Minor improvements of runtime code.

I ran coverity on the jdk11 hotspot sources and want to propose the following fixes. I scanned the linux x86_64 build. Some issues are similar to previous parfait fixes (check for NULL, add guarantees etc.) I also identified some issues I consider real problems:

Real issues:
------------

jvmtiEnvBase.cpp
  || should be &&.
  Attention, this is the only change that really will change behaviour.
  But if thr == NULL we will see a crash below.

perfMemory_linux.cpp:
  Wrong buffer length used.

systemDictionary.cpp:
  Move code dereferencing ik under if (ik != NULL).

virtualspace.cpp
  Initialization is missing. Moved constructor up to the other
  constructors.


Useful code improvements:
-------------------------

vm_version_ext_x86.cpp
  Assure buffer is not accessed at offset -1.

os_linux.cpp
  Numa_max_node returns int, and a -1 in some cases.

moduleEntry.cpp
  name might be NULL. Just a fix for tracing.

systemDictionaryShared.cpp
  clearify code.
  It would be wrong if only entry == NULL would hold, one
  would hit the assertion below.

verifier.cpp
  Fix tracing.
  Illegal opcode is -1 and should not be passed to name array.

logOutput.cpp
  If n_selections == 0, best_selection would be NULL.
  Move up the assertion and turn into a guarantee.

filemap.cpp
  Either base can be NULL, or parts of the code before are dead.

metaspace.cpp
  We now an exception is pending.

klassVtable.cpp
  Coverity does not like the format in a variable.
  Anyways this is quite rough coding, transformed to use stringStream
  as with other similar exceptions.

jvmFlag.cpp
  match might be NULL.

writableFlags.cpp
  name might be NULL.

ostream.cpp
  If ftell returns error code -1, we need not continue.
  Especially we should not fseek(-1).

logTestUtils.inline.hpp
  ftell returns -1.

test_metachunk.cpp
  wrong datatype.