"libata-dev-2.6" is a new patch queue for SATA features which are under
development, or need additional testing before being deployed into the
upstream-bound libata-2.[46] queues.
It currently contains:
1) A workaround for problems on multiple SATA controllers when using
PATA disks attached via SATA bridges.
2) New driver AHCI. This should currently be considered a technology
preview or early beta test. Simply because it is a young driver on
complicated hardware, using a new "FIS-based" delivery method (read:
brand new hotpath), caution is advised. Do not use in production yet.
I have updated my SATA status page to reflect these updates:
http://linux.yyz.us/sata/sata-status.html
BK users:
bk pull bk://gkernel.bkbits.net/libata-dev-2.6
Patch (NOTE: relative to 2.6.9-rc2-bk1-libata1.patch):
http://www.kernel.org/pub/linux/kernel/people/jgarzik/libata/2.6.9-rc2-bk1-libata1-dev1.patch.bz2
This will update the following files:
drivers/scsi/Kconfig | 8
drivers/scsi/Makefile | 1
drivers/scsi/ahci.c | 973 +++++++++++++++++++++++++++++++++++++++++++++
drivers/scsi/libata-core.c | 37 +
include/linux/ata.h | 1
include/linux/libata.h | 1
6 files changed, 1018 insertions(+), 3 deletions(-)
through these ChangeSets:
<[email protected]> (04/09/15 1.1905)
[libata] add AHCI driver
<[email protected]> (04/09/15 1.1902.1.1)
[PATCH] libata basic detection and errata for PATA->SATA bridges
This patch works around an issue with WD drives (and possibly others) over SiL PATA->SATA Bridges on SATA controllers locking up with transfers > 200 sectors.
Signed-off-by: Brad Campbell <[email protected]>
Jeff Garzik wrote:
> I have updated my SATA status page to reflect these updates:
> http://linux.yyz.us/sata/sata-status.html
Good stuff :-)
Is there any news on Marvell progress? I'll get my hands dirty on a
shipment with servers using the chips on friday, I may be able to sneak
in some testing before deployment (they will be rewired to use ICH5-R
for now)
Hmm, I wonder if there is any AHCI PCI(-X/E) plug-in boards planned.
That would be nice indeed. Seems like SATAII-like features on PCI-boards
are bound to Marvell or hardware RAID cards like 3ware 9xxx or newer LSI
MegaRAID currently.