Optimize bulk writing of all blocks of bloom filter

[viirya]

Which issue does this PR close?

Part of #3320.

Rationale for this change

write_batch primitive/4096 values primitive with bloom filter                                                      
                        time:   [4.2750 ms 4.2905 ms 4.3059 ms]                                                    
                        thrpt:  [40.995 MiB/s 41.143 MiB/s 41.291 MiB/s]                                           
                 change:                                                                                           
                        time:   [-18.394% -16.070% -13.847%] (p = 0.00 < 0.05)                                     
                        thrpt:  [+16.072% +19.147% +22.541%]                              
                        Performance has improved.                                                                  
write_batch primitive/4096 values primitive non-null with bloom filter     
                        time:   [4.4500 ms 4.5848 ms 4.7319 ms]                 
                        thrpt:  [36.582 MiB/s 37.756 MiB/s 38.900 MiB/s]
                 change:                                                                                           
                        time:   [-21.355% -17.148% -12.445%] (p = 0.00 < 0.05)                                     
                        thrpt:  [+14.214% +20.697% +27.153%]                                                       
                        Performance has improved.                                                                  
write_batch primitive/4096 values string with bloom filter
                        time:   [945.94 µs 946.62 µs 947.37 µs]
                        thrpt:  [84.060 MiB/s 84.126 MiB/s 84.186 MiB/s]
                 change:
                        time:   [-26.414% -25.196% -23.997%] (p = 0.00 < 0.05)
                        thrpt:  [+31.573% +33.682% +35.895%]
                        Performance has improved.
write_batch primitive/4096 values string dictionary with bloom filter
                        time:   [494.68 µs 495.13 µs 495.60 µs]
                        thrpt:  [97.137 MiB/s 97.230 MiB/s 97.318 MiB/s]
                 change:
                        time:   [-19.467% -18.637% -17.900%] (p = 0.00 < 0.05)
                        thrpt:  [+21.803% +22.905% +24.173%]
                        Performance has improved.
write_batch primitive/4096 values string non-null with bloom filter
                        time:   [1.0266 ms 1.0275 ms 1.0284 ms]
                        thrpt:  [76.484 MiB/s 76.553 MiB/s 76.618 MiB/s]
                 change:
                        time:   [-28.422% -28.120% -27.852%] (p = 0.00 < 0.05)
                        thrpt:  [+38.604% +39.121% +39.709%]
                        Performance has improved.

What changes are included in this PR?

Are there any user-facing changes?

[tustvold]

I wonder if we should be encouraging users to wrap File in a BufWriter? I.e. add this higher up. We likely don't want BufWriter when writing to Vec<u8> though... 🤔

[viirya]

Yea, seems other parts of the parquet writer could have benefit from BufWriter if we can add it higher.

[ursabot]

Benchmark runs are scheduled for baseline = 46b2848 and contender = 2749dcc. 2749dcc 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