List of fake vs. real SATA drives.
sos at DeepCore.dk
Tue Nov 23 10:44:09 GMT 2004
David Gilbert wrote:
>>>>>>"Søren" == Søren Schmidt <sos at DeepCore.dk> writes:
> Søren> David Gilbert wrote:
>>>>>>>>"João" == João Carlos Mendes Luís <jonny at jonny.eng.br> writes:
>>>João> IIF they really have command queueing, I do believe. So, a
>>>João> bridged SATA drive will not have command queuing, right?
>>>Well... from what I've read, the WD bridged drives do have queueing
>>>because they had an ATA-100 implementation of it.
>>>João> Does FreeBSD already take advantage of this? How could I
>>>check João> if my SATA drivers have command queueing or not?
>>>I'd like to know as well.
> Søren> Currently the ATA driver does not support neither TCQ nor NCQ.
> Søren> For this to work at all both disk and controller needs to
> Søren> support the mode in question. I've just gotten my hands on one
> Søren> of the new Promise controllers that supports NCQ, but I still
> Søren> need disks to get it going. I do have a few Raptor's that
> Søren> support the TCQ mode but initial testing shows little benefit
> Søren> from it so it moved to the backburner...
> Do you need direct access to a drive or is access to a machine with a
> drive/controller in it sufficient. Are promise controllers the only
> ones to support NCQ so far? Promise doesn't always have the best
> reputation. We may be able to provide a test machine with remote
> access to a drive/controller combination.
Usually I'd need "hands on" HW for things like this, as there might be
alot of resetting/powertoggle in there until things gets twisted the
right way, its just so much easier to deal with. Lets see how things
turns out, if I cannot get my hands on real disks here in the lab, I'll
take you up on the offer..
More information about the freebsd-hackers