求助各位，实在找不到解决方案,不知道大家有没有遇到过类似的问题，

问题描述

我是在一个 arm 环境的 ubuntu 工控机上跑一个 springboot 些的 java 后端程序，会不时遇到 jvm 崩溃，而且多在调用接口的时候，hs_err_pid*.log 文件显示是在[libpthread.so.0+0x755c]start_thread+0xbc 这个位置, 分析 core dump 文件也是在这个位置崩溃的，

可疑的诱因：我这个后端程序会调用一个第三方提供的 so 库文件连接局域网一个相机设备，当相机不在线时我的程序就会变得容易崩溃，但我单独使用同一个 jdk 写一个定时调用该 so 库测试的 demo 时候又没有崩溃的情况发生，不知道是不是我的后端程序有更多复杂业务的原因

使用的 java 版本

java version "17.0.10" 2024-01-16 LTS Java(TM) SE Runtime Environment (build 17.0.10+11-LTS-240) Java HotSpot(TM) 64-Bit Server VM (build 17.0.10+11-LTS-240, mixed mode, sharing)

ubuntu 版本 20.04.5 LTS ，

Linux linux 5.10.110 #2 SMP Mon May 6 16:55:18 CST 2024 aarch64 aarch64 aarch64 GNU/Linux

gdb 分析 core dump 崩溃点的结果：

(gdb) bt
#0  __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:50
#1  0x0000007fa0695aac in __GI_abort () at abort.c:79
#2  0x0000007f9ffafc38 in os::abort(bool, void*, void const*) () from /userdata/hifi/jdk/lib/server/libjvm.so
#3  0x0000007fa01a5f3c in VMError::report_and_die(int, char const*, char const*, std::__va_list, Thread*, unsigned char*, void*, void*, char const*, int, unsigned long)
    () from /userdata/hifi/jdk/lib/server/libjvm.so
#4  0x0000007fa01a6908 in VMError::report_and_die(Thread*, unsigned int, unsigned char*, void*, void*, char const*, ...) () from /userdata/hifi/jdk/lib/server/libjvm.so
#5  <signal handler called>
#6  0x0000007f9ff9b2f0 in os::is_first_C_frame(frame*) () from /userdata/hifi/jdk/lib/server/libjvm.so
#7  0x0000007fa01a0f34 in next_frame(frame, Thread*) () from /userdata/hifi/jdk/lib/server/libjvm.so
#8  0x0000007fa01a1fd8 in VMError::print_native_stack(outputStream*, frame, Thread*, char*, int) () from /userdata/hifi/jdk/lib/server/libjvm.so
#9  0x0000007fa01a3738 in VMError::report(outputStream*, bool) () from /userdata/hifi/jdk/lib/server/libjvm.so
#10 0x0000007fa01a5e30 in VMError::report_and_die(int, char const*, char const*, std::__va_list, Thread*, unsigned char*, void*, void*, char const*, int, unsigned long)
    () from /userdata/hifi/jdk/lib/server/libjvm.so
#11 0x0000007fa01a6908 in VMError::report_and_die(Thread*, unsigned int, unsigned char*, void*, void*, char const*, ...) () from /userdata/hifi/jdk/lib/server/libjvm.so
#12 0x0000007fa00623bc in JVM_handle_linux_signal () from /userdata/hifi/jdk/lib/server/libjvm.so
#13 <signal handler called>
#14 start_thread (arg=<error reading variable: Cannot access memory at address 0x4a28>) at pthread_create.c:478
Backtrace stopped: Cannot access memory at address 0x49c8

cppc

38 天前

分享一点经验：

## C 风格字符串

需要注意 C 风格字符串是靠 null-byte(0x0)标记字符串长度

## 结构体传参

要注意是 ByReference 还是 ByValue ，核对 C 端的函数原型

## 结构体内存布局

要注意对齐方式,仔细核对 C 端的结构体声明

## 结构体 memory dump

开启 `-Djna.dump_memory=true`

> Normally when you invoke toString on a Structure, it will print each defined field with its calculated memory offset. If when launching the VM, you pass it "-Djna.dump_memory=true", toString will also dump the contents of the corresponding native memory. This is useful to determine if you've added or omitted a field, or chosen an incorrect size. Viewing the memory as bytes usually makes it clear where field boundaries should be, assuming the memory has been initialized by native code.

## 日志

对 C 端的函数调用做一下封装,调用前后记录日志，先把崩溃的位置找到

[求助]关于一个 jvm 崩溃的问题

问题描述