Date: Wed, 30 Jul 1997 00:24:36 +0200 (MEST) From: Søren Schmidt <sos@sos.freebsd.dk> To: se@FreeBSD.ORG (Stefan Esser) Cc: freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches Message-ID: <199707292224.AAA00624@sos.freebsd.dk> In-Reply-To: <19970729221036.52544@mi.uni-koeln.de> from Stefan Esser at "Jul 29, 97 10:10:36 pm"
next in thread | previous in thread | raw e-mail | index | archive | help
In reply to Stefan Esser who wrote:
> On Jul 29, Søren Schmidt <sos@sos.freebsd.dk> wrote:
> >
> > Sure, on my dev machine (P6@233Mhz-Natoma/64MB/2*Maxtor 84000A6)
>
> This seems to be an OEM version of the 83840a6, if I
> interpret the data sheets on their web page correctly.
> AFAIK the fastest EIDE drive series on the market today!
Well, I think they are mainline drives, you can buy them on every
corner here in DK land for about US$250.
> And the result is apparently *higher* CPU load, at least
> if you (did) trust the numbers reported by Bonnie ...
Well its most visible in the "per char" mesurements, I guess
time is spent processing the inividual chars through the I/O
system, the increase in used time is almost linear.
> But the "per char" read performance improvement indicates
> what's really going on: CPU cycles spent in the interrupt
> handler probably account for the 4.5MB/s mark in the PIO
> case, while raw data rate of the disk is approached with
> the DMA driver.
Exactly!
> But writes are still significantly slower than reads, which
> makes me think, that one revolution of the media is lost per
> 64KB written ...
I think it might be some kind of interaction with the drives
cache, but our driver might be guilty too..
> > Not bad, for a "simple" software upgrade :)
>
> True! But there is still room for improvement ... ;)
Well, I don't think we are going to get much more performance
out of it, at least not in an order like this, but there sure
is room for cleaning up the driver, which I'll eventually get
to if/when I get the time...
> > It has improved my worldstone by ~10% if that counts for real world use..
>
> Yes, one thing that is often underestimated is the fact,
> that while PIO mode transfers consume only a small fraction
> of the cycles of a fast CPU, but they tend to consume them
> exactly at the time, when the CPU is highly loaded anyway.
Yep. Now we only need ATA drives that can overlap commands. It seems they
are getting there, but its a mess still. When that settles (and we get
new drives :) ), we can get even a wee bit more performance...
> Thanks for sending those very interesting Bonnie results!
You're welcome :)
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Søren Schmidt (sos@FreeBSD.org) FreeBSD Core Team
Even more code to hack -- will it ever end
..
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199707292224.AAA00624>