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, 212 insertions, 0 deletions

diff --git a/net/9p/mod.c b/net/9p/mod.c
new file mode 100644
index 000000000..85160b52d
--- /dev/null
+++ b/net/9p/mod.c
@@ -0,0 +1,212 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  9P entry point
+ *
+ *  Copyright (C) 2007 by Latchesar Ionkov <lucho@ionkov.net>
+ *  Copyright (C) 2004 by Eric Van Hensbergen <ericvh@gmail.com>
+ *  Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov>
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/module.h>
+#include <linux/kmod.h>
+#include <linux/errno.h>
+#include <linux/sched.h>
+#include <linux/moduleparam.h>
+#include <net/9p/9p.h>
+#include <linux/fs.h>
+#include <net/9p/client.h>
+#include <net/9p/transport.h>
+#include <linux/list.h>
+#include <linux/spinlock.h>
+
+#ifdef CONFIG_NET_9P_DEBUG
+unsigned int p9_debug_level;	/* feature-rific global debug level  */
+EXPORT_SYMBOL(p9_debug_level);
+module_param_named(debug, p9_debug_level, uint, 0);
+MODULE_PARM_DESC(debug, "9P debugging level");
+
+void _p9_debug(enum p9_debug_flags level, const char *func,
+	       const char *fmt, ...)
+{
+	struct va_format vaf;
+	va_list args;
+
+	if ((p9_debug_level & level) != level)
+		return;
+
+	va_start(args, fmt);
+
+	vaf.fmt = fmt;
+	vaf.va = &args;
+
+	if (level == P9_DEBUG_9P)
+		pr_notice("(%8.8d) %pV", task_pid_nr(current), &vaf);
+	else
+		pr_notice("-- %s (%d): %pV", func, task_pid_nr(current), &vaf);
+
+	va_end(args);
+}
+EXPORT_SYMBOL(_p9_debug);
+#endif
+
+/* Dynamic Transport Registration Routines */
+
+static DEFINE_SPINLOCK(v9fs_trans_lock);
+static LIST_HEAD(v9fs_trans_list);
+
+/**
+ * v9fs_register_trans - register a new transport with 9p
+ * @m: structure describing the transport module and entry points
+ *
+ */
+void v9fs_register_trans(struct p9_trans_module *m)
+{
+	spin_lock(&v9fs_trans_lock);
+	list_add_tail(&m->list, &v9fs_trans_list);
+	spin_unlock(&v9fs_trans_lock);
+}
+EXPORT_SYMBOL(v9fs_register_trans);
+
+/**
+ * v9fs_unregister_trans - unregister a 9p transport
+ * @m: the transport to remove
+ *
+ */
+void v9fs_unregister_trans(struct p9_trans_module *m)
+{
+	spin_lock(&v9fs_trans_lock);
+	list_del_init(&m->list);
+	spin_unlock(&v9fs_trans_lock);
+}
+EXPORT_SYMBOL(v9fs_unregister_trans);
+
+static struct p9_trans_module *_p9_get_trans_by_name(const char *s)
+{
+	struct p9_trans_module *t, *found = NULL;
+
+	spin_lock(&v9fs_trans_lock);
+
+	list_for_each_entry(t, &v9fs_trans_list, list)
+		if (strcmp(t->name, s) == 0 &&
+		    try_module_get(t->owner)) {
+			found = t;
+			break;
+		}
+
+	spin_unlock(&v9fs_trans_lock);
+
+	return found;
+}
+
+/**
+ * v9fs_get_trans_by_name - get transport with the matching name
+ * @s: string identifying transport
+ *
+ */
+struct p9_trans_module *v9fs_get_trans_by_name(const char *s)
+{
+	struct p9_trans_module *found = NULL;
+
+	found = _p9_get_trans_by_name(s);
+
+#ifdef CONFIG_MODULES
+	if (!found) {
+		request_module("9p-%s", s);
+		found = _p9_get_trans_by_name(s);
+	}
+#endif
+
+	return found;
+}
+EXPORT_SYMBOL(v9fs_get_trans_by_name);
+
+static const char * const v9fs_default_transports[] = {
+	"virtio", "tcp", "fd", "unix", "xen", "rdma",
+};
+
+/**
+ * v9fs_get_default_trans - get the default transport
+ *
+ */
+
+struct p9_trans_module *v9fs_get_default_trans(void)
+{
+	struct p9_trans_module *t, *found = NULL;
+	int i;
+
+	spin_lock(&v9fs_trans_lock);
+
+	list_for_each_entry(t, &v9fs_trans_list, list)
+		if (t->def && try_module_get(t->owner)) {
+			found = t;
+			break;
+		}
+
+	if (!found)
+		list_for_each_entry(t, &v9fs_trans_list, list)
+			if (try_module_get(t->owner)) {
+				found = t;
+				break;
+			}
+
+	spin_unlock(&v9fs_trans_lock);
+
+	for (i = 0; !found && i < ARRAY_SIZE(v9fs_default_transports); i++)
+		found = v9fs_get_trans_by_name(v9fs_default_transports[i]);
+
+	return found;
+}
+EXPORT_SYMBOL(v9fs_get_default_trans);
+
+/**
+ * v9fs_put_trans - put trans
+ * @m: transport to put
+ *
+ */
+void v9fs_put_trans(struct p9_trans_module *m)
+{
+	if (m)
+		module_put(m->owner);
+}
+EXPORT_SYMBOL(v9fs_put_trans);
+
+/**
+ * init_p9 - Initialize module
+ *
+ */
+static int __init init_p9(void)
+{
+	int ret;
+
+	ret = p9_client_init();
+	if (ret)
+		return ret;
+
+	p9_error_init();
+	pr_info("Installing 9P2000 support\n");
+
+	return ret;
+}
+
+/**
+ * exit_p9 - shutdown module
+ *
+ */
+
+static void __exit exit_p9(void)
+{
+	pr_info("Unloading 9P2000 support\n");
+
+	p9_client_exit();
+}
+
+module_init(init_p9)
+module_exit(exit_p9)
+
+MODULE_AUTHOR("Latchesar Ionkov <lucho@ionkov.net>");
+MODULE_AUTHOR("Eric Van Hensbergen <ericvh@gmail.com>");
+MODULE_AUTHOR("Ron Minnich <rminnich@lanl.gov>");
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Plan 9 Resource Sharing Support (9P2000)");