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, 262 insertions, 0 deletions

diff --git a/fs/befs/debug.c b/fs/befs/debug.c
new file mode 100644
index 000000000..02fa66fb8
--- /dev/null
+++ b/fs/befs/debug.c
@@ -0,0 +1,262 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ *  linux/fs/befs/debug.c
+ *
+ * Copyright (C) 2001 Will Dyson (will_dyson at pobox.com)
+ *
+ * With help from the ntfs-tng driver by Anton Altparmakov
+ *
+ * Copyright (C) 1999  Makoto Kato (m_kato@ga2.so-net.ne.jp)
+ *
+ * debug functions
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+#ifdef __KERNEL__
+
+#include <linux/stdarg.h>
+#include <linux/string.h>
+#include <linux/spinlock.h>
+#include <linux/kernel.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+
+#endif				/* __KERNEL__ */
+
+#include "befs.h"
+
+void
+befs_error(const struct super_block *sb, const char *fmt, ...)
+{
+	struct va_format vaf;
+	va_list args;
+
+	va_start(args, fmt);
+	vaf.fmt = fmt;
+	vaf.va = &args;
+	pr_err("(%s): %pV\n", sb->s_id, &vaf);
+	va_end(args);
+}
+
+void
+befs_warning(const struct super_block *sb, const char *fmt, ...)
+{
+	struct va_format vaf;
+	va_list args;
+
+	va_start(args, fmt);
+	vaf.fmt = fmt;
+	vaf.va = &args;
+	pr_warn("(%s): %pV\n", sb->s_id, &vaf);
+	va_end(args);
+}
+
+void
+befs_debug(const struct super_block *sb, const char *fmt, ...)
+{
+#ifdef CONFIG_BEFS_DEBUG
+
+	struct va_format vaf;
+	va_list args;
+
+	va_start(args, fmt);
+	vaf.fmt = fmt;
+	vaf.va = &args;
+	pr_debug("(%s): %pV\n", sb->s_id, &vaf);
+	va_end(args);
+
+#endif				//CONFIG_BEFS_DEBUG
+}
+
+void
+befs_dump_inode(const struct super_block *sb, befs_inode *inode)
+{
+#ifdef CONFIG_BEFS_DEBUG
+
+	befs_block_run tmp_run;
+
+	befs_debug(sb, "befs_inode information");
+
+	befs_debug(sb, "  magic1 %08x", fs32_to_cpu(sb, inode->magic1));
+
+	tmp_run = fsrun_to_cpu(sb, inode->inode_num);
+	befs_debug(sb, "  inode_num %u, %hu, %hu",
+		   tmp_run.allocation_group, tmp_run.start, tmp_run.len);
+
+	befs_debug(sb, "  uid %u", fs32_to_cpu(sb, inode->uid));
+	befs_debug(sb, "  gid %u", fs32_to_cpu(sb, inode->gid));
+	befs_debug(sb, "  mode %08x", fs32_to_cpu(sb, inode->mode));
+	befs_debug(sb, "  flags %08x", fs32_to_cpu(sb, inode->flags));
+	befs_debug(sb, "  create_time %llu",
+		   fs64_to_cpu(sb, inode->create_time));
+	befs_debug(sb, "  last_modified_time %llu",
+		   fs64_to_cpu(sb, inode->last_modified_time));
+
+	tmp_run = fsrun_to_cpu(sb, inode->parent);
+	befs_debug(sb, "  parent [%u, %hu, %hu]",
+		   tmp_run.allocation_group, tmp_run.start, tmp_run.len);
+
+	tmp_run = fsrun_to_cpu(sb, inode->attributes);
+	befs_debug(sb, "  attributes [%u, %hu, %hu]",
+		   tmp_run.allocation_group, tmp_run.start, tmp_run.len);
+
+	befs_debug(sb, "  type %08x", fs32_to_cpu(sb, inode->type));
+	befs_debug(sb, "  inode_size %u", fs32_to_cpu(sb, inode->inode_size));
+
+	if (S_ISLNK(fs32_to_cpu(sb, inode->mode))) {
+		befs_debug(sb, "  Symbolic link [%s]", inode->data.symlink);
+	} else {
+		int i;
+
+		for (i = 0; i < BEFS_NUM_DIRECT_BLOCKS; i++) {
+			tmp_run =
+			    fsrun_to_cpu(sb, inode->data.datastream.direct[i]);
+			befs_debug(sb, "  direct %d [%u, %hu, %hu]", i,
+				   tmp_run.allocation_group, tmp_run.start,
+				   tmp_run.len);
+		}
+		befs_debug(sb, "  max_direct_range %llu",
+			   fs64_to_cpu(sb,
+				       inode->data.datastream.
+				       max_direct_range));
+
+		tmp_run = fsrun_to_cpu(sb, inode->data.datastream.indirect);
+		befs_debug(sb, "  indirect [%u, %hu, %hu]",
+			   tmp_run.allocation_group,
+			   tmp_run.start, tmp_run.len);
+
+		befs_debug(sb, "  max_indirect_range %llu",
+			   fs64_to_cpu(sb,
+				       inode->data.datastream.
+				       max_indirect_range));
+
+		tmp_run =
+		    fsrun_to_cpu(sb, inode->data.datastream.double_indirect);
+		befs_debug(sb, "  double indirect [%u, %hu, %hu]",
+			   tmp_run.allocation_group, tmp_run.start,
+			   tmp_run.len);
+
+		befs_debug(sb, "  max_double_indirect_range %llu",
+			   fs64_to_cpu(sb,
+				       inode->data.datastream.
+				       max_double_indirect_range));
+
+		befs_debug(sb, "  size %llu",
+			   fs64_to_cpu(sb, inode->data.datastream.size));
+	}
+
+#endif				//CONFIG_BEFS_DEBUG
+}
+
+/*
+ * Display super block structure for debug.
+ */
+
+void
+befs_dump_super_block(const struct super_block *sb, befs_super_block *sup)
+{
+#ifdef CONFIG_BEFS_DEBUG
+
+	befs_block_run tmp_run;
+
+	befs_debug(sb, "befs_super_block information");
+
+	befs_debug(sb, "  name %s", sup->name);
+	befs_debug(sb, "  magic1 %08x", fs32_to_cpu(sb, sup->magic1));
+	befs_debug(sb, "  fs_byte_order %08x",
+		   fs32_to_cpu(sb, sup->fs_byte_order));
+
+	befs_debug(sb, "  block_size %u", fs32_to_cpu(sb, sup->block_size));
+	befs_debug(sb, "  block_shift %u", fs32_to_cpu(sb, sup->block_shift));
+
+	befs_debug(sb, "  num_blocks %llu", fs64_to_cpu(sb, sup->num_blocks));
+	befs_debug(sb, "  used_blocks %llu", fs64_to_cpu(sb, sup->used_blocks));
+	befs_debug(sb, "  inode_size %u", fs32_to_cpu(sb, sup->inode_size));
+
+	befs_debug(sb, "  magic2 %08x", fs32_to_cpu(sb, sup->magic2));
+	befs_debug(sb, "  blocks_per_ag %u",
+		   fs32_to_cpu(sb, sup->blocks_per_ag));
+	befs_debug(sb, "  ag_shift %u", fs32_to_cpu(sb, sup->ag_shift));
+	befs_debug(sb, "  num_ags %u", fs32_to_cpu(sb, sup->num_ags));
+
+	befs_debug(sb, "  flags %08x", fs32_to_cpu(sb, sup->flags));
+
+	tmp_run = fsrun_to_cpu(sb, sup->log_blocks);
+	befs_debug(sb, "  log_blocks %u, %hu, %hu",
+		   tmp_run.allocation_group, tmp_run.start, tmp_run.len);
+
+	befs_debug(sb, "  log_start %lld", fs64_to_cpu(sb, sup->log_start));
+	befs_debug(sb, "  log_end %lld", fs64_to_cpu(sb, sup->log_end));
+
+	befs_debug(sb, "  magic3 %08x", fs32_to_cpu(sb, sup->magic3));
+
+	tmp_run = fsrun_to_cpu(sb, sup->root_dir);
+	befs_debug(sb, "  root_dir %u, %hu, %hu",
+		   tmp_run.allocation_group, tmp_run.start, tmp_run.len);
+
+	tmp_run = fsrun_to_cpu(sb, sup->indices);
+	befs_debug(sb, "  indices %u, %hu, %hu",
+		   tmp_run.allocation_group, tmp_run.start, tmp_run.len);
+
+#endif				//CONFIG_BEFS_DEBUG
+}
+
+#if 0
+/* unused */
+void
+befs_dump_small_data(const struct super_block *sb, befs_small_data *sd)
+{
+}
+
+/* unused */
+void
+befs_dump_run(const struct super_block *sb, befs_disk_block_run run)
+{
+#ifdef CONFIG_BEFS_DEBUG
+
+	befs_block_run n = fsrun_to_cpu(sb, run);
+
+	befs_debug(sb, "[%u, %hu, %hu]", n.allocation_group, n.start, n.len);
+
+#endif				//CONFIG_BEFS_DEBUG
+}
+#endif  /*  0  */
+
+void
+befs_dump_index_entry(const struct super_block *sb,
+		      befs_disk_btree_super *super)
+{
+#ifdef CONFIG_BEFS_DEBUG
+
+	befs_debug(sb, "Btree super structure");
+	befs_debug(sb, "  magic %08x", fs32_to_cpu(sb, super->magic));
+	befs_debug(sb, "  node_size %u", fs32_to_cpu(sb, super->node_size));
+	befs_debug(sb, "  max_depth %08x", fs32_to_cpu(sb, super->max_depth));
+
+	befs_debug(sb, "  data_type %08x", fs32_to_cpu(sb, super->data_type));
+	befs_debug(sb, "  root_node_pointer %016LX",
+		   fs64_to_cpu(sb, super->root_node_ptr));
+	befs_debug(sb, "  free_node_pointer %016LX",
+		   fs64_to_cpu(sb, super->free_node_ptr));
+	befs_debug(sb, "  maximum size %016LX",
+		   fs64_to_cpu(sb, super->max_size));
+
+#endif				//CONFIG_BEFS_DEBUG
+}
+
+void
+befs_dump_index_node(const struct super_block *sb, befs_btree_nodehead *node)
+{
+#ifdef CONFIG_BEFS_DEBUG
+
+	befs_debug(sb, "Btree node structure");
+	befs_debug(sb, "  left %016LX", fs64_to_cpu(sb, node->left));
+	befs_debug(sb, "  right %016LX", fs64_to_cpu(sb, node->right));
+	befs_debug(sb, "  overflow %016LX", fs64_to_cpu(sb, node->overflow));
+	befs_debug(sb, "  all_key_count %hu",
+		   fs16_to_cpu(sb, node->all_key_count));
+	befs_debug(sb, "  all_key_length %hu",
+		   fs16_to_cpu(sb, node->all_key_length));
+
+#endif				//CONFIG_BEFS_DEBUG
+}