T.R | Title | User | Personal Name | Date | Lines |
---|
124.1 | more info please ... | EICMFG::KAV30_SUPP | | Mon May 09 1994 11:50 | 15 |
| Hi Marino ,
since you have the latest revision of the KAV30, the problem I was
thinking about we once had with VME-irq's and SCSI are already solved.
Therefore I currently have no idea what your problem could be...
could you post the following information:
- ebuild .dat file
- complete configuration in the crate (e.g. slot #, irq-level, vector
and base address of all the boards you have in there ?
Thanks and best regards,
Thomas
|
124.2 | More data on the problem | TRN01::FORMIA | | Tue May 10 1994 11:19 | 29 |
| Hi!
following is the ebuild file
characteristic /connect_time=300 /remote_cli /remote_term /norotate /net_device=EZA /objects=20000 /ports=200 -
/jobs=24 /subprocesses=64 /debug=none /loader=30000 /separate_loading /p0_virtual_size=25600 -
/p1_virtual_size=4096 /io_region=10000
characteristic /interrupt_stack=1024 /target=37 /sys1_param=%X870000FF /sys2_param=%XA0006
node /lat=active
program BOOTER$FIL:BOOTER /debug
program ELN$:EDISPLAY.EXE /norun
device EZA /vector=%X130 /net_def
device IC40 /vector=%XA00 /noautoload
device ANCHORCOMMUNICATIONDEVICEMASTER /register=%O122 /vector=%O4064 /noautoload
device ANCHORDEVICEMASTER /vector=%X830 /noautoload
device ANCHORDEVICESLAVE /vector=%X838 /noautoload
device LMCOMMUNICATIONDEVICEMASTER /vector=%X834 /noautoload
device LMDEVICEMASTER /vector=%X830 /noautoload
device LMDEVICESLAVE /vector=%X838 /noautoload
device DIGIO /vector=%X800 /file=CAGIS$SYS:DIGIO.EXE
device DUA /vector=%X550 /priority=6 /file=KRDRIVER /device_dependent=%X7FF
IRQ used are 1 for Digital I/O
2 for IC40
IC40 is a Frame grabber (OS/9 board handled by KAV30 driver as I/O port
for the application.
ciao, MARINO
|
124.3 | ... | EICMFG::KAV30_SUPP | | Wed May 18 1994 23:30 | 12 |
| Hi,
we can not see anything unusual with this, so the only explanation we
have is the following:
VME-irq's have higher priority than SCSI, so if you have an VME irq,
SCSI processing is stopped - but once the VME-irq is done, SCSI should
resume operation. Normally this is transparent to the user and works
fimne... How many VME-irq's do you get pwr second ?
Thomas
|
124.4 | more on Digital I/O driver | TRN01::FORMIA | | Fri May 20 1994 00:02 | 8 |
| Thanks Thomas,
I spent time to review and review the code (I also tried to change from
autoload to no autoload the driver, change the vector and more ...) but
with no better result; as soon as I get an interrupt from one channel
of digital I/O board I go into the problem.
ciao, Marino
|
124.5 | lab test required... | EICMFG::KAV30_SUPP | | Fri May 20 1994 11:07 | 10 |
| Hi Marino,
we will try to reproduce your problem in our lab... currently we have
no idea what the problem is - we will contact you by mail in order to
get more details of your application. I hope that we can solve your
problem soon!
best regards,
Thomas
|
124.6 | Not reproducible using IRQ/SCSI. | EICMFG::KAV30_SUPP | | Tue May 31 1994 11:28 | 5 |
| The problem described was tested with several combinations of VME IRQs
but the problem is not reproducible.
Julian.
|