在规范步骤中以编程方式失败的RPM安装 [英] Failing an RPM install programmatically in a spec step

问题描述

我正在制作RPM.这个特定的RPM具有不能表示为RPM前提条件的要求，因此可以将它们称为特定的文件系统/磁盘配置.目前，失败是在安装后，运行时未满足要求的情况下发生的.

I'm making an RPM. This particular RPM has requirements that can't be expressed as RPM prerequisites, lets call them a particular filesystem/disk configuration. Currently the failure happens after install, at runtime, when the requirements aren't met.

我可以在脚本的％install，部分中检查所需的先决条件.但是，如果满足某些条件，我无法弄清楚如何使安装失败.是否可以通过％install(或其他)部分中的某些触发器在运行时使rpm安装失败?

I can check for the required prerequisites in the %install, section of my script. However, I can't figure out how to fail the install if certain criteria are met. Is it possible to fail an rpm install at runtime via some trigger in the %install (or some other) section?

在.spec文件中，示例如下所示:

An example would look something like so, in a .spec file:

%install
if [ -f /some/file ]
then
    FAIL_RPM_INSTALL ## What is this command?
fi

推荐答案

事实证明，如果退出%pre阶段，则rpm安装将失败.

It turns out that if you exit in the %pre stage the rpm install will fail.

%pre
if [ -f /some/file ]
then
    echo "/some/file exists, it shouldn't"
    exit 1
fi

参考: https://fedoraproject.org/wiki/Packaging:ScriptletSnippets

这篇关于在规范步骤中以编程方式失败的RPM安装的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！