dubbogo / gost Goto Github PK

[func NewDecFromFloatForTest(f float64) *Decimal {](

What happened:
A total of two defective components with vulnerabilities were introduced into the project. The following are some of the main information：

infomation: https://www.oscs1024.com/cd/1530636116259217408?sign=c65eceeb
What you expected to happen:

How to reproduce it (as minimally and precisely as possible):

Anything else we need to know?:

What would you like to be added:

根据 gost 各个 package 的意义，改进 readme 内容。

Why is this needed:

方便用户使用 gost 库。

CPrintfln方法中fmt.Fprint应为fmt.Fprintf

What happened:

What you expected to happen:

How to reproduce it (as minimally and precisely as possible):

Anything else we need to know?:

../../../go/goworks/pkg/mod/dubbo.apache.org/dubbo-go/[email protected]/remoting/nacos/builder.go:52:60: cannot use sc (variable of type []"github.com/nacos-group/nacos-sdk-go/common/constant".ServerConfig) as []"github.com/nacos-group/nacos-sdk-go/v2/common/constant".ServerConfig value in argument to nacosClient.NewNacosConfigClient
..
What happened:

What you expected to happen:

How to reproduce it (as minimally and precisely as possible):

Anything else we need to know?:

What happened:
in database/kv/zk/client.go, gost ignores the error of node exists, and when I try to create some node (like dubbo/config) I didn't know it exists with CreateWithValue, I got no error and I thought it worked, but in fact, it failed.

What you expected to happen:
When I try to create some node that exists, I want to know it, and then I can use SetContent to modify it.

How to reproduce it (as minimally and precisely as possible):

Anything else we need to know?:

What would you like to be added:

When the current AcquireBytes function misses a byte slice in the pool, it will return a byte slice with a length of 0 and a cap of size, and when it hits, it will return a slice with both length and cap of size. I think consistency is more user friendly

Why is this needed:

When the current AcquireBytes function misses a byte slice in the pool, it will return a byte slice with a length of 0 and a cap of size, and when it hits, it will return a slice with both length and cap of size. I think consistency is more user friendly

What happened:
DefaultHandler#HandleZkEvent，这段代码有以下问题，看注释：

func (d *DefaultHandler) HandleZkEvent(z *ZookeeperClient) {
	var (
		state int
		event zk.Event
	)
	for {
		select {
		case event = <-z.Session:
                        // 直接用 event.Stata 就可以对比，int 多余
			switch (int)(event.State) {
			case (int)(zk.StateDisconnected):
				atomic.StoreUint32(&z.valid, 0)
                        // 第一个和第三个 case 都是用状态码来匹配，但这一个 case 是用的事件类型来匹配，逻辑有大问题，
                        // 但是 go-zookeeper 并没有定义连接成功的状态码（3），而 EventNodeDataChanged 的值也是 3，因此正好进入这个分支
			case (int)(zk.EventNodeDataChanged), (int)(zk.EventNodeChildrenChanged):
				z.eventRegistryLock.RLock()
				for p, a := range z.eventRegistry {
					if strings.HasPrefix(p, event.Path) {
						for _, e := range a {
							*e <- struct{}{}
						}
					}
				}
				z.eventRegistryLock.RUnlock()
			case (int)(zk.StateConnecting), (int)(zk.StateConnected), (int)(zk.StateHasSession):
				if state == (int)(zk.StateHasSession) {
					continue
				}
				if event.State == zk.StateHasSession {
					atomic.StoreUint32(&z.valid, 1)
					//if this is the first connection, don't trigger reconnect event
					if !atomic.CompareAndSwapUint32(&z.initialized, 0, 1) {
						close(z.reconnectCh)
						z.reconnectCh = make(chan struct{})
					}
				}
				z.eventRegistryLock.RLock()
				if a, ok := z.eventRegistry[event.Path]; ok && 0 < len(a) {
					for _, e := range a {
						*e <- struct{}{}
					}
				}
				z.eventRegistryLock.RUnlock()
			}
			state = (int)(event.State)
		}
	}
}

What you expected to happen:

How to reproduce it (as minimally and precisely as possible):

Anything else we need to know?:

What happened:

WARNING: DATA RACE
Write at 0x000000722aa8 by goroutine 6:
  github.com/dubbogo/gost/sync.restore.func1()
      /home/travis/gopath/src/github.com/dubbogo/gost/sync/deadlock_test.go:162 +0x4c
  github.com/dubbogo/gost/sync.TestNoDeadlocks()
      /home/travis/gopath/src/github.com/dubbogo/gost/sync/deadlock_test.go:62 +0x23e
  testing.tRunner()
      /home/travis/.gimme/versions/go1.12.linux.amd64/src/testing/testing.go:865 +0x163
Previous read at 0x000000722aa8 by goroutine 56:
  github.com/dubbogo/gost/sync.lock.func1()
      /home/travis/gopath/src/github.com/dubbogo/gost/sync/deadlock.go:184 +0x76

What you expected to happen:
No race

How to reproduce it (as minimally and precisely as possible):
go test ./... -race

Anything else we need to know?:

What happened:go get fail

What you expected to happen:go get success

How to reproduce it (as minimally and precisely as possible): go get github.com/dubbogo/[email protected]

Anything else we need to know?:go get github.com/dubbogo/[email protected]
go get: github.com/dubbogo/[email protected] requires
github.com/coreos/[email protected]: invalid version: unknown revision 000000000000