lkml.org 
[lkml]   [1998]   [Oct]   [12]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [ME TOO] Re: Linux-2.1.125 ... pre-2.2 imminent - SCSI issues
On Mon, 12 Oct 1998, Rogier Wolff wrote:

> Gerard Roudier wrote:
> >
> > My experience 4 years ago using a IBM S12:
> >
> > 1542CF: 2.2 MB/second (with some bus on / bus off hack).
>
> Hmm. That's odd. In the manual it says that it can do up to 4Mb per second.

Under the right circumstances a 154x could do even more than that AFAIK,
but...

Don't expect a PCI motherboard to do it! Or at least not any modern PCI
board (I've seen a few early VLB/PCI combos, where the PCI part was really
crappy, but the VLB/ISA part was OK).


Given one of the few MB where you could run it reliable at 12.5 MHz ISA
bus *and* the 8 or 10 MB/s DMA timing settings (not the usable DMA
bandwidth, this is before off/bus-on modifications!) it should be possible
to go beyond 5 MB/s with the right bus-on/bus-off settings.

(I'm fairly sure the 10 MB/s setting needed at *least* 12.5 MHz ISA bus,
quite possibly higher!, in fact I can't recall seeing anyone run it
reliably, but I have seen people claim that it worked)


> (*) For those that require an explanation:
> That means they don't even advertize that the thing can do better, no
> they even SAY that it can only do 4Mb per second (sync). That was
> really great stuff back then, but nowadays, you really need to be
> able to do 10Mb per second AT LEAST.

The only thing I can find about it on Adaptec's current Web pages is:
Host Bus Burst Data Rate: 10 MB/s
(none of the other data in the DataSheet or the TechSpec seems relevant).

Now, 'Host Bus' ought to be the ISA bus, and while I don't have access to
any 154x's any longer, they might actually be correct. Not that the 10
MB/s setting is actually usable with any MB I've seen (8/6.7 is usually
the highest that works), but it *IS* there AFAIK?

And even under the best circumstances the burst data rate is significantly
higher than the sustained transfer-rate, due to bus-on/bus-off timings.


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:44    [W:0.113 / U:0.520 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site