tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: SCSI status 0x3



On Tue, Dec 31, 2013 at 12:00:31PM -0500, Thor Lancelot Simon wrote:
> It looked to me like the change you made to mpii might use more
> descriptors than the device can actually handle (typically more
> than fit in one page or one instance of some other relevant size from
> the device's point of view) to describe a transfer.  Since I have no
> datasheet I don't know what the actual device limitations are but you
> bumped the max from 32 to 512, right?  That might cause this.

Yes I did that change, because with 32, it fails very early with EFBIG.
With 512 it I get NMI and invalid SGL reply. With 4096, no NMI, but
still invalid SGL reply.

Of course it would be easier if I knew the right value to use. Linux
seems to use a configurable value (SCSI_MAX_SG_SEGMENTS), with max
as 256. But they might configure the thing differently.

-- 
Emmanuel Dreyfus
manu%netbsd.org@localhost


Home | Main Index | Thread Index | Old Index