|
|
|
|
| 遇见调试中最不想遇见的问题了--不稳定。各位大侠帮小弟看看! |
 遇见调试中最不想遇见的问题了--不稳定。各位大侠帮小弟看看! - smarteasy [ 2006-03-28 22:13 | 823 byte(s)]
 Re: 遇见调试中最不想遇见的问题了--不稳定。各位大侠帮小弟看看! - BNN [ 2006-03-29 22:00 | 61 byte(s)]
 Re: 遇见调试中最不想遇见的问题了--不稳定。各位大侠帮小弟看看! - smarteasy [ 2006-03-29 22:33 | 841 byte(s)]
|
|
|
|
[Original]
[Print]
[Top]
|
背景::
SOC的芯片,CPU是MIPS4000,243MHz, 16M flash, 64M DDR
BSP起来后(一个带类似内核的BSP),运行时经常会出exception。
现象:
比如通过串口执行Help, Ping等(有简单的网卡驱动),执行几次就会出exception;
每种运行情况出现的EPC地址都不一样,比如执行Help时的EPC地址经常出现在0X83DF9F40, 执行Ping时的EPC地址经常出现在0x83ff2034.
如果不执行任何任务,不会出EPC
调试:
开始以为可能是DDR不稳定,于是对DDR进行严格的读写测试,结果发现DDR工作非常稳定,没有出错。BSP的版本很成熟,代码不会出那么多的问题。
从EPC地址和CAUSE查的方法,感觉不适合这种情况。
有人认为问题的根源可能是某些硬件部分不稳定,我也这样担心,只是不稳定的硬件同样不好查。
不知道各位朋友有没有碰到这种类似的“不稳定”问题,大家共同探讨。
|
|
|
[Original]
[Print]
[Top]
|
|
[Original]
[Print]
[Top]
|
Thanks for your great advice BNN!
Actually the power&ground was the first part we debugged. Yet here are one or two not-so-good PCB design, like the ground layer was intrusived by a small plane of 2.5v power supply. We cannot make sure that this is the cause. The power signal probed was not bad (noise within acceptable range). Maybe we will redesign the PCB if we cannot found any other causes. But this is the last and unwanted choice.
A new problem I find today is that the PLL software configuration cannot be actived, that is the PLL
configration register, Active bit, cannot be set. I'm checking on this issue.
Any other advice, we'll be very appreciated!!
Best Regards!
Smarteasy
|
|
|
[Original]
[Print]
[Top]
|
|
|