these often fail on PGO builds.

Too many build tool chains do not properly support it, including building
and linking an executable fine that simply segfaults when you try to run
it (such as debian jessie 8.5's gcc 4.9).  On others where it does appear
to build (ubuntu 16.04's gcc 5.4) there are still test_distutils and test_gdb
failures to deal with.

We're not going to spend time attempting to maintain a complicated white list
of what does and doesn't work in our configure.ac file.

StackNavigationTests.test_up_at_top and PyBtTests.test_threads.

Unfortunately we have no way of identifying if we were compiled profile-opt
or not from what is in sysconfig.get_config_vars() so we just disable it for
all optimized builds.  This test suite crazy fragile.

Issue #21201: Improves readability of multiprocessing error message from server to client for certain exceptions

required for PGO linking - to be a configure time error rather than
make time when --with-optimizations is enabled.  Also improve our
ability to find the llvm-profdata tool on MacOS and some Linuxes.

between runs given the same Grammar.txt input regardless of the hash
randomization setting.

Backport of 186bb8dc5540 from 3.5.  Done in 2.7 per the lib2to3 exemption.

It is unused. It also forces a 16-byte alignment, which creates problems because
Python's allocator only uses 8-byte alignment.

Apparently the test succeeds on XP.

This test should always raise RuntimeError.

The tests no longer attempt to figure out if a soundcard or particular
system sounds are available.  Instead, it just tries everything and
accepts RuntimeError as a flavor of success.

Also, don't run it if 'gui' is not requested.  Patch by Xiang Zhang.

(Omitted for 2.7 from previous patch.)

Issue #27934: Use float.__repr__ instead of plain repr when JSON-encoding an instance of a float subclass. Thanks Eddie James.