Hello,
A quick update on this... With the help of Mark Lord from the hdparm team I found the following code in libata-core.c which should do what I need:
if (!tried_spinup && (id[2] == 0x37c8 || id[2] == 0x738c)) {
tried_spinup = 1;
/*
* Drive powered-up in standby mode, and requires a specific
* SET_FEATURES spin-up subcommand before it will accept
* anything other than the original IDENTIFY command.
*/
err_mask = ata_dev_set_feature(dev, SETFEATURES_SPINUP, 0);
if (err_mask && id[2] != 0x738c) {
rc = -EIO;
reason = "SPINUP failed";
goto err_out;
}
/*
* If the drive initially returned incomplete IDENTIFY info,
* we now must reissue the IDENTIFY command.
*/
if (id[2] == 0x37c8)
goto retry;
}
I've verified that this is in 2.6.32 but it isn't working on my system which uses Silicon Image port multipliers to connect 45 drives. On a hunch, I downgraded to Debian 4 with 2.6.26 and everything working as expected... This means that either something broke between 2.6.26 and 2.6.32 *or* the Debian team is compiling the kernel without support for this.
Does anyone have a guess as to what's going on here?
Thanks!
Tim
On Sep 1, 2011, at 2:34 PM, Tim Nufire wrote:
> Hello,
>
> I'm using 3TB Hitachi Deskstar 5K3000 drives in a custom 45 drive high density enclosure (http://blog.backblaze.com/2011/07/20/petabytes-on-a-budget-v2-0revealing-more-secrets) and need to limit the inrush current at startup by stagger the drive spin-up. Since I can't control drive power through by backplanes, the best way to do this appears to be by enabling the Power-Up In Standby feature on the drives via 'hdparm -s1'. Unfortunately this is not working for me....
>
> I believe the problem I'm having is that the Hitachi Deskstar 5K3000 drives require a SET FEATURES subcommand to spin-up to active state when the device has powered-up into Standby. I'm getting this from section 9.1 in the following spec:
>
> http://www.hitachigst.com/tech/techlib.nsf/techdocs/FFDF1FA949853DBD8825785A005CBC55/$file/DS5K3000_US5K3000_OEMSpecRev1.1.pdf
>
> Section 4.17 in the ATA spec that I found at http://t13.org/Documents/UploadedDocuments/docs2006/D1699r3f-ATA8-ACS.pdf says the following:
>
> A device may implement a SET FEATURES subcommand that notifies the device to spin-up to the Active state when the
> device has powered-up into Standby. If the device implements this SET FEATURES subcommand and power-up into Standby
> is enabled, the device shall remain in Standby until the SET FEATURES subcommand is received. If the device implements
> this SET FEATURES subcommand, the fact that the feature is implemented is reported in the IDENTIFY DEVICE or IDENTIFY
> PACKET DEVICE data.
> .
> .
> .
> If the device does not implement the SET FEATURES subcommand to spin-up the device after power-up and power-up into
> Standby is enabled, the device shall spin-up upon receipt of the first command that requires the device to access the
> media.
>
> Section 7.47.8 then says:
>
> Subcommand code 07h shall cause a device that has powered-up into Standby to go to the Active state (see 4.17 and figure 7).
>
> From this I'm concluding the following:
>
> 1) Not all drives require the spin up subcommand which explains why this works for some drives
>
> 2) Our Hitachi drives require this command and the libATA drivers don't appear to support it
>
> Is this right? If so, is there a reason why the libATA does not support this feature? I'm running Debian 5 with a 2.6.32 kernel... Has this been implemented upstream? Would this be a feature hard to add/backport?
>
> Thanks,
> Tim--
> To unsubscribe from this list: send the line "unsubscribe linux-ide" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
