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, 161 insertions, 0 deletions

diff --git a/fs/sysfs/dir.c b/fs/sysfs/dir.c
new file mode 100644
index 000000000..ffdcd4153
--- /dev/null
+++ b/fs/sysfs/dir.c
@@ -0,0 +1,161 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * fs/sysfs/dir.c - sysfs core and dir operation implementation
+ *
+ * Copyright (c) 2001-3 Patrick Mochel
+ * Copyright (c) 2007 SUSE Linux Products GmbH
+ * Copyright (c) 2007 Tejun Heo <teheo@suse.de>
+ *
+ * Please see Documentation/filesystems/sysfs.rst for more information.
+ */
+
+#define pr_fmt(fmt)	"sysfs: " fmt
+
+#include <linux/fs.h>
+#include <linux/kobject.h>
+#include <linux/slab.h>
+#include "sysfs.h"
+
+DEFINE_SPINLOCK(sysfs_symlink_target_lock);
+
+void sysfs_warn_dup(struct kernfs_node *parent, const char *name)
+{
+	char *buf;
+
+	buf = kzalloc(PATH_MAX, GFP_KERNEL);
+	if (buf)
+		kernfs_path(parent, buf, PATH_MAX);
+
+	pr_warn("cannot create duplicate filename '%s/%s'\n", buf, name);
+	dump_stack();
+
+	kfree(buf);
+}
+
+/**
+ * sysfs_create_dir_ns - create a directory for an object with a namespace tag
+ * @kobj: object we're creating directory for
+ * @ns: the namespace tag to use
+ */
+int sysfs_create_dir_ns(struct kobject *kobj, const struct ns_common *ns)
+{
+	struct kernfs_node *parent, *kn;
+	kuid_t uid;
+	kgid_t gid;
+
+	if (WARN_ON(!kobj))
+		return -EINVAL;
+
+	if (kobj->parent)
+		parent = kobj->parent->sd;
+	else
+		parent = sysfs_root_kn;
+
+	if (!parent)
+		return -ENOENT;
+
+	kobject_get_ownership(kobj, &uid, &gid);
+
+	kn = kernfs_create_dir_ns(parent, kobject_name(kobj), 0755, uid, gid,
+				  kobj, ns);
+	if (IS_ERR(kn)) {
+		if (PTR_ERR(kn) == -EEXIST)
+			sysfs_warn_dup(parent, kobject_name(kobj));
+		return PTR_ERR(kn);
+	}
+
+	kobj->sd = kn;
+	return 0;
+}
+
+/**
+ *	sysfs_remove_dir - remove an object's directory.
+ *	@kobj:	object.
+ *
+ *	The only thing special about this is that we remove any files in
+ *	the directory before we remove the directory, and we've inlined
+ *	what used to be sysfs_rmdir() below, instead of calling separately.
+ */
+void sysfs_remove_dir(struct kobject *kobj)
+{
+	struct kernfs_node *kn = kobj->sd;
+
+	/*
+	 * In general, kobject owner is responsible for ensuring removal
+	 * doesn't race with other operations and sysfs doesn't provide any
+	 * protection; however, when @kobj is used as a symlink target, the
+	 * symlinking entity usually doesn't own @kobj and thus has no
+	 * control over removal.  @kobj->sd may be removed anytime
+	 * and symlink code may end up dereferencing an already freed node.
+	 *
+	 * sysfs_symlink_target_lock synchronizes @kobj->sd
+	 * disassociation against symlink operations so that symlink code
+	 * can safely dereference @kobj->sd.
+	 */
+	spin_lock(&sysfs_symlink_target_lock);
+	kobj->sd = NULL;
+	spin_unlock(&sysfs_symlink_target_lock);
+
+	if (kn) {
+		WARN_ON_ONCE(kernfs_type(kn) != KERNFS_DIR);
+		kernfs_remove(kn);
+	}
+}
+
+int sysfs_rename_dir_ns(struct kobject *kobj, const char *new_name,
+			const struct ns_common *new_ns)
+{
+	struct kernfs_node *parent;
+	int ret;
+
+	parent = kernfs_get_parent(kobj->sd);
+	ret = kernfs_rename_ns(kobj->sd, parent, new_name, new_ns);
+	kernfs_put(parent);
+	return ret;
+}
+
+int sysfs_move_dir_ns(struct kobject *kobj, struct kobject *new_parent_kobj,
+		      const struct ns_common *new_ns)
+{
+	struct kernfs_node *kn = kobj->sd;
+	struct kernfs_node *new_parent;
+
+	new_parent = new_parent_kobj && new_parent_kobj->sd ?
+		new_parent_kobj->sd : sysfs_root_kn;
+
+	return kernfs_rename_ns(kn, new_parent, NULL, new_ns);
+}
+
+/**
+ * sysfs_create_mount_point - create an always empty directory
+ * @parent_kobj:  kobject that will contain this always empty directory
+ * @name: The name of the always empty directory to add
+ */
+int sysfs_create_mount_point(struct kobject *parent_kobj, const char *name)
+{
+	struct kernfs_node *kn, *parent = parent_kobj->sd;
+
+	kn = kernfs_create_empty_dir(parent, name);
+	if (IS_ERR(kn)) {
+		if (PTR_ERR(kn) == -EEXIST)
+			sysfs_warn_dup(parent, name);
+		return PTR_ERR(kn);
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(sysfs_create_mount_point);
+
+/**
+ *	sysfs_remove_mount_point - remove an always empty directory.
+ *	@parent_kobj: kobject that will contain this always empty directory
+ *	@name: The name of the always empty directory to remove
+ *
+ */
+void sysfs_remove_mount_point(struct kobject *parent_kobj, const char *name)
+{
+	struct kernfs_node *parent = parent_kobj->sd;
+
+	kernfs_remove_by_name_ns(parent, name, NULL);
+}
+EXPORT_SYMBOL_GPL(sysfs_remove_mount_point);