RFR(M) JDK-8031819: Remove legacy jdk checks and code

[David Holmes]

Hi Harold,

thread.cpp:

You don't need the comment regarding the change in method to be called 
as again there is no context that tells you there was a different method 
in an earlier release. So this:

1742     // JSR-166: change call from ThreadGroup.uncaughtException to
1743     // java.lang.Thread.dispatchUncaughtException

can simply be:

1742     // Invoke java.lang.Thread.dispatchUncaughtException

You are still checking if dispatchUncaughtException exists but that is 
not necessary so this block can go:

1748  CallInfo callinfo;
1749  KlassHandle thread_klass(THREAD,SystemDictionary::Thread_klass());
1750  LinkResolver::resolve_virtual_call(callinfo, threadObj, 
recvrKlass, thread_klass,
1751             vmSymbols::dispatchUncaughtException_name(),
1752             vmSymbols::throwable_void_signature(),
1753             KlassHandle(), false, false, THREAD);
1754  CLEAR_PENDING_EXCEPTION;
1755  methodHandle method = callinfo.selected_method();
1756  assert(method.not_null(), "Thread.dispatchUncaughtException() not 
found");

---

A few follow ups on your other responses ...

On 10/06/2014 1:03 AM, harold seigel wrote:
>>> In classLoader.* is it worth renaming/absorbing the is_rt_jar13 and
>>> related "13" items?
> Good suggestion, but I think that this is a lot of change without a
> significant benefit.

The benefit is code clarity, as noone will understand what the 13 
indicates. It seems to me that the real_jzentry12 and real_jzfile12 
definitions are now unused. So the "13" variants can just drop the 13 
and is_rt_jar13 becomes is_rt_jar.

>>> systemDictionary.hpp:
>>>
>>> The check_klass_Opt_Only_JDK* functions seem to be unused and only
>>> generated assertion failures when they were used.
>>>
> I was unsure about this change so I did not remove the
> check_klass_Opt_Only_JDK* functions.

Unsure in what sense - that they are dead code? They seem as dead as 
other things you are deleting :)

Thanks,
David