Jump to content

Dave

Administrators
  • Content Count

    2,046
  • Joined

  • Last visited

Everything posted by Dave

  1. Best to use PIC16F628.h, the P prefix ones are for the old C2C compiler. If you include#<system.h> then it automatically includes the PIC prefixed header files. Please let us know if there is a problem in PIC16F628.h so we can correct it. Regards Dave
  2. Tom, I think I jump to my conclusion too quickly. Behind the scenes linker renames static functions so that they only get linked into the file they are declared in. I would expect both to fail to link in both cases because of my explanation of the static keyword when applied to a function. What version of the compiler do you use? Regards Dave
  3. It has been said that in 'C' static is an overused keyword. In the case of a function it means that it has static linkage, which means it is only visible in the file in which it was declared. So not a bug, just the way designers of the 'C' language thought it should work. Regards Dave
  4. Dave

    List File

    joli, A successful build in any version creates the following files: .lst (asm listing), .asm (assembly file), .casm ( c code and asm listing interleaved) and .map (memory map). Regards Dave
  5. What is surprising is that I'm not seeing any of the AV's, even with a similar hardware setup. That's a shame because it makes the whole thing much more difficult to nail. Just re-capping on my test system a clean built machine with the following: AMD Athlon XP2200+ 256MB RAM, Geforce MX440 graphics card. Windows 2000 SP4 (not pro version) IE 6.0 Without other user reports of this issue and with being able to reproduce the problem we are kind of stuck. Maybe Pavel has some thoughts ??? Regards Dave
  6. It all compiles and links just fine for me. I noticed a difference in our command lines: .... "C:\Program Files\SourceBoost\Lib\libc.pic16.lib" -O1 -p "test" -t 16F877 -ld "C:\Program Files\SourceBoost\Lib" .... "C:\Program Files\SourceBoost\Lib\libc.pic16.lib" -O1 -p "test" -t 16F877 -ld C:\Program Files\SourceBoost\Lib Looks like yours has the quotes missed from the last part. I'm using MPLAB V8.33, maybe that makes difference. Regards Dave
  7. I was referring to the BoostC license key, not the Win2k license key. Now I don't think will make any difference My experience is this: No access violations at all. But alot of the time registration of the compilers fail, even the command window displays success but the preg program indicates a failure in it's log window as well - maybe this is a related issue. I used your activation key QQA5YZE.... and that made no difference. I also just ran a batch file to key registering the PIC16 compiler and it was successful 51 out of 100 times. Regards Dave
  8. Try now, I;ve added rar as a supported attachment type. Regards Dave
  9. Might be handy if you zip up your project and add the attachment to your next forum post. Regards Dave
  10. Nonplussed, I now have a system setup with AMD Athlon XP 2200+ with 256MB RAM, running windows 2000. I have succesfully installed SourceBoost V6.97 and registered a Pro licence key, without any issue. From the distance past I seem to remember that sometimes activation keys don't work on some operationg systems (the less commonly used ones). So the answer will probably to request a new activation key. Regards Dave
  11. I think what you are talking about is some of the Flowcode functionality. It looks like from what you are saying that the macro substitution only works with with the first character after the %, any more characters are consider part of the text that will be kept. I don't really know Flowcode so this is a guess from what you say. Sounds like a question for the Matrix Multimedia Flowcode forum. Regards Dave
  12. Jean-Marie, In your MPLAB project file Affichage.c is never rebuilt, just linked in as an .obj file as the Affichage.c file is not in the MPLAB project but Affichage.obj is. Rebuilding this file resolves the issue (for me at least), so it looks like that file make have been built with a different version of the compiler. Please add Affichage.c to the project and remove Affichage.obj from the project. Regards Dave
  13. schneiderj, Which version of MPLAB are you using? This is an internal error, so you should never get to see it. The only thing I can think of is the the order of linkage is different between MPLAB and SourceBoost IDE, but even so this message should never appear. Please send all the project files to support@sourceboost.com so we can take a look. Regards Dave
  14. Your are right there must be two copies floating around. Duplicate variable definitions at fixed addresses is allowed under the BoostC compiler as long as the variables have the same fixed addresses. Normally when writing a library it is none target specific and hence won't have any fixed addresses in it. You are right to delare the cpu register variables used as extern, then the project they are linked in to will determine their addresses. Regards Dave
  15. tmd_63, When you purchase Flowcode you get a BoostC compiler license that can only be used from within the Flowcode environment. If you want to use standalone BoostC compiler then you can either use the free lience version, or buy the desired license.License details can be found here: http://www.sourceboost.com/Products/BoostC...imitations.html Regards Dave
  16. Consider baud rate accuracy, with RS232 comms the two ends are not synchronised with a common clock, they both free run at their own speed. If the clock is slightly wrong on one or both ends then there is potential for them to go in and out of phase. This could explain initially data is good (in phase), then goes bad (out of phase) and then good again(in phase). Check accurately timing of the serial data. Regards Dave
  17. What versions of software do you have? Regards Dave
  18. This target is not in the sourceboost package yet (but will be very soon), but you can get the all required files. See this post: http://forum.sourceboost.com/index.php?s=&...ost&p=16422 Regards Dave
  19. Once you have a program that compiles, press the bug symbol button on the tool bar to enter the debugger/simulator. Regards Dave
  20. Thanks for checking into this, support for COD file was dropped in favour of COF in the SourceBoost debugger as COD had so many limitations. MPASM output passed through MPLINK can generate a COF file. But when I tried this SourceBoost debugger would not load the resulting file - I haven't found out why yet. This item is now on our todo list. Regards Dave
  21. No, that particular device has the 12bit instruction opcodes and 2 deep stack (like PIC12's). The BoostC compiler only supports PICs with 14bit and 16bit instruction opcodes. Regards Dave
  22. Ted, Looks like MPASM no longer generates a debug output file: To get a COD or COFF file the .o file(s) from MPASM need to be pushed through MPLINK.exe. Regards Dave
  23. Disabling interrupts around the code that uses the shared variable will prevent corruption: void interrupt() { if(intcon.TMR0IF) { TickCounter++; intcon.TMR0IF=0; tmr0l=7; //reset timer0 } } void main() { while( 1 ) { if( someEvent ) { // prevent corruption by disabling interrupts intcon.gie = 0; TickCounter = 0; intcon.gie = 1; } } } Regards Dave
  24. Memory locations on PIC16 are multiplied by 2 to get the address required in the hex file, so 0x4200 is actually writing to 0x2100. It appears this all came about because Microchip uses 14 bits for the opcode and the opcode resides at a single address. Intel hex writes 8 bit data to an address, so each microchip opcode location needs two addresses. I will admit it is quite confusing at times. Take a look at the produced .asm and .lst files, they have everything in them in a more readily understandable form. If you push the .asm file through MPASM you will get the same hex file as BoostC produces. Regards Dave
  25. Looks like some duplication in the header file: #define _CONFIG1 0x8007 #define _CONFIG2 0x8008 .... ///////////////////////////////////////////////// // Config Register ///////////////////////////////////////////////// #define _CONFIG 0x2007 Thanks for pointing this out. Regards Dave
×
×
  • Create New...