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 --- samples/pidfd/pidfd-metadata.c | 120 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 120 insertions(+) create mode 100644 samples/pidfd/pidfd-metadata.c (limited to 'samples/pidfd/pidfd-metadata.c') diff --git a/samples/pidfd/pidfd-metadata.c b/samples/pidfd/pidfd-metadata.c new file mode 100644 index 000000000..c459155da --- /dev/null +++ b/samples/pidfd/pidfd-metadata.c @@ -0,0 +1,120 @@ +// SPDX-License-Identifier: GPL-2.0 + +#define _GNU_SOURCE +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#ifndef CLONE_PIDFD +#define CLONE_PIDFD 0x00001000 +#endif + +#ifndef __NR_pidfd_send_signal +#define __NR_pidfd_send_signal -1 +#endif + +static int do_child(void *args) +{ + printf("%d\n", getpid()); + _exit(EXIT_SUCCESS); +} + +static pid_t pidfd_clone(int flags, int *pidfd) +{ + size_t stack_size = 1024; + char *stack[1024] = { 0 }; + +#ifdef __ia64__ + return __clone2(do_child, stack, stack_size, flags | SIGCHLD, NULL, pidfd); +#else + return clone(do_child, stack + stack_size, flags | SIGCHLD, NULL, pidfd); +#endif +} + +static inline int sys_pidfd_send_signal(int pidfd, int sig, siginfo_t *info, + unsigned int flags) +{ + return syscall(__NR_pidfd_send_signal, pidfd, sig, info, flags); +} + +static int pidfd_metadata_fd(pid_t pid, int pidfd) +{ + int procfd, ret; + char path[100]; + + snprintf(path, sizeof(path), "/proc/%d", pid); + procfd = open(path, O_DIRECTORY | O_RDONLY | O_CLOEXEC); + if (procfd < 0) { + warn("Failed to open %s\n", path); + return -1; + } + + /* + * Verify that the pid has not been recycled and our /proc/ handle + * is still valid. + */ + ret = sys_pidfd_send_signal(pidfd, 0, NULL, 0); + if (ret < 0) { + switch (errno) { + case EPERM: + /* Process exists, just not allowed to signal it. */ + break; + default: + warn("Failed to signal process\n"); + close(procfd); + procfd = -1; + } + } + + return procfd; +} + +int main(int argc, char *argv[]) +{ + int pidfd = -1, ret = EXIT_FAILURE; + char buf[4096] = { 0 }; + pid_t pid; + int procfd, statusfd; + ssize_t bytes; + + pid = pidfd_clone(CLONE_PIDFD, &pidfd); + if (pid < 0) + err(ret, "CLONE_PIDFD"); + if (pidfd == -1) { + warnx("CLONE_PIDFD is not supported by the kernel"); + goto out; + } + + procfd = pidfd_metadata_fd(pid, pidfd); + close(pidfd); + if (procfd < 0) + goto out; + + statusfd = openat(procfd, "status", O_RDONLY | O_CLOEXEC); + close(procfd); + if (statusfd < 0) + goto out; + + bytes = read(statusfd, buf, sizeof(buf)); + if (bytes > 0) + bytes = write(STDOUT_FILENO, buf, bytes); + close(statusfd); + ret = EXIT_SUCCESS; + +out: + (void)wait(NULL); + + exit(ret); +} -- cgit v1.3.1