speedup dynamicwstream

- move bytesWritten calculation to query the tail, allowing write() to be faster since it doesn't have to update anything extra per-write.
- enforce that all blocks are multiple-of-4 bytes big
- update the minimum block size to 4K

Before: 30ms
After:  23ms for non-4-bytes writes
        13ms for 4-bytes writes

BUG=skia:

Change-Id: Id06ecad3b9fe426747e02accf1393595e3356ce3
Reviewed-on: https://skia-review.googlesource.com/6087
Reviewed-by: Mike Klein <mtklein@chromium.org>
Commit-Queue: Mike Reed <reed@google.com>
3 files changed