[Bug]: A failed command exits with 0 when core_exit.sh is downloaded on exit

未关闭
  • 哪里不对？
#5,191 0 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

维护者通常 1 天内回复

已经有一个关联 PR 被合并了。

  • #5192 来自 @borzaka —— 已合并

评估

难度
3/5
预计耗时
1-2 天
新手友好度
35/100
Issue 类型
缺陷
描述清晰度
描述清楚
活跃度
停滞
技术栈
bash, linux, shell
领域
backend, cli

调研方向

Read core_modules.sh and core_dl.sh, focusing on core_exit.sh(), fn_fetch_module, and fn_fetch_file around the local exitcode declaration. Reproduce with a fresh instance and the reported start command inside tmux, then compare its exit code before and after core_exit.sh is present. Done means the first failing run returns its original nonzero code; linked pull request #5192 is merged.

由索引模型根据 Issue 内容生成。

描述

command: start distro: Debian engine: null game: Teamspeak 3 info: tmux reproducible: always severity: low status: fixed in develop type: bug
Severity

severity: low

Reproducibility

reproducible: always

Regression

regression: no

Affects latest release

latest-release: yes

User story

As a server admin who runs LinuxGSM from scripts or cron, I want a failed command to return a non-zero exit code on a new instance too, so that the failure is not reported as success.

Script name

ts3server (any server is affected)

Game

Teamspeak 3

Linux distro

Debian 13

Command

command: start

Expected behavior

The command exits with the code it set, as it does on later runs (1 for the tmuxception error below).

Actual behavior

If lgsm/modules/core_exit.sh has not been downloaded yet, it is downloaded on exit and the command exits with 0. The same command exits with 1 once the file exists.

Further information

core_exit.sh() in core_modules.sh loads the module with fn_fetch_module, and fn_fetch_file (core_dl.sh) sources it inside its own function. When the file is missing, fn_fetch_file first declares local exitcode="" (line 505), which the curl result and fn_script_log_pass "Downloading ..." set to 0, so core_exit.sh exits with that local instead of the caller's code. Once the file exists the download branch is skipped, so only the exit that downloads core_exit.sh is affected, usually the first failing command on a new instance. The pattern is already in v21.3.2; #4159 only changed local exitcode=$? to local exitcode="". A fix is to download core_exit.sh without running it, keep exitcode, and then load it as before. Related but different exit code fixes: #5060, #5109, #5113.

Pre-checks
  • I ran update and validate before reporting this issue.

Not run: found by code review and reproduced on a new develop instance, which has nothing to update or validate yet.

Relevant log output
# New ts3server instance from develop (11103431), Debian 13 (WSL2), run inside a tmux session.
$ ./ts3server start; echo $?
[ FAIL ] Starting ts3server: tmuxception error: Sorry Cobb you cannot start a tmux session inside of a tmux session.
...
fetching GitHub [ core_exit.sh ] ... OK
0
$ ./ts3server start; echo $?
[ FAIL ] Starting ts3server: tmuxception error: Sorry Cobb you cannot start a tmux session inside of a tmux session.
...
1
Steps to reproduce
  1. Create a new instance: curl -Lo linuxgsm.sh https://raw.githubusercontent.com/GameServerManagers/LinuxGSM/develop/linuxgsm.sh && chmod +x linuxgsm.sh && bash linuxgsm.sh ts3server.
  2. Inside a tmux session, run LGSM_GITHUBBRANCH=develop ./ts3server start; echo $?. It fails with the tmuxception error, fetches core_exit.sh last and prints 0.
  3. Run the same command again. It prints 1.

🤖 Generated with Claude Code

主要语言
Shell
星标
4.9k
派生
870
平均合并
18 小时 44 分钟
30 天内合并 PR
148

环境准备

在 Codespaces 中打开

在浏览器里用你自己的 GitHub 账号启动这个项目的开发容器。

从这里开始

  1. 先读完整个 Issue，再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库，在一个分支上完成修改。
  4. 提交 Pull Request，并在描述里引用这个 Issue 编号。

GameServerManagers/LinuxGSM 的其他 Issue

查看 GameServerManagers/LinuxGSM 的全部 Issue

相似的 Issue

更多 Shell/Bash Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。