[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Good design recognition
- To: Mailing List Recipients <pci-sig-request@znyx.com>
- Subject: Good design recognition
- From: "Philip Ronzone" <Philip.Ronzone@eng.efi.com>
- Date: Fri, 1 May 1998 13:16:56 -0700
- Resent-Date: Fri, 1 May 1998 14:43:47 -0700
- Resent-From: pci-sig-request@znyx.com
- Resent-Message-ID: <"zXgIC2.0.re7.JsYIr"@electra.znyx.com>
- Resent-Sender: pci-sig-request@znyx.com
> Date: Thu, 30 Apr 1998 23:05:33 -0700
> From: Jochen Roth <jochen@znyx.com>
> Subject: Re: Big and little endian issue
>
> At 10:14 AM 4/28/98 -0700, Philip Ronzone wrote:
> >
> >Because people designing bridges do not do very good designs.
> >
>
> Aside from throwing out features you see no use for -- can you think
> of some you would add to make the first ever 'very good' bridge?
Well, I'm not an aerospace engineer either. But if I see the airplanes
keep falling from the sky, I *am* qualified to say that the design
has some obvious deficiencies.
Which, BTW, the "not good designs" is in the context of BE/LE issues.
But, rather than leave you with no example, a company that will remain
nameless built a mutli-function chip, that among other things had
a PCI-to-ISA bridge.
After much sweating and bug-hunting, I discovered that this bridge
did not allow I/O port accesses through to the ISA bus while any
PCI bus mastering operations from the ISA bus were going on!
Actually, it wasn't a case of not allowing - if it happened the whole
device went crazy and crashed.
The chip designers ACTUALLY asked why it was a problem! They said I
should just not do any ISA I/O port operations while any DMA/bus-mastering
was going on!!!
So while this chip was "good" in the sense it had a lot of functions
crammed into it and was relatively cheap (etc. ...), it was NOT a good
design from a software use-ability standpoint.
Reminds me of design status of the eventual Trey-1 jet engine, an engine
designed to meet the need for very high thrust decades ago:
"... So when we said that no matter what the cost, and whatever the
weight, it would set out to build the highest thrust jet engine ever made,
I am pleased to announce we've met the cost and weight goals ..."
-------------------------------------------------------------------------------
Philip K. Ronzone (RON-zone), extension 8300 ..................................
-------------------------------------------------------------------------------
DIR:(415) 286-8300 (work email) Electronics for Imaging, Inc.
OPER:(415) 286-8600 Philip.Ronzone@eng.efi.com 2855 Campus Drive
FAX:(415) 286-8663 San Mateo, CA