DragonFly users List (threaded) for 2009-07
Re: Sili/achi and Silicon Image 3512 chip
:I am considering to do another update to 2.3-DEV soon after Matt done
:with achi/sili tests and HAMMER V2 marked as stable.
:My 320Gb Seagate ST3300831AS 3.01 SATA drive is connected via:
:atapci0: <SiI 3512 SATA150 controller> port
:mem 0xfeaffc00-0xfeaffdff irq 6 at device 12.0 on pci2
:Are there any possibility to benefit from switching to achi or sili
:driver for it?
:Is it supported by any of these drivers?
I think only by the old NATA driver. I believe the 3512 emulates an
IDE controller for the most part. It isn't compatible with our 3132
:Is it ok to change access mode in BIOS from compatibility to ACHI?
:Will I need any re-partitioning of drive or something?
If you have an AHCI controller then including the ahci driver in the
kernel build or loading it as a module will cause it to probe the
controller before NATA. Disks recognized by the AHCI driver will
be named da0, da1, etc... and not ad0, ad1, etc.
So if the AHCI driver recognizes the disk you have to fix up /etc/fstab
and possibly /etc/rc.conf w/ the new names. You do not have to
repartition the drive or anything like that.
It's easy to get stuck here. If you do a full world & kernel build
and install the boot loader will have a menu option to disable the
AHCI driver. You would want to first boot with it enabled (the default)
to see if it recognizes the device. It would then pop into a mountroot>
menu since the normal root mount will not be found. You can then
reboot and use the menu option to disable the AHCI driver so you boot
with the old NATA driver, login, and adjust /etc/fstab and /etc/rc.conf
appropriately to reference the AHCI driver designation (da0, da1, etc)
instead of the NATA driver designation.