Re: IDE-TNG what to do ?

John Jones (little.jones.family@ntlworld.com)
Tue, 20 Aug 2002 17:04:15 +0100


Hey Alan

I know that SATA can be attached to any IDE device because the
controller has all the "legacy crap"
I am not saying that you cant do it just that IDE-TNG should "bah
humbug" it and let the main IDE deal with it

I am quite happy to hear you say PCI IDE is looking better but really
it's going to be gone in the next 2 years as PCI just has not got the
bandwidth for the new drives
(yeah I know PCIX and vax's are still around in terms of legacy machine
people hang onto but I am sticking my head in the sand and
singing...SCSI or SAS )

really I am suggesting that you have IDE-TNG just for a few controllers
and drives
(real world testing is easier)

and the way to restrict the number is to say Serial ATA only

bad or (partly)good ?

regards

John Jones

On Tuesday, August 20, 2002, at 04:41 pm, Alan Cox wrote:

> On Tue, 2002-08-20 at 16:26, John Jones wrote:
>> IDE-TNG should ONLY deal with Serial ATA and ONLY chipset support not
>> PCI based implementations and should be a config option (keep it simple
>> as possible).
>
> SATA bridges attach to any IDE device and any IDE controller. SATA still
> uses PIO. SATA still has to deal with all the other legacy crap.
>
> The view that new IDE somehow gets rid of the old cruft is alas not born
> out by the reality.
>
> In terms of what works. As far as I can tell right now all the PCI stuff
> works in the current -ac cleanup code. There is a weird ide-scsi report
> and one person whose drives have run off somewhere and hidden. Other
> than that its working as well - and in some cases better.
>
> The 2.5 side is about getting the new request queue/bio logic right
> which is something I've not looked into
>
> Alan
>

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/