[svlug] summary of nvidia/sil soft raid & Linux soft raid

Ivan Sergio Borgonovo mail at webthatworks.it
Tue Mar 14 04:03:22 PST 2006

On Mon, 13 Mar 2006 17:56:37 -0800
Rick Moen <rick at linuxmafia.com> wrote:


> All of this is covered, at least to some degree, on my page.


> I think Garzik must have gotten tired of being asked about exactly that
> sort of question, because eventually he snapped and wrote this rather
> hilarious FAQ:  http://linux-ata.org/faq-sata-raid.html  (You probably
> have to have a rather dry sense of humour, but *I* thought it was funny,
> anyway.)

I read both. I even read the code in the kernel... and at a first sight sata_nv should support queueing (kubuntu 2.6.15-17-amd64-k8).
I have to admit that I'm not sure about it anymore. But considering I finally decided to connect my hds to the nvidia ctl I think I did it considering the benchmark I saw and the fact that I checked sata_nv support queueing.

> > I can't understand clearly if sil or nvidia are in category 2 or 3.

> Category 2.  Also, "reconstructing sets" would be primarily handled by
> the host CPU, to the best of my understanding.  All the BIOS does is
> boot logic and establishing stripe sets for blank disks.  All work
> involved with actual disk writing is CPU-directed.

> > BTW... I think I already wrote it in previous emails... you were
> > asking about nvidia ctl...  Another point it has over sil... is that
> > the driver support queueing.
> Sure about that?  The latest libata status report Garzik has put on the
> Web (dtd 2006-01-26) says:

OK let me check the code once more ;)
.can_queue is set to 1 in all except one sata_* drivers.
can_queue is tested just in a way that could be related to sata just in ./drivers/scsi/scsi_lib.c and not to other specific scsi drivers.
Anyway... sata_nv.c has one reference to queue stuff. sata_sil.c none, and sata_sil24.c several.
That seems to agree to what is written in Garzik's page.

> Frankly, I'd definitely favour the newer SiI family over Nvidia's stuff,
> based on everything I've heard.

Me too... if I had a sil 3124. I have a geforce4 and a sil 3114.

> [1] He might mean an NDA with an expiration date aligned with the
> intended driver release date.  The XFree86/X.org people have been known
> to negotiate such things.

Well... I wonder why and how he wrote: "that permits implementation of NCQ support."
At least there should be someone that looked at the documents to say so.
Anyone is aware of any OS that have nvidia drivers not written by nvidia?
Or nvidia just said to Garzik: hey if you sign this NDA you'll be able to write drivers that support NCQ, trust us ;)

More information about the svlug mailing list