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, 78 insertions, 0 deletions

diff --git a/drivers/w1/w1_internal.h b/drivers/w1/w1_internal.h
new file mode 100644
index 000000000..cb5a74e99
--- /dev/null
+++ b/drivers/w1/w1_internal.h
@@ -0,0 +1,78 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * Copyright (c) 2004 Evgeniy Polyakov <zbr@ioremap.net>
+ */
+
+#ifndef __W1_H
+#define __W1_H
+
+#include <linux/w1.h>
+
+#include <linux/completion.h>
+#include <linux/mutex.h>
+
+#define W1_SLAVE_ACTIVE		0
+#define W1_SLAVE_DETACH		1
+
+/**
+ * struct w1_async_cmd - execute callback from the w1_process kthread
+ * @async_entry: link entry
+ * @cb: callback function, must list_del and destroy this list before
+ * returning
+ *
+ * When inserted into the w1_master async_list, w1_process will execute
+ * the callback.  Embed this into the structure with the command details.
+ */
+struct w1_async_cmd {
+	struct list_head	async_entry;
+	void (*cb)(struct w1_master *dev, struct w1_async_cmd *async_cmd);
+};
+
+int w1_create_master_attributes(struct w1_master *master);
+void w1_destroy_master_attributes(struct w1_master *master);
+void w1_search(struct w1_master *dev, u8 search_type,
+	       w1_slave_found_callback cb);
+void w1_search_devices(struct w1_master *dev, u8 search_type,
+		       w1_slave_found_callback cb);
+/* call w1_unref_slave to release the reference counts w1_search_slave added */
+struct w1_slave *w1_search_slave(struct w1_reg_num *id);
+/*
+ * decrements the reference on sl->master and sl, and cleans up if zero
+ * returns the reference count after it has been decremented
+ */
+int w1_unref_slave(struct w1_slave *sl);
+void w1_slave_found(struct w1_master *dev, u64 rn);
+void w1_search_process_cb(struct w1_master *dev, u8 search_type,
+			  w1_slave_found_callback cb);
+struct w1_slave *w1_slave_search_device(struct w1_master *dev,
+					struct w1_reg_num *rn);
+struct w1_master *w1_search_master_id(u32 id);
+
+/* Disconnect and reconnect devices in the given family.  Used for finding
+ * unclaimed devices after a family has been registered or releasing devices
+ * after a family has been unregistered.  Set attach to 1 when a new family
+ * has just been registered, to 0 when it has been unregistered.
+ */
+void w1_reconnect_slaves(struct w1_family *f, int attach);
+int w1_attach_slave_device(struct w1_master *dev, struct w1_reg_num *rn);
+/* 0 success, otherwise EBUSY */
+int w1_slave_detach(struct w1_slave *sl);
+
+void __w1_remove_master_device(struct w1_master *dev);
+
+void w1_family_put(struct w1_family *f);
+void __w1_family_get(struct w1_family *f);
+struct w1_family *w1_family_registered(u8 fid);
+
+extern struct device_driver w1_master_driver;
+extern struct device w1_master_device;
+extern int w1_max_slave_count;
+extern int w1_max_slave_ttl;
+extern struct list_head w1_masters;
+extern struct mutex w1_mlock;
+extern spinlock_t w1_flock;
+
+int w1_process_callbacks(struct w1_master *dev);
+int w1_process(void *data);
+
+#endif /* __W1_H */