Hello @QuChen88, the idea makes sense, but on the other hand, it is odd that we don't have handling for config directives that are strings in the SDS form: we have plenty of them. So because this fix is non critical, I'm not sure it makes any sense to add more custom code for new options instead of fixing the problem that we don't handle SDS string options. Pinging @madolson for her POV as well.

@antirez Yes I think the suggestion makes sense. I can take a look into supporting binary SDS string options in a more general way here.

Oh look, no more antirez :D. I still think his original suggestion made sense. I think we should extend the createStringconfig() to support createSDSConfig(). It's a different type, so I think it should have a different variable. I would look at the way numerics are handled for an example.

Oh look, no more antirez :D. I still think his original suggestion made sense. I think we should extend the createStringconfig() to support createSDSConfig(). It's a different type, so I think it should have a different variable. I would look at the way numerics are handled for an example.

My soul keep watching on you.

There is a segfault in the test run, can you take a look at that?

I was not able to reproduce the test failure / crash locally on my Mac. All the tests passed.

I ran make test on a Ubuntu instance launched in EC2, I was able to reproduce this error after I rebased on the latest redis:unstable. Taking a look.

This CR is getting too outdated and have merge-conflicts and test failures. I created another branch and opened a more updated PR. Closing this one in favor of #8200