[clug] USB 2.0 vs Firewire CPU overhead query.

[Brad Hards]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, 19 Dec 2003 08:34 am, Telek, John wrote:
> Is it true that there is more CPU overhead with usb than firewire ?
The question is not well formed. Do you mean the UHCI host controller in USB, 
the OHCI host controller in USB, EHCI host controller in USB, OHCI host 
controller in Firewire or some non-standard variant? With which functions? To 
what device?

> If so, why cant we have like a hardware enhaced usb that has it's own
> local memory and DMA support ???
Read the USB host controller specs. Read the Firewire host controller specs. 
Figure out where the CPU load is.

> The reason I ask is I have noticed that my USB/FW external drive seems a
> lot quicker when I use FW as opposed to USB 2.0.
This is probably a driver implementation quality issue. Kernel versions in use 
are probably the biggest issue. Also, are you sure that you are using USB 
high speed? Does /proc/bus/usb/devices show 480 for the speed?

Brad
- -- 
http://lca2004.linux.org.au - I'm registered. Are you?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.3 (GNU/Linux)

iD8DBQE/4mesGwwszQ/PZzgRAlVOAJ498sRRD43g0VS5//yk98fRsscWdACgjdKI
jQCSxFMNCy0dS/W506Bqzzg=
=kAS4
-----END PGP SIGNATURE-----