k1te@lectroid.sw.stratus.com (Bradshaw Lupton) (07/24/90)
I am trying to interface to a scsi disk controller that provides neither parity nor ATN signals. (lucky me) The device information for the scsi chip on the mac gives register values that will allow parity to be shut off and ATN to be shut off. Is there a technote, or a method of changing the initial scsi register values? tnx, Bradshaw -- | Bradshaw B. Lupton Jr. |================| | Stratus Computer, Inc. |k1te@stratus.com| | Marlboro, Ma 01752 |================| | 508 490 6484 / 480 8219(f) |/\/\/\/\/\/\/\/\|
minow@mountn.dec.com (Martin Minow) (07/25/90)
In article <1786@lectroid.sw.stratus.com> k1te@.stratus.com (Bradshaw Lupton) writes: > >I am trying to interface to a scsi disk controller that provides neither >parity nor ATN signals. (lucky me) > The device "should" both generate and detect parity errors. If it doesn't detect parity errors, you run a risk of undetected data errors. This may be acceptable for a one-off piece of lab gear, but I wouldn't recommend selling a product that didn't support parity. If the device doesn't generate parity, it will probably not work on the Mac (the Mac will see parity errors.) ATN is generated by the Mac and the device "should" respond to it. I don't think the Mac currently actually uses ATN, but I would suspect that a future release of either A/UX or Mac OS will, as it is essential for increased throughput. > >The device information for the scsi chip on the mac gives register values >that will allow parity to be shut off and ATN to be shut off. > >Is there a technote, or a method of changing the initial scsi register >values? > You need the NCR 5380 hardware documentation, and knowledge of where the chip exists in the Mac address space (which, I would imagine, varies for different cpu's). I DO NOT RECOMMEND that you fiddle with the configuration registers: "Doctor, it hurts when I do that." "Don't do that." Martin Minow minow@bolt.enet.dec.com The above does not represent the position of Digital Equipment Corporation
wiseman@tellabs.com (Jeff Wiseman) (07/26/90)
In article <1788@mountn.dec.com> minow@bolt.enet.dec.com (Martin Minow) writes: >The device "should" both generate and detect parity errors. If it doesn't >detect parity errors, you run a risk of undetected data errors. This >may be acceptable for a one-off piece of lab gear, but I wouldn't recommend >selling a product that didn't support parity. If the device doesn't >generate parity, it will probably not work on the Mac (the Mac will >see parity errors.) Ok folks, I know that this is not what Martin was talking about when he made the above statement, but when I read it, I just couldn't help but remember all the times people have had queries about swapping SIMMS between macs and PCs. Seems you sometimes can't put a Mac's SIMMS in a PC because the SIMMS that Apple ships with their equipment are only 8 bit and don't support parity :-) Isn't that a bit like wearing only one boot when you go to play in the snow? At least ONE foot won't get wet! :-) -- Jeff Wiseman: ....uunet!tellab5!wiseman OR wiseman@TELLABS.COM