Bug 29318

Summary: [scsi] [patch] Exabyte 8200 needs SA_QUIRK_1FM and SA_QUIRK_NODREAD
Product: Base System Reporter: James Stephen Larson <jim>
Component: kernAssignee: Matt Jacob <mjacob>
Status: Closed FIXED    
Severity: Affects Only Me CC: jim
Priority: Normal    
Version: 4.3-RELEASE   
Hardware: Any   
OS: Any   
Attachments:
Description Flags
file.diff none

Description James Stephen Larson 2001-07-30 05:50:01 UTC 
When I upgraded from FreeBSD 2.2.8 to FreeBSD 4.3, my Exabyte 8200
didn't work until I added the quirks below.
Comment 1 Matt Jacob freebsd_committer freebsd_triage 2001-10-02 06:42:35 UTC 
Responsible Changed
From-To: freebsd-bugs->mjacob

maintainer
Comment 2 Matt Jacob freebsd_committer freebsd_triage 2001-10-02 06:43:17 UTC 
State Changed
From-To: open->feedback

the NODREAD I could understand. why was the 1FM necessary?
Comment 3 Nate Lawson 2002-10-02 22:45:51 UTC 
Last message on this was a year ago.  Matt was wondering why 1FM was
needed and not just NODREAD.  If the tape drive is not working, please let
us know.  The requested quirk has NOT been added.

I'll close this in 2 weeks if no further input.

-Nate
Comment 4 James Stephen Larson 2002-10-07 01:20:21 UTC 
In message <Pine.BSF.4.21.0210021442470.89977-100000@root.org> you write:
>Last message on this was a year ago.  Matt was wondering why 1FM was
>needed and not just NODREAD.  If the tape drive is not working, please let
>us know.  The requested quirk has NOT been added.
>
>I'll close this in 2 weeks if no further input.

Thanks for bringing this up again!

I need to move the tape drive to another machine (got laid off last
week and have to return the company's box), but I'll get that done
within the two week deadline and let you know how it works with
a new box and FreeBSD 4.6.2.

On the plus side, the layoff means that I'll have more time to devote
to this!

Jim
Comment 5 Nate Lawson 2002-10-17 18:23:05 UTC 
Have you gotten the tape drive again?

-Nate
Comment 6 James Stephen Larson 2002-10-17 18:43:34 UTC 
In message <Pine.BSF.4.21.0210171022180.42958-100000@root.org> you write:
>Have you gotten the tape drive again?

Working on it today.  I'll send you something this afternoon.

Jim
Comment 7 James Stephen Larson 2002-10-23 20:17:21 UTC 
In message <200210171743.g9HHhY109197@lefse.jetcafe.org> you write:
>In message <Pine.BSF.4.21.0210171022180.42958-100000@root.org> you write:
>>Have you gotten the tape drive again?
>
>Working on it today.  I'll send you something this afternoon.

Sorry, things came up.  I'm working on this today.  I'll re-run
all previously-requested experiments and permutations and send you
the results.

Jim
Comment 8 James Stephen Larson 2002-10-24 22:52:56 UTC 
In message <Pine.BSF.4.21.0210021442470.89977-100000@root.org> you write:
>Last message on this was a year ago.  Matt was wondering why 1FM was
>needed and not just NODREAD.  If the tape drive is not working, please let
>us know.  The requested quirk has NOT been added.

I upgraded to FreeBSD 4.6.2 and re-ran the experiments.
It *seems* like we only need the NODREAD quirk for this
device, but I'll let the experts make that call.

Here's the recap:

Exabyte 8200 SCSI quirks roundup.

When I upgraded from FreeBSD 2.2.8 to 4.3 I had to add the following
to /usr/src/sys/cam/scsi/scsi_sa.c:

+       {       /* jim@jetcafe.org */
+               { T_SEQUENTIAL, SIP_MEDIA_REMOVABLE, "EXABYTE",
+                 "EXB-8200", "425A"}, SA_QUIRK_1FM|SA_QUIRK_NODREAD, 0
+       },

I've now upgraded to FreeBSD 4.6.2.

The SCSI-relevant section of my kernel config file is:

># SCSI Controllers
>device                ahc             # AHA2940 and onboard AIC7xxx devices
>
># SCSI peripherals
>device          scbus           # SCSI bus (required)
>device          da              # Direct Access (disks)
>device          sa              # Sequential Access (tape etc)
>device          cd              # CD
>device          pass            # Passthrough device (direct SCSI access)
>
>options         CAMDEBUG
>options         CAM_DEBUG_BUS=-1
>options         CAM_DEBUG_TARGET=-1
>options         CAM_DEBUG_LUN=-1
>options         CAM_DEBUG_FLAGS="CAM_DEBUG_INFO|CAM_DEBUG_TRACE|CAM_DEBUG_CDB"

Inserted tape #1 (containing tar output):

lutefisk# camcontrol devlist
<EXABYTE EXB-8200 425A>            at scbus0 target 2 lun 0 (pass0,sa0)
lutefisk# camcontrol debug -Ic 0:2:0
Debugging enabled for 0:2:0
lutefisk# mt -f /dev/nsa0 status
Mode      Density              Blocksize      bpi      Compression
Current:  default              variable       0        unsupported
---------available modes---------
0:        default              variable       0        unsupported
1:        default              variable       0        unsupported
2:        default              variable       0        unsupported
3:        default              variable       0        unsupported
---------------------------------
Current Driver State: at rest.
---------------------------------
File Number: 0  Record Number: 0        Residual Count 12

/var/log/messages says:
Oct 23 18:00:00 lutefisk newsyslog[165]: logfile turned over
Oct 24 13:59:56 lutefisk /kernel: (xpt0:xpt0:0:-1:-1): xpt_action
Oct 24 14:00:34 lutefisk /kernel: (xpt0:ahc0:0:2:0): xpt_compile_path
Oct 24 14:00:34 lutefisk /kernel: (xpt0:ahc0:0:2:0): xpt_setup_ccb
Oct 24 14:00:34 lutefisk /kernel: (xpt0:ahc0:0:2:0): xpt_action
Oct 24 14:00:34 lutefisk /kernel: (xpt0:ahc0:0:2:0): debugging flags now 9
Oct 24 14:01:00 lutefisk /kernel: (sa0:ahc0:0:2:0): saopen(0): dev=0x0 softc=0x0
Oct 24 14:01:00 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): RES
ERVE(06). CDB: 16 0 0 0 0 0 
Oct 24 14:01:01 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 0 dxfer_len 0
Oct 24 14:01:01 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): TES
T UNIT READY. CDB: 0 0 0 0 0 0 
Oct 24 14:01:01 lutefisk /kernel: (sa0:ahc0:0:2:0): LOAD UNLOAD. CDB: 1b 0 0 0 1
 0 
Oct 24 14:01:29 lutefisk /kernel: (sa0:ahc0:0:2:0): READ(06). CDB: 8 0 0 20 0 0 
Oct 24 14:01:29 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 8192 dxfer_len 8192
Oct 24 14:01:29 lutefisk /kernel: (sa0:ahc0:0:2:0): REWIND. CDB: 1 0 0 0 0 0 
Oct 24 14:01:29 lutefisk /kernel: (sa0:ahc0:0:2:0): READ BLOCK LIMITS. CDB: 5 0 
0 0 0 0 
Oct 24 14:01:30 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 f 0 1c 0 
Oct 24 14:01:30 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 28 dxfer_len 28
Oct 24 14:01:30 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SENSE(06). CDB: 1a 0 10
 0 1c 0 
Oct 24 14:01:30 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 28 dxfer_len 28
Oct 24 14:01:30 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SENSE(06). CDB: 1a 0 0 
0 c 0 
Oct 24 14:01:30 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:31 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:31 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:32 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:32 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:32 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x06 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:32 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:32 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:32 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:33 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:33 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:33 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0xc3 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:33 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:34 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:34 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:34 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:34 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:34 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x03 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:35 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:35 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:35 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:36 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:36 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:36 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x02 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:36 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:37 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:37 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:37 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:38 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:38 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x01 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:38 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:38 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:38 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:39 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:39 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x26 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x22 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:40 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x11 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x10 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x0f 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x05 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:41 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x84 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x04 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x12 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 12 dxfer_len 12
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): Err22, Mode Select Data= 0x0
0 0x00 0x10 0x08 0x29 0x00 0x00 0x00 0x00 0x00 0x04 0x00
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): MODE SELECT(06). CDB: 15 0 0
 0 c 0 
Oct 24 14:01:42 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:43 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x04 0x00
Oct 24 14:01:43 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SELECT(06). CDB: 15 0 0 0 c 0 
Oct 24 14:01:43 lutefisk /kernel: (sa0:ahc0:0:2:0): Err0, Mode Select Data= 0x00
 0x00 0x10 0x08 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
Oct 24 14:01:43 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): PRE
VENT ALLOW MEDIUM REMOVAL. CDB: 1e 0 0 0 1 0 
Oct 24 14:01:43 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): MOD
E SENSE(06). CDB: 1a 0 0 0 c 0 
Oct 24 14:01:43 lutefisk /kernel: (sa0:ahc0:0:2:0): Mode Sense Data= 0x10 0x85 0
x10 0x08 0x00 0x23 0x01 0x20 0x00 0x00 0x00 0x00
Oct 24 14:01:43 lutefisk /kernel: (sa0:ahc0:0:2:0): saclose(0): dev=0x0 softc=0x
100d
Oct 24 14:01:43 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): PRE
VENT ALLOW MEDIUM REMOVAL. CDB: 1e 0 0 0 0 0 

No errors yet.  Change tapes to tape #2, containing three filesystem
dump(8)s, each as a separate tape "file", using the mt terminology.

lutefisk# mt -f /dev/nsa0 status
mt: /dev/nsa0: Input/output error

/var/log/messages says:
Oct 24 14:18:25 lutefisk /kernel: (sa0:ahc0:0:2:0): saopen(0): dev=0x0 softc=0x1
000
Oct 24 14:18:25 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): RES
ERVE(06). CDB: 16 0 0 0 0 0 
Oct 24 14:18:26 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x6 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 0 dxfer_len 0
Oct 24 14:18:26 lutefisk /kernel: (sa0:ahc0:0:2:0): RESERVE(06). CDB: 16 0 0 0 0
 0 
Oct 24 14:18:26 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 0 dxfer_len 0
Oct 24 14:18:26 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): TES
T UNIT READY. CDB: 0 0 0 0 0 0 
Oct 24 14:18:26 lutefisk /kernel: (sa0:ahc0:0:2:0): LOAD UNLOAD. CDB: 1b 0 0 0 1
 0 
Oct 24 14:18:54 lutefisk /kernel: (sa0:ahc0:0:2:0): READ(06). CDB: 8 0 0 20 0 0 
Oct 24 14:19:00 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x0 ASC/ASCQ 0x0 0x0 fla
gs 0x20 resid -2048 dxfer_len 8192
Oct 24 14:19:00 lutefisk /kernel: (sa0:ahc0:0:2:0): REWIND. CDB: 1 0 0 0 0 0 
Oct 24 14:19:00 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x0 ASC/ASCQ 0x0 0x0 fla
gs 0x20 resid -2048 dxfer_len 0
Oct 24 14:19:00 lutefisk /kernel: (sa0:ahc0:0:2:0): REWIND. CDB: 1 0 0 0 0 0 
Oct 24 14:19:01 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x0 ASC/ASCQ 0x0 0x0 fla
gs 0x20 resid -2048 dxfer_len 0
Oct 24 14:19:01 lutefisk /kernel: (sa0:ahc0:0:2:0): unable to rewind after test 
read
Oct 24 14:19:01 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): REL
EASE(06). CDB: 17 0 0 0 0 0 
Oct 24 14:19:01 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x0 resid 0 dxfer_len 0

Issue "mt off" to switch back to original tape.

lutefisk# mt off
mt: /dev/nsa0: Input/output error

/var/log/messages says:
Oct 24 14:26:36 lutefisk /kernel: (sa0:ahc0:0:2:0): saopen(0): dev=0x0 softc=0x0
Oct 24 14:26:37 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): RES
ERVE(06). CDB: 16 0 0 0 0 0 
Oct 24 14:26:37 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x0 resid 0 dxfer_len 0
Oct 24 14:26:37 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): TES
T UNIT READY. CDB: 0 0 0 0 0 0 
Oct 24 14:26:37 lutefisk /kernel: (sa0:ahc0:0:2:0): LOAD UNLOAD. CDB: 1b 0 0 0 1
 0 
Oct 24 14:27:04 lutefisk /kernel: (sa0:ahc0:0:2:0): READ(06). CDB: 8 0 0 20 0 0 
Oct 24 14:27:10 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x0 ASC/ASCQ 0x0 0x0 fla
gs 0x20 resid -2048 dxfer_len 8192
Oct 24 14:27:10 lutefisk /kernel: (sa0:ahc0:0:2:0): REWIND. CDB: 1 0 0 0 0 0 
Oct 24 14:27:10 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x0 ASC/ASCQ 0x0 0x0 fla
gs 0x20 resid -2048 dxfer_len 0
Oct 24 14:27:10 lutefisk /kernel: (sa0:ahc0:0:2:0): REWIND. CDB: 1 0 0 0 0 0 
Oct 24 14:27:11 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x0 ASC/ASCQ 0x0 0x0 fla
gs 0x20 resid -2048 dxfer_len 0
Oct 24 14:27:11 lutefisk /kernel: (sa0:ahc0:0:2:0): unable to rewind after test 
read
Oct 24 14:27:11 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): REL
EASE(06). CDB: 17 0 0 0 0 0 
Oct 24 14:27:12 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x0 resid 0 dxfer_len 0

Since "mt off" doesn't want to work, hit manual eject button.
Switch back to tape #1.

lutefisk# mt -f /dev/nsa0 status
mt: /dev/nsa0: Input/output error

/var/log/messages says:
Oct 24 14:30:59 lutefisk /kernel: (sa0:ahc0:0:2:0): saopen(0): dev=0x0 softc=0x0
Oct 24 14:31:00 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): RES
ERVE(06). CDB: 16 0 0 0 0 0 
Oct 24 14:31:00 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x6 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 0 dxfer_len 0
Oct 24 14:31:00 lutefisk /kernel: (sa0:ahc0:0:2:0): RESERVE(06). CDB: 16 0 0 0 0
 0 
Oct 24 14:31:00 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x40 resid 0 dxfer_len 0
Oct 24 14:31:00 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): TES
T UNIT READY. CDB: 0 0 0 0 0 0 
Oct 24 14:31:00 lutefisk /kernel: (sa0:ahc0:0:2:0): LOAD UNLOAD. CDB: 1b 0 0 0 1
 0 
Oct 24 14:31:23 lutefisk /kernel: (sa0:ahc0:0:2:0): READ(06). CDB: 8 0 0 20 0 0 
Oct 24 14:31:28 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x0 ASC/ASCQ 0x0 0x0 fla
gs 0x20 resid -2048 dxfer_len 8192
Oct 24 14:31:29 lutefisk /kernel: (sa0:ahc0:0:2:0): REWIND. CDB: 1 0 0 0 0 0 
Oct 24 14:31:29 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x0 ASC/ASCQ 0x0 0x0 fla
gs 0x20 resid -2048 dxfer_len 0
Oct 24 14:31:29 lutefisk /kernel: (sa0:ahc0:0:2:0): REWIND. CDB: 1 0 0 0 0 0 
Oct 24 14:31:30 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x0 ASC/ASCQ 0x0 0x0 fla
gs 0x20 resid -2048 dxfer_len 0
Oct 24 14:31:30 lutefisk /kernel: (sa0:ahc0:0:2:0): unable to rewind after test 
:read
Oct 24 14:31:30 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): REL
EASE(06). CDB: 17 0 0 0 0 0 
Oct 24 14:31:30 lutefisk /kernel: (sa0:ahc0:0:2:0): Key 0x5 ASC/ASCQ 0x0 0x0 fla
gs 0x0 resid 0 dxfer_len 0

I'm not seeing the error message that earlier led me to add the
"1FM" quirk.  Let me know if you need me to perform any further
experiments.

Jim
Comment 9 mjacob 2002-10-24 23:15:11 UTC 
Would it have been possible upgrade to 4.7 which contains changes to the
SA driver that make it more in sync with freebsd-current?
Comment 10 James Stephen Larson 2002-10-24 23:53:41 UTC 
In message <Pine.BSF.4.21.0210241514420.1676-100000@beppo> you write:
>Would it have been possible upgrade to 4.7 which contains changes to the
>SA driver that make it more in sync with freebsd-current?

I guess I can do that.  It would be sweet if I could upgrade just
the SA driver - do you know if the 4.6.2 -> 4.7 changes to that file
can stand on their own? Otherwise it might be another day or two.

Jim
Comment 11 mjacob 2002-10-24 23:58:04 UTC 
> In message <Pine.BSF.4.21.0210241514420.1676-100000@beppo> you write:
> >Would it have been possible upgrade to 4.7 which contains changes to the
> >SA driver that make it more in sync with freebsd-current?
> 
> I guess I can do that.  It would be sweet if I could upgrade just
> the SA driver - do you know if the 4.6.2 -> 4.7 changes to that file
> can stand on their own? Otherwise it might be another day or two.

Umm- should. See if it compiles at least.
Comment 12 James Stephen Larson 2002-10-25 22:14:44 UTC 
In message <Pine.BSF.4.21.0210241557470.1676-100000@beppo> you write:
>[Jim writes:]
>> In message <Pine.BSF.4.21.0210241514420.1676-100000@beppo> you write:
>> >Would it have been possible upgrade to 4.7 which contains changes to the
>> >SA driver that make it more in sync with freebsd-current?
>> 
>> I guess I can do that.  It would be sweet if I could upgrade just
>> the SA driver - do you know if the 4.6.2 -> 4.7 changes to that file
>> can stand on their own? Otherwise it might be another day or two.
>
>Umm- should. See if it compiles at least.

Okay, I grabbed the 4.7-RELEASE version of scsi_sa.c, rebuilt
the kernel (compiled just fine) and reran some experiments.

lutefisk# camcontrol devlist
<EXABYTE EXB-8200 425A>            at scbus0 target 2 lun 0 (pass0,sa0)
lutefisk# camcontrol debug -Ic 0:2:0
Debugging enabled for 0:2:0

<pop in tape with 3 dump images as separate "files">

lutefisk# mt -f /dev/nsa0 status
Mode      Density              Blocksize      bpi      Compression
Current:  default              variable       0        unsupported
---------available modes---------
0:        default              variable       0        unsupported
1:        default              variable       0        unsupported
2:        default              variable       0        unsupported
3:        default              variable       0        unsupported
---------------------------------
Current Driver State: at rest.
---------------------------------
File Number: 0  Record Number: 0        Residual Count 0
lutefisk# mt fsf
lutefisk# restore -iv

all works, generates volumes of output.

lutefisk# mt off

This looks cool.  Turn off debugging with "camcontrol debug off" and
switch tapes.

lutefisk# tar t
tar: can't open /dev/sa0 : Input/output error

/var/log/messages says
Oct 25 12:35:12 lutefisk /kernel: (sa0:ahc0:0:2:0): unable to rewind after test 
read

Okay, turn debugging back on.
lutefisk# camcontrol debug -Ic 0:2:0
lutefisk# tar t
...generates output...
Now it works!

Turn debugging off...
lutefisk# camcontrol debug off
Debugging turned off
lutefisk# tar t
...generates output...
Hmm.

lutefisk# mt off

load another tape

lutefisk# mt -f /dev/nsa0 status
mt: /dev/nsa0: Input/output error

/var/log/messages says:
Oct 25 14:01:14 lutefisk /kernel: (sa0:ahc0:0:2:0): unable to rewind after test read

lutefisk# camcontrol debug -Ic 0:2:0
Debugging enabled for 0:2:0
lutefisk# mt -f /dev/nsa0 status
mt: /dev/nsa0: Input/output error

/var/log/messages says:
Oct 25 14:02:55 lutefisk /kernel: (xpt0:ahc0:0:2:0): debugging flags now 9
Oct 25 14:03:06 lutefisk /kernel: (sa0:ahc0:0:2:0): saopen(0): dev=0x0 softc=0x0
Oct 25 14:03:06 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): RES
ERVE(06). CDB: 16 0 0 0 0 0 
Oct 25 14:03:06 lutefisk /kernel: (sa0:ahc0:0:2:0): CDB[0]=0x16 Key 0x5 ASC/ASCQ
 0x0/0x0 CAM STATUS 0xc flags 0x0 resid 0 dxfer_len 0
Oct 25 14:03:06 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): TES
T UNIT READY. CDB: 0 0 0 0 0 0 
Oct 25 14:03:06 lutefisk /kernel: (sa0:ahc0:0:2:0): LOAD UNLOAD. CDB: 1b 0 0 0 1
 0 
Oct 25 14:03:34 lutefisk /kernel: (sa0:ahc0:0:2:0): READ(06). CDB: 8 0 0 20 0 0 
Oct 25 14:03:40 lutefisk /kernel: (sa0:ahc0:0:2:0): CDB[0]=0x8 Key 0x0 ASC/ASCQ 
0x0/0x0 CAM STATUS 0xc flags 0x20 resid -2048 dxfer_len 8192
Oct 25 14:03:40 lutefisk /kernel: (sa0:ahc0:0:2:0): REWIND. CDB: 1 0 0 0 0 0 
Oct 25 14:03:40 lutefisk /kernel: (sa0:ahc0:0:2:0): CDB[0]=0x1 Key 0x0 ASC/ASCQ 
0x0/0x0 CAM STATUS 0xc flags 0x20 resid -2048 dxfer_len 0
Oct 25 14:03:40 lutefisk /kernel: (sa0:ahc0:0:2:0): REWIND. CDB: 1 0 0 0 0 0 
Oct 25 14:03:41 lutefisk /kernel: (sa0:ahc0:0:2:0): CDB[0]=0x1 Key 0x0 ASC/ASCQ 
0x0/0x0 CAM STATUS 0xc flags 0x20 resid -2048 dxfer_len 0
Oct 25 14:03:41 lutefisk /kernel: (sa0:ahc0:0:2:0): unable to rewind after test 
read
Oct 25 14:03:42 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): REL
EASE(06). CDB: 17 0 0 0 0 0 
Oct 25 14:03:42 lutefisk /kernel: (sa0:ahc0:0:2:0): CDB[0]=0x17 Key 0x5 ASC/ASCQ
 0x0/0x0 CAM STATUS 0xc flags 0x0 resid 0 dxfer_len 0

lutefisk# camcontrol debug off
Debugging turned off
lutefisk# mt -f /dev/nsa0 status
mt: /dev/nsa0: Input/output error

/var/log/messages says:
Oct 25 14:08:36 lutefisk /kernel: (sa0:ahc0:0:2:0): unable to rewind after test
read

lutefisk# camcontrol debug -Ic 0:2:0
Debugging enabled for 0:2:0
lutefisk# mt -f /dev/nsa0 status
mt: /dev/nsa0: Input/output error

/var/log/messages says:
Oct 25 14:11:01 lutefisk /kernel: (xpt0:ahc0:0:2:0): debugging flags now 9
Oct 25 14:11:08 lutefisk /kernel: (sa0:ahc0:0:2:0): saopen(0): dev=0x0 softc=0x0
Oct 25 14:11:08 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): RES
ERVE(06). CDB: 16 0 0 0 0 0 
Oct 25 14:11:08 lutefisk /kernel: (sa0:ahc0:0:2:0): CDB[0]=0x16 Key 0x5 ASC/ASCQ
 0x0/0x0 CAM STATUS 0xc flags 0x0 resid 0 dxfer_len 0
Oct 25 14:11:08 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): TES
T UNIT READY. CDB: 0 0 0 0 0 0 
Oct 25 14:11:08 lutefisk /kernel: (sa0:ahc0:0:2:0): LOAD UNLOAD. CDB: 1b 0 0 0 1
 0 
Oct 25 14:11:36 lutefisk /kernel: (sa0:ahc0:0:2:0): READ(06). CDB: 8 0 0 20 0 0 
Oct 25 14:11:42 lutefisk /kernel: (sa0:ahc0:0:2:0): CDB[0]=0x8 Key 0x0 ASC/ASCQ 
0x0/0x0 CAM STATUS 0xc flags 0x20 resid -2048 dxfer_len 8192
Oct 25 14:11:42 lutefisk /kernel: (sa0:ahc0:0:2:0): REWIND. CDB: 1 0 0 0 0 0 
Oct 25 14:11:42 lutefisk /kernel: (sa0:ahc0:0:2:0): CDB[0]=0x1 Key 0x0 ASC/ASCQ 
0x0/0x0 CAM STATUS 0xc flags 0x20 resid -2048 dxfer_len 0
Oct 25 14:11:42 lutefisk /kernel: (sa0:ahc0:0:2:0): REWIND. CDB: 1 0 0 0 0 0 
Oct 25 14:11:43 lutefisk /kernel: (sa0:ahc0:0:2:0): CDB[0]=0x1 Key 0x0 ASC/ASCQ 
0x0/0x0 CAM STATUS 0xc flags 0x20 resid -2048 dxfer_len 0
Oct 25 14:11:43 lutefisk /kernel: (sa0:ahc0:0:2:0): unable to rewind after test 
read
Oct 25 14:11:43 lutefisk /kernel: (sa0:ahc0:0:2:0): sastart(sa0:ahc0:0:2:0): REL
EASE(06). CDB: 17 0 0 0 0 0 
Oct 25 14:11:43 lutefisk /kernel: (sa0:ahc0:0:2:0): CDB[0]=0x17 Key 0x5 ASC/ASCQ
 0x0/0x0 CAM STATUS 0xc flags 0x0 resid 0 dxfer_len 0

Okay, I think that's enough.

It looks like there is a problem when you load the *second* tape.
Toggling the "camcontrol debug" status seemed to reset things so
it could operate successfully until you change tapes again, but it
doesn't seem to do this consistently.

Awaiting further instructions.

Jim
Comment 13 mjacob 2002-10-27 16:48:18 UTC 
> Okay, I think that's enough.
> 
> It looks like there is a problem when you load the *second* tape.
> Toggling the "camcontrol debug" status seemed to reset things so
> it could operate successfully until you change tapes again, but it
> doesn't seem to do this consistently.
> 
> Awaiting further instructions.
> 

Is this with the NODREAD quirk? 

I've also lost some original notes- what's the SCSI revision of this
device (per boot messages- didn't see this in the PR)

I've been considering making some 'pessimism' changes for all devices
less than SCSI-2. If this reports as a SCSI-2 device, then we'll just
quirk it and move on.


-matt
Comment 14 James Stephen Larson 2002-10-27 20:49:34 UTC 
In message <Pine.BSF.4.21.0210270845260.15449-100000@beppo> you write:
>> Okay, I think that's enough.
>> 
>> It looks like there is a problem when you load the *second* tape.
>> Toggling the "camcontrol debug" status seemed to reset things so
>> it could operate successfully until you change tapes again, but it
>> doesn't seem to do this consistently.
>> 
>> Awaiting further instructions.
>
>Is this with the NODREAD quirk? 

No, this is without any quirks.

>I've also lost some original notes- what's the SCSI revision of this
>device (per boot messages- didn't see this in the PR)

sa0 at ahc0 bus 0 target 2 lun 0
sa0: <EXABYTE EXB-8200 425A> Removable Sequential Access SCSI-CCS device 
sa0: 3.300MB/s transfers

>I've been considering making some 'pessimism' changes for all devices
>less than SCSI-2. If this reports as a SCSI-2 device, then we'll just
>quirk it and move on.

The "CCS" implies pre-SCSI-2.

I took out the CAMDEBUG stuff and added the "NODREAD" quirk.
It then was able to read from tapes without problems.
I then attempted to write a tape, and tripped over this
error:

(sa0:ahc0:0:2:0): unable to backspace over one of double filemarks at end of tap
e
(sa0:ahc0:0:2:0): it is possible that this device needs a SA_QUIRK_1FM quirk set
 for it
(sa0:ahc0:0:2:0): tape is now frozen- use an OFFLINE, REWIND or MTEOM command to
 clear this state.

This is what led me to my original quirk submission:
SA_QUIRK_1FM | SA_QUIRK_NODREAD.

If you have some pessimism changes, I'd be happy to beta them for you.

Jim
Comment 15 mjacob 2002-10-27 21:23:08 UTC 
> >> Okay, I think that's enough.
> >> 
> >> It looks like there is a problem when you load the *second* tape.
> >> Toggling the "camcontrol debug" status seemed to reset things so
> >> it could operate successfully until you change tapes again, but it
> >> doesn't seem to do this consistently.
> >> 
> >> Awaiting further instructions.
> >
> >Is this with the NODREAD quirk? 
> 
> No, this is without any quirks.
> 
> >I've also lost some original notes- what's the SCSI revision of this
> >device (per boot messages- didn't see this in the PR)
> 
> sa0 at ahc0 bus 0 target 2 lun 0
> sa0: <EXABYTE EXB-8200 425A> Removable Sequential Access SCSI-CCS device 
> sa0: 3.300MB/s transfers
> 
> >I've been considering making some 'pessimism' changes for all devices
> >less than SCSI-2. If this reports as a SCSI-2 device, then we'll just
> >quirk it and move on.
> 
> The "CCS" implies pre-SCSI-2.

Yes.


> 
> I took out the CAMDEBUG stuff and added the "NODREAD" quirk.
> It then was able to read from tapes without problems.
> I then attempted to write a tape, and tripped over this
> error:
> 
> (sa0:ahc0:0:2:0): unable to backspace over one of double filemarks at end of tap
> e
> (sa0:ahc0:0:2:0): it is possible that this device needs a SA_QUIRK_1FM quirk set
>  for it
> (sa0:ahc0:0:2:0): tape is now frozen- use an OFFLINE, REWIND or MTEOM command to
>  clear this state.
> 
> This is what led me to my original quirk submission:
> SA_QUIRK_1FM | SA_QUIRK_NODREAD.
> 
> If you have some pessimism changes, I'd be happy to beta them for you.

Try a 'mt seteotmodel 1' and try your tests again.

The pessimims would do the NODREAD case- but I'm still really peeved
about the double filemark issue.

This is clearly still and active problem.

-matt
Comment 16 James Stephen Larson 2002-10-27 21:31:28 UTC 
In message <Pine.BSF.4.21.0210271321110.15449-100000@beppo> you write:
>Try a 'mt seteotmodel 1' and try your tests again.

With any quirks enabled?  If so, which ones?  Do you want the
CAMDEBUG output too?

>The pessimims would do the NODREAD case- but I'm still really peeved
>about the double filemark issue.

What's so bad about the double filemark, of you don't mind my asking?
There's other equipment in the quirks table with the 1FM quirk.

>This is clearly still and active problem.

Yes.

Jim
Comment 17 mjacob 2002-10-27 21:52:56 UTC 
On Sun, 27 Oct 2002, Jim Larson wrote:

> In message <Pine.BSF.4.21.0210271321110.15449-100000@beppo> you write:
> >Try a 'mt seteotmodel 1' and try your tests again.
> 
> With any quirks enabled?  If so, which ones?  Do you want the
> CAMDEBUG output too?

NODREAD, and I don't think CAMDEBUG is telling me much at this point.
> 
> >The pessimims would do the NODREAD case- but I'm still really peeved
> >about the double filemark issue.
> 
> What's so bad about the double filemark, of you don't mind my asking?
> There's other equipment in the quirks table with the 1FM quirk.

This is a big topic- some of it having to do compatibility, some having
to do with various belief systems about 'emulating' devices that have
don't have an End of Data indicator (namely, 9 track 1/2" tape).

Some devices *can't* physically do double filemarks- QIC devices,
primarily. All other devices *should* be able to do double filemarks and
sort it out. Apparently this f/w flavor for the 8200 doesn't.

> 
> >This is clearly still and active problem.
> 
> Yes.
> 
> Jim
>
Comment 18 James Stephen Larson 2002-10-28 20:31:52 UTC 
In message <Pine.BSF.4.21.0210271349550.15449-100000@beppo> you write:
>On Sun, 27 Oct 2002, Jim Larson wrote:
>> In message <Pine.BSF.4.21.0210271321110.15449-100000@beppo> you write:
>> >Try a 'mt seteotmodel 1' and try your tests again.
>> 
>> With any quirks enabled?  If so, which ones?  Do you want the
>> CAMDEBUG output too?
>
>NODREAD, and I don't think CAMDEBUG is telling me much at this point.

Okay.  It looks like it's happy with just "NODREAD" as a quirk
and "mt seteotmodel 1".

Jim
Comment 19 mjacob 2002-10-28 20:44:14 UTC 
> > In message <Pine.BSF.4.21.0210271349550.15449-100000@beppo> you write:
> > >On Sun, 27 Oct 2002, Jim Larson wrote:
> > >> In message <Pine.BSF.4.21.0210271321110.15449-100000@beppo> you write:
> > >> >Try a 'mt seteotmodel 1' and try your tests again.
> > >> 
> > >> With any quirks enabled?  If so, which ones?  Do you want the
> > >> CAMDEBUG output too?
> > >
> > >NODREAD, and I don't think CAMDEBUG is telling me much at this point.
> > 
> > Okay.  It looks like it's happy with just "NODREAD" as a quirk
> > and "mt seteotmodel 1".
> 
> The mt command does the same thing as the 1FM quirk, setting 1 file mark.

Yes- of course.

My thinking here is still to take pre-SCSI2 devices and implicitly
NODREAD them amongst other things. But the 1FM issue I *can't* change
because of user expectations. 

What I *can* do is say, in 'known bugs', that the Exabyte 8200 with X
f/w may require this change.

The reason I'm doing this rather than  just adding a quirk is that we
should be doing everything we can to *reduce* quirks- not grow
them. Quirks are a tail chasing exercise and wrong. 

-matt
Comment 20 James Stephen Larson 2002-10-28 20:59:52 UTC 
In message <Pine.BSF.4.21.0210281241290.47358-100000@beppo> you write:
>> > In message <Pine.BSF.4.21.0210271349550.15449-100000@beppo> you write:
>> > Okay.  It looks like it's happy with just "NODREAD" as a quirk
>> > and "mt seteotmodel 1".
>> 
>> The mt command does the same thing as the 1FM quirk, setting 1 file mark.
>
>Yes- of course.
>
>My thinking here is still to take pre-SCSI2 devices and implicitly
>NODREAD them amongst other things. But the 1FM issue I *can't* change
>because of user expectations. 
>
>What I *can* do is say, in 'known bugs', that the Exabyte 8200 with X
>f/w may require this change.
>
>The reason I'm doing this rather than  just adding a quirk is that we
>should be doing everything we can to *reduce* quirks- not grow
>them. Quirks are a tail chasing exercise and wrong. 

As long as there's *some* way I can use my tape drive without making
custom changes to the driver, I'm happy.  An extra user-level step
is fine.  BTW, what's the persistence of the "seteotmodel"?  Until
the tape is ejected, or until a reboot?

Jim
Comment 21 mjacob 2002-10-28 21:05:59 UTC 
> As long as there's *some* way I can use my tape drive without making
> custom changes to the driver, I'm happy.  An extra user-level step
> is fine.  BTW, what's the persistence of the "seteotmodel"?  Until
> the tape is ejected, or until a reboot?

reboot or device detach/attach due to rescan
Comment 22 Peter da Silva 2002-12-21 05:28:00 UTC 
My DLT lost its brain, so I've borrowed an Exabyte to hold me over, and 
lo and behold it got the same message... *but* only after I'd 
successfully written half a dozen tapes and read a couple back. Here's 
my dmesg output...

FreeBSD 4.7-RELEASE #2: Fri Dec 13 14:54:43 CST 2002
...
ahc0: <Adaptec 2940 Ultra SCSI adapter> port 0xd000-0xd0ff mem 
0xdb000000-0xdb00
0fff irq 11 at device 8.0 on pci0
aic7880: Ultra Single Channel A, SCSI Id=7, 16/253 SCBs
...
(ahc0:A:5:0): refuses synchronous negotiation. Using asynchronous transfers
sa0 at ahc0 bus 0 target 5 lun 0
sa0: <EXABYTE EXB-8200 045H> Removable Sequential Access SCSI-CCS device
sa0: 3.300MB/s transfers

[[ This is the only device on the bus ]]

(sa0:ahc0:0:5:0): SPACE. CDB: 11 1 ff ff ff 0
(sa0:ahc0:0:5:0): HARDWARE FAILURE info?:ffffa000 asc:15,1
(sa0:ahc0:0:5:0): Mechanical positioning error
(sa0:ahc0:0:5:0): unable to backspace over one of double filemarks at 
end of tape
(sa0:ahc0:0:5:0): it is possible that this device needs a SA_QUIRK_1FM 
quirk set for it
(sa0:ahc0:0:5:0): tape is now frozen- use an OFFLINE, REWIND or MTEOM 
command to clear this state.

I'm not sure that this is a real problem, though, because the tape was 
very full... it may have actually run out of tape in which case an error 
like this seems plausible to me.

I don't know how to check the firmware level of the drive, but if 
there's anything you'd like me to check out just give a holler.
Comment 23 James Stephen Larson 2002-12-22 00:42:36 UTC 
In message <3E03FBE0.4070000@abbnm.com> you write:
>(sa0:ahc0:0:5:0): SPACE. CDB: 11 1 ff ff ff 0
>(sa0:ahc0:0:5:0): HARDWARE FAILURE info?:ffffa000 asc:15,1
>(sa0:ahc0:0:5:0): Mechanical positioning error
>(sa0:ahc0:0:5:0): unable to backspace over one of double filemarks at 
>end of tape
>(sa0:ahc0:0:5:0): it is possible that this device needs a SA_QUIRK_1FM 
>quirk set for it
>(sa0:ahc0:0:5:0): tape is now frozen- use an OFFLINE, REWIND or MTEOM 
>command to clear this state.

If you don't hear anything from someone more learned in the way
of SCSI, you might be able to make some progress with this.
Matthew Jacob (mjacob@feral.com) pointed out to me, you
can do a "mt seteotmodel 1" instead of setting SA_QUIRK_1FM.

Jim
Comment 24 Mark Linimon freebsd_committer freebsd_triage 2004-08-26 03:38:35 UTC 
Responsible Changed
From-To: mjacob->freebsd-bugs

With bugmeister hat on, reassign from inactive committer.
Comment 25 Mark Linimon freebsd_committer freebsd_triage 2004-08-26 07:07:48 UTC 
Responsible Changed
From-To: freebsd-bugs->mjacob

mjacob has reactived his commit bit.  mea culpa for the bogus reassignment.
Comment 26 Matt Jacob freebsd_committer freebsd_triage 2006-06-25 05:51:49 UTC 
State Changed
From-To: feedback->closed

Stale at this point.