On Tue, Jul 02, 2002 at 12:05:15AM +0900, Sung-Won Chung wrote:
> Hi,
> 
> During netbsd souce using cvs. I met strange SCSI error in Ultra 1 170E.
> Fortunately, Ultra 1 seemed to work in spite of this SCSI error.
> I would appreciate if anyone let me know the meaning of this error message.
> 
> The SCSI error message is
> 
>   esp0: 
> error:csr=b2930a13<INT,ERR,DRAINING=0,IEN,ENDMA,DSBL_SCSI_DRN,BURST=0,TCI
>   esp0: DMA error; resetting
>   esp0: SCSI bus reset
> 
> Info. from dmesg about SCSI HDD is,
> 
>   sd0 at scsibus0 target 0 lun 0: <SEAGATE, ST318406LC, 010A> SCSI3 0/direct 
> fixed
>   sd0: 17501 MB, 26302 cyl, 2 head, 681 sec, 512 bytes/sect x 35843670 
> sectors
>   sd0: sync (100.0ns offset 15), 16-bit (20.000MB/s) transfers, tagged 
> queueing
> 
> Info. from dmesg about SCSI controller is,
> 
>   esp0 at sbus0 slot 14 offset 0x8800000 vector 20 ipl 3: dma rev fas
>   : FAS366/HME, 40MHz, SCSI ID 7
> 
> I used 20020526-1.6A snapshot.
> 

I see similar failure only when I have 2 drives and copy a large file from one
disk to another. Turning tagged queueing off heals the problem. I can also add
that in my case timing is involved so if I turn on debug output the problem
disappears.

	Andrey