dkms: continue on error

[cshirron]

In the postinst script, if we hit an error building for one kernel, record the error and continue on to the next kernel. When done, exit with the value of any recorded error. This allows us to do a best-effort build for all installed kernels.

[anbe42]

I like the idea, but there should be a way to toggle this behavior (framework.conf setting, environment variable, ...). If the script is going to exit with an error, it needs to report a list of kernels where building has failed. Reading the scrollback is not feasible.

The correct synopsis prefix for the commit would be 'common.postinst:', not 'dkms:'

[scaronni]

The correct synopsis prefix for the commit would be 'common.postinst:', not 'dkms:'

@anbe42 if you think we should have these guidelines in a CONTRIBUTING.md file or something similar, feel free. Currently we have no guidelines in this regard, I also don't follow this convention.

[cshirron]

I have taken your suggestions and made the appropriate changes.

[scaronni]

@cshirron sorry for the delay. Can you please make the necessary adjustments so that the pipelines pass? Thanks.