Remove unnecessary flush calls on TrackedWrite

[viirya]

Which issue does this PR close?

Closes #.

Rationale for this change

We have BufWriter inside TrackedWrite which buffers output and flushes if necessary. We don't need to call flush in the middle before closing the writer.

It seems to be a simple win without regression.

write_batch primitive/4096 values primitive non-null
                        time:   [639.52 µs 642.46 µs 646.64 µs]
                        thrpt:  [267.70 MiB/s 269.44 MiB/s 270.68 MiB/s]
                 change:
                        time:   [-7.1374% -6.3056% -5.1252%] (p = 0.00 < 0.05)
                        thrpt:  [+5.4021% +6.7299% +7.6860%]
                        Performance has improved.
write_batch primitive/4096 values bool non-null
                        time:   [81.192 µs 83.273 µs 86.123 µs]
                        thrpt:  [7.7071 MiB/s 7.9708 MiB/s 8.1752 MiB/s]
                 change:
                        time:   [-6.9304% -4.3398% -1.8032%] (p = 0.00 < 0.05)
                        thrpt:  [+1.8363% +4.5367% +7.4465%]
                        Performance has improved.
write_batch primitive/4096 values string non-null
                        time:   [447.68 µs 448.90 µs 450.81 µs]
                        thrpt:  [174.48 MiB/s 175.23 MiB/s 175.71 MiB/s]
                 change:
                        time:   [-9.6784% -9.1649% -8.6186%] (p = 0.00 < 0.05)
                        thrpt:  [+9.4314% +10.090% +10.715%]
                        Performance has improved.
write_batch nested/4096 values primitive list non-null
                        time:   [1.2610 ms 1.2647 ms 1.2708 ms]
                        thrpt:  [149.95 MiB/s 150.67 MiB/s 151.12 MiB/s]
                 change:
                        time:   [-4.5901% -3.9472% -3.2749%] (p = 0.00 < 0.05)
                        thrpt:  [+3.3858% +4.1094% +4.8109%]
                        Performance has improved.

What changes are included in this PR?

Are there any user-facing changes?

[tustvold]

Approved, with the caveat that this isn't necessary to flush Thrift encoder state

[tustvold]

Is there some need to flush the protocol before drop, or does this simply flush the inner writer

[viirya]

It simply flushes the inner writer (a TWriteTransport: Write) which is TrackedWrite.

[ursabot]

Benchmark runs are scheduled for baseline = 8b84d4d and contender = c1c97f1. c1c97f1 is a master commit associated with this PR. Results will be available as each benchmark for each run completes.
Conbench compare runs links:
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ec2-t3-xlarge-us-east-2] ec2-t3-xlarge-us-east-2
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on test-mac-arm] test-mac-arm
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ursa-i9-9960x] ursa-i9-9960x
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ursa-thinkcentre-m75q] ursa-thinkcentre-m75q
Buildkite builds:
Supported benchmarks:
ec2-t3-xlarge-us-east-2: Supported benchmark langs: Python, R. Runs only benchmarks with cloud = True
test-mac-arm: Supported benchmark langs: C++, Python, R
ursa-i9-9960x: Supported benchmark langs: Python, R, JavaScript
ursa-thinkcentre-m75q: Supported benchmark langs: C++, Java