kern/143462: hw.ata.ata_dma_check_80pin no longer seems to work

Sean Winn sean at gothic.net.au
Tue Feb 2 01:00:11 UTC 2010


>Number:         143462
>Category:       kern
>Synopsis:       hw.ata.ata_dma_check_80pin no longer seems to work
>Confidential:   no
>Severity:       non-critical
>Priority:       medium
>Responsible:    freebsd-bugs
>State:          open
>Quarter:        
>Keywords:       
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Tue Feb 02 01:00:10 UTC 2010
>Closed-Date:
>Last-Modified:
>Originator:     Sean Winn
>Release:        8.0-STABLE
>Organization:
>Environment:
FreeBSD queen.gothic.net.au 8.0-STABLE FreeBSD 8.0-STABLE #23: Tue Feb  2 07:15:17 EST 2010
>Description:
PATA-SATA bridge directly in IDE socket, obviously wired incorrectly seeing it doesn't report an 80-pin 'cable', but there is no cable, and it works perfectly well up to UDMA-133 speeds (Condor MP20330-2 is the device).

FreeBSD 7, no problems, just enable hw.ata.ata_dma_check_80pin and most of the problems go away; still a slow PC with faster disks, but it all works.

Recently, I've had to try this with FreeBSD 8 and both with/without ATA_CAM, it no longer seems to apply.

>From -v probe...

(aprobe0:ata0:0:0:0): SIGNATURE: 0000
ata0: DMA limited to UDMA33, controller found non-ATA66 cable
(aprobe0:ata0:0:1:0): SIGNATURE: 0000
ata0: DMA limited to UDMA33, controller found non-ATA66 cable

loader.conf definitely has
hw.ata.ata_dma_check_80pin=0


So that check shouldn't even be happening. 

>How-To-Repeat:
Use ancient hardware and a dodgy bridge to make it useful

>Fix:


>Release-Note:
>Audit-Trail:
>Unformatted:


More information about the freebsd-bugs mailing list