Optimize the content of README and code

Merge pull request #9 from shengyanli1982/dev

Optimize the content of README and code

Improved code remarks and optimized GC optimization of the pop-up node in Queue

1. Improved code remarks
2. Optimized GC optimization of the pop-up node in Queue

Redesign code structure, optimize overall performance, and interface forward compatibility.

Provides more than 4x performance commissions, compatible with previous call interfaces.

1. Redesign the code architecture to maintain the original double buffering and reduce unnecessary Goroutine.
2. Replace the deque with a single necklace watch without lock.
3. Abandon sync.Pool and redesign the ultra-small object buffer pool.
4. Run poller with a single Goroutine to reduce unnecessary sync.Mutex calls.

Fixed buffio bugs and update queue implementation

Fixed buffio bugs and update queue implementation

Fixed the bufferIoWriterRefresh generation not updating timers

Fixed the bufferIoWriterRefresh generation not updating timers

Fix bugs and optimize code and documentation

Merge pull request #1 from shengyanli1982/dev

完成代码编写和文档设计