HP state does not support PM.
Installed HP BIOS is J06 06/06/2014.
Used a USB key with a small build of FreeBSD.
Can try in AHCI mode on Gen8.
Can try in non-AHCI i.e. RAID mode Gen8.
Can test to see if device reports PM supported.
DL'd FreeBSD v10.1 (iso image) for use with USB Stick
So...
Changed BIOS settings to RAID. Was at AHCI.
This loads RAID v3.54.0 at BIOS load.
This loads HP AHCI SATA Controller v0.84 at BIOS load.
Ran FreeBSD v10.1 Live CD.
dmesg | grep -i ah
> ahci0: AHCI v1.30 with 6 6Gbps ports, Port Multiplier supported
reports the chip supports Port Multiplier
Changed BIOS settings to AHCI. Was at RAID.
This loads HP AHCI SATA Controller v0.90 at BIOS load.
This loads HP AHCI SATA Controller v0.84 at BIOS load.
Ran FreeBSD v10.1 Live CD.
dmesg | grep -i ah
> ahci0: AHCI v1.30 with 6 6Gbps ports, Port Multiplier supported
reports the chip supports Port Multiplier
Conclusions
Is bit 17 optional in the CAP register so as to, in effect make use of
PMs optional by the vendor, should they seek to do this?
For example, see reported issues with Supermicro microcloud storage and FreeBSD.
For the gen8, The issue of AHCI vs. RAID seems thus superfluous as they
both report PM can be used but this must be, it seems clear to me, for ports 4-5 only per datasheet.
The datasheet refers to not used by AHCI as opposed to, AHCI only.
So... seems reasonable to assert, AHCI ignores some register settings in the PCH.
If anyone is running, e.g. freeBSD/freeNAS as installed OS they will be able to determine the
caps, caps2 and EM caps settings for the controller in AHCI and/or RAID mode.
If caps, caps2 values reports FBS, eSATA et al, can conclude ports 4-5 may indeed be
used for FIS based swithing re - PM etc.
Can someone confirm please?
I only briefly tested with LiveCD. This is not conclusive.
Installed HP BIOS is J06 06/06/2014.
Used a USB key with a small build of FreeBSD.
Can try in AHCI mode on Gen8.
Can try in non-AHCI i.e. RAID mode Gen8.
Can test to see if device reports PM supported.
DL'd FreeBSD v10.1 (iso image) for use with USB Stick
So...
Changed BIOS settings to RAID. Was at AHCI.
This loads RAID v3.54.0 at BIOS load.
This loads HP AHCI SATA Controller v0.84 at BIOS load.
Ran FreeBSD v10.1 Live CD.
dmesg | grep -i ah
> ahci0: AHCI v1.30 with 6 6Gbps ports, Port Multiplier supported
reports the chip supports Port Multiplier
Changed BIOS settings to AHCI. Was at RAID.
This loads HP AHCI SATA Controller v0.90 at BIOS load.
This loads HP AHCI SATA Controller v0.84 at BIOS load.
Ran FreeBSD v10.1 Live CD.
dmesg | grep -i ah
> ahci0: AHCI v1.30 with 6 6Gbps ports, Port Multiplier supported
reports the chip supports Port Multiplier
Conclusions
Is bit 17 optional in the CAP register so as to, in effect make use of
PMs optional by the vendor, should they seek to do this?
For example, see reported issues with Supermicro microcloud storage and FreeBSD.
For the gen8, The issue of AHCI vs. RAID seems thus superfluous as they
both report PM can be used but this must be, it seems clear to me, for ports 4-5 only per datasheet.
The datasheet refers to not used by AHCI as opposed to, AHCI only.
So... seems reasonable to assert, AHCI ignores some register settings in the PCH.
If anyone is running, e.g. freeBSD/freeNAS as installed OS they will be able to determine the
caps, caps2 and EM caps settings for the controller in AHCI and/or RAID mode.
If caps, caps2 values reports FBS, eSATA et al, can conclude ports 4-5 may indeed be
used for FIS based swithing re - PM etc.
Can someone confirm please?
I only briefly tested with LiveCD. This is not conclusive.