Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 1 Dec 1998 11:45:47 -0700 (MST)
From:      "Kenneth D. Merry" <ken@plutotech.com>
To:        bsdx@spawnet.com (Adam McDougall)
Cc:        scsi@FreeBSD.ORG
Subject:   Re: tosha reports random errors after transition to CAM
Message-ID:  <199812011845.LAA21068@panzer.plutotech.com>
In-Reply-To: <3663EFA0.BD143233@spawnet.com> from Adam McDougall at "Dec 1, 98 08:31:12 am"

next in thread | previous in thread | raw e-mail | index | archive | help
Adam McDougall wrote...
> Matthew Hunt wrote:
> > 
> > Since upgrading from 2.2-STABLE to 3.0-CURRENT, and hence CAM, I
> > have observed random errors when using tosha to read CD audio data
> > from my CD-ROM drive.  The drive is branded IBM, but when I purchased
> > it, I was told that it is a Toshiba XM-4101.
> > 
> > When I try to rip CD audio, tosha emits the following error:
> > 
> > wopr:~/tmp$ tosha -v -t 1 -o out.pcm
> > Device: /dev/cd0c IBM CDRM00201 !F 0724
> > tosha: WARNING: Drive type not recognized.
> > Output file: out.pcm
> > 
> >  track   playing  start    end     raw size  mp3 size   # of
> >  number   time    sector  sector   in bytes  128 kbps  frames
> > --------------------------------------------------------------
> >     1    4:22'18       0   19667   46259136   4196675   10039
> > error returned from CD-DA read command:
> > (pass1:ahc0:0:3:0): Vendor Specific Command. CDB: d8
> > (pass1:ahc0:0:3:0): ILLEGAL REQUEST asc:bf,0
> > (pass1:ahc0:0:3:0): Vendor Specific ASC
> > 
> > The error is always the same, but it occurs at random places and does
> > not appear to depend on what CD or track I am ripping.  I never had
> > any problems pre-CAM.  I have (obviously) recompiled tosha to support
> > the CAM architecture.
> > 
> > Can anyone provide illumination?
> > Matt
> > 
> > --
> > Matthew Hunt <mph@pobox.com> * Inertia is a property of matter.
> > http://www.pobox.com/~mph/pgp.key for PGP public key 0x67203349.
> > 
> > To Unsubscribe: send mail to majordomo@FreeBSD.org
> > with "unsubscribe freebsd-scsi" in the body of the message
> 
> 
> I think I notice the same thing too. (well, similar)
> It started happening when I changed HD's from a Atlas II to a Viking UW. 

That could be your problem...I'll explain below.

> Seems like my symptoms are more drastic though, because when I run tosha
> (even when not storing the data being read on da0) I will get something
> like this:
>     1   63:07'00      33  284057  668026800  60592835  144972
>   (output file: track01.pcm)
> (da0:dpt0:0:0:0): WRITE(06). CDB: a 9 e0 40 10 0
> (da0:dpt0:0:0:0): ILLEGAL REQUEST asc:24,0
> (da0:dpt0:0:0:0): Invalid field in CDB field replaceable unit: 80
> sks:c0,9
> (da0:dpt0:0:0:0): READ(06). CDB: 8 2 95 58 8 0
> (da0:dpt0:0:0:0): ILLEGAL REQUEST asc:24,0
> (da0:dpt0:0:0:0): Invalid field in CDB field replaceable unit: 80
> sks:c0,9
> swap_pager: I/O error - pagein failed; blkno 46424, size 4096, error 22
> vm_fault: pager read error, pid 96 (syslogd)
> (da0:dpt0:0:0:0): WRITE(06). CDB: a 1 0 50 10 0
> (da0:dpt0:0:0:0): ILLEGAL REQUEST asc:24,0
> (da0:dpt0:0:0:0): Invalid field in CDB field replaceable unit: 80
> sks:c0,9
> ^C
> (da0:dpt0:0:0:0): READ(06). CDB: 8 2 9 b0 28 0
> (da0:dpt0:0:0:0): ILLEGAL REQUEST asc:24,0
> (da0:dpt0:0:0:0): Invalid field in CDB field replaceable unit: 80
> sks:c0,9

This is rather bizzare, since the drive is complaining about byte 9 of a 6
byte CDB.

Justin suspects that the drive may not be handling 6-byte reads and writes
properly.

> dpt0: <DPT Caching SCSI RAID Controller> rev 0x02 int a irq 12 on
> pci0.17.0
> dpt0: DPT PM2144UW FW Rev. 07LY, 1 channel, 64 CCBs
> da0: <QUANTUM VIKING 4.5 WSE 8808> Fixed Direct Access SCSI2 device 
> da0: Tagged Queueing Enabled
> da0: 4345MB (8899736 512 byte sectors: 255H 63S/T 553C)

Try this --  go into sys/cam/scsi/scsi_da.c, and on about line 1079, adjust
the minimum_cmd_size for the read and write commands to 10 bytes.

It could be that the disk isn't handling 6-byte reads and writes.  It could
also be some problem with the DPT driver, or the DPT firmware.  (Simon
would probably have a better idea on this one.)  I suspect, though, that
it may be a drive firmware problem.

Ken
-- 
Kenneth Merry
ken@plutotech.com

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-scsi" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199812011845.LAA21068>