- Epsilon is now understood as relative to expected value.
- Improve error messages for epsilon checks.

This is needed for examining the values of user counters (needed
for #348). It is also needed for checking the values of standard
benchmark results like items_processed or complexities (for example,
checking the standard deviation is needed for unit testing #357
as discussed in #362).

The tests are still missing a way to check actual validity of
numerical results; this will be done next. As they currently are,
the tests pass, but the problem detected with #378 is still
standing and the results with non-standard counters are wrong.

The problem was that the call to Finish() the user counters was
lost in a big merge. If I had already written the tests for the
user counters, this would probably have been catched earlier.

* fix android compilation

* checking __GLIBCXX__ and __GLIBCPP__ macro in addition to __ANDROID__

* using vsnprintf instead of std::vsnprintf to compile on Android

* removed __GLIBCPP__ check on Android

* StringPrintF instead of std::to_string for Android

When using CPU time to determine the correct number of iterations the
library additionally checks if the benchmark has consumed 5x the minimum
required time according to the wall clock. This prevents benchmarks
with low CPU usage from running for much longer than actually intended.

However when a benchmark uses a manual timer this heuristic isn't helpful
and likely isn't correct since we don't know what the manual timer actually
measures.

This patch removes the above restriction when a benchmark specifies a manual
timer.

* Add Benchmark::Iterations for explicitly specifying the number of iterations to use.

* Document that benchmark::Iterations should not be used to limit benchmark runtimes

In non-debug builds CMake automatically adds -DNDEBUG, this means
that uses of `assert` in the tests are disabled for non-debug builds.
Obviously we want these tests to run, regardless of configuration.

This patch strips -DNDEBUG during non-debug builds and adds
-UNDEBUG just to be sure.

Using CMAKE_SOURCE_DIR benchmark cannot be built as part of a project.
This change allows to add benchmark using add_subdirectory to a bigger
project.

- Remove target_include_directories of ${PROJECT_SOURCE_DIR}/include to
  fix error: Target "benchmark" INTERFACE_INCLUDE_DIRECTORIES property
  contains path which is prefixed in the source directory.

* [tools] python 2/3 support

* update authors/contributors

* Add BENCHMARK_BUILD_32_BITS option and add builders to test it

* Attempt to fix travis configuration

* Make add_required_cxx_compiler_flag cause an error when the flag isn't supported

* add gcc-multilib dependancy on travis

* attempt to fix travis.yml parsing error

* Require g++-multilib instead of gcc-multilib

* Add 32 bit release configurations

* Attempt to fix libc++ travis build w/ 32 bits

* Work around CMake configuration failure on Travis

* Fix reading CPU info from file

Macro CHECK do nothing for release mode, meaning it doesn't invoke the
arguments

* Add myself to AUTHORS and CONTRIBUTORS

Previously the constants used for converting between
different units of time were declared using int64_t. However
we should only use explicitly sized integer types when they
are required, and should use 'int' everwhere else, and there is
no good reason to use int64_t here.

For that reason this patch changes the type of the constants.
This should help address issue #354 as well.

fixes #354

The build fails with ICC17 because of warnings and Werror. What is the correct solution to fix it?
Should a patch

disable Werror for ICC (or maybe all non known compilers)
disable the false postive warnings for all files. This could be done using:
add_cxx_compiler_flag(-wd2102) #ICC17u2: Many false positives for Wstrict-aliasing
add_cxx_compiler_flag(-wd2259) #ICC17u2: non-pointer conversion from "long" to "int" may lose significant bits (even for explicit static cast, sleep.cc(44))
add_cxx_compiler_flag(-wd654) #ICC17u2: overloaded virtual function "benchmark::Fixture::SetUp" is only partially overridden (because of deprecated overload)
disable warnings at file level or some other granularity

I recently learned Windows provides a function called _ReadWriteBarrier
which is literally ClobberMemory under a different name. This patch
uses it to implement ClobberMemory under MSVC.

Previously benchmark_api.h wasn't allowed to include standard library
headers. For this reason SetLabel had a hack to accept std::string
without including <string>. The hack worked by attempting to detect
the injected class name `basic_string`. However Clang has changed
it's behavior regarding injected class names so this hack no longer
works.

This patch removes the hack and replaces it with a function that
actually names std::string. However we still cannot pass std::string
across the dylib boundary because of libstdc++'s dual C++11 ABI.