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, 181 insertions, 0 deletions

diff --git a/fs/ecryptfs/super.c b/fs/ecryptfs/super.c
new file mode 100644
index 000000000..686b2b4a9
--- /dev/null
+++ b/fs/ecryptfs/super.c
@@ -0,0 +1,181 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * eCryptfs: Linux filesystem encryption layer
+ *
+ * Copyright (C) 1997-2003 Erez Zadok
+ * Copyright (C) 2001-2003 Stony Brook University
+ * Copyright (C) 2004-2006 International Business Machines Corp.
+ *   Author(s): Michael A. Halcrow <mahalcro@us.ibm.com>
+ *              Michael C. Thompson <mcthomps@us.ibm.com>
+ */
+
+#include <linux/fs.h>
+#include <linux/mount.h>
+#include <linux/key.h>
+#include <linux/slab.h>
+#include <linux/seq_file.h>
+#include <linux/file.h>
+#include <linux/statfs.h>
+#include <linux/magic.h>
+#include "ecryptfs_kernel.h"
+
+struct kmem_cache *ecryptfs_inode_info_cache;
+
+/**
+ * ecryptfs_alloc_inode - allocate an ecryptfs inode
+ * @sb: Pointer to the ecryptfs super block
+ *
+ * Called to bring an inode into existence.
+ *
+ * Only handle allocation, setting up structures should be done in
+ * ecryptfs_read_inode. This is because the kernel, between now and
+ * then, will 0 out the private data pointer.
+ *
+ * Returns a pointer to a newly allocated inode, NULL otherwise
+ */
+static struct inode *ecryptfs_alloc_inode(struct super_block *sb)
+{
+	struct ecryptfs_inode_info *inode_info;
+	struct inode *inode = NULL;
+
+	inode_info = alloc_inode_sb(sb, ecryptfs_inode_info_cache, GFP_KERNEL);
+	if (unlikely(!inode_info))
+		goto out;
+	ecryptfs_init_crypt_stat(&inode_info->crypt_stat);
+	mutex_init(&inode_info->lower_file_mutex);
+	atomic_set(&inode_info->lower_file_count, 0);
+	inode_info->lower_file = NULL;
+	inode = &inode_info->vfs_inode;
+out:
+	return inode;
+}
+
+static void ecryptfs_free_inode(struct inode *inode)
+{
+	struct ecryptfs_inode_info *inode_info;
+	inode_info = ecryptfs_inode_to_private(inode);
+
+	kmem_cache_free(ecryptfs_inode_info_cache, inode_info);
+}
+
+/**
+ * ecryptfs_destroy_inode
+ * @inode: The ecryptfs inode
+ *
+ * This is used during the final destruction of the inode.  All
+ * allocation of memory related to the inode, including allocated
+ * memory in the crypt_stat struct, will be released here.
+ * There should be no chance that this deallocation will be missed.
+ */
+static void ecryptfs_destroy_inode(struct inode *inode)
+{
+	struct ecryptfs_inode_info *inode_info;
+
+	inode_info = ecryptfs_inode_to_private(inode);
+	BUG_ON(inode_info->lower_file);
+	ecryptfs_destroy_crypt_stat(&inode_info->crypt_stat);
+}
+
+/**
+ * ecryptfs_statfs
+ * @dentry: The ecryptfs dentry
+ * @buf: The struct kstatfs to fill in with stats
+ *
+ * Get the filesystem statistics. Currently, we let this pass right through
+ * to the lower filesystem and take no action ourselves.
+ */
+static int ecryptfs_statfs(struct dentry *dentry, struct kstatfs *buf)
+{
+	struct dentry *lower_dentry = ecryptfs_dentry_to_lower(dentry);
+	int rc;
+
+	if (!lower_dentry->d_sb->s_op->statfs)
+		return -ENOSYS;
+
+	rc = lower_dentry->d_sb->s_op->statfs(lower_dentry, buf);
+	if (rc)
+		return rc;
+
+	buf->f_type = ECRYPTFS_SUPER_MAGIC;
+	rc = ecryptfs_set_f_namelen(&buf->f_namelen, buf->f_namelen,
+	       &ecryptfs_superblock_to_private(dentry->d_sb)->mount_crypt_stat);
+
+	return rc;
+}
+
+/**
+ * ecryptfs_evict_inode
+ * @inode: The ecryptfs inode
+ *
+ * Called by iput() when the inode reference count reached zero
+ * and the inode is not hashed anywhere.  Used to clear anything
+ * that needs to be, before the inode is completely destroyed and put
+ * on the inode free list. We use this to drop out reference to the
+ * lower inode.
+ */
+static void ecryptfs_evict_inode(struct inode *inode)
+{
+	truncate_inode_pages_final(&inode->i_data);
+	clear_inode(inode);
+	iput(ecryptfs_inode_to_lower(inode));
+}
+
+/*
+ * ecryptfs_show_options
+ *
+ * Prints the mount options for a given superblock.
+ * Returns zero; does not fail.
+ */
+static int ecryptfs_show_options(struct seq_file *m, struct dentry *root)
+{
+	struct super_block *sb = root->d_sb;
+	struct ecryptfs_mount_crypt_stat *mount_crypt_stat =
+		&ecryptfs_superblock_to_private(sb)->mount_crypt_stat;
+	struct ecryptfs_global_auth_tok *walker;
+
+	mutex_lock(&mount_crypt_stat->global_auth_tok_list_mutex);
+	list_for_each_entry(walker,
+			    &mount_crypt_stat->global_auth_tok_list,
+			    mount_crypt_stat_list) {
+		if (walker->flags & ECRYPTFS_AUTH_TOK_FNEK)
+			seq_printf(m, ",ecryptfs_fnek_sig=%s", walker->sig);
+		else
+			seq_printf(m, ",ecryptfs_sig=%s", walker->sig);
+	}
+	mutex_unlock(&mount_crypt_stat->global_auth_tok_list_mutex);
+
+	seq_printf(m, ",ecryptfs_cipher=%s",
+		mount_crypt_stat->global_default_cipher_name);
+
+	if (mount_crypt_stat->global_default_cipher_key_size)
+		seq_printf(m, ",ecryptfs_key_bytes=%zd",
+			   mount_crypt_stat->global_default_cipher_key_size);
+	if (mount_crypt_stat->flags & ECRYPTFS_GLOBAL_ENCRYPT_FILENAMES) {
+		seq_printf(m, ",ecryptfs_fn_cipher=%s",
+			   mount_crypt_stat->global_default_fn_cipher_name);
+		if (mount_crypt_stat->global_default_fn_cipher_key_bytes)
+			seq_printf(m, ",ecryptfs_fn_key_bytes=%zd",
+				   mount_crypt_stat->global_default_fn_cipher_key_bytes);
+	}
+	if (mount_crypt_stat->flags & ECRYPTFS_PLAINTEXT_PASSTHROUGH_ENABLED)
+		seq_printf(m, ",ecryptfs_passthrough");
+	if (mount_crypt_stat->flags & ECRYPTFS_XATTR_METADATA_ENABLED)
+		seq_printf(m, ",ecryptfs_xattr_metadata");
+	if (mount_crypt_stat->flags & ECRYPTFS_ENCRYPTED_VIEW_ENABLED)
+		seq_printf(m, ",ecryptfs_encrypted_view");
+	if (mount_crypt_stat->flags & ECRYPTFS_UNLINK_SIGS)
+		seq_printf(m, ",ecryptfs_unlink_sigs");
+	if (mount_crypt_stat->flags & ECRYPTFS_GLOBAL_MOUNT_AUTH_TOK_ONLY)
+		seq_printf(m, ",ecryptfs_mount_auth_tok_only");
+
+	return 0;
+}
+
+const struct super_operations ecryptfs_sops = {
+	.alloc_inode = ecryptfs_alloc_inode,
+	.destroy_inode = ecryptfs_destroy_inode,
+	.free_inode = ecryptfs_free_inode,
+	.statfs = ecryptfs_statfs,
+	.evict_inode = ecryptfs_evict_inode,
+	.show_options = ecryptfs_show_options
+};