Very Computer

Suse 9.1, 2.6.5-7.95-default, cdrecord version: cdrecord-2.01a27-21

A long time ago I had trouble getting md5sum's of .iso files and the
burned data CD-R to agree.  Then I discovered that the drive needed to
be reset by opening/closing the drawer.  After that, it always worked.

Now I have a new hardware setup, with a drive from Sony:

Vendor_info    : 'SONY    '
Identifikation : 'CD-RW  CRX230ED '
Revision       : '4YS1'

Whereas before I had a LiteON.  I haven't tried putting a LiteON in this
machine to see if the problems go away.

Basically what happens now is like this:

Here is the output of my backup script:

ef81ce67788429be829a9f8e7337fc49  /home/user1/cd-rw/datasheets.backup.iso
e41a295d467515cf6403ba93d858f39f  /dev/cdrecorder

NOTE that sometimes the result is different, and instead of inconsistent
md5sums, I get an IO error on the read of /dev/cdrecorder.  This time it
gave a wrong sum.  Actually, the reading bailed out after a very short
time, but didn't report an IO error like usual.

Now I mount the drive, and everything "looks" fine; all my files are there.

Now I run md5sum by hand again to make sure:

e41a295d467515cf6403ba93d858f39f  /dev/cdrecorder

Now I eject and close AGAIN (I already did this in the script):

ef81ce67788429be829a9f8e7337fc49  /dev/cdrecorder

Any tips on why the drive is being so stupid?

Here's my script:

#!/bin/bash
echo We are in the datasheets.CD-RW.backup script!
source ~/backup.scripts/backup.settings
mkisofs -o ~/cd-rw/datasheets.backup.iso -R -J -v -graft-points ~/datasheets
echo \n
du -h ~/cd-rw/datasheets.backup.iso
echo \n
echo Press Enter to continue or Ctrl-Cto exit...
read REPLY
cdrecord dev=$CDRW_SCSI_DEV -v blank=fast -dao ~/cd-rw/datasheets.backup.iso
cdrecord dev=$CDRW_SCSI_DEV -eject
cdrecord dev=$CDRW_SCSI_DEV -load
echo \n
md5sum ~/cd-rw/datasheets.backup.iso
md5sum $CDRW_DEV_PATH
echo \n
echo Press Enter or close when done...
read REPLY
exit

Here's the settings file:

#echo We are in the backup.settings file
CDRW_SCSI_DEV=ATA:3,0,0
CDRW_DEV_PATH=/dev/cdrecorder

--
_____________________
Christopher R. Carlen

SuSE 9.1 Linux 2.6.5

I took their script and added an argument error check.

http://groups.google.com/advanced_group_search
rawread_cd              in the first box
bit twister             in the author box

try this.

$ isoinfo -d -i  $CDRW_DEV_PATH

CD-ROM is ISO 9660 format
System id: LINUX
Volume id: XXXXXX
.......
Logical block size is: 2048
Volume size is: 349916

El Torito VD version 1 found, boot catalog is in sector 492...
.......

read and set the values
# Logical block size is:
BSIZE=2048

# Volume size is:
VSIZE=349916

$ dd if=$CDRW_DEV_PATH  bs=$BSIZE count=$VSIZE conv=notrunc,noerror |
md5sum

122ac8924883a455ee37ff473a691617 -

// moma
    http://www.futuredesktop.org/how2burn.html#mirrors
    http://www.futuredesktop.org/hpc_linux.html

When I*discoverd this effect I dod go down to a file-by-file
md5 somparison several times. I never found any problems.

Arno
--
For email address: lastname AT tik DOT ee DOT ethz DOT ch
GnuPG: ID:1E25338F FP:0C30 5782 9D93 F785 E79C  0296 797F 6B50 1E25 338F
  "The more corrupt the state, the more numerous the laws" - Tacitus

Before using the drive as a burner, I mounted and unmounted a disc, then
did some md5sums directly and using your way.  No problems.

Then I did a burn.

23f82c3254ac40cc3a08d0fdec324a06  /home/user1/cd-rw/full.backup.iso
md5sum: /dev/cdrecorder: Input/output error

[snip]
Logical block size is: 2048
Volume size is: 292006
[snip]

conv=notrunc,noerror | md5sum
dd: reading `/dev/cdrecorder': Input/output error
102112+0 records in
102112+0 records out
dd: reading `/dev/cdrecorder': Input/output error
102112+0 records in
102112+0 records out
etc.
etc.
etc.

Maybe I need to get a non-tweaked cdrecord:

ZY$?: Operation not permitted. WARNING: Cannot set RR-scheduler
ZY$?: Permission denied. WARNING: Cannot set priority using setpriority().
ZY$?: WARNING: This causes a high risk for buffer underruns.
Cdrecord-Clone-dvd 2.01a27 (i686-suse-linux) Copyright (C) 1995-2004
J?rg Schilling
Note: This version is an unofficial (modified) version with DVD support
Note: and therefore may have bugs that are not present in the original.
Note: Please send bug reports or support requests to
http://www.suse.de/feedback
Note: The author of cdrecord should not be bothered with problems in
this version.
cdrecord: No such file or directory. WARNING: Cannot do mlockall(2).
cdrecord: WARNING: This causes a high risk for buffer underruns.
cdrecord: Warning: not running as root user, fs= option ignored.
scsidev: 'ATA:3,0,0'
devname: 'ATA'
scsibus: 3 target: 0 lun: 0
Linux sg driver version: 3.5.27
Using libscg version 'schily-0.8'.
cdrecord: Warning: using inofficial libscg transport code version

1.80 04/03/08 Copyright 1997 J. Schilling').
Device type    : Removable CD-ROM
Version        : 0
Response Format: 2
Capabilities   :
Vendor_info    : 'SONY    '
Identifikation : 'CD-RW  CRX230ED '
Revision       : '4YS1'
Device seems to be: Generic mmc CD-RW.
Using generic SCSI-3/mmc   CD-R/CD-RW driver (mmc_cdr).
Driver flags   : MMC-3 SWABAUDIO BURNFREE FORCESPEED
Supported modes: TAO PACKET SAO SAO/R96P SAO/R96R RAW/R16 RAW/R96P RAW/R96R

--
_____________________
Christopher R. Carlen

SuSE 9.1 Linux 2.6.5

You clearly need superuser rights to succeed.

Please try again with other activities (such as
networking) shut off and with root privileges.

--

Tauno Voipio
tauno voipio (at) iki fi

However, it is not cdrecord that is actually failing.  It is the read of
the drive *afterward*.  That gives IO errors.

Here's what happened:

1.  Start Linux, do md5sum on /dev/cdrecorder -- success, no errors.
2.  Burn with cdrecord -- success, no errors and reports from cdrecord
(not shown in my post) indicated no buffer underruns, despite the warning.
3.  Do md5sum on /dev/cdrecorder -- errors, though the same disk after a
reboot will md5sum just fine.

Conclusion:  I think cdrecord is doing something at the driver level,
that is not getting reset, leaving the drive in a strange state.  Oddly,
I CAN mount the disk and read the files, without errors though I should
qualify this by saying that I haven't tried reading all the files' data,
rather I just see the directory structure intact, and a randomly
selected file reads OK.  But since the disk always md5sums correctly
after a reboot, I think this proves that the data is in fact intact.

Furthermore, on a different machine with a different drive, the same
*software* setup doesn't have a problem with md5sums after the burns.
This other machine also gives the warnings from cdrecord seen above.
(See example below.)

I think I will do the following:

1.  try running cdrecord as root just to see if it makes a difference

2.  try a different physical drive in the problem machine.  If that
makes the problem go away, I'd be convinced that there is a driver level
or cdrecord interaction with driver problem, that is specific to that
hardware.

Example of operation on different machine, also not as root (some excess
output snipped):

 > cdrecord dev=$CDRW_SCSI_DEV -v blank=fast -dao ~/cd-rw/full.backup.iso
ZY$?: Operation not permitted. WARNING: Cannot set RR-scheduler
ZY$?: Permission denied. WARNING: Cannot set priority using setpriority().
ZY$?: WARNING: This causes a high risk for buffer underruns.
Cdrecord-Clone-dvd 2.01a27 (i686-suse-linux) Copyright (C) 1995-2004
J?rg Schilling
Note: This version is an unofficial (modified) version with DVD support
Note: and therefore may have bugs that are not present in the original.
Note: Please send bug reports or support requests to
http://www.suse.de/feedback
Note: The author of cdrecord should not be bothered with problems in
this version.
TOC Type: 1 = CD-ROM
cdrecord: No such file or directory. WARNING: Cannot do mlockall(2).
cdrecord: WARNING: This causes a high risk for buffer underruns.
cdrecord: Warning: not running as root user, fs= option ignored.
scsidev: 'ATA:3,0,0'
devname: 'ATA'
scsibus: 3 target: 0 lun: 0
Linux sg driver version: 3.5.27
Using libscg version 'schily-0.8'.
cdrecord: Warning: using inofficial libscg transport code version

1.80 04/03/08 Copyright 1997 J. Schilling').
SCSI buffer size: 64512
atapi: -1
Device type    : Removable CD-ROM
Version        : 0
Response Format: 2
Capabilities   :
Vendor_info    : 'LITE-ON '
Identifikation : 'LTR-52327S      '
Revision       : 'QS53'
Device seems to be: Generic mmc CD-RW.
Current: 0x000A
Profile: 0x000A (current)
Profile: 0x0009
Profile: 0x0008
Profile: 0x0002
Using generic SCSI-3/mmc   CD-R/CD-RW driver (mmc_cdr).
Driver flags   : MMC-3 SWABAUDIO BURNFREE FORCESPEED
Supported modes: TAO PACKET SAO SAO/R96P SAO/R96R RAW/R16 RAW/R96P RAW/R96R
Drive buf size : 1419264 = 1386 KB
FIFO size      : 4194304 = 4096 KB
Track 01: data   560 MB
Total size:      644 MB (63:48.20) = 287115 sectors
Lout start:      644 MB (63:50/15) = 287115 sectors
Current Secsize: 2048
ATIP info from disk:
   Indicated writing power: 1
   Reference speed: 0
   Is not unrestricted
   Is erasable
   Disk sub type: Ultra High speed Rewritable media (2)
   ATIP start of lead in:  -11076 (97:34/24)
   ATIP start of lead out: 336075 (74:43/00)
   1T speed low: 16 1T speed high: 16
   2T speed low:  8 2T speed high: 24
   power mult factor: 4 5
   recommended erase/write power: 1
   A1 values: 66 4A 99
   A2 values: 38 80 00
   A3 values: 04 C4 A0
Disk type:    Phase change
Manuf. index: 11
Manufacturer: Mitsubishi Chemical Corporation
cdrecord: Cannot get next writable address for 'invisible' track.
cdrecord: This means that we are checking recorded media.
cdrecord: This media cannot be written in streaming mode anymore.
cdrecord: If you like to write to 'preformatted' RW media, try to blank
a first.
Forcespeed is OFF.
Starting to write CD/DVD at speed  24.0 in real SAO mode for single ses
Last chance to quit, starting real write    0 seconds. Operation starts
Waiting for reader process to fill input buffer ... input buffer ready.
Performing OPC...
Blanking PMA, TOC, pregap
Blanking time:   11.491s
BURN-Free is OFF.
Performing OPC...
Sending CUE sheet...
Writing pregap for track 1 at -150
Starting new track at sector: 0
Track 01:  560 of  560 MB written (fifo 100%) [buf  99%]  24.6x.
Track 01: Total bytes read/written: 588011520/588011520 (287115 sectors
Writing  time:  171.403s
Average write speed  22.3x.
Min drive buffer fill was 83%
Fixating...
Fixating time:    4.051s
cdrecord: fifo had 9262 puts and 9262 gets.
cdrecord: fifo was 0 times empty and 7206 times full, min fill was 85%.

 > cdrecord dev=$CDRW_SCSI_DEV -eject
[output snipped]
 > cdrecord dev=$CDRW_SCSI_DEV -load
[output snipped]

md5sum ~/cd-rw/full.backup.iso
c1a34ee315712828ca4c5b712d679f24  /home/crcarle/cd-rw/full.backup.iso
md5sum $CDRW_DEV_PATH
c1a34ee315712828ca4c5b712d679f24  /dev/cdrecorder

Press Enter or close when done...

Good day!

--
_______________________________________________________________________
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA

Sorry for not reading with the brain in gear.

Is jettisoning the drive out of question. It reminds me of the LG drives
killing their internal Flash when asked for a simple ATA standard request.

Quote:> Good day!

--

Tauno Voipio
tauno voipio (at) iki fi