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 --- fs/ntfs/time.h | 87 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 87 insertions(+) create mode 100644 fs/ntfs/time.h (limited to 'fs/ntfs/time.h') diff --git a/fs/ntfs/time.h b/fs/ntfs/time.h new file mode 100644 index 000000000..aeed195bd --- /dev/null +++ b/fs/ntfs/time.h @@ -0,0 +1,87 @@ +/* SPDX-License-Identifier: GPL-2.0-or-later */ +/* + * NTFS time conversion functions. + * + * Copyright (c) 2001-2005 Anton Altaparmakov + */ + +#ifndef _LINUX_NTFS_TIME_H +#define _LINUX_NTFS_TIME_H + +#include +#include /* For do_div(). */ + +#define NTFS_TIME_OFFSET ((s64)(369 * 365 + 89) * 24 * 3600) + +/* + * utc2ntfs - convert Linux UTC time to NTFS time + * @ts: Linux UTC time to convert to NTFS time + * + * Convert the Linux UTC time @ts to its corresponding NTFS time and return + * that in little endian format. + * + * Linux stores time in a struct timespec64 consisting of a time64_t tv_sec + * and a long tv_nsec where tv_sec is the number of 1-second intervals since + * 1st January 1970, 00:00:00 UTC and tv_nsec is the number of 1-nano-second + * intervals since the value of tv_sec. + * + * NTFS uses Microsoft's standard time format which is stored in a s64 and is + * measured as the number of 100-nano-second intervals since 1st January 1601, + * 00:00:00 UTC. + */ +static inline __le64 utc2ntfs(const struct timespec64 ts) +{ + /* + * Convert the seconds to 100ns intervals, add the nano-seconds + * converted to 100ns intervals, and then add the NTFS time offset. + */ + return cpu_to_le64((u64)(ts.tv_sec + NTFS_TIME_OFFSET) * 10000000 + + ts.tv_nsec / 100); +} + +/* + * get_current_ntfs_time - get the current time in little endian NTFS format + * + * Get the current time from the Linux kernel, convert it to its corresponding + * NTFS time and return that in little endian format. + */ +static inline __le64 get_current_ntfs_time(void) +{ + struct timespec64 ts; + + ktime_get_coarse_real_ts64(&ts); + return utc2ntfs(ts); +} + +/* + * ntfs2utc - convert NTFS time to Linux time + * @time: NTFS time (little endian) to convert to Linux UTC + * + * Convert the little endian NTFS time @time to its corresponding Linux UTC + * time and return that in cpu format. + * + * Linux stores time in a struct timespec64 consisting of a time64_t tv_sec + * and a long tv_nsec where tv_sec is the number of 1-second intervals since + * 1st January 1970, 00:00:00 UTC and tv_nsec is the number of 1-nano-second + * intervals since the value of tv_sec. + * + * NTFS uses Microsoft's standard time format which is stored in a s64 and is + * measured as the number of 100 nano-second intervals since 1st January 1601, + * 00:00:00 UTC. + */ +static inline struct timespec64 ntfs2utc(const __le64 time) +{ + struct timespec64 ts; + s32 t32; + + /* Subtract the NTFS time offset. */ + s64 t = le64_to_cpu(time) - NTFS_TIME_OFFSET * 10000000; + /* + * Convert the time to 1-second intervals and the remainder to + * 1-nano-second intervals. + */ + ts.tv_sec = div_s64_rem(t, 10000000, &t32); + ts.tv_nsec = t32 * 100; + return ts; +} +#endif /* _LINUX_NTFS_TIME_H */ -- cgit v1.3.1