Hi
Could you please review following fix which stop test execution if JVMTI
function returns error. The test fails anyway however using potentially
bad data in JVMTI function might cause misleading crash failures. The
hs_err will contains the stacktrace not with problem function but with
function called with corrupted data. Most of tests already has such
behavior but not all. Also I fixed a couple of tests to finish if they
haven't managed to suspend thread.
I've updated only tests which try to use corrupted data in JVMTI
functions after errors. I haven't updated tests which just compare/print
values from erroring JVMTI functions. The crash in strcmp/println is not
so misleading and might be point to real issue.
webrev: http://cr.openjdk.java.net/~lmesnik/8242891/webrev.00/
bug: https://bugs.openjdk.java.net/browse/JDK-8242891
Leonid
