kskelm@happy.colorado.edu (04/06/91)
In article <21741@shlump.nac.dec.com>, plouff@kali.enet.dec.com (Wes Plouff) writes: > > Have there been any published examples of Amiga peripherals using DMA? > Or, as an alternative, are schematics available in technical or repair > manuals for any of the popular DMA hard drive controllers? Enquiring > minds want to know the relative hardware complexity of taking over the > bus. So does THIS enquiring mind! It seems like I read somewhere in the docs that one should NOT take over the bus [for any length of time?]. If I built a card with a DMA controller, can I be certain that I am not goofing up, say, some OTHER device (like a HD) by taking the bus over for, say, (worst case) 128,000 cycles (a large RAM transfer)? My kingdom for some docs..... Kevin
ckp@grebyn.com (Checkpoint Technologies) (04/06/91)
In article <1991Apr5.110637.1@happy.colorado.edu> kskelm@happy.colorado.edu writes: > It seems like I read somewhere in the docs that one should NOT take >over the bus [for any length of time?]. If I built a card with a DMA >controller, can I be certain that I am not goofing up, say, some OTHER device >(like a HD) by taking the bus over for, say, (worst case) 128,000 cycles (a >large RAM transfer)? "Should not" is really in the light of being a good neighbor. There's nothing in the Amiga system that I've seen that gets hurt if you keep the bus permanently, except that no further work gets done. You can't take over the chip bus, that's always owned by the custom chips, your DMA device only steals some cycles. There are no refresh cycles for DRAM sent out on the bus, DRAM cards must generate their own, so they're OK. Most DMA hard drive adapters (significant exception: A2090A) must tolerate being unable to get on the bus for LONG periods, because of the chip RAM contention thing. And the motherboard has no logic for timing out a takeover and throwing you off (this might have changed for the A3000, I'm not sure). -- First comes the logo: C H E C K P O I N T T E C H N O L O G I E S / / ckp@grebyn.com \\ / / Then, the disclaimer: All expressed opinions are, indeed, opinions. \ / o Now for the witty part: I'm pink, therefore, I'm spam! \/
daveh@cbmvax.commodore.com (Dave Haynie) (04/16/91)
In article <1991Apr6.144654.10699@grebyn.com> ckp@grebyn.com (Checkpoint Technologies) writes: >In article <1991Apr5.110637.1@happy.colorado.edu> kskelm@happy.colorado.edu writes: >> It seems like I read somewhere in the docs that one should NOT take >>over the bus [for any length of time?]. >"Should not" is really in the light of being a good neighbor. There's >nothing in the Amiga system that I've seen that gets hurt if you keep >the bus permanently, except that no further work gets done. You can miss interrupts if you keep the bus too long. Ideally, anything that needs immediate interrupt response has an IOP to manage it, but that's not always true in practice. The on-board serial port and some of the Zorro II bus serial ports need pretty regular interrupt servicing or they fail. >Most DMA hard drive adapters (significant exception: A2090A) must >tolerate being unable to get on the bus for LONG periods, because of the >chip RAM contention thing. And the motherboard has no logic for timing >out a takeover and throwing you off (this might have changed for the A3000, >I'm not sure). There's nothing you can legally do to throw a Zorro II bus master off the bus, even if it's being a hog. So we recommend that Zorro II bus masters don't act like hogs. A Zorro III bus master gets time scheduled for it by the bus master, and can be gated on or off on a cycle by cycle basis. >First comes the logo: C H E C K P O I N T T E C H N O L O G I E S / / -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "That's me in the corner, that's me in the spotlight" -R.E.M.