Runtime API deprecation warnings

[ducky64]

Add runtime warnings for use of deprecated Chisel methods. This is done using a macro that takes the message from a @deprecated annotation, and adds a call to Builder.deprecated.

Reasoning is that by default, Scala doesn't print all deprecations, and that it's somewhat tricky to notice them - yet some support questions revolve around the use of deprecated and terribad API. This now prints warnings for uses of deprecated functions at runtime, and aggregates them by error and line to avoid spam. Also included is convenient information on enabling scalac deprecations.

This also changes how line numbers for Chisel's error facility is determined, using prefix string comparison of the stack trace element classnames, instead of checking if the class is a subtype of UserModule. The previous one (specifically, calls to Class.forName) seems to interact badly with reflection-based cloneType when called at scale. This should also give more accurate reporting of errors that are in user code but outside of a UserModule.

It turns out that @deprecated on macro functions don't do anything, so this changes the tags to the functions that the macros point to, which seems to work properly. It also turns out that there's a bunch of uses of deprecated functions in chiselTests which needs to be fixed.

Not all @deprecated functions are also annotated with @chiselRuntimeDeprecation, because they're still used in Chisel internals, and we can't track whether they're called by the user or by Chisel and it will give a misleading error. These are a small amount of functions.

• Related issue (if applicable)

• Type of change

  • Bug report
  • Feature request
  • Other enhancement

• Impact

  • no functional change
  • API addition (no impact on existing code)
  • API modification

• Development Phase

  • proposal
  • implementation

• Release Notes

Add runtime deprecation warnings and improve error reporting.

[ducky64]

Updated to remove most deprecation warnings during runtime.

Only deprecation message that pops up now during a rocket run is:

[deprecated] Debug.scala:928 (1 calls): $bang$eq is deprecated: "Use \'=/=\', which avoids potential precedence problems"

Side note: the nasty code introduced in ReadyValidIO and Valid needs to be removed eventually. Because those are in chisel3._, they are under new semantics even if invoked from compatibility code, which is why the Builder.currentModule detection is needed. Ideally rocket would fix this at the source.

[chick]

This looks pretty good to me. A couple of questions:

• Is there any kind of test that would show this in action and/or make sure code is run?
• Is there any corresponding wiki page showing example output?
• Does this represent a methodology for deprecating things in the future and if so can we make a wiki page for that?

[ducky64]

This is solely a best-effort deprecation scheme that tries to make deprecations a bit more noisy than what they are in scalac by default, and also tell them how to turn on detailed deprecation warnings in scalac. It also makes it clear to the user that deprecated functions may go away and that they should fix their code.

Tests: this is hard to do, because it's all printlns. It also shouldn't affect the output at all, so I'm less concerned about testing.
Wiki page: this is meant to be self-explanatory, basically taking the scalac @deprecated annotation and also making sure messages are reported during runtime
General deprecations: Use @deprecated(message, since) and stack this runtime deprecation annotation on top of it to also make it complain during runtime (if that's what you want).

[ucbjrl]

lgtm

[jackkoenig]

Did this change deprecate != in compatibility code? It's not that big of a deal, most compatibility code uses =/= anyway

[ducky64]

Yeah, !=, even in compatibility mode, ends up triggering runtime deprecations because it was already triggering compile time deprecations.

[jackkoenig]

What I mean to say is that I have code that didn't get the compile-time deprecation warnings before, and now does. I'm not sure why that is but it was only 3 places so 🤷‍♀️

[ducky64]

Ah. Yeah, I think this was one of those cases where the deprecated annotation was on the macro (for SourceInfoTransform) instead of the macro target, so it didn't fire before. But this PR also corrects that, so it fires now.

[jackkoenig]

Anyway, this lgtm!