Kommentar: Einfügen von HTML im Kommentar: Link einfügen: <a href="LINKURL" target="_blank">LINKTITEL</a> Bild einfügen: <img src="BILDURL"> Text formatieren: <b>fetter Text</b> <i>kursiver Text</i> <u>unterstrichener Text</u> Kombinationen sind auch möglich z.B.: <b><i>fetter & kursiver Text</i></b> C Quellcode formatieren: <code>Quellcode</code> BASIC Quellcode formatieren: <basic>Quellcode</basic> (Innerhalb eines Quellcodeabschnitts ist kein html möglich.) Wichtig: Bitte mache Zeilenumbrüche, bevor Du am rechten Rand des Eingabefeldes ankommst ! -> I > > > Hi all, > > > > > > I am trying to find out if and how assembly routines are interrupted > > > by the bytecode interpreter. The reason is that when my Probot128 > > > records or plays sounds, it remains in an assembly routine for many > > > seconds. If possible it should at the same time keep track of > > > the pulses from the wheel encoders via external interrupts. > > > > > > Now I found out that during an assembly routine, when interrupts are > > > not disabled with cli, when an interrupt occurs, the byte code > > > increases a counter associated with that particular interrupt. (It is > > > that counter that is read and reset with Irq_GetCount.) It seems that > > > is the only thing that happens. In any case, no user written ISR is > > > carried out. > > > > > > If interrupts are disabled with cli, any interrupt that occurs is > > > remembered by the hardware. When the assembly routine is left and > > > interrupts are enabled again with sei, the bytecode interpreter > > > increases the counter and jumps to the user defined ISR. > > > > > > Is that correct? > > > > Hallo Pepe, > > > > I think the bytecode-interpreter only checks the > > value of the counter-variable and the actual interrupt-routine > > does increament this variable. > > > > > > > > Would it be possible to locate the address of the routine that > > > increases the interrupt counter? > > > > Surely, because the address of the interrupt-routine should be > > placed in the interrupt-vector-table. But we must know where > > the table is. Is it at the start of the flash memory or at the > > the beginning of the boot loader section of the flash. > > Maybe we could check the IVSEL-BIT (interrupt vector select bit) > > in the MCU-Control-Register. > > The interrupt table in the bootloader is active, but that code uses > the vectors at the begin of the interpreter. > > > > > > > > > Would it be possible to replace that routine with another one, e.g. > > > by changing the interrupt vector? > > > > I would warn urgently against that unless you knows the system > > very exactly. For polling the interrupt flag you need only few > > system clocks. > > You are right, changing the interrupt vectors can have ugly side effects. > But except the timer interrupt that is used for multitasking, servo and clock, > experimenting with the other interrupts should not be fatal. > > Regards, > > Peter > > > > > Gruß Torsten > > > > > > > > I would be grateful for your ideas and insights. > > > > > > Regards, > > > Pepe > > > > > >
