Port-vax archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Anyone with a QDSS equipped machine



Den 2021-12-05 kl. 20:06, skrev Mouse:
     stray interrupt: vector 0x10, ipl 31
I don't know whether this is the case here, but SCB vector 0x10 is
"Privileged Instruction fault".
IIRC the "stray interrupt" prints out the offset from beginning of
the SCB.
So, it might be something completely different?
I don't have the source on hand for that version.  But what I do have
on hand says

scb_stray(void *arg)
{
...
	vector = ((int) arg) & ~3;
...
		printf("stray interrupt: vector 0x%x, ipl %d\n", vector, ipl);

so, assuming that code remains operationally unchanged, the value
printed appears to be the byte offset from the SCBB.

My VARM transcription says that offset 0x10 is for "reserved or
privileged instruction" fault.  If the SCB is seen as array of
longwords, the vector at [0x10] is at offset 0x40, which is the CHMK
handler.  I find the former far more plausible, not to mention a better
fit for the code probably responsible for the printf.
Actually, no.  The vector printed is as documented in the VARM, so 0x10 is privileged instruction fault.

But - here is the odd thing - this vector cannot be a "stray interrupt", since it always has a vector (strays are for detection of device interrupts).
So something odd is happening here.

Best way to debug it would probably be to do something like add a halt in the stray routine and then try to find out where it traps...

-- R


Home | Main Index | Thread Index | Old Index