machinst x64: refactor to use types::[type] everywhere

[abrown]

This change is a pure refactoring--no change to functionality. It removes use crate::ir::types::* imports and uses instead types::I32, e.g., throughout the x64 code. Though it increases code verbosity, this change makes it more clear where the type identifiers come from (they are generated by cranelif-codegen-meta so without a prefix it is difficult to find their origin), avoids IDE confusion (e.g. CLion flags the un-prefixed identifiers as errors), and avoids importing unwanted identifiers into the namespace.

[github-actions]

Subscribe to Label Action

cc @bnjbvr

Details

This issue or pull request has been labeled: "cranelift", "cranelift:area:x64"

Thus the following users have been cc'd because of the following labels:

• bnjbvr: cranelift

To subscribe or unsubscribe from this label, edit the .github/subscribe-to-label.json configuration file.

Learn more.

[cfallin]

Sure, this seems reasonable. @julian-seward1 or @bnjbvr may have different opinions here, but (i) many of the locations this refactor touches were inconsistent in the first place (there were some ir::types::X as well as bare X instances), and (ii) IMHO, a little extra explicitness is good when it's not so much to be cumbersome. So, +1 from me. Thanks!

[abrown]

cc: @jlb6740

[jlb6740]

@abrown Thanks, agree with everything said. As @cfallin said .. the use of types:: prefix is inconsistent so this does improve on that. As you said it does increase verbosity which I think can sometimes reduce readability when it is too redundant, but maybe this allows ide's and the programming to have to parse and deal with less ambiguity. So pros and cons .. I personally wouldn't mind either way. Interested in what others prefer.

[abrown]

I plan on merging this before it gets too out-of-date (especially since #2100 removes the x64 CI check--I ran the x64 tests locally on the rebased version and the unrebased commit already passed the x64 CI check).