完善资料让更多小伙伴认识你,还能领取20积分哦, 立即完善>
扫一扫,分享给好友
我很难让EBI与PIC32MZ2048ECH144芯片一起工作。我使用了所有最新的开发工具,据我所知:MPLAB X V3.40XC32V1.42HARMonyV1.09Harmony配置正确,我所做的就是配置振荡器EBI,还有一个I/O引脚和用于调试的UART。振荡器通过I/O引脚和UART定时确认是正确的。当我停止调试以免试图读取指向我映射的图形控制器芯片的指针时,EBI寄存器看起来配置正确;配置为4MbSRAM@物理地址0x20000,使用多达地址线A21,16位数据总线。EBICS0=0x20000000EBIMSK0=0x00000027EBISMT0=0x07000001EBIFTRPD=0x000C8EBISMCON=0x00000001CFGEBIA=0x0003FFFCFGEBIC=0x01000313I正试图从偏移0和偏移1处的前两个映射位置读取数据,这两个位置应该提供两个16 16 16指示图形控制器存在的位值。我试图从0xE0000000(不可缓存)和0xC0000000(可缓存)的基本地址读取,但是没有任何结果。我可以从UART中看到正确的指针值,但是一旦读操作被调用,处理器就不会进入任何人的领地。当调试MPLAB X变得与我的真实ICE不同步时,如果我对芯片进行编程并在不调试的情况下运行它,则会得到相同的结果,适当的指针值来自UART,但之后什么也没有。re UART输出,在异常陷阱中设置的任何断点都不会打印字符串,调试也不会停止。被编译器征服了,谢谢任何我能得到的帮助!
以上来自于百度翻译 以下为原文 I am having great difficulty in getting the EBI to work with the PIC32MZ2048ECH144 chip. I am using all the latest dev tools as far as I can determine: MPLAB X V3.40 XC32 V1.42 Harmony V1.09 Harmony is configured properly as far as I can tell, all I have done is configure the oscillator the EBI, and one I/O pin and a UART for debugging. The osciallator is confirmed is proper thru both an I/O pin and proper UART timing. The EBI registers look like they are configured properly when I stop debugging just shy of trying to read a pointer to the graphics controller chip I have mapped; configured as 4Mb SRAM @ physical address 0x20000000, using up to address line A21, 16 bit data bus. EBICS0 = 0x20000000 EBIMSK0 = 0x00000027 EBISMT0 = 0x07000001 EBIFTRPD = 0x000000C8 EBISMCON = 0x00000001 CFGEBIA = 0x003FFFFF CFGEBIC = 0x01000313 I am attempting to read from the first two mapped locations at offset 0 and offset 1 which should provide two 16 bit values indicating that the graphics controller is present. I have tried to read from a base address of 0xE0000000 (non-cacheable) and 0xC0000000 (cacheable) with no results at all. I can see the proper pointer values coming out the UART but once the read action is called the processor goes off into no man's land. When debugging MPLAB X becomes out of synch with my REAL ICE, if I program the chip and run it without debugging I get the same results, proper pointer values coming our of the UART but nothing after that. No exceptions are being thrown, I am using the Harmony exception framework to which I added more UART output, no strings are ever printed nor does the debugging stop at any breakpoint I set in the exception trap. The EBI documentation mentions that the TLB and MMU need to be configured properly but there is little or no documentation on those features, they are supposed to be properly initialized by the compiler. Thanks for any help I can get! |
|
相关推荐
11个回答
|
|
嗨,EBI被用于几个图形示例中,因为在一些演示中需要外部SRAM作为帧缓冲。如果您正在启动项目,请考虑切换到PIC32MZ...EF,它是与PIC32MZ...EC针脚兼容的设备。除了ADC,它是完全SW兼容的。EF版本的硅缺陷更少,A/D转换器更好,它具有相同的外围设备(除了ADC)并且增加了FPU。
以上来自于百度翻译 以下为原文 Hi, EBI is used in several graphics examples because external SRAM is needed for frame buffer in some demos. If you are starting your project, please consider switching to the PIC32MZ...EF which is a pin to pin compatible device with PIC32MZ...EC. It is fully SW compatible except ADC. EF version has much less silicon bugs, the A/D converter is much better, it has same peripherals (except ADC) and it adds FPU. Regards |
|
|
|
是的,正如我们所说的HTTP://www. MyCHIP.COM/FUMMS/M9550.ASPX,TLB和MMU在链接器文件和编译器中处理的东西是…相当复杂/复杂。关于在文档中可以找到的TLB条目的小信息是不够的。
以上来自于百度翻译 以下为原文 yeah, as we said http://www.microchip.com/forums/m955010.aspx That thing that TLB and MMU are handled in the linker file and hence by the compiler is... quite complicated/complicating. And the little info about TLB entries that you can find in the Docs are not enough. One would need some examples on how to map... |
|
|
|
谢谢大家,RISC:我正在研究最初的原型,我已经考虑转向EF部件,特别是现在我知道EC部件上的硅有问题。最初的处理器决定是在一段时间以前,现在EC部件已经被标记为不推荐用于新的设计,这可能是我需要的方式,只需要交换芯片并更改projectDario:唯一的例子是在PIC32MZ上的EBI文档中,它声明了假设TLB和MMU配置正确。现在我坚信我所写的都是正确的,所以可能是硅问题。我会换成EF部分并汇报。
以上来自于百度翻译 以下为原文 Thanks all, RISC: I am working on the initial prototypes and I have considered moving to the EF parts, expecially now that I know that the silicon on the EC parts have issues. The initial processor decision was a while ago, and now the EC parts have been flagged as not recommended for new designs, that may be the way I need to go and simply have the chips swapped out and change the project Dario: The only example is in the EBI docs on the PIC32MZ's and it states that is assumes that the TLB and MMU are configured correctly. I firmly believe now that all that I have written is correct, so perhaps it is a silicon issue. I will change to the EF parts and report back. |
|
|
|
是的,我的问题是不同的,因为我可以使这个示例工作(使用默认链接器脚本等),但我需要使用4个EBI“端口”。但就是这样…
以上来自于百度翻译 以下为原文 Yeah, my issue is different since I can make that example work (using default linker script etc) but I need to use 4 EBI "ports". But that's it... |
|
|
|
我甚至不能走那么远,我只跟一个芯片用~CS0交谈。你是在使用EF还是EC芯片?
以上来自于百度翻译 以下为原文 I can't even get that far, I am only talking to one chip with ~CS0. Are you running with the EF or EC chips? |
|
|
|
好,这给了我一点希望,我可以让我的单个设备EBI工作…我注意到,与EBI配置和控制引脚选择没有改变引脚的配置,因为它做了当你配置振荡器或UART或SPI端口。你是否需要在那里进行更改以使第一台设备正常工作?我计划确保CS和其他线路被设置为输出或输入根据需要…
以上来自于百度翻译 以下为原文 Good, that gives me a small amount of hope that I can get my single device EBI working ... I did notice that with the EBI configured and control pins selected nothing changes in the pin configuration as it does when you configure the oscillator or UART or SPI port. Did you have to make changes there to get the first device working? I am planning to make sure the CS and other lines are set as outputs or inputs as required ... |
|
|
|
嗯,现在我不确定,因为它不是我测试真正的董事会:但所有的工作根据PIC的数据表和EBI文档…是的,尝试他们!
以上来自于百度翻译 以下为原文 Hmmm now I'm not sure 'cause it was not me testing the real board: but everything worked according to PIC's datasheet and EBI doc... Yeah, try them too! |
|
|
|
没有设置I/O引脚配置的乐趣,处理器进入NANA土地的相同结果…我不认为我的问题是外在的,但值得一试。我把控制线和地址线设置为除了RDY线之外的输出,而数据线则保留为数字输入。模块:系统总线当权限访问被启用时,不被允许的发起者的任何访问都不会成功,但是状态寄存器可能不能准确地报告违规行为。工作周围--------------------------------------------------------------------------------------------------------。我想知道PC是否进入了除了普通e之外的其他错误陷阱。XCEP处理程序?顺便说一下,勘误表不在EF芯片中。
以上来自于百度翻译 以下为原文 No joy on setting the I/O pin configuration, same result of the processor going off into nana land ... I didn't think my issue was external but it was worth a shot. I set the control lines and address lines to outputs except for the RDY line, that and the data lines were left as digital inputs. I did see something interesting in the errata for the EC chip: 17. Module: System Bus When Permissions Access is enabled, any access by an initiator that is not allowed will not succeed; however the status registers may not accurately report the violations. Work Around --------------- None. I wonder if the PC goes off into some other error trap other than the general exception handler? BTW that errata is not in the EF chip. |
|
|
|
我还没有准确地使用参考代码,我的EBI配置是不同的,因为我不想与SRAM芯片只与CS线,我使用的BSO,BS1,和一个现成的线路。我可能会尝试他们的确切代码(在我的环境中当然)只是为了看看是否有用。即使我的外部图形芯片没有正确地响应控制线(它不应该),如果处理器不运行到没有人的土地,我得到一些垃圾数据,我可以说,这可能是一个EBI配置问题。我的EF芯片是在一夜之间,董事会将之后改变,但我保留第一板与EC芯片只是为了比较,现在我进入另一个项目,或两个,而芯片正在改变。我会让你知道我看到的情况下,它有助于你的EBI问题…
以上来自于百度翻译 以下为原文 I haven't used the Reference code exactly, my EBI configuration is different as I am not trying to interface an SRAM chip with only a CS line, I am using the BSO, BS1, and one of the ready lines. I might try their exact code (within my environment of course) just to see if something works. Even if my external graphics chip does not respond properly to the control lines (which it shouldn't) if the processor does not run off into no-mans land and I get some garbage data I could then say that it might be an EBI configuration issue. My EF chips are coming in overnight and the boards will be changed after that but I am retaining this first board with the EC chip just for comparison, right now I am off into another project or two while the chips are getting changed. I'll let you know what I see in case it helps your EBI issues ... |
|
|
|
好啊!IMO,你不应该得到例外(没有人的土地是好的:)如果你试图解决错误的或“陌生”的地址,不管引脚设置等…
以上来自于百度翻译 以下为原文 Ok! IMO, you should not get an exception (no man's land is nice :) ) if you're trying to address a wrong or "strange" address, regardless of Pins settings etc... |
|
|
|
确切地!去引用一个空指针会产生一个数据总线保护故障,除以0会产生一个算术故障等等。如果没有一个异常处理程序,处理器会重置以清除错误。我不是说根据个人经验,我听说它应该这样工作;-)试图读取一个我没有权限的存储器地址,很可能会导致数据总线故障,而不是将处理器发送到某个不可恢复的地方。errata向我暗示,从数据指针读取不应该给出的数据可能不会给我一个正确的错误代码或PC值,但是它仍然应该被捕获。
以上来自于百度翻译 以下为原文 Exactly! Dereferencing a null pointer should generate a data bus protection fault, dividing by 0 generate an arithmetic fault, etc. And you can get the PC at which it occurred. And if you don't have an exception handler in place the processor resets to clear the fault. Not that I am talking from personal experience, that's how I heard it is supposed to work ;-) Trying to read a memory address for which I do not have the permission would most probably result in a data bus fault, not send the processor off somewhere non-recoverable. The errata implies to me that reading from a data pointer that I am not supposed to might not give me a proper error code or PC value but it still should be trapped. |
|
|
|
只有小组成员才能发言,加入小组>>
5250 浏览 9 评论
2037 浏览 8 评论
1958 浏览 10 评论
请问是否能把一个ADC值转换成两个字节用来设置PWM占空比?
3218 浏览 3 评论
请问电源和晶体值之间有什么关系吗?PIC在正常条件下运行4MHz需要多少电压?
2266 浏览 5 评论
788浏览 1评论
680浏览 1评论
有偿咨询,关于MPLAB X IPE烧录PIC32MX所遇到的问题
609浏览 1评论
PIC Kit3出现目标设备ID(00000000)与预期的设备ID(02c20000)不匹配。是什么原因
685浏览 0评论
582浏览 0评论
小黑屋| 手机版| Archiver| 电子发烧友 ( 湘ICP备2023018690号 )
GMT+8, 2024-12-29 02:24 , Processed in 6.349383 second(s), Total 68, Slave 62 queries .
Powered by 电子发烧友网
© 2015 bbs.elecfans.com
关注我们的微信
下载发烧友APP
电子发烧友观察
版权所有 © 湖南华秋数字科技有限公司
电子发烧友 (电路图) 湘公网安备 43011202000918 号 电信与信息服务业务经营许可证:合字B2-20210191 工商网监 湘ICP备2023018690号