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 --- include/uapi/asm-generic/statfs.h | 84 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 84 insertions(+) create mode 100644 include/uapi/asm-generic/statfs.h (limited to 'include/uapi/asm-generic/statfs.h') diff --git a/include/uapi/asm-generic/statfs.h b/include/uapi/asm-generic/statfs.h new file mode 100644 index 000000000..bd35c7619 --- /dev/null +++ b/include/uapi/asm-generic/statfs.h @@ -0,0 +1,84 @@ +/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */ +#ifndef _UAPI_GENERIC_STATFS_H +#define _UAPI_GENERIC_STATFS_H + +#include + + +/* + * Most 64-bit platforms use 'long', while most 32-bit platforms use '__u32'. + * Yes, they differ in signedness as well as size. + * Special cases can override it for themselves -- except for S390x, which + * is just a little too special for us. And MIPS, which I'm not touching + * with a 10' pole. + */ +#ifndef __statfs_word +#if __BITS_PER_LONG == 64 +#define __statfs_word __kernel_long_t +#else +#define __statfs_word __u32 +#endif +#endif + +struct statfs { + __statfs_word f_type; + __statfs_word f_bsize; + __statfs_word f_blocks; + __statfs_word f_bfree; + __statfs_word f_bavail; + __statfs_word f_files; + __statfs_word f_ffree; + __kernel_fsid_t f_fsid; + __statfs_word f_namelen; + __statfs_word f_frsize; + __statfs_word f_flags; + __statfs_word f_spare[4]; +}; + +/* + * ARM needs to avoid the 32-bit padding at the end, for consistency + * between EABI and OABI + */ +#ifndef ARCH_PACK_STATFS64 +#define ARCH_PACK_STATFS64 +#endif + +struct statfs64 { + __statfs_word f_type; + __statfs_word f_bsize; + __u64 f_blocks; + __u64 f_bfree; + __u64 f_bavail; + __u64 f_files; + __u64 f_ffree; + __kernel_fsid_t f_fsid; + __statfs_word f_namelen; + __statfs_word f_frsize; + __statfs_word f_flags; + __statfs_word f_spare[4]; +} ARCH_PACK_STATFS64; + +/* + * IA64 and x86_64 need to avoid the 32-bit padding at the end, + * to be compatible with the i386 ABI + */ +#ifndef ARCH_PACK_COMPAT_STATFS64 +#define ARCH_PACK_COMPAT_STATFS64 +#endif + +struct compat_statfs64 { + __u32 f_type; + __u32 f_bsize; + __u64 f_blocks; + __u64 f_bfree; + __u64 f_bavail; + __u64 f_files; + __u64 f_ffree; + __kernel_fsid_t f_fsid; + __u32 f_namelen; + __u32 f_frsize; + __u32 f_flags; + __u32 f_spare[4]; +} ARCH_PACK_COMPAT_STATFS64; + +#endif /* _UAPI_GENERIC_STATFS_H */ -- cgit v1.3.1