Adds a good test for 32-bit binary float values.

[tgregg]

Issue #, if available:
#11

Description of changes:
The contents of the file is as follows:

e0 01 00 ea    // IVM
44 00 00 00 00 // 0e0
44 80 00 00 00 // -0e0
44 40 86 66 66 // 4.2
44 c0 86 66 66 // -4.2
44 ff 80 00 00 // -inf
44 7f 80 00 00 // +inf
44 ff 7f ff ff // minimum 32-bit float
44 7f 7f ff ff // maximum 32-bit float
44 7f ff ff ff // nan

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.