RFR(S) 8177901: JDWP exit error JVMTI_ERROR_WRONG_PHASE(112): on checking for an interface

[serguei.spit...@oracle.com]

Please, review a fix for:   https://bugs.openjdk.java.net/browse/JDK-8177901 Webrev:   http://cr.openjdk.java.net/~sspitsyn/webrevs/2017/hotspot/8177901-jdi-wrong_phase.1 Summary:     The approach to fix this is to introduce a new lock vmDeathLock in eventHelper.c   and use it to sync the commandLoop with the JVMTI event callback cbVMDeath   the same way as it was done for debugLoop_run.   (see the fix of: https://bugs.openjdk.java.net/browse/JDK-8134103)   One potential issue with it is that the commandLoop() transitively uses some helper   functions (e.g. from util.c) that use the macro EXIT_ERROR, and so, can abort.   It seems, in such a case the vmDeathLock will remain locked, and so,   the cbVMDeath() will block on it causing a deadlock.   Note, that these helper functions can be also called from different contexts/threads   (not from the commandLoop thread only). In such contexts the commandLoop vmDeathLock   is not being entered, and so, should not be exited.   To fix this potential issue, new function, commandLoop_exitVmDeathLockOnError(),   is introduced, and it is called from the debugInit_exit().   The commandLoop_exitVmDeathLockOnError() always checks if current thread is   the commandLoop thread and only in such a case unlocks the vmDeathLock. Testing:    Ran the nsk.jdi, nsk.jdwp and jtreg jdk_jdi for both release and fastdebug builds.   All tests are passed. Thanks, Serguei