Enable more warning flags.

- /W4 on MSVC
- -Wall -Wextra + others on gcc/clang
- New NVBench_ENABLE_WERROR option to toggle "warnings as errors"
- Mark the nlohmann_json library as IMPORTED to switch to system includes
- Rename nvbench_main -> nvbench.main to follow target name conventions
- Explicitly suppress some cudafe warnings when compiling templates in
  nlohmann_json headers.
- Explicitly suppress some warnings from Thrust headers.
- Various fixes for warnings exposed by new flags.
- Disable CUPTI on CTK < 11.3 (See #52).

testing/range.cu

@@ -64,11 +64,12 @@ void test_fp_tolerance()
   // Make sure that the range is padded a bit for floats to prevent rounding
   // errors from skipping `end`. This test will trigger failures without
   // the padding.
-  const nvbench::float32_t start  = 0.1;
-  const nvbench::float32_t stride = 1e-4;
+  const nvbench::float32_t start  = 0.1f;
+  const nvbench::float32_t stride = 1e-4f;
   for (std::size_t size = 1; size < 1024; ++size)
   {
-    const nvbench::float32_t end = start + stride * (size - 1);
+    const nvbench::float32_t end =
+      start + stride * static_cast<nvbench::float32_t>(size - 1);
     ASSERT_MSG(nvbench::range(start, end, stride).size() == size,
                "size={}", size);
   }