Skip to content

Conversation

@cyphar
Copy link
Member

@cyphar cyphar commented Nov 6, 2016

The previous config generation produced a config which would cause runC
to run the container using the host filesystem (with no namespaces set
up). While we should try to be as agnostic as possible, we don't want to
generate config.json payloads that are known to be bad (and useless).

Fixes #76

Signed-off-by: Aleksa Sarai [email protected]

The previous config generation produced a config which would cause runC
to run the container using the host filesystem (with no namespaces set
up). While we should try to be as agnostic as possible, we don't want to
generate config.json payloads that are known to be bad (and useless).

Signed-off-by: Aleksa Sarai <[email protected]>
s.Version = specs.Version
// we should at least apply the default spec, otherwise this is totally useless
s.Process.Terminal = true
// TODO: Get this default from somewhere, rather than hardcoding it here.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I suggest you get the default from a vendored runtime-tools/generate.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I ended up implementing cyphar/umoci#13 so I'll probably end up just merging that here and dropping the old code.

@coolljt0725
Copy link
Member

I'm +1 with this

@vbatts
Copy link
Member

vbatts commented Feb 8, 2017

I'm +1 for this, but it looks like make lint is failing.

@zhouhao3
Copy link

zhouhao3 commented Oct 12, 2017

@cyphar I passed the test after I rebased it. Can you rebase it?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants