RT-Thread-f407+lwip2.0.2 +rtt4.0.0+tcp 服务器死机问题（已解决）RT-Thread问答社区

f407+lwip2.0.2 +rtt4.0.0+tcp 服务器死机问题（已解决）

发布于 2019-02-28 14:00:52 浏览：3099 订阅该版

* 本帖最后由 sync 于 2019-3-1 09:32 编辑 *

```
```**测试环境：**f407+8720 接入办公室路由器做tcp服务器，用自己的笔记本做tcp客户端

``` \ | /
- RT - Thread Operating System
/ | \ 4.0.0 build Feb 28 2019
2006 - 2018 Copyright by rt-thread team
lwIP-2.0.2 initialized!```**故障现象：**
开机进程如下
```thread   pri  status      sp     stack size max used left tick  error
-------- ---  ------- ---------- ----------  ------  ---------- ---
tshell    20  running 0x00000084 0x00000800    25%   0x00000009 000
phy       30  suspend 0x00000094 0x00000400    19%   0x00000002 000
tcpip      6  suspend 0x000000c4 0x00000400    59%   0x0000000e 000
etx        8  suspend 0x00000094 0x00000600    09%   0x00000010 000
erx        7  suspend 0x00000094 0x00000600    35%   0x0000000f 000
tidle0    31  ready   0x00000054 0x00000100    32%   0x0000001c 000
main      10  suspend 0x0000008c 0x00000800    36%   0x0000000e 000```将官方例程tcp服务器 放入单独线程tcpserv中 
然后开始测试 tcp客服端每间隔200ms发送一次数据
```[13:38:30.818]发→◇This is TCP Server from RT-Thread.□
[13:38:30.825]收←◆This is TCP Server from RT-Thread.
[13:38:31.021]发→◇This is TCP Server from RT-Thread.□
[13:38:31.029]收←◆This is TCP Server from RT-Thread.
[13:38:31.228]发→◇This is TCP Server from RT-Thread.□
[13:38:31.237]收←◆This is TCP Server from RT-Thread.
[13:38:31.432]发→◇This is TCP Server from RT-Thread.□
[13:38:31.441]收←◆This is TCP Server from RT-Thread.
[13:38:31.639]发→◇This is TCP Server from RT-Thread.□
[13:38:31.649]收←◆This is TCP Server from RT-Thread.
[13:38:31.833]发→◇This is TCP Server from RT-Thread.□
[13:38:31.845]收←◆This is TCP Server from RT-Thread.
[13:38:32.036]发→◇This is TCP Server from RT-Thread.□
[13:38:32.049]收←◆This is TCP Server from RT-Thread.
[13:38:32.240]发→◇This is TCP Server from RT-Thread.□
[13:38:32.247]收←◆This is TCP Server from RT-Thread.
[13:38:32.444]发→◇This is TCP Server from RT-Thread.□
[13:38:32.451]收←◆This is TCP Server from RT-Thread.
[13:38:32.650]发→◇This is TCP Server from RT-Thread.□
[13:38:32.659]收←◆This is TCP Server from RT-Thread.
[13:38:32.850]发→◇This is TCP Server from RT-Thread.□
[13:38:32.857]收←◆This is TCP Server from RT-Thread.
[13:38:33.056]发→◇This is TCP Server from RT-Thread.□
[13:38:33.065]收←◆This is TCP Server from RT-Thread.
[13:38:33.257]发→◇This is TCP Server from RT-Thread.□
[13:38:33.264]收←◆This is TCP Server from RT-Thread.
[13:38:33.459]发→◇This is TCP Server from RT-Thread.□
[13:38:33.471]收←◆This is TCP Server from RT-Thread.```**经过若干时间后，十几分钟，或者1小时，或者2小时后，笔记本的tcp客户端就收不到F407上Tcp服务器的返回数据了。**
```[13:47:57.609]发→◇This is TCP Server from RT-Thread.□
[13:47:57.820]发→◇This is TCP Server from RT-Thread.□
[13:47:58.024]发→◇This is TCP Server from RT-Thread.□
[13:47:58.224]发→◇This is TCP Server from RT-Thread.□
[13:47:58.426]发→◇This is TCP Server from RT-Thread.□
[13:47:58.629]发→◇This is TCP Server from RT-Thread.□
[13:47:58.830]发→◇This is TCP Server from RT-Thread.□
[13:47:59.037]发→◇This is TCP Server from RT-Thread.□
[13:47:59.239]发→◇This is TCP Server from RT-Thread.□
[13:47:59.449]发→◇This is TCP Server from RT-Thread.□
[13:47:59.642]发→◇This is TCP Server from RT-Thread.□
[13:47:59.845]发→◇This is TCP Server from RT-Thread.□
[13:48:00.051]发→◇This is TCP Server from RT-Thread.□
[13:48:00.251]发→◇This is TCP Server from RT-Thread.□
[13:48:00.461]发→◇This is TCP Server from RT-Thread.□
[13:48:00.661]发→◇This is TCP Server from RT-Thread.□
[13:48:00.862]发→◇This is TCP Server from RT-Thread.□
[13:48:01.064]发→◇This is TCP Server from RT-Thread.□
[13:48:01.274]发→◇This is TCP Server from RT-Thread.□
[13:48:01.476]发→◇This is TCP Server from RT-Thread.□
[13:48:01.675]发→◇This is TCP Server from RT-Thread.□
[13:48:01.881]发→◇This is TCP Server from RT-Thread.□
[13:48:02.078]发→◇This is TCP Server from RT-Thread.□
[13:48:02.287]发→◇This is TCP Server from RT-Thread.□
[13:48:02.492]发→◇This is TCP Server from RT-Thread.□
[13:48:02.691]发→◇This is TCP Server from RT-Thread.□
[13:48:02.896]发→◇This is TCP Server from RT-Thread.□```

**原因分析：**
经过多次测试发现，每次故障发生后，
命令ps结果如下：
```thread pri status sp stack size max used left tick error
-------- --- ------- ---------- ---------- ------ ---------- ---
tcpserv 5 suspend 0x000001bc 0x00000400 77% 0x00000005 000
tshell 20 running 0x00000084 0x00000800 25% 0x00000007 000
phy 30 suspend 0x00000094 0x00000400 19% 0x00000002 000
tcpip 6 suspend 0x000000c4 0x00000400 64% 0x00000001 000
etx 8 suspend 0x00000094 0x00000600 09% 0x0000000b 000
erx 7 ready 0x0000007c 0x00000600 35% 0x00000010 000
tidle0 31 ready 0x00000054 0x00000100 32% 0x00000019 000
main 10 suspend 0x0000008c 0x00000800 36% 0x0000000e 000```

list_mailbox,命令返回结果
如下：mailbox  entry size suspend thread
-------- ----  ---- --------------
mbox3    0000  0010 1:tcpserv
mbox2    0000  0010 0
mbox0    0000  0016 1:tcpip
etxmb    0000  0016 1:etx
**erxmb    0010  0016 0**

正常时，应该如下：
```mailbox  entry size suspend thread
-------- ----  ---- --------------
mbox3    0000  0010 1:tcpserv
mbox2    0000  0010 0
mbox0    0000  0016 1:tcpip
etxmb    0000  0016 1:etx
<b>erxmb    0000  0016 1:erx</b>```

我将8720和笔记本的网口用网线直连，都使用固定ip
这次发生故障后，打印信息如下：
```msh >list_mailbox
mailbox entry size suspend thread
-------- ---- ---- --------------
mbox3 0000 0010 1:tcpserv
mbox2 0000 0010 0
mbox0 0008 0016 0
etxmb 0000 0016 1:etx
erxmb 0000 0016 1:erx
msh >ps
thread pri status sp stack size max used left tick error
-------- --- ------- ---------- ---------- ------ ---------- ---
tcpserv 5 suspend 0x000001bc 0x00000400 77% 0x00000004 000
tshell 20 running 0x00000084 0x00000800 25% 0x00000001 000
phy 30 suspend 0x00000094 0x00000400 19% 0x00000002 000
tcpip 6 ready 0x000000ac 0x00000400 64% 0x00000004 -02
etx 8 suspend 0x00000094 0x00000600 09% 0x00000002 000
erx 7 suspend 0x00000094 0x00000600 25% 0x00000003 000
tidle0 31 ready 0x00000044 0x00000100 32% 0x00000014 000
main 10 suspend 0x0000008c 0x00000800 27% 0x0000000e 000```

**疑问：**
**1.erx线程不是应该一致等待邮箱信号吗？什么情况下会造成线程始终处于挂起状态，不再等待邮箱信号了呢？****2.tcpip线程也有挂起的情况，且不再等待邮箱信号了，这是啥原因呢？**

**解决
**昨天修改了程序这处，结果稳定运行了一晚上:lol

![1551402219(1).png](/uploads/201903/01/092903lrtrlqy91s4ryy13.png)测试发现线程死机时，rt_mb_recv里面的参数timeout等于1，我增加上面红框内的语句后，让timeout最小等于10，就可以正常运行了
问题算是解决了，但原因却不明白，哪位大佬能解释下吗？