RFR: 8081833: Clean up JVMFlag getter/setter code

[Gerard Ziemski]

On Tue, 15 Sep 2020 05:01:13 GMT, Ioi Lam <iklam at openjdk.org> wrote:

> This is a follow to [JDK-8243208 Clean up JVMFlag implementation](https://bugs.openjdk.java.net/browse/JDK-8243208).
> 
> I try to use templates and inheritance (and a bit of macros) to get rid of the large amount of duplicated code in the
> JVMFlag getters/setters.
> Stefan Karlsson and I have tried this several times already, but I think the current attempt probably has the least
> amount of hacks and most functionality:
> - Type safety is enforced up to the point of loading/storing `JVMFlag::_addr`, without any unchecked typecasts.
> - Runtime type checking is implemented by the `JVM_FLAG_TYPE(t)` macro (see jvmFlagAccess.hpp), by passing along an
>   integer (`JVMFlag::TYPE_int`, etc). This is more efficient than previous attempts that passed along a function pointer
>   (see http://cr.openjdk.java.net/~stefank/8081833/webrev.04/)
> - The old `JVMFlag::xxxAtPut` functions were undocumented. There are actually two groups that have different usage. I
>   added documentation in jvmFlagAccess.hpp to explain this.
> - I got rid of the `JVMFlagEx` class and moved the code into `JVMFlag`. This is possible with C++11 which allows forward
>   declaration of `enum JVMFlagsEnum : int;`
> - I changed `JVMFlag::_type` from a string into an enum. I am surprised this had lasted for over 20 years.
> - I also changed `JVMFlag` from a struct into a class and made some fields private. I probably will do more cleanup in a
>   separate RFE.
> 
> Please start with jvmFlagAccess.hpp, jvmFlagAccess.hpp and then globals_extension.hpp.
> 
> Probably not everyone is in love with the `JVM_FLAG_TYPE(t)` macro. I'll be happy to try out alternatives.

Changes requested by gziemski (Committer).

src/hotspot/share/runtime/flags/jvmFlagLimit.cpp line 147:

> 145: };
> 146:
> 147: JVMFlagsEnum JVMFlagLimit::_last_checked = static_cast<JVMFlagsEnum>(-1);

Can we add some "unchecked" enum value to JVMFlagsEnum enums (ex. UNCHECKED_FLAG) and use that instead of hardcoding
to -1 ?

src/hotspot/share/runtime/flags/jvmFlagAccess.hpp line 78:

> 76:       if (flag->type() != type_enum) {
> 77:         return JVMFlag::WRONG_FORMAT;
> 78:       }

Can this normally happen at runtime? Maybe replace it with assert_type() instead?

src/hotspot/share/runtime/flags/jvmFlagAccess.cpp line 91:

> 89: public:
> 90:   JVMFlag::Error set_impl(JVMFlag* flag, void* value_addr, JVMFlag::Flags origin) const {
> 91:     bool verbose = !JVMFlagLimit::validated_after_ergo();

Why only verbose when not AfterErgo? Maybe add comment explaining that.

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

PR: https://git.openjdk.java.net/jdk/pull/163