+			unsigned long arg)
+{
+	struct compat_atm_iobuf __user *iobuf32 = compat_ptr(arg);
+	u32 data;
+
+	if (get_user(data, &iobuf32->buffer))
+		return -EFAULT;
+
+	return atm_getnames(&iobuf32->length, compat_ptr(data));
+}
+
+static int do_atmif_sioc(struct socket *sock, unsigned int cmd,
+			 unsigned long arg)
+{
+	struct compat_atmif_sioc __user *sioc32 = compat_ptr(arg);
+	int number;
+	u32 data;
+
+	if (get_user(data, &sioc32->arg) || get_user(number, &sioc32->number))
+		return -EFAULT;
+	return atm_dev_ioctl(cmd, compat_ptr(data), &sioc32->length, number, 0);
+}
+
+static int do_atm_ioctl(struct socket *sock, unsigned int cmd32,
+			unsigned long arg)
+{
+	int i;
+	unsigned int cmd = 0;
+
+	for (i = 0; i < NR_ATM_IOCTL; i++) {
+		if (cmd32 == atm_ioctl_map[i].cmd32) {
+			cmd = atm_ioctl_map[i].cmd;
+			break;
+		}
+	}
+	if (i == NR_ATM_IOCTL)
+		return -EINVAL;
+
+	switch (cmd) {
+	case ATM_GETNAMES:
+		return do_atm_iobuf(sock, cmd, arg);
+
+	case ATM_GETLINKRATE:
+	case ATM_GETTYPE:
+	case ATM_GETESI:
+	case ATM_GETCIRANGE:
+	case ATM_SETCIRANGE:
+	case ATM_SETESI:
+	case ATM_SETESIF:
+	case ATM_GETSTAT:
+	case ATM_GETSTATZ:
+	case ATM_GETLOOP:
+	case ATM_SETLOOP:
+	case ATM_QUERYLOOP:
+		return do_atmif_sioc(sock, cmd, arg);
+	}
+
+	return -EINVAL;
+}
+
+int vcc_compat_ioctl(struct socket *sock, unsigned int cmd,
+		     unsigned long arg)
+{
+	int ret;
+
+	ret = do_vcc_ioctl(sock, cmd, arg, 1);
+	if (ret != -ENOIOCTLCMD)
+		return ret;
+
+	return do_atm_ioctl(sock, cmd, arg);
+}
+#endif