refactor: make s2n_stuffer_read_hex match s2n_stuffer_read

[lrstewart]

Description of changes:

@maddeleine called out that s2n_stuffer_read_hex has its arguments reversed compared to s2n_stuffer_read. In s2n_stuffer_read, the data is read from a stuffer and written to a blob. In s2n_stuffer_read_hex, the data is read from a blob and written to a stuffer.

The primary benefit of s2n_stuffer_read_hex reading from a blob (fixed size) and writing to a stuffer (potentially flexible size) is that the caller doesn't need to know the size of the data that a given hex string will produce. That math is easy (hex_size / 2), but I still liked that the method hid it.

However, I think maddeleine is right and that benefit doesn't outweigh the cost of the confusing discrepancy with s2n_stuffer_read. Switching the arguments in existing code wasn't difficult, suggesting that setting the size of the output blob isn't a serious barrier to using the method. We also often don't use growable or oversized stuffers in production code (outside of IO buffers), so the caller needs to know the size of the output anyway.

Testing:

Updated tests + cbmc proof.

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