1230scsi.device Access Fault????

[CFou!]

I have a technical curiosity....

ON 68060 Blizzard+ Quantum SCSI fireball+IDE HD+32MO

I have all times this test IN 1230scsi.device

tst.w $ffff0

Have you an idea of the reason of this test at the chip memory?

PS: SCSI Loading are really slow.... more slow with IDE drive
it's the reason?


See MUGuardianAngel, LOG

WORD READ from 000FFFF0 PC: 760022BA
USP : 760131A8 SR: 0000 (U0)(-)(-) TCB: 76010274
Data: 0000000A 00000007 00000000 0000FFFF 00000000 00000000 00000000 C0000000
Addr: 7601036C 76010310 760470B0 76014AA0 00EA8000 760101EC 00002130 001FFFF0
Stck: 4EF900F9 F5224EF9 00F9F522 4EF900F9 F5224EF9 00F9F522 4EF900F9 F5224EF9
Stck: 00F9F522 4EF900F9 F5224EF9 00F9F522 4EF900FA 3AE64EF9 00FA3A7C 4EF900FA
7600229a : 0001 0018 ori.b #$18,d1
7600229e : 1968 0002 0030 move.b $2(a0),$30(a4)
760022a4 : 1968 0005 0014 move.b $5(a0),$14(a4)
760022aa : 41e9 005a lea.l $5a(a1),a0
760022ae : 3018 move.w (a0)+,d0
760022b0 : 33fc 4000 00df f09a move.w #$4000,$dff09a
760022b8 : 4e71 nop
760022ba : *4a79 000f fff0 tst.w $ffff0
760022c0 : 4e71 nop
760022c2 : 082d 0002 1e6c btst.b #$2,$1e6c(a5)
760022c8 : 6642 bne.s $7600230c
760022ca : 1229 0046 move.b $46(a1),d1
760022ce : b23c 0041 cmp.b #$41,d1
760022d2 : 6706 beq.s $760022da
760022d4 : 1969 0044 0008 move.b $44(a1),$8(a4)
Name: "1230scsi.device"

[NOB]

A tst.w followed by a nop instruction
is curios.
Normally it is followed by a branch-intruction.
IMO it makes no sense.(Except it is a mini wait instruction.)

[keropi]

scsi troubles eh....?
me too... damn, I am tempted for a fastata...

[musashi5150]

More importantly:- what's at $ffff0 ?? Did you get a dump of this area too CFOU! ?

[mrodfr]

hello,

I don't know if It's well related to the cfou's question but if you have hits from 1230scsi.device when testing software there is a bugs on 1230scsi.device and can be corrected by using:

run <>NIL: fixcybaccess or the other one (don't remember the name). here this one used with my 1260. read the doc/guide for the answer.

and found on the mmulib.lha archive on aminet by thomas richter (so sad that thomas stop amiga because amiga died).

again hope It's the good anwer of the question. sorry if not.

[Chain]

um..its version 8.5 ?

[mrodfr]

hello,

yes, with a1234.rom 8.5 and 8.1 I have.

if someone have A1234.ROM newer than 8.5, I'm verry interested :-)

[CFou!]

Originally Posted by mrodfr

hello,

yes, with a1234.rom 8.5 and 8.1 I have.

if someone have A1234.ROM newer than 8.5, I'm verry interested :-)

have you read/write speed problem on your SCSI drive?

[mrodfr]

Originally Posted by CFOU!

have you read/write speed problem on your SCSI drive?

no, 3.8mo/s with sysspeed on my old quantum vikingII.
never tested on my seagate 160go with acard ide to scsi.

have you tested the little fix with your hardware ???? theorically the slow down doesn't come from the hit of 1230scsi.device. (just annoying mmuforce hit allways).

maybe value too high on mask and maxtransfer ???
here: mask:7ff.....c max:fffe00 (value came from the author of sfs)

maybe bad terminated scsi chain ????

hope that help.