|
Hallo all, I am a new in DSP programming using ADSP 21065L. And I have some question regarding on it. Now I am using Visual DSP++ in my project, in order to make my program easier, since I can write the program in C++ and just compile it. But recently everytime I want to debug my compiled program (it can be compiled, so it means there is no error in my program, isnt it?). There is always error message saying that "Communication timed out. Board is in the unknown state". And it asked me to reset processor. What does it mean? As you know, this error message doesnt appear directly right after I stepped in to the program. It normally appears after some steps. Is there any limited cycle that can be run in Visual DSP++? If yes, how to make it longer? Thank you very much in advance. Best regards, Suhadi |
|
|
Communication timed out in Visual DSP++
Started by ●January 16, 2002
Reply by ●January 16, 20022002-01-16
|
Hi, thanks for the quick reply .... But I think it is a problem with Visual DSP++, since it is also happened in my assembly program, even in the example program from ADSP itself. Can I see and change the setting of the Visual DSP++ esp for the communication? I asked my colleages, they also have the same problem. If the problem is unsolved, is it possible to use Visual DSP++ as the simulator, so I dont need to upload my asm or C++ program to the processor? If yes, I would like to know how to do it. Thank you very much in advance. Regards, Suhadi >From: Jaime Andres Aranguren Cardona <> >Reply-To: >To: "Suhadi ." <> >Subject: Re: [adsp] Communication timed out in Visual DSP++ >Date: Wed, 16 Jan 2002 06:53:13 -0800 (PST) > Hi, >Maybe you are overwritting something (a timer, an interrupt, etc...) you >need for debugging through a monitor program. Check out the EzKit (I think >you are using it) documentation, you'll find something. >JaaC _________________________________________________________________ MSN Photos is the easiest way to share and print your photos: http://photos.msn.com/support/worldwide.aspx |
Reply by ●January 16, 20022002-01-16
|
I've had similar and other problems with VDSP++. I also have a 21065l
EZKit and am running VDSP++ 2.0 with version 2.00.3 of the EZKit. Older versions of the EZKit do not work with VDSP++ 2.0. The "connection" from the board to the host (debugger) is not a very stable one. I am referring to the S/W. It is easily upset. If you mess with the UART you will cause problems or if you access certain interrupts that should be left alone. The manual covers this. However, since you said it happens with example programs as well it may be some setting in the VDSP++ environment. I have had problems when retrieving "data/memory". Right now, I have to upload my programs twice to get them to run. For example, I take the TT program and change in the input from MIC to LINE in the source code. Re-build it and upload it. The jumpers on the board are set for LINE and I have a 500Hz test signal from a function generator on the LINE input. The output goes to a scope. If I upload the modified TT program once and run it, I see nothing on the scope. If I stop the running program and do a re-load and run it, it works fine. Always takes two uploads for things to work right. I contacted AD and they could not duplicate it, so there was nothing they could/would do. I set up a second PC with the older version of VDSP and the tools and it only takes one upload to work right. Both PCs run the same OS (Win98 SE) with the same upgrades and S/W and ! the test is done on the same 21065l board (I have only one). I frequently crash VDSP++. It sometimes asks me to manually reset the board if the connection times out and then it crashes. All this with their own TT program. I find VDSP++ quite flaky. Maybe it's just with the 21065l toolkit added on. Has anyone had these types of problems with the 21061 toolkit? Marcell |
Reply by ●January 16, 20022002-01-16
|
Hi! Marcell Stoer wrote: > I frequently crash VDSP++. It sometimes asks me to manually reset the board if the connection times out and then it crashes. All this with their own TT program. I find VDSP++ quite flaky. Maybe it's just with the 21065l toolkit added on. Has anyone had these types of problems with the 21061 toolkit? Yes, VDSP++ _is_ quite flaky indeed. Though I haven't tested it with any 21x61 hw. Similar symptoms for me. I'm using it with a 218x ez-ice. It crashes frequently, and I even have to reboot this computer quite often (NT 4.0 SP6). It especially hates when I turn off and on the board that I'm testing (though, I guess it would be supposed to resync -- anything but definitely not crashing). When it goes out of sense, it sometimes even removes the 218x ez-ice from the session list, together with the session's settings. It makes me mad when I'm in a hurry. L. |
Reply by ●January 16, 20022002-01-16
|
> Similar symptoms for me. I'm using it with a 218x ez-ice. It
crashes > frequently, and I even have to reboot this computer quite often (NT 4.0 > SP6). It especially hates when I turn off and on the board that I'm > testing (though, I guess it would be supposed to resync -- anything but > definitely not crashing). When it goes out of sense, it sometimes even > removes the 218x ez-ice from the session list, together with the session's > settings. I used to run NT4SP6, but found it too slow on my older PC and put Win98SE on it. It runs much better now. I generally have no problems running Win98SE and have fewer reboots that with NT. I have been very careful not to push buttons I should not and VDSP++ has still crashed. I tried to make it crash by removing the serial connector once the TT program was running (I wanted to know if a flaky connection was the problem). The program kept running and when I reconnected it, VDSP++ was even able to halt it. When I reloaded TT it crashed though. The jury's til out on that one. Marcell |
Reply by ●January 17, 20022002-01-17
|
Hi Marcell and all, for the communication time out, I dont know if by lowering the baud rate can solve the problem. But recently I tried to make the baud rate slower into 9600. And it is worked though I have to wait a bit longer to upload and download my program. And now my new question is how to turn off the sixth LED (Flag 9th). I can turn the other LED off, but the sixth LED. I tried to turn it off, but as the result my program always crash. My code to turn the sixth LED off is the following: asm("R0 = DM(IOSTAT); R0 = BSET R0 BY 5; DM(IOSTAT) = R0;"); I think it is already right, isnt it? or is there other register that I should configure? Thank you in advance. Regards, Suhadi >From: Marcell Stoer <> >Reply-To: Marcell Stoer <> >To: >Subject: Re: [adsp] Communication timed out in Visual DSP++ >Date: Wed, 16 Jan 2002 8:51:40 PST > >I've had similar and other problems with VDSP++. I also have a 21065l >EZKit and am running VDSP++ 2.0 with version 2.00.3 of the EZKit. Older >versions of the EZKit do not work with VDSP++ 2.0. The "connection" from >the board to the host (debugger) is not a very stable one. I am referring >to the S/W. It is easily upset. If you mess with the UART you will cause >problems or if you access certain interrupts that should be left alone. >The manual covers this. > >However, since you said it happens with example programs as well it may be >some setting in the VDSP++ environment. I have had problems when >retrieving "data/memory". Right now, I have to upload my programs twice to >get them to run. For example, I take the TT program and change in the >input from MIC to LINE in the source code. Re-build it and upload it. The >jumpers on the board are set for LINE and I have a 500Hz test signal from a >function generator on the LINE input. The output goes to a scope. If I >upload the modified TT program once and run it, I see nothing on the scope. > If I stop the running program and do a re-load and run it, it works fine. > Always takes two uploads for things to work right. I contacted AD and >they could not duplicate it, so there was nothing they could/would do. I >set up a second PC with the older version of VDSP and the tools and it only >takes one upload to work right. Both PCs run the same OS (Win98 SE) with >the same upgrades and S/W and ! >the test is done on the same 21065l board (I have only one). > >I frequently crash VDSP++. It sometimes asks me to manually reset the >board if the connection times out and then it crashes. All this with their >own TT program. I find VDSP++ quite flaky. Maybe it's just with the >21065l toolkit added on. Has anyone had these types of problems with the >21061 toolkit? > >Marcell >_____________________________________ >Note: If you do a simple "reply" with your email client, only the author of >this message will receive your answer. You need to do a "reply all" if you >want your answer to be distributed to the entire group. > >_____________________________________ >About this discussion group: > >To Join: Send an email to > >To Post: Send an email to > >To Leave: Send an email to > >Archives: http://groups.yahoo.com/group/adsp > >Other Groups: http://www.dsprelated.com/groups.php3 >">http://docs.yahoo.com/info/terms/ _________________________________________________________________ Chat with friends online, try MSN Messenger: http://messenger.msn.com |
Reply by ●January 17, 20022002-01-17
|
That looks alright to me. That LED is set by the kernel after the POST is completed and the kernel enters the "halt loop". That's where the kernel waits for commands from the host. I am not an assembly language programmer, so I would not be able to read through the kernel code and tell you whether this LED is used as a diagnostic or not. You could look at it yourself to see if there is a problem with that LED or not (the kernel source code came with the tool kit). Marcell On Thu, 17 Jan 2002 20:56:34 +0700 wrote: > Hi Marcell and all, > > for the communication time out, I dont know if by lowering the baud rate > can solve the problem. But recently I tried to make the baud rate slower > into 9600. And it is worked though I have to wait a bit longer to upload > and download my program. > And now my new question is how to turn off the sixth LED (Flag 9th). I can > turn the other LED off, but the sixth LED. I tried to turn it off, but as > the result my program always crash. > My code to turn the sixth LED off is the following: > > asm("R0 = DM(IOSTAT); R0 = BSET R0 BY 5; DM(IOSTAT) = R0;"); > > I think it is already right, isnt it? or is there other register that I > should configure? > > Thank you in advance. > > Regards, > Suhadi |
Reply by ●January 18, 20022002-01-18
|
I must come to the conclusion that this fancy development system is just a large heap of sh*t with sugar on its top. It keeps frustrating me. I can't run my damned code since half an hour because of the """EASY""" ICE stuff's irritating stupid cryptic error messages. Whether I reset the ICE or not, reset the board or not, restart the computer or not or should I do _ANYTHING_. I know that it's not the right place to tell my opinion but I would have expected more useability and less consumer gaining fluff from such an expensive product. (Sorry guys, I'm pissed off.) L. |
|
|
Reply by ●January 18, 20022002-01-18
|
Hi!, source-ASIC wrote: > > Harsfalvi, I'm not a DSP engineer however the feedback I've gotten from > others is > that Visual DSP++ can be a slow dog unless you have a screaming 64-bit > workstation, > and depending if you are running other apps concurrently. Well I cooled down a bit. I'm not angry about the application's speed; could be faster, but runs reasonably well IMHO. Well, when it does. Best regards, Levente |
Reply by ●January 18, 20022002-01-18
|
If I've followed this thread correctly you are using 218X EZE-ICE. I have your problems too (as well as lack of documentation issues with the Visual DSP C compiler but that's another story.) I seem to have found a sequence of operations that keeps the EZE-ICE working (although I feel that for the money it should be a lot less fussy.) I'm running Windows 2000. Visual DSP 2.0, service pack 1, 2.5V EZE-ICE with 2189M target. To make it work: Start your PC (or exit Visual DSP if its running). Connect the EZE-ICE to target. Power up the target. Power up the EZE-ICE (I just push the power plug in ) If the little orange Activity light glows its Ok else power off target and ICE and try again. Start Visual DSP and it always works for me. If you try and reload code while the target is running you will need to exit VDSP and power off ICE and target to get it back. If your target is not working then VDSP won't start up properly if the ICE is the project target. (God knows why but that's what happens.) I hope this is useful. Let me know. Michael Kellett MK Electronics Ltd, 92A Howbury Street, Bedford, MK40 3QS Telephone: 01234 271640 Mobile: 07968 111148 Fax: 01234 272025 Email: -----Original Message----- From: Harsfalvi Levente [mailto:] Sent: 18 January 2002 13:26 To: Subject: Re: [adsp] Communication timed out in Visual DSP++ I must come to the conclusion that this fancy development system is just a large heap of sh*t with sugar on its top. It keeps frustrating me. I can't run my damned code since half an hour because of the """EASY""" ICE stuff's irritating stupid cryptic error messages. Whether I reset the ICE or not, reset the board or not, restart the computer or not or should I do _ANYTHING_. I know that it's not the right place to tell my opinion but I would have expected more useability and less consumer gaining fluff from such an expensive product. (Sorry guys, I'm pissed off.) L. _____________________________________ Note: If you do a simple "reply" with your email client, only the author of this message will receive your answer. You need to do a "reply all" if you want your answer to be distributed to the entire group. _____________________________________ About this discussion group: To Join: Send an email to To Post: Send an email to To Leave: Send an email to Archives: http://groups.yahoo.com/group/adsp Other Groups: http://www.dsprelated.com/groups.php3 ">http://docs.yahoo.com/info/terms/ |
