Re: SCSI tape I/O errors with pre2.0.7 and sendmail zombies?

Info acct. (kernel@henge1.henge.com)
Sun, 26 May 1996 23:28:21 -0600 (MDT)


I am also getting error messages while trying to read/write with my
Exabyte 8500. I have CONFIG_SCSI_NCR53C7xx, and have tried it with
CONFIG_SCSI_NCR53C7xx_FAST on and off, but have not yet tried it with
CONFIG_SCSI_NCR53C7xx_DISCONNECT on. I have attached portions of the
messages I received to the bottom of this message.

On Sun, 26 May 1996, William M. Perkins wrote:

> With linux-pre2.0.7 I am getting a SCSI subsystem lockup while
> trying to read a SCSI QIC-250 tape with tar. The tape is read for
> awhile and than tar freezes with a reported I/O error. I can
> still switch windows and the mouse pointer still moves, but no
> new programs can be started without an I/O error message being
> displayed. I had to hard reset the system to reboot. I reran
> the tar file restore with linux-pre2.0.6, and experienced no
> problems. I am using a ncr53c810 controller with a Conner 2150
> tape drive.
>
> Bill
>
> PS: I also have lockups when trying to edit/remove file on a minix
> floppy with pre2.0.6.
>
> --
> William M. Perkins Internet - wperkins@us.net
> The Greenwood or - bill@cais.com
> Commodore is dead. Long lives the Amiga! (AmigaOS/Linux/NetBSD)
>
>

here are the errors:

d EOM Current error st09:00: sense key Unit Attention
Additional sense indicates Power on, reset, or bus device reset occurred
scsi : aborting command due to timeout : pid 256784, scsi0, channel 0, id 0, lun 0 Read (6) 01 d5 d0 0e 00
scsi0 : DANGER : command running, can not abort.
scsi : aborting command due to timeout : pid 256785, scsi0, channel 0, id 0, lun 0 Read (6) 01 d5 d6 02 00
scsi0 : found command 256785 in Linux issue queue
scsi : aborting command due to timeout : pid 256787, scsi0, channel 0, id 0, lun 0 Write (6) 11 40 5a 02 00
scsi0 : found command 256787 in Linux issue queue
scsi : aborting command due to timeout : pid 256784, scsi0, channel 0, id 0, lun 0 Read (6) 01 d5 d0 0e 00
scsi0 : DANGER : command running, can not abort.
SCSI host 0 abort (pid 256784) timed out - resetting
SCSI bus is being reset for host 0 channel 0.
scsi0 : DCMD|DBC=0x820b0000, DNAD=0x8485c (virt 0x0008485c)
DSA=0x5dc1e8 (virt 0x005dc1e8)
DSPS=0x8484c, TEMP=0x5dc0ac (virt 0x005dc0ac), DMODE=0x88
SXFER=0x0, SCNTL3=0x3
BSY phase=CMDOUT, 0 bytes in SCSI FIFO
STEST0=0x7
scsi0 : DSP 0x84854 (virt 0x00084854) ->
0x84854 (virt 0x00084854) : 0x820b0000 0x0008484c (virt 0x0008484c)
0x8485c (virt 0x0008485c) : 0x8f0b0000 0x00084a04 (virt 0x00084a04)
0x84864 (virt 0x00084864) : 0x9e0b0000 0x00000000 (virt 0x00000000)
0x8486c (virt 0x0008486c) : 0x800b0000 0x0008488c (virt 0x0008488c)
0x84874 (virt 0x00084874) : 0x810b0000 0x000848e4 (virt 0x000848e4)
0x8487c (virt 0x0008487c) : 0x830b0000 0x00084c0c (virt 0x00084c0c)
scsi0 : connected (SDID=0x2, SSID=0x0)
scsi0 : dsa at phys 0x5dc1e8 (virt 0x005dc1e8)
+ 64 : dsa_msgout length = 1, data = 0x5dc02c (virt 0x005dc02c)
Identify disconnect not allowed lun 0
+ 60 : select_indirect = 0x3020000
+ 56 : dsa_cmnd = 0x82410 result = 0xffff, target = 2, lun = 0, cmd = Write Filemarks 00 00 00 01 00
+ 48 : dsa_next = 0x0
scsi0 target 2 : sxfer_sanity = 0x0, scntl3_sanity = 0x3
script : 0x78030300 0x0 0x78050000 0x0 0x90080000 0x0 0x0 0x0
scsi0 : saved data pointer at offset 0
scsi0 : active data pointer at offset 0
scsi0 : issue queue
scsi0 : dsa at phys 0x968080 (virt 0x00968080)
+ 64 : dsa_msgout length = 542180, data = 0x0 (virt 0x00000000)
+ 60 : select_indirect = 0xc0000004
+ 56 : dsa_cmnd = 0x0
+ 48 : dsa_next = 0x0
scsi0 : dsa at phys 0x14ec080 (virt 0x014ec080)
+ 64 : dsa_msgout length = 542180, data = 0x0 (virt 0x00000000)
+ 60 : select_indirect = 0xc0000004
+ 56 : dsa_cmnd = 0x0
+ 48 : dsa_next = 0x0
scsi0 : schedule dsa array :
scsi0 : dsa at phys 0x14b8080 (virt 0x014b8080)
+ 64 : dsa_msgout length = 542180, data = 0x86220 (virt 0x00086220)
+ 60 : select_indirect = 0xc0000004
+ 56 : dsa_cmnd = 0x0
+ 48 : dsa_next = 0x0
scsi0 : end schedule dsa array
scsi0 : reconnect_dsa_head :
scsi0 : end reconnect_dsa_head
The sti() implicit in a printk() prevents hangs
scsi0 : target 0 accepting asynchronous SCSI
scsi0 : setting target 0 to asynchronous SCSI
scsi : aborting command due to timeout : pid 256785, scsi0, channel 0, id 0, lun 0 Read (6) 01 d5 d6 02 00
scsi0 : did this command ever run?
scsi : aborting command due to timeout : pid 256787, scsi0, channel 0, id 0, lun 0 Write (6) 11 40 5a 02 00
scsi0 : did this command ever run?
scsi : aborting command due to timeout : pid 256786, scsi0, channel 0, id 2, lun 0 Write Filemarks 00 00 00 01 00
scsi0 : did this command ever run?
scsi0 : target 2 accepting asynchronous SCSI
scsi0 : setting target 2 to asynchronous SCSI
st0: Error with sense data: extra data not valid EOM Current error st09:00: sense key Unit Attention
Additional sense indicates Power on, reset, or bus device reset occurred
st0: Error on write filemark.