fix SIOCGSTAMP with new kernels, linux, socketcan

quick fixing OPCUA-1743

LogIt single component cleanup

should have OPCUA-1640 fixed
pending verification with multithreaded server

bugfix and vcan

• fix OPCUA-1597: added feature so that you can use "vcan" now for socketcan@linux as well, and tested it this time as well.
• fix OPCISEG-119: user code called createBus() as well, so that method got called twice. Two problems solved:
  (1) only one reply-handler (from the first call) should be connected, the second call (and all subsequent calls for the same bus) to createBus() should be ignored.
  (2) the user code should not call createBus() explicitly but rather follow the API as documented. The createBus() has to be public unfortunately, by design. I am open to suggestions how to isolate this virtual public method from the user API. But with (1) already fixed, CanModule will only connect one reply-handler anyway , so the code will run correctly.
• documentation updated accordingly

bugfix build chain

OPCUA-1602 bugfix

from 1.1.9.7:
suppressed c-wrapper

ubuntu & static libs

• Builds -static and normal shared libs seperatedly
• specify CANMODULE_AS_STATIC_AS_POSSIBLE for the -static libs
• tested on a physical ubuntu system: all 3 vendors pass
  -inofficial support for ubuntu is OK therefore-
• CANT-39 solved

sw disconnect

• protect global connection maps with mutexes against parallel access
• correct software close/open behaviour as possible without reloading the library. Delays are needed
• prefer reloading the library when software close/open
• works in general, but differences between modules/OS are to be noted, see Documentation/sphinx-result html tree
• some bugs corrected related to reconnect
• tested thouroughly with extra test suite (simpletest) for all vendors and linux/windows2016
• lots of commits as always because of X-build chain ;-)

sync flag suppressed

• suppressed the sync flag from CanModule: it was only used in anagate and had no effect there. Anagate does not implement anything related to that flag: this was a leftover "for future use" from CanModule

• highSpeed default == true for anagate as well

• Changelog.md knocked into shape as well

cleanup & test

• cleaned up and pruned documentation

• tested in lab: OK

• cwrapper should be suppressed, depending on what ISEG says and our discussion

• CANX: uses stats as well

bugfix-OPCUA-1474

HEADER path was wrong for systec@windows build
https://its.cern.ch/jira/browse/OPCUA-1474

bugfix-OPCUA-1469

some vendors' CMakeLists had the wrong variales for link and include pahts. Must have slipped in during a complex merge I did. Thanks for spotting and reporting that on. It is understood and fixed in the build chain.