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, 71 insertions, 0 deletions

diff --git a/fs/btrfs/lru_cache.h b/fs/btrfs/lru_cache.h
new file mode 100644
index 000000000..07f1bb1c6
--- /dev/null
+++ b/fs/btrfs/lru_cache.h
@@ -0,0 +1,71 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#ifndef BTRFS_LRU_CACHE_H
+#define BTRFS_LRU_CACHE_H
+
+#include <linux/types.h>
+#include <linux/maple_tree.h>
+#include <linux/list.h>
+
+/*
+ * A cache entry. This is meant to be embedded in a structure of a user of
+ * this module. Similar to how struct list_head and struct rb_node are used.
+ *
+ * Note: it should be embedded as the first element in a struct (offset 0), and
+ * this module assumes it was allocated with kmalloc(), so it calls kfree() when
+ * it needs to free an entry.
+ */
+struct btrfs_lru_cache_entry {
+	struct list_head lru_list;
+	u64 key;
+	/*
+	 * Optional generation associated to a key. Use 0 if not needed/used.
+	 * Entries with the same key and different generations are stored in a
+	 * linked list, so use this only for cases where there's a small number
+	 * of different generations.
+	 */
+	u64 gen;
+	/*
+	 * The maple tree uses unsigned long type for the keys, which is 32 bits
+	 * on 32 bits systems, and 64 bits on 64 bits systems. So if we want to
+	 * use something like inode numbers as keys, which are always a u64, we
+	 * have to deal with this in a special way - we store the key in the
+	 * entry itself, as a u64, and the values inserted into the maple tree
+	 * are linked lists of entries - so in case we are on a 64 bits system,
+	 * that list always has a single entry, while on 32 bits systems it
+	 * may have more than one, with each entry having the same value for
+	 * their lower 32 bits of the u64 key.
+	 */
+	struct list_head list;
+};
+
+struct btrfs_lru_cache {
+	struct list_head lru_list;
+	struct maple_tree entries;
+	/* Number of entries stored in the cache. */
+	unsigned int size;
+	/* Maximum number of entries the cache can have. */
+	unsigned int max_size;
+};
+
+#define btrfs_lru_cache_for_each_entry_safe(cache, entry, tmp)		\
+	list_for_each_entry_safe_reverse((entry), (tmp), &(cache)->lru_list, lru_list)
+
+static inline struct btrfs_lru_cache_entry *btrfs_lru_cache_lru_entry(
+					      struct btrfs_lru_cache *cache)
+{
+	return list_first_entry_or_null(&cache->lru_list,
+					struct btrfs_lru_cache_entry, lru_list);
+}
+
+void btrfs_lru_cache_init(struct btrfs_lru_cache *cache, unsigned int max_size);
+struct btrfs_lru_cache_entry *btrfs_lru_cache_lookup(struct btrfs_lru_cache *cache,
+						     u64 key, u64 gen);
+int btrfs_lru_cache_store(struct btrfs_lru_cache *cache,
+			  struct btrfs_lru_cache_entry *new_entry,
+			  gfp_t gfp);
+void btrfs_lru_cache_remove(struct btrfs_lru_cache *cache,
+			    struct btrfs_lru_cache_entry *entry);
+void btrfs_lru_cache_clear(struct btrfs_lru_cache *cache);
+
+#endif