Jeff Garzik wrote:
Jeff Garzik wrote:
The proposed sata_nv patch does the opposite -- guarantees we must
support the continually problematic legacy IDE interface ad infinitum.
Such patches are OK for the test lab, but in this specific case users
/suffer/ when not running AHCI mode.
Just to reinforce...
sata_nv support and bug fixes are primarily done right now through the
valiant efforts of Robert Hancock (with assists from Alan, Tejun, and
others).
Robert's job is difficult, because he has no hardware documentation[1],
and NVIDIA does not seem to be helping out much with driver bug reports
on the lists or in bugzillas.
Right, I don't have anything. Unless the original incomplete ADMA driver
release from NVIDIA counts as documentation, lol.
And yes, I've CC'ed NVIDIA people about a few ADMA-related issues and
been met with silence. It would be nice if they were as responsive about
ADMA issues as I must say Kuan and Peer have been on the SWNCQ side of
things..
As far as I know, I am the only one in the universe outside of NVIDIA
with any SATA docs at all, and those docs _only_ cover ADMA registers
and DMA structures, no PCI config info, no errata, nothing on SWNCQ or
legacy IDE (well, half a page).
NVIDIA has indeed become more engaged in sata_nv in recent times, and
that's a positive sign. You, Kuon and Ayaz have all been noticeably
more responsive in email. Thanks. Users have definitely benefited,
particularly from your help addressing a couple SWNCQ issues.
But at this point in time, being asked to choose between sata_nv and
ahci is no choice at all. One has public documentation, wide industry
support and little-or-no bugs. The other has several open issues, no
documentation, and support obstacles.
They're not even equivalent interfaces in this case, in the proposed
AHCI legacy mode patch these controllers are supported in the default
SFF mode only, no ADMA or SWNCQ, so you don't get any NCQ support..
--
Robert Hancock Saskatoon, SK, Canada
To email, remove "nospam" from hancockr@xxxxxxxxxxxxx
Home Page: http://www.roberthancock.com/
-
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
