Feature: Simd Selection of column filter

[sundy-li]

Summary

Description for this feature.

Now filter works with memcpy one by one, we can use Vec.extend function to copy a batch of at most 64 items one time.

https://github.com/datafuselabs/databend/blob/dc058c9d22baa9e61763661f77cd10ec62c87c48/common/datavalues/src/columns/primitive/mod.rs#L187-L204

Some useful tools:

• Bitmap's Chunk function:

impl Bitmap {
    /// Returns an iterator over bits in chunks of `T`, which is useful for
    /// bit operations.
    pub fn chunks<T: BitChunk>(&self) -> BitChunks<T> {
        BitChunks::new(&self.bytes, self.offset, self.length)
    }
}

• Function: __builtin_ctz

  int __builtin_ctz (unsigned int x)
Returns the number of trailing 0-bits in x, starting at the least significant bit position. If x is 0, the result is undefined.

Rust's https://doc.rust-lang.org/std/primitive.u32.html#method.trailing_zeros

• Blog: https://lemire.me/blog/2018/02/21/iterating-over-set-bits-quickly/

[platoneko]

/assignme