Hi,
according to spru189f NMIE cannot be cleard manually. But there exists
IRQ_nmidisable(). How does this work?
TIA
Gustl
C6713: NMI, IRQ_nmidisable()
Started by ●February 1, 2007
Reply by ●February 2, 20072007-02-02
Bernhard Gustl Bauer schrieb:
> according to spru189f NMIE cannot be cleard manually. But there exists
> IRQ_nmidisable(). How does this work?
I had a look at the assembler code. It tries to clear the NMIE bit. But
it can't. I have tried it!
So why does TI write such a function? To confuse people?
Gustl
> according to spru189f NMIE cannot be cleard manually. But there exists
> IRQ_nmidisable(). How does this work?
I had a look at the assembler code. It tries to clear the NMIE bit. But
it can't. I have tried it!
So why does TI write such a function? To confuse people?
Gustl
Reply by ●February 5, 20072007-02-05
Gustl-
> > according to spru189f NMIE cannot be cleard manually. But there exists
> > IRQ_nmidisable(). How does this work?
>
> I had a look at the assembler code. It tries to clear the NMIE bit. But
> it can't. I have tried it!
>
> So why does TI write such a function? To confuse people?
I might guess it works on at least one device somewhere, maybe not even C5000 or C6000 series, and TI software developers are trying to keep CSL API consistent, so they made an "educated guess" without being
device experts :-)
-Jeff
> > according to spru189f NMIE cannot be cleard manually. But there exists
> > IRQ_nmidisable(). How does this work?
>
> I had a look at the assembler code. It tries to clear the NMIE bit. But
> it can't. I have tried it!
>
> So why does TI write such a function? To confuse people?
I might guess it works on at least one device somewhere, maybe not even C5000 or C6000 series, and TI software developers are trying to keep CSL API consistent, so they made an "educated guess" without being
device experts :-)
-Jeff
