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 --- drivers/crypto/ccp/tee-dev.h | 116 +++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 116 insertions(+) create mode 100644 drivers/crypto/ccp/tee-dev.h (limited to 'drivers/crypto/ccp/tee-dev.h') diff --git a/drivers/crypto/ccp/tee-dev.h b/drivers/crypto/ccp/tee-dev.h new file mode 100644 index 000000000..c23416cb7 --- /dev/null +++ b/drivers/crypto/ccp/tee-dev.h @@ -0,0 +1,116 @@ +/* SPDX-License-Identifier: MIT */ +/* + * Copyright (C) 2019,2021 Advanced Micro Devices, Inc. + * + * Author: Rijo Thomas + * Author: Devaraj Rangasamy + * + */ + +/* This file describes the TEE communication interface between host and AMD + * Secure Processor + */ + +#ifndef __TEE_DEV_H__ +#define __TEE_DEV_H__ + +#include +#include + +#define TEE_DEFAULT_CMD_TIMEOUT (10 * MSEC_PER_SEC) +#define TEE_DEFAULT_RING_TIMEOUT 10 +#define MAX_BUFFER_SIZE 988 + +/** + * struct tee_init_ring_cmd - Command to init TEE ring buffer + * @low_addr: bits [31:0] of the physical address of ring buffer + * @hi_addr: bits [63:32] of the physical address of ring buffer + * @size: size of ring buffer in bytes + */ +struct tee_init_ring_cmd { + u32 low_addr; + u32 hi_addr; + u32 size; +}; + +#define MAX_RING_BUFFER_ENTRIES 32 + +/** + * struct ring_buf_manager - Helper structure to manage ring buffer. + * @ring_start: starting address of ring buffer + * @ring_size: size of ring buffer in bytes + * @ring_pa: physical address of ring buffer + * @wptr: index to the last written entry in ring buffer + */ +struct ring_buf_manager { + struct mutex mutex; /* synchronizes access to ring buffer */ + void *ring_start; + u32 ring_size; + phys_addr_t ring_pa; + u32 wptr; +}; + +struct psp_tee_device { + struct device *dev; + struct psp_device *psp; + void __iomem *io_regs; + struct tee_vdata *vdata; + struct ring_buf_manager rb_mgr; +}; + +/** + * enum tee_cmd_state - TEE command states for the ring buffer interface + * @TEE_CMD_STATE_INIT: initial state of command when sent from host + * @TEE_CMD_STATE_PROCESS: command being processed by TEE environment + * @TEE_CMD_STATE_COMPLETED: command processing completed + */ +enum tee_cmd_state { + TEE_CMD_STATE_INIT, + TEE_CMD_STATE_PROCESS, + TEE_CMD_STATE_COMPLETED, +}; + +/** + * enum cmd_resp_state - TEE command's response status maintained by driver + * @CMD_RESPONSE_INVALID: initial state when no command is written to ring + * @CMD_WAITING_FOR_RESPONSE: driver waiting for response from TEE + * @CMD_RESPONSE_TIMEDOUT: failed to get response from TEE + * @CMD_RESPONSE_COPIED: driver has copied response from TEE + */ +enum cmd_resp_state { + CMD_RESPONSE_INVALID, + CMD_WAITING_FOR_RESPONSE, + CMD_RESPONSE_TIMEDOUT, + CMD_RESPONSE_COPIED, +}; + +/** + * struct tee_ring_cmd - Structure of the command buffer in TEE ring + * @cmd_id: refers to &enum tee_cmd_id. Command id for the ring buffer + * interface + * @cmd_state: refers to &enum tee_cmd_state + * @status: status of TEE command execution + * @res0: reserved region + * @pdata: private data (currently unused) + * @res1: reserved region + * @buf: TEE command specific buffer + * @flag: refers to &enum cmd_resp_state + */ +struct tee_ring_cmd { + u32 cmd_id; + u32 cmd_state; + u32 status; + u32 res0[1]; + u64 pdata; + u32 res1[2]; + u8 buf[MAX_BUFFER_SIZE]; + u32 flag; + + /* Total size: 1024 bytes */ +} __packed; + +int tee_dev_init(struct psp_device *psp); +void tee_dev_destroy(struct psp_device *psp); +int tee_restore(struct psp_device *psp); + +#endif /* __TEE_DEV_H__ */ -- cgit v1.3.1