RFR: 8241502: Migrate x86_64.ad to MacroAssembler [v6]

[John Tortugo]

On Thu, 25 Mar 2021 22:42:32 GMT, Dean Long <dlong at openjdk.org> wrote:

>>> > I guess you counted an instruction validated if it was generated at least once (so at least 1 possible encoding)? On RISC architectures that would probably be enough, but I know on x86/x64, register choice can change the instruction encoding and length, thanks to prefixes, etc.
>>> 
>>> One possibility is to write a native test (using gtest) and exercise some encoding methods by enumerating all the possible input value locations. Probably, too much for a regularly executed test (there's a lot of stuff missing right now to support such scenario), but something which can be quickly hacked for a one-time run.
>> 
>> I suggest to do that as separate RFE. This changes could be tested by running in all our tiers testing.
>> 
>> Cleaning unused code could be also done separately.
>
>> Can you elaborate, please, what you mean here? I don't see "enc_classes and related code" in the latest patch.
> 
> I was confused by this at first too.  I believe the question is, should unused code such as "enc_class cdql_enc" be removed as part of this PR?

Thank you all for reviewing!

> Can you elaborate, please, what you mean here? I don't see "enc_classes and related code" in the latest patch.

@dean-long got my wording right. I'll do the cleaning as a separate RFE.

> This changes could be tested by running in all our tiers testing.

Right, this is how I tested: hotspot_all, tier1, tier2, tier3, different GCs, different OS's.

-------------

PR: https://git.openjdk.java.net/jdk/pull/2420