From a5f7a5bb3b7f28ba7e4fa246775b29a0e5537255 Mon Sep 17 00:00:00 2001 From: Jinke Han Date: Tue, 8 Sep 2026 15:37:42 +0800 Subject: x86/kprobes: Fix crash when probing CS CALL instructions 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) Signed-off-by: Jinke Han Signed-off-by: Ingo Molnar Reviewed-by: Masami Hiramatsu (Google) Acked-by: Yafang Shao Acked-by: Borislav Petkov Cc: Peter Zijlstra Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000 --- Documentation/crypto/crypto_engine.rst | 80 ++++++++++++++++++++++++++++++++++ 1 file changed, 80 insertions(+) create mode 100644 Documentation/crypto/crypto_engine.rst (limited to 'Documentation/crypto/crypto_engine.rst') diff --git a/Documentation/crypto/crypto_engine.rst b/Documentation/crypto/crypto_engine.rst new file mode 100644 index 000000000..7ef850e28 --- /dev/null +++ b/Documentation/crypto/crypto_engine.rst @@ -0,0 +1,80 @@ +.. SPDX-License-Identifier: GPL-2.0 + +Crypto Engine +============= + +Overview +-------- +The crypto engine (CE) API is a crypto queue manager. + +Requirement +----------- +You must put, at the start of your transform context your_tfm_ctx, the structure +crypto_engine: + +:: + + struct your_tfm_ctx { + struct crypto_engine engine; + ... + }; + +The crypto engine only manages asynchronous requests in the form of +crypto_async_request. It cannot know the underlying request type and thus only +has access to the transform structure. It is not possible to access the context +using container_of. In addition, the engine knows nothing about your +structure "``struct your_tfm_ctx``". The engine assumes (requires) the placement +of the known member ``struct crypto_engine`` at the beginning. + +Order of operations +------------------- +You are required to obtain a struct crypto_engine via ``crypto_engine_alloc_init()``. +Start it via ``crypto_engine_start()``. When finished with your work, shut down the +engine using ``crypto_engine_stop()`` and destroy the engine with +``crypto_engine_exit()``. + +Before transferring any request, you have to fill the context enginectx by +providing functions for the following: + +* ``prepare_cipher_request``/``prepare_hash_request``: Called before each + corresponding request is performed. If some processing or other preparatory + work is required, do it here. + +* ``unprepare_cipher_request``/``unprepare_hash_request``: Called after each + request is handled. Clean up / undo what was done in the prepare function. + +* ``cipher_one_request``/``hash_one_request``: Handle the current request by + performing the operation. + +Note that these functions access the crypto_async_request structure +associated with the received request. You are able to retrieve the original +request by using: + +:: + + container_of(areq, struct yourrequesttype_request, base); + +When your driver receives a crypto_request, you must to transfer it to +the crypto engine via one of: + +* crypto_transfer_aead_request_to_engine() + +* crypto_transfer_akcipher_request_to_engine() + +* crypto_transfer_hash_request_to_engine() + +* crypto_transfer_kpp_request_to_engine() + +* crypto_transfer_skcipher_request_to_engine() + +At the end of the request process, a call to one of the following functions is needed: + +* crypto_finalize_aead_request() + +* crypto_finalize_akcipher_request() + +* crypto_finalize_hash_request() + +* crypto_finalize_kpp_request() + +* crypto_finalize_skcipher_request() -- cgit v1.3.1