Change Kshell sorting precision to full precision

[ye-luo]

Proposed changes

Having modk2 in full precision seems fixing kshell counting issue.
fixes #5384

What type(s) of changes does this code introduce?

• Bugfix

Does this introduce a breaking change?

• No

What systems has this change been tested on?

epyc-serer

Checklist

• Yes. This PR is up to date with current the current state of 'develop'

[PDoakORNL]

It's going to be a bunch of work for me to unwrap all my dependent changes from my fix and encapsulation of kcontainer. Just finish your review of my work, I didn't ask you to play coding golf with this and I've been held up for a very long time in review. My performance is not evaluated based on how small a change you can eventually figure out how to make.

[ye-luo]

It's going to be a bunch of work for me to unwrap all my dependent changes from my fix and encapsulation of kcontainer. Just finish your review of my work, I didn't ask you to play coding golf with this and I've been held up for a very long time in review. My performance is not evaluated based on how small a change you can eventually figure out how to make.

We can probably use git trick to simply revert the change #5389 included in #5406. I can facilitate the work. Are there any others PRs or branches affected by #5389?

[PDoakORNL]

I do not support this reversion.

[ye-luo]

I do not support this reversion.

This PR is left pending as it was. Will revise and clean up once your estimates get in.

[jtkrogel]

Is this going anywhere?

[ye-luo]

@jtkrogel the changes is PR is still crucial to keep minimal code with mixed precision support.
I will follow up on this PR once @PDoakORNL finishes the dependent estimator.

[PDoakORNL]

I will revisit this once the application level energy density and offload per particle CoulombAA work is merged.