2002-01-08 01:31:19

by Andre Hedrick

[permalink] [raw]
Subject: U133/48-bit (IDE command queueing. (fwd))


Not doable w/o infrastructure of patch.

---------- Forwarded message ----------
Date: Tue, 18 Dec 2001 08:51:50 +1300
From: Stuart W <[email protected]>
To: [email protected]
Subject: IDE command queueing.


Greetings from New Zealand.

to cut to the chase, I just though I'd point out:
http://www.storagereview.com/jive/sr/thread.jsp?forum=5&thread=20936
most specifically, the mention of some IBM IDE drives doing command queing.

I'm sure you know about this (much more than I do), however I was wondering
if the current linux IDE implementation can take advantage of this?

I'm a VERY strong proponent of both IDE and Linux, which as I'm sure you know
is an up hill battle against many SCSI bigots.. Many of them can't see beyond
the $ signs that they think mean SCSI must be better.

I'm truly impressed with the latest 2.4 kernels, and have seen some major
performance improvements under IDE, but I am hoping for more.

I find it interesting that such sited as storage review only do windows
testing, even for 'server' products, completely ignoring the fact that the
host OS can have a major effect on driver performance, and that windows is
not nescessarily the best in these cases.

Anyway, I hope perhaps there is something usefull to you in the above link.

Regards,
Stuart.