Subject: Re: Wrong Machine identification (was ser.c and ttnread)
To: Arthur Hoffmann <hoffmann@it.ntu.edu.au>
From: Michael L. Hitch <osymh@gemini.oscs.montana.edu>
List: amiga-dev
Date: 05/15/1994 22:12:18
On May 16, 12:57pm, Arthur Hoffmann wrote:
> Now I commented out all the scsi drivers in my GENERIC file, except
> ahsc0, also I suped the very latest 16.05.1994 stuff and did make
> depend;  make again, now The kernel had a problem with sd.c (the
> one in /usr/src/sys/scsi, not the one in /usr/src/sys/arch/amiga/dev)

  The amiga/dev/sd.c (and st.c) is no longer being used.

> loadbsd resulted in an undefined function
> dk_establish
> I found it in the /usr/src/sys/arch/amiga/gtsc.c file, so I just
> copied it into the sd.c file and the compile worked.
> The function only contains:
> 
> int
> dk_establish()
> {
>         return(-1);
> }

  Hmmm.  I'm not sure where that routine is supposed to be, but it
certain shouldn't be in the driver file that may not be included.

> ahsc0 targ5 lun0 <Quantum LP105S 9101094053.1> SCSI2 direct fixed
> sd5 at scsibus1: 100MB, 1211 cyl, 4 head, 42 sec, 512 bytes/sec
> ahsc0 targ6 lun0 <Quantum LP240S GM240S01X6.4> SCSI2 direct fixed
> sd5 at scsibus1: 234MB, 1818 cyl, 4 head, 65 sec, 512 bytes/sec
> zthreebus0 at mainbus0
> 
> here the machine hangs with the HD light on steady. When I press a key
> I get a panic with the PC at 0001A34E.

  Maybe you are the first on to test the A3000 driver code?  I don't
know if Chris Hopps had anyone try out the A3000 or A2091 driver.
Neither he nor I have an A3000 to test A3000-specific things on.

  Oh oh - I found what may be the problem:  a3000dmaintr() calls
sbicintr, but passes the controller number instead of the sbic_softc
pointer.  Try changing "sbicintr(i)" to "sbicintr(dev)".  The A2091
driver has the same bug (atzsc.c).

  The panic is probably caused by a bug when a keyboard interrupt
is generated before the console device has been opened.

> Is this related to the dk_establish function?

  I doubt it.

> And why SCSI2? I thought A3000 are SCSI1. Just in case it matters my
> SCSI chip is the A version rev 00-08.

  SCSI2 is the value the drive reports in the inquiry data.  It has 
nothing to do with the host adapter.

Michael

------------------------------------------------------------------------------