Just to clarify one thing ...
I may have been unclear. What I meant was: there is a repeated pattern
where directly in xxxDCmd::execute() a VMOp is prepped and fired. I
was vaguely thinking that if we give the DCmd a "needs-safepoint"
attribute, whoever fires the command could then also create t
Please review the changes that fix a build failure for Jvmti tests.
The fix replaces printf "%ld" format with JLONG_FORMAT macro defined in the
same way as it is defined in the hotspot
(src/hotspot//share/utilities/globalDefinitions.hpp and
src/hotspot//share/utilities/globalDefinitions_gcc.hpp
This is a review request for a previously closed test bug.
The test was recently moved to the open repos, and the
proposed fix is in the open code.
Webrev: http://cr.openjdk.java.net/~gadams/8169718/webrev/
Forwarded Message
Subject: RFR: JDK-8169718: nsk/jdb/locals/locals0
Hi Frederic,
>>
>> I am not saying that is what we should do, I am just saying it is not
>> difficult to implement.
>
> You’re right, this implementation is not as complex as I thought.
> However, the way this code works silently add a new constraint to
> diagnostic commands: a diagnostic comman
