optimize buffer build

[FelixYBW]

There are 3 optimizations included:

1. Added prefetch to buffer_build, assumption is if we append one value, we may append the second one soon
2. Since most of the value isn't NULL, we reset the null bitmap buffer to 0xffffffff once we allocate it. Later we skip the bitset if value is true. Now we still need to update the bitlength, and one branch. Two more optimizations can be done: 1) skip memory allocation until first false value met. 2) in binary_builder etc, skip the bitmap append until first false value
3. specify int32_t for typedbufferbuild, in which we use *(pos)=value, instead of memcpy

[github-actions]

Thanks for opening a pull request!

If this is not a minor PR. Could you open an issue for this pull request on JIRA? https://issues.apache.org/jira/browse/ARROW

Opening JIRAs ahead of time contributes to the Openness of the Apache Arrow project.

Then could you also rename pull request title in the following format?

ARROW-${JIRA_ID}: [${COMPONENT}] ${SUMMARY}

or

MINOR: [${COMPONENT}] ${SUMMARY}

See also:

• Other pull requests
• Contribution Guidelines - How to contribute patches