Reply by Bende Georg ●January 19, 20042004-01-19
I use the following "monkey trap" routine for stack space control:

asm{move SP,i};
if (i>STACK_LIMIT)asm(debug); //monkey trap

Actually the #define of STACK limit depends on your linker.cmd file, you should
check what "normal" and "out of control values are.

Georg Bende
Softwareentwickler
Abteilung Elektronik
Dr. Fritz Faulhaber GmbH & Co KG
Daimlerstr. 23
71101 Schaich
Tel: +49 7031 638294 > -----Ursprgliche Nachricht-----
> Von: Bruno Tremblay [mailto:]
> Gesendet: Freitag, 16. Januar 2004 20:18
> An:
> Cc: Rajarama Puranika;
> Betreff: RE: AW: [motoroladsp] Programming Data Flash on a F805 >
> I've done some tests with Jacob Christ routine for
> programming flash. So for
> now, I don't use the SDK drivers. For what I can tell, the flash erase
> routine work fine.
>
> void dflash_erase (UInt16 addr)
> {
> UInt16 readsyscntl;
> ArchIO.DataFlash.IntSourceReg = 0;
>
> ArchIO.DataFlash.ProgramReg = 0;
> ArchIO.DataFlash.EraseReg = 0;
> ArchIO.DataFlash.ControlReg = 0;
>
> while ((readsyscntl = ArchIO.DataFlash.ControlReg) & 0x8000 );
> ArchIO.DataFlash.EraseReg = 0x4000 + ( ( addr >> 8) & 0x007f);
> *(UInt16 *) addr = 0;
> while ((readsyscntl = ArchIO.DataFlash.ControlReg) & 0x8000 );
> ArchIO.DataFlash.EraseReg = 0;
> ArchIO.DataFlash.ControlReg = 0;
> }
>
> But the software hang on the flash program routine.
>
> void dflash_program (UInt16 addr, UInt16 data)
> {
> UInt16 readsyscntl;
> ArchIO.DataFlash.IntSourceReg = 0;
> ArchIO.DataFlash.ProgramReg = 0;
> ArchIO.DataFlash.EraseReg = 0;
> ArchIO.DataFlash.ControlReg = 0;
> while ((readsyscntl = ArchIO.DataFlash.ControlReg) & 0x8000 );
> ArchIO.DataFlash.ProgramReg = 0x4000 + (( addr >> 5) & 0x03ff);
> *(UInt16 *) addr = data;
> while ((readsyscntl = ArchIO.DataFlash.ControlReg) & 0x8000 );
> ArchIO.DataFlash.ProgramReg = 0;
> }
>
> For my tests, I tried to erase the first page of dataflash. I placed a
> breakpoint just after dflash_erase (0x1000) to verify. At
> that point it
> worked fine. All the data were 0xFFFF. But it went to an
> unhanded interrupt
> when I tried to program the first word of flash with the
> dflash_program
> (0x1000, Calib[0]). I thought that by using archDisableInt(), it was
> disabling all interrupts and getting rid of the interrupt problem.
>
> static void ProgFlash(void)
> {
> archDisableInt();
>
> dflash_erase (0x1000);
>
> dflash_program (0x1000, Calib[0]);
>
> archEnableInt();
> }
>
> When rechecking the archive, I found message 1793 (sep 30,
> 2003 7:06 am)
> where Jacob Christ spoke about being out of stack space. Can
> some one put a
> bit of light on this? If it's the stack space, how can I change it?
>
> Thanks
>
> Bruno > -----Message d'origine-----
> De : Rajarama Puranika [mailto:]
> Envoy: 15 janvier, 2004 22:18
> : Bende Georg
> Cc : ;
> Objet : Re: AW: [motoroladsp] Programming Data Flash on a F805 > Even I had the same problem, when I tried to write the
> first 1k of data flash. The processor may store
> constant data, in first 1k of data flash. So it may
> convert the segment into read only memory and when you
> try to write, it may generate an interrupt which is
> unhandled.
> So just avoid writing into first 1k segment.Start from
> next segment it may work fine. And also you need to
> block all the interrupts while writing into dataflash
> ,and should be the highest priority task.
>
> Thanks,
> Rajaram > --- Bende Georg <> wrote:
> > I don't know that the SDK solution is, but I do it
> > as it is described in the user's manual from page
> > 161 through 164. You have to erase one page of the
> > data flash, then write that part. You have to know
> > EXACTLY which page contains your data. I created an
> > array of 256 values which is put in the beginning of
> > dFlash for this purpose.
> > Your application cannot access the data flash
> > during erase/write operations, this means you cannot
> > access your data constants and variables that were
> > stored there.
> > Problem might be this in your case. If it is badly
> > needed, store your constant table in the program
> > flash. (or if you aren't using it, boot flash)
> > Interrupt activation and deactivation goes through
> > setting/clearing the SR bits I0 and I1 (p139-140 in
> > the manual)
> >
> > Georg Bende
> > Softwareentwickler
> > Abteilung Elektronik
> > Dr. Fritz Faulhaber GmbH & Co KG
> > Daimlerstr. 23
> > 71101 Schaich
> > Tel: +49 7031 638294
> >
> >
> > > -----Ursprgliche Nachricht-----
> > > Von: Bruno Tremblay [mailto:]
> > > Gesendet: Mittwoch, 14. Januar 2004 20:03
> > > An:
> > > Betreff: [motoroladsp] Programming Data Flash on a
> > F805
> > >
> > >
> > >
> > > Hi,
> > > I know that is a old subject on this group but I
> > did't found
> > > the answers to
> > > my questions in the archive. I'm trying to
> > programm data
> > > flash on a F805 to
> > > save some calibration variables. I'm using the SDK
> > Flash
> > > drivres. For now, I
> > > can read what's in the data flash but when I try
> > to write in
> > > it, the DSP
> > > hang. From what I read and understand, I must not
> > de access from an
> > > interrupt while erasing or writing. But does it
> > mean that I should
> > > deactivate all the interrupts (PWM, Quad Timers
> > and SCI)
> > > during writing? If
> > > yes, how can the interrupts be reactivate
> > automaticaly after
> > > the write.
> > >
> > > Beside The SDK, is there a easy way to use the
> > data flash to
> > > save about 30
> > > words of data?
> > >
> > > Bruno >
> _____________________________________
> 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:
>
> To Post:
>
> To Leave:
>
> Archives: http://www.yahoogroups.com/group/motoroladsp
>
> More Groups: http://www.dsprelated.com/groups.php3 >