知识点
goroutine为什么比线程快
- 创建调度与销毁: goroutine的创建调度与销毁都在用户态，线程的在内核态
- 内存分配: goroutine栈内存为2-4KB，线程1MB
- 切换: goroutine切换时仅需要保护少数寄存器，而线程则需要保护大量的寄存器，goroutine无需操作系统线程切换
GMP
- 两级线程模型，内核调度实体kse与线程的对应关系为m:n
- kse -> m -> p ->g
- m: machine 关联一个内核线程
- p: processor 作为上下文，衔接m和g
- g: goroutine 轻量化线程，调用栈、调用信息， 例如channel等
https://github.com/fengyuan-liang/notes/blob/main/GoLang/golang%E5%A4%A7%E6%9D%80%E5%99%A8GMP%E6%A8%A1%E5%9E%8B.md
https://www.jianshu.com/p/4afa0679851d
slice的扩容
golang 1.21.4
- 对于小slice进行1.25倍扩展 逐渐过度
- 对于大slice进行2倍扩展
src/runtime/slice.go
newcap := oldCap
doublecap := newcap + newcap
if newLen > doublecap {
newcap = newLen
} else {
const threshold = 256
if oldCap < threshold {
newcap = doublecap
} else {
// Check 0 < newcap to detect overflow
// and prevent an infinite loop.
for 0 < newcap && newcap < newLen {
// Transition from growing 2x for small slices
// to growing 1.25x for large slices. This formula
// gives a smooth-ish transition between the two.
newcap += (newcap + 3*threshold) / 4
}
// Set newcap to the requested cap when
// the newcap calculation overflowed.
if newcap <= 0 {
newcap = newLen
}
}
}
g0
持有调度栈的goroutine，作用有
- goroutine的创建
- 大内存分配
- CGO函数的执行
三色标记法
https://community.apinto.com/d/34057-golang-gc
